Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

On 10/30/15, 3:19 PM, "Justin Mclean" <ju...@classsoftware.com> wrote:

>Hi,
>
>> Hmm, I was hoping more PMC folks would respond.  Remember that,
>>according
>> to the release process, the PMC folks planning to vote are supposed to
>>be
>> running tests now. In theory, the only new test to be run  after we
>>start
>> the vote is whether the PGP signature is valid.
>
>We’re continually trying to test a moving target which involves a greater
>time commitment that I currently have available. You’ve never quite sure
>if the version you testing is going to be the version in the release
>candidate and unless you very carefully follow all of the commits it’s
>not obvious what needs to be retested at two different time intervals.

Hmm, pleading is working so maybe I’ll try guilt-tripping.

Yes, the nightly builds are a moving target.  IMO, we all want to grow the
community by attracting customers and hopefully convert a few to be
committers and the only way I know to do it is to keep making the code
better and releasing the best release we can in the most efficient manner.
 IMO, freezing a branch and not allowing important bug fixes that might
make a difference in whether someone becomes more active in our community
doesn’t make sense.  Taking the time to build out an RC and post it and
start a vote thread in order to finally get some testing isn’t very
efficient either.

Historically, when we produced an RC and immediately started a release
vote, bugs would be found and we’d cancel the RC and roll out another one.
The goal of the release process we voted in, IMO, was to reduce this
overhead of posting RCs, opening and closing vote threads, etc. so we can
more efficiently achieve the goal of serving our customers and attracting
some of them to becoming committers so we can have more people find bugs
sooner by working with the develop branch.

Recently, I’ve spent several days on improving build and approval scripts
so testing what is in development takes less time.  In theory, you can now
start up the approval script which will pull down the bits, answer a few
questions, then go do something else for 5 to 25 minutes and then come
back and poke at it.  I would have rather spent that time on features for
our customers, but I gambled that this would help us get the release out
sooner.  I’m not sure that paid off.


I don’t have any other ideas on how to make it easier for those of you who
contribute in your spare time to stay up on the commits and bugs.  It
should be ok to take any nightly build and run it through your tests and
report your findings.  Ideally, you would be up to date on the commits and
bugs and other discussions to know whether what you find is a duplicate or
not, but at this point, I don’t care if you report a duplicate.  At least
that means you verified a bunch of other code paths worked for you.  I
don’t know how other Apache projects with really active code bases do it.

It is certainly fine to be too busy to vote on a release.  I was hoping to
get more folks to poke at the bits before starting a vote because it will
be a waste of community energy to start a vote and then have a bunch of
PMC voters jump in and start reporting important bugs.  But I think the
community has waited too long already, so I am going to start a vote soon,
and Peter and I will vote and maybe Josh and/or Harbs and we’ll be good to
go.  Hopefully any others jumping in late won’t find release blockers and
we’ll just make another release later.


-Alex

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

There are many scripts.

ApproveFlexJS.xml takes you through the steps required to approve a
release.
Installer.xml converts a binary release into an IDE-compatible folder
structure.

In build.xml, there is an ‘all’ target that attempts to clone all the
upstream repositories and build them in the correct order.
Also in build.xml, the main target will detect that you are running from
an IDE-compatible folder structure and prompt you for a new empty folder
to use to clone all of the upstream repositories and build them in the
correct order.

I’m not quite sure I understand what scenario you are trying to resolve.
Most folks trying to use FlexJS instead of Flex to build a web app should
just be able to use the same Apache Flex Installer app most folks use to
install Flex and will get all of the dependencies and folder structures
required to use the results in Flash Builder and maybe IntelliJ.  And use
the same monkey-patch techniques to workaround bugs.

The scripts target in build.xml are for those trying to work from our git
repos to fix bugs and add features to FlexJS.

Thanks,
-Alex

On 11/5/15, 3:47 PM, "jude" <fl...@gmail.com> wrote:

>Ok, the reason why I was so confused is because it's "ApproveFlexJS"
>build.
>:P I thought it was the install and run script. Is there a another ant
>build file?
>
>On Thu, Nov 5, 2015 at 3:21 PM, jude <fl...@gmail.com> wrote:
>
>> Alex,
>> I'm going to go ahead and put together an app so that you can install
>> FlexJS with one click and compile. I want it to be as simple as possible
>> for anyone (new web developers). I've downloaded the build script and
>> looked over it but there is some code in it for voting and release. I'm
>> sure you will be repeating yourself again but here's my requirements:
>>
>> • button to validate FlexJS requirements (paths, etc)
>> • button to set paths for the user
>> • one button to download FlexJS
>> • another button to run FlexJS on a project app (without an IDE)
>>
>> Are there targets to do each of these (or where should I start)? Since
>> it's an AIR app I want to run each target independently. BTW I'm
>>bundling
>> ant with the app so that the user doesn't have to install anything
>>extra.
>> If someone has FB or IntelliJ installed does the install directory need
>>to
>> change? I'll put the project on github or post it to the group as soon
>>as
>> it's usable.
>>
>>
>> On Thu, Nov 5, 2015 at 6:32 AM, Alex Harui <ah...@adobe.com> wrote:
>>
>>>
>>>
>>> On 11/5/15, 3:56 AM, "Harbs" <ha...@gmail.com> wrote:
>>>
>>> >I’m a bit confused about the release process.
>>> >
>>> >I thought we were creating release branches in git for each release to
>>> >“freeze” the code, so we do not have a wildly moving target. It does
>>>not
>>> >seem like that’s happening, so I’m not sure if I just misunderstood.
>>>
>>> You understand correctly.  I’m cheating right now because it is just
>>>more
>>> work to set up release branches, and there isn’t a lot of non-critical
>>> work going on the develop branch.  Peter and I are working on the back
>>> port from JS to AS in a separate branch.  Only important fixes are
>>>being
>>> pushed to develop.  If we had more folks contributing more often, then
>>>I
>>> would have used a release branch.
>>>
>>> Even release branches have historically moved because folks don’t start
>>> testing until late in the game and find important bugs.
>>>
>>> -Alex
>>>
>>> >
>>> >Harbs
>>> >
>>> >On Nov 3, 2015, at 12:04 AM, Alex Harui <ah...@adobe.com> wrote:
>>> >
>>> >>
>>> >>
>>> >> On 10/30/15, 3:19 PM, "Justin Mclean" <ju...@classsoftware.com>
>>> wrote:
>>> >>
>>> >>> Hi,
>>> >>>
>>> >>>> Hmm, I was hoping more PMC folks would respond.  Remember that,
>>> >>>> according
>>> >>>> to the release process, the PMC folks planning to vote are
>>>supposed
>>> to
>>> >>>> be
>>> >>>> running tests now. In theory, the only new test to be run  after
>>>we
>>> >>>> start
>>> >>>> the vote is whether the PGP signature is valid.
>>> >>>
>>> >>> We’re continually trying to test a moving target which involves a
>>> >>>greater
>>> >>> time commitment that I currently have available. You’ve never quite
>>> >>>sure
>>> >>> if the version you testing is going to be the version in the
>>>release
>>> >>> candidate and unless you very carefully follow all of the commits
>>>it’s
>>> >>> not obvious what needs to be retested at two different time
>>>intervals.
>>> >>
>>> >> Hmm, pleading is working so maybe I’ll try guilt-tripping.
>>> >>
>>> >> Yes, the nightly builds are a moving target.  IMO, we all want to
>>>grow
>>> >>the
>>> >> community by attracting customers and hopefully convert a few to be
>>> >> committers and the only way I know to do it is to keep making the
>>>code
>>> >> better and releasing the best release we can in the most efficient
>>> >>manner.
>>> >> IMO, freezing a branch and not allowing important bug fixes that
>>>might
>>> >> make a difference in whether someone becomes more active in our
>>> >>community
>>> >> doesn’t make sense.  Taking the time to build out an RC and post it
>>>and
>>> >> start a vote thread in order to finally get some testing isn’t very
>>> >> efficient either.
>>> >>
>>> >> Historically, when we produced an RC and immediately started a
>>>release
>>> >> vote, bugs would be found and we’d cancel the RC and roll out
>>>another
>>> >>one.
>>> >> The goal of the release process we voted in, IMO, was to reduce this
>>> >> overhead of posting RCs, opening and closing vote threads, etc. so
>>>we
>>> >>can
>>> >> more efficiently achieve the goal of serving our customers and
>>> >>attracting
>>> >> some of them to becoming committers so we can have more people find
>>> bugs
>>> >> sooner by working with the develop branch.
>>> >>
>>> >> Recently, I’ve spent several days on improving build and approval
>>> >>scripts
>>> >> so testing what is in development takes less time.  In theory, you
>>>can
>>> >>now
>>> >> start up the approval script which will pull down the bits, answer a
>>> few
>>> >> questions, then go do something else for 5 to 25 minutes and then
>>>come
>>> >> back and poke at it.  I would have rather spent that time on
>>>features
>>> >>for
>>> >> our customers, but I gambled that this would help us get the release
>>> out
>>> >> sooner.  I’m not sure that paid off.
>>> >>
>>> >>
>>> >> I don’t have any other ideas on how to make it easier for those of
>>>you
>>> >>who
>>> >> contribute in your spare time to stay up on the commits and bugs.
>>>It
>>> >> should be ok to take any nightly build and run it through your tests
>>> and
>>> >> report your findings.  Ideally, you would be up to date on the
>>>commits
>>> >>and
>>> >> bugs and other discussions to know whether what you find is a
>>>duplicate
>>> >>or
>>> >> not, but at this point, I don’t care if you report a duplicate.  At
>>> >>least
>>> >> that means you verified a bunch of other code paths worked for you.
>>> I
>>> >> don’t know how other Apache projects with really active code bases
>>>do
>>> >>it.
>>> >>
>>> >> It is certainly fine to be too busy to vote on a release.  I was
>>>hoping
>>> >>to
>>> >> get more folks to poke at the bits before starting a vote because it
>>> >>will
>>> >> be a waste of community energy to start a vote and then have a
>>>bunch of
>>> >> PMC voters jump in and start reporting important bugs.  But I think
>>>the
>>> >> community has waited too long already, so I am going to start a vote
>>> >>soon,
>>> >> and Peter and I will vote and maybe Josh and/or Harbs and we’ll be
>>>good
>>> >>to
>>> >> go.  Hopefully any others jumping in late won’t find release
>>>blockers
>>> >>and
>>> >> we’ll just make another release later.
>>> >>
>>> >>
>>> >> -Alex
>>> >
>>>
>>>
>>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[jude <fl...@gmail.com>]

Ok, the reason why I was so confused is because it's "ApproveFlexJS" build.
:P I thought it was the install and run script. Is there a another ant
build file?

On Thu, Nov 5, 2015 at 3:21 PM, jude <fl...@gmail.com> wrote:

> Alex,
> I'm going to go ahead and put together an app so that you can install
> FlexJS with one click and compile. I want it to be as simple as possible
> for anyone (new web developers). I've downloaded the build script and
> looked over it but there is some code in it for voting and release. I'm
> sure you will be repeating yourself again but here's my requirements:
>
> • button to validate FlexJS requirements (paths, etc)
> • button to set paths for the user
> • one button to download FlexJS
> • another button to run FlexJS on a project app (without an IDE)
>
> Are there targets to do each of these (or where should I start)? Since
> it's an AIR app I want to run each target independently. BTW I'm bundling
> ant with the app so that the user doesn't have to install anything extra.
> If someone has FB or IntelliJ installed does the install directory need to
> change? I'll put the project on github or post it to the group as soon as
> it's usable.
>
>
> On Thu, Nov 5, 2015 at 6:32 AM, Alex Harui <ah...@adobe.com> wrote:
>
>>
>>
>> On 11/5/15, 3:56 AM, "Harbs" <ha...@gmail.com> wrote:
>>
>> >I’m a bit confused about the release process.
>> >
>> >I thought we were creating release branches in git for each release to
>> >“freeze” the code, so we do not have a wildly moving target. It does not
>> >seem like that’s happening, so I’m not sure if I just misunderstood.
>>
>> You understand correctly.  I’m cheating right now because it is just more
>> work to set up release branches, and there isn’t a lot of non-critical
>> work going on the develop branch.  Peter and I are working on the back
>> port from JS to AS in a separate branch.  Only important fixes are being
>> pushed to develop.  If we had more folks contributing more often, then I
>> would have used a release branch.
>>
>> Even release branches have historically moved because folks don’t start
>> testing until late in the game and find important bugs.
>>
>> -Alex
>>
>> >
>> >Harbs
>> >
>> >On Nov 3, 2015, at 12:04 AM, Alex Harui <ah...@adobe.com> wrote:
>> >
>> >>
>> >>
>> >> On 10/30/15, 3:19 PM, "Justin Mclean" <ju...@classsoftware.com>
>> wrote:
>> >>
>> >>> Hi,
>> >>>
>> >>>> Hmm, I was hoping more PMC folks would respond.  Remember that,
>> >>>> according
>> >>>> to the release process, the PMC folks planning to vote are supposed
>> to
>> >>>> be
>> >>>> running tests now. In theory, the only new test to be run  after we
>> >>>> start
>> >>>> the vote is whether the PGP signature is valid.
>> >>>
>> >>> We’re continually trying to test a moving target which involves a
>> >>>greater
>> >>> time commitment that I currently have available. You’ve never quite
>> >>>sure
>> >>> if the version you testing is going to be the version in the release
>> >>> candidate and unless you very carefully follow all of the commits it’s
>> >>> not obvious what needs to be retested at two different time intervals.
>> >>
>> >> Hmm, pleading is working so maybe I’ll try guilt-tripping.
>> >>
>> >> Yes, the nightly builds are a moving target.  IMO, we all want to grow
>> >>the
>> >> community by attracting customers and hopefully convert a few to be
>> >> committers and the only way I know to do it is to keep making the code
>> >> better and releasing the best release we can in the most efficient
>> >>manner.
>> >> IMO, freezing a branch and not allowing important bug fixes that might
>> >> make a difference in whether someone becomes more active in our
>> >>community
>> >> doesn’t make sense.  Taking the time to build out an RC and post it and
>> >> start a vote thread in order to finally get some testing isn’t very
>> >> efficient either.
>> >>
>> >> Historically, when we produced an RC and immediately started a release
>> >> vote, bugs would be found and we’d cancel the RC and roll out another
>> >>one.
>> >> The goal of the release process we voted in, IMO, was to reduce this
>> >> overhead of posting RCs, opening and closing vote threads, etc. so we
>> >>can
>> >> more efficiently achieve the goal of serving our customers and
>> >>attracting
>> >> some of them to becoming committers so we can have more people find
>> bugs
>> >> sooner by working with the develop branch.
>> >>
>> >> Recently, I’ve spent several days on improving build and approval
>> >>scripts
>> >> so testing what is in development takes less time.  In theory, you can
>> >>now
>> >> start up the approval script which will pull down the bits, answer a
>> few
>> >> questions, then go do something else for 5 to 25 minutes and then come
>> >> back and poke at it.  I would have rather spent that time on features
>> >>for
>> >> our customers, but I gambled that this would help us get the release
>> out
>> >> sooner.  I’m not sure that paid off.
>> >>
>> >>
>> >> I don’t have any other ideas on how to make it easier for those of you
>> >>who
>> >> contribute in your spare time to stay up on the commits and bugs.  It
>> >> should be ok to take any nightly build and run it through your tests
>> and
>> >> report your findings.  Ideally, you would be up to date on the commits
>> >>and
>> >> bugs and other discussions to know whether what you find is a duplicate
>> >>or
>> >> not, but at this point, I don’t care if you report a duplicate.  At
>> >>least
>> >> that means you verified a bunch of other code paths worked for you.  I
>> >> don’t know how other Apache projects with really active code bases do
>> >>it.
>> >>
>> >> It is certainly fine to be too busy to vote on a release.  I was hoping
>> >>to
>> >> get more folks to poke at the bits before starting a vote because it
>> >>will
>> >> be a waste of community energy to start a vote and then have a bunch of
>> >> PMC voters jump in and start reporting important bugs.  But I think the
>> >> community has waited too long already, so I am going to start a vote
>> >>soon,
>> >> and Peter and I will vote and maybe Josh and/or Harbs and we’ll be good
>> >>to
>> >> go.  Hopefully any others jumping in late won’t find release blockers
>> >>and
>> >> we’ll just make another release later.
>> >>
>> >>
>> >> -Alex
>> >
>>
>>
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

On 11/5/15, 10:04 PM, "jude" <fl...@gmail.com> wrote:

>Yep, that's the plan. I'm still trying to catch up on what web developers
>are doing these days and I haven't heard IDE's come up much (anyone
>wishing
>to share I would appreciate it). It seems like people are using code
>editors like Brackets and vim. There are some tools that run preprocessing
>or post processing on CSS, others that minify JS, others that do other
>stuff. Some people have terminal open watching a directory to run these
>things. I wish I could tell you more. So, to me it would be nice to create
>the Flex JS project directory, then have a button to compile or watch the
>project. Maybe a launch button or something.

Sounds good.

>
>By contrast, if we were writing HTML you would edit the page, cmd + s and
>then refresh in the browser. With FlexJS we have edit, save, compile,
>refresh. There's that one extra step. With a lightweight FlexJS app we
>could compile on a different thread in the background and return errors
>and
>warnings. They could still use the same workflow. It wouldn't be much but
>something someone can use to get started.

FWIW, one of reasons Falcon and FalconJX emit MXML as data instead of code
like the current MXMLC compiler does for Flex is to open the potential of
skipping the compile step for certain kinds of changes.  In theory, this
mini-IDE you write could connect to a running app and as you edit certain
MXML properties it could alter the data in the app.  It won’t work for
everything like changing code in event handlers, but could work for
positioning-related properties, which is where I think it would be a big
win.

>
>I'm glad to hear the Apache Installer takes care of installing FlexJS. Can
>someone post a link to the ant build file (that compiles a project) I
>should be studying and nightly source for the Apache installer (that has
>the ant build that installs FJS)?
>

See examples/build_example.xml for building the example apps.  Each
example folder has a build.xml that calls build_example.xml.

See installer.xml for the ant script that the Apache Flex Installer runs.
When the Apache Flex installer is installing FlexJS, the installer isn’t
much more than a GUI front end on Ant.

Sounds very interesting.  Best of luck.  And of course, it would be best
if you could manage the code in an ASF repository, but not a requirement.

-Alex

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Justin Mclean <ju...@classsoftware.com>]

Hi,

> I haven't heard IDE's come up much (anyone wishing
> to share I would appreciate it). It seems like people are using code
> editors like Brackets and vim.

Most common editors I’ve seen used in that space recently are TextMate, Sublime, Atom, and Brackets. Perhaps WebStorm (IntelliJ) is also worth a look?

Justin

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[jude <fl...@gmail.com>]

Yep, that's the plan. I'm still trying to catch up on what web developers
are doing these days and I haven't heard IDE's come up much (anyone wishing
to share I would appreciate it). It seems like people are using code
editors like Brackets and vim. There are some tools that run preprocessing
or post processing on CSS, others that minify JS, others that do other
stuff. Some people have terminal open watching a directory to run these
things. I wish I could tell you more. So, to me it would be nice to create
the Flex JS project directory, then have a button to compile or watch the
project. Maybe a launch button or something.

By contrast, if we were writing HTML you would edit the page, cmd + s and
then refresh in the browser. With FlexJS we have edit, save, compile,
refresh. There's that one extra step. With a lightweight FlexJS app we
could compile on a different thread in the background and return errors and
warnings. They could still use the same workflow. It wouldn't be much but
something someone can use to get started.

People would use their own editors. I'm not sure if I mentioned it but
there is a wrapper class
<https://github.com/monkeypunch3/flexcapacitor/blob/master/MainLibrary/src/com/flexcapacitor/controls/AceEditor.as>
for the Ace editor <http://ace.c9.io/>. You can embed that in your AIR
apps. There is also a web version. We could use that for view source. Check
it out. There's a lot you can do with it. It's possible to support AS3 and
MXML code completion but someone has to write the language files for it or
at least hook into the code completion extension points dynamically. I
looked into it briefly and it was over my head. FYI I haven't supported the
full Ace API but it's getting there. It was half, now it's about 3/4. It's
not hard to add more just other priorities. There's an example project that
shows how to use it. I might be able to post it here if we can post zips.
For now, I would be focusing on just making a few buttons create, compile
and launch projects (maybe also install).

I'm glad to hear the Apache Installer takes care of installing FlexJS. Can
someone post a link to the ant build file (that compiles a project) I
should be studying and nightly source for the Apache installer (that has
the ant build that installs FJS)?


On Thu, Nov 5, 2015 at 5:10 PM, Alex Harui <ah...@adobe.com> wrote:

> Jude,
>
> Are you envisioning some sort of a mini-IDE?  One button creates a new
> project, another button compiles it, another button cross-compiles it?
> That would be cool.  Especially for folks who don’t have FB or IntelliJ.
> Not sure what to do about code editing and debugging though.
>
> Another related idea that has been in the back of my mind is that we don’t
> have a way to do “Source View” without an IDE.  Of course, you can see the
> output JS in any browser debugger, but I wonder if packaging and viewing
> the MXML and AS for the app would help with adoption.  IMO, the compiler
> should be mostly capable of doing it by adding a new emitter so that’s a
> project for anyone wanting to code in Java, but on the other hand, if a we
> didn’t package up annotated HTML of the MXML and AS and had some sort of
> AIR app that would colorize the source automatically, that would be cool
> too, but you would need some sort of AST to do it, I think.
>
> -Alex
>
> On 11/5/15, 4:52 PM, "Josh Tynjala" <jo...@gmail.com> wrote:
>
> >As Alex said, most people who want to try out FlexJS 0.5.0 will simply use
> >the Apache Flex SDK Installer. It'll be super easy for them. In fact, you
> >can already install the nightlies and RC1 using the installer if you
> >right-click and choose Show Dev Builds. Once 0.5.0 is released, the build
> >will show up on the regular list, obviously.
> >
> >People who are having trouble with building are trying to build from the
> >source distribution and/or the Git repository. Things are a little more
> >complicated that way, but the vast majority of people won't need to deal
> >with that type of build.
> >
> >- Josh
> >
> >On Thu, Nov 5, 2015 at 3:21 PM, jude <fl...@gmail.com> wrote:
> >
> >> Alex,
> >> I'm going to go ahead and put together an app so that you can install
> >> FlexJS with one click and compile. I want it to be as simple as possible
> >> for anyone (new web developers). I've downloaded the build script and
> >> looked over it but there is some code in it for voting and release. I'm
> >> sure you will be repeating yourself again but here's my requirements:
> >>
> >> • button to validate FlexJS requirements (paths, etc)
> >> • button to set paths for the user
> >> • one button to download FlexJS
> >> • another button to run FlexJS on a project app (without an IDE)
> >>
> >> Are there targets to do each of these (or where should I start)? Since
> >>it's
> >> an AIR app I want to run each target independently. BTW I'm bundling ant
> >> with the app so that the user doesn't have to install anything extra. If
> >> someone has FB or IntelliJ installed does the install directory need to
> >> change? I'll put the project on github or post it to the group as soon
> >>as
> >> it's usable.
> >>
> >>
> >> On Thu, Nov 5, 2015 at 6:32 AM, Alex Harui <ah...@adobe.com> wrote:
> >>
> >> >
> >> >
> >> > On 11/5/15, 3:56 AM, "Harbs" <ha...@gmail.com> wrote:
> >> >
> >> > >I’m a bit confused about the release process.
> >> > >
> >> > >I thought we were creating release branches in git for each release
> >>to
> >> > >“freeze” the code, so we do not have a wildly moving target. It does
> >>not
> >> > >seem like that’s happening, so I’m not sure if I just misunderstood.
> >> >
> >> > You understand correctly.  I’m cheating right now because it is just
> >>more
> >> > work to set up release branches, and there isn’t a lot of non-critical
> >> > work going on the develop branch.  Peter and I are working on the back
> >> > port from JS to AS in a separate branch.  Only important fixes are
> >>being
> >> > pushed to develop.  If we had more folks contributing more often,
> >>then I
> >> > would have used a release branch.
> >> >
> >> > Even release branches have historically moved because folks don’t
> >>start
> >> > testing until late in the game and find important bugs.
> >> >
> >> > -Alex
> >> >
> >> > >
> >> > >Harbs
> >> > >
> >> > >On Nov 3, 2015, at 12:04 AM, Alex Harui <ah...@adobe.com> wrote:
> >> > >
> >> > >>
> >> > >>
> >> > >> On 10/30/15, 3:19 PM, "Justin Mclean" <ju...@classsoftware.com>
> >> wrote:
> >> > >>
> >> > >>> Hi,
> >> > >>>
> >> > >>>> Hmm, I was hoping more PMC folks would respond.  Remember that,
> >> > >>>> according
> >> > >>>> to the release process, the PMC folks planning to vote are
> >>supposed
> >> to
> >> > >>>> be
> >> > >>>> running tests now. In theory, the only new test to be run  after
> >>we
> >> > >>>> start
> >> > >>>> the vote is whether the PGP signature is valid.
> >> > >>>
> >> > >>> We’re continually trying to test a moving target which involves a
> >> > >>>greater
> >> > >>> time commitment that I currently have available. You’ve never
> >>quite
> >> > >>>sure
> >> > >>> if the version you testing is going to be the version in the
> >>release
> >> > >>> candidate and unless you very carefully follow all of the commits
> >> it’s
> >> > >>> not obvious what needs to be retested at two different time
> >> intervals.
> >> > >>
> >> > >> Hmm, pleading is working so maybe I’ll try guilt-tripping.
> >> > >>
> >> > >> Yes, the nightly builds are a moving target.  IMO, we all want to
> >>grow
> >> > >>the
> >> > >> community by attracting customers and hopefully convert a few to be
> >> > >> committers and the only way I know to do it is to keep making the
> >>code
> >> > >> better and releasing the best release we can in the most efficient
> >> > >>manner.
> >> > >> IMO, freezing a branch and not allowing important bug fixes that
> >>might
> >> > >> make a difference in whether someone becomes more active in our
> >> > >>community
> >> > >> doesn’t make sense.  Taking the time to build out an RC and post it
> >> and
> >> > >> start a vote thread in order to finally get some testing isn’t very
> >> > >> efficient either.
> >> > >>
> >> > >> Historically, when we produced an RC and immediately started a
> >>release
> >> > >> vote, bugs would be found and we’d cancel the RC and roll out
> >>another
> >> > >>one.
> >> > >> The goal of the release process we voted in, IMO, was to reduce
> >>this
> >> > >> overhead of posting RCs, opening and closing vote threads, etc. so
> >>we
> >> > >>can
> >> > >> more efficiently achieve the goal of serving our customers and
> >> > >>attracting
> >> > >> some of them to becoming committers so we can have more people find
> >> bugs
> >> > >> sooner by working with the develop branch.
> >> > >>
> >> > >> Recently, I’ve spent several days on improving build and approval
> >> > >>scripts
> >> > >> so testing what is in development takes less time.  In theory, you
> >>can
> >> > >>now
> >> > >> start up the approval script which will pull down the bits, answer
> >>a
> >> few
> >> > >> questions, then go do something else for 5 to 25 minutes and then
> >>come
> >> > >> back and poke at it.  I would have rather spent that time on
> >>features
> >> > >>for
> >> > >> our customers, but I gambled that this would help us get the
> >>release
> >> out
> >> > >> sooner.  I’m not sure that paid off.
> >> > >>
> >> > >>
> >> > >> I don’t have any other ideas on how to make it easier for those of
> >>you
> >> > >>who
> >> > >> contribute in your spare time to stay up on the commits and bugs.
> >>It
> >> > >> should be ok to take any nightly build and run it through your
> >>tests
> >> and
> >> > >> report your findings.  Ideally, you would be up to date on the
> >>commits
> >> > >>and
> >> > >> bugs and other discussions to know whether what you find is a
> >> duplicate
> >> > >>or
> >> > >> not, but at this point, I don’t care if you report a duplicate.  At
> >> > >>least
> >> > >> that means you verified a bunch of other code paths worked for
> >>you.  I
> >> > >> don’t know how other Apache projects with really active code bases
> >>do
> >> > >>it.
> >> > >>
> >> > >> It is certainly fine to be too busy to vote on a release.  I was
> >> hoping
> >> > >>to
> >> > >> get more folks to poke at the bits before starting a vote because
> >>it
> >> > >>will
> >> > >> be a waste of community energy to start a vote and then have a
> >>bunch
> >> of
> >> > >> PMC voters jump in and start reporting important bugs.  But I think
> >> the
> >> > >> community has waited too long already, so I am going to start a
> >>vote
> >> > >>soon,
> >> > >> and Peter and I will vote and maybe Josh and/or Harbs and we’ll be
> >> good
> >> > >>to
> >> > >> go.  Hopefully any others jumping in late won’t find release
> >>blockers
> >> > >>and
> >> > >> we’ll just make another release later.
> >> > >>
> >> > >>
> >> > >> -Alex
> >> > >
> >> >
> >> >
> >>
>
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

Jude,

Are you envisioning some sort of a mini-IDE?  One button creates a new
project, another button compiles it, another button cross-compiles it?
That would be cool.  Especially for folks who don’t have FB or IntelliJ.
Not sure what to do about code editing and debugging though.

Another related idea that has been in the back of my mind is that we don’t
have a way to do “Source View” without an IDE.  Of course, you can see the
output JS in any browser debugger, but I wonder if packaging and viewing
the MXML and AS for the app would help with adoption.  IMO, the compiler
should be mostly capable of doing it by adding a new emitter so that’s a
project for anyone wanting to code in Java, but on the other hand, if a we
didn’t package up annotated HTML of the MXML and AS and had some sort of
AIR app that would colorize the source automatically, that would be cool
too, but you would need some sort of AST to do it, I think.

-Alex

On 11/5/15, 4:52 PM, "Josh Tynjala" <jo...@gmail.com> wrote:

>As Alex said, most people who want to try out FlexJS 0.5.0 will simply use
>the Apache Flex SDK Installer. It'll be super easy for them. In fact, you
>can already install the nightlies and RC1 using the installer if you
>right-click and choose Show Dev Builds. Once 0.5.0 is released, the build
>will show up on the regular list, obviously.
>
>People who are having trouble with building are trying to build from the
>source distribution and/or the Git repository. Things are a little more
>complicated that way, but the vast majority of people won't need to deal
>with that type of build.
>
>- Josh
>
>On Thu, Nov 5, 2015 at 3:21 PM, jude <fl...@gmail.com> wrote:
>
>> Alex,
>> I'm going to go ahead and put together an app so that you can install
>> FlexJS with one click and compile. I want it to be as simple as possible
>> for anyone (new web developers). I've downloaded the build script and
>> looked over it but there is some code in it for voting and release. I'm
>> sure you will be repeating yourself again but here's my requirements:
>>
>> • button to validate FlexJS requirements (paths, etc)
>> • button to set paths for the user
>> • one button to download FlexJS
>> • another button to run FlexJS on a project app (without an IDE)
>>
>> Are there targets to do each of these (or where should I start)? Since
>>it's
>> an AIR app I want to run each target independently. BTW I'm bundling ant
>> with the app so that the user doesn't have to install anything extra. If
>> someone has FB or IntelliJ installed does the install directory need to
>> change? I'll put the project on github or post it to the group as soon
>>as
>> it's usable.
>>
>>
>> On Thu, Nov 5, 2015 at 6:32 AM, Alex Harui <ah...@adobe.com> wrote:
>>
>> >
>> >
>> > On 11/5/15, 3:56 AM, "Harbs" <ha...@gmail.com> wrote:
>> >
>> > >I’m a bit confused about the release process.
>> > >
>> > >I thought we were creating release branches in git for each release
>>to
>> > >“freeze” the code, so we do not have a wildly moving target. It does
>>not
>> > >seem like that’s happening, so I’m not sure if I just misunderstood.
>> >
>> > You understand correctly.  I’m cheating right now because it is just
>>more
>> > work to set up release branches, and there isn’t a lot of non-critical
>> > work going on the develop branch.  Peter and I are working on the back
>> > port from JS to AS in a separate branch.  Only important fixes are
>>being
>> > pushed to develop.  If we had more folks contributing more often,
>>then I
>> > would have used a release branch.
>> >
>> > Even release branches have historically moved because folks don’t
>>start
>> > testing until late in the game and find important bugs.
>> >
>> > -Alex
>> >
>> > >
>> > >Harbs
>> > >
>> > >On Nov 3, 2015, at 12:04 AM, Alex Harui <ah...@adobe.com> wrote:
>> > >
>> > >>
>> > >>
>> > >> On 10/30/15, 3:19 PM, "Justin Mclean" <ju...@classsoftware.com>
>> wrote:
>> > >>
>> > >>> Hi,
>> > >>>
>> > >>>> Hmm, I was hoping more PMC folks would respond.  Remember that,
>> > >>>> according
>> > >>>> to the release process, the PMC folks planning to vote are
>>supposed
>> to
>> > >>>> be
>> > >>>> running tests now. In theory, the only new test to be run  after
>>we
>> > >>>> start
>> > >>>> the vote is whether the PGP signature is valid.
>> > >>>
>> > >>> We’re continually trying to test a moving target which involves a
>> > >>>greater
>> > >>> time commitment that I currently have available. You’ve never
>>quite
>> > >>>sure
>> > >>> if the version you testing is going to be the version in the
>>release
>> > >>> candidate and unless you very carefully follow all of the commits
>> it’s
>> > >>> not obvious what needs to be retested at two different time
>> intervals.
>> > >>
>> > >> Hmm, pleading is working so maybe I’ll try guilt-tripping.
>> > >>
>> > >> Yes, the nightly builds are a moving target.  IMO, we all want to
>>grow
>> > >>the
>> > >> community by attracting customers and hopefully convert a few to be
>> > >> committers and the only way I know to do it is to keep making the
>>code
>> > >> better and releasing the best release we can in the most efficient
>> > >>manner.
>> > >> IMO, freezing a branch and not allowing important bug fixes that
>>might
>> > >> make a difference in whether someone becomes more active in our
>> > >>community
>> > >> doesn’t make sense.  Taking the time to build out an RC and post it
>> and
>> > >> start a vote thread in order to finally get some testing isn’t very
>> > >> efficient either.
>> > >>
>> > >> Historically, when we produced an RC and immediately started a
>>release
>> > >> vote, bugs would be found and we’d cancel the RC and roll out
>>another
>> > >>one.
>> > >> The goal of the release process we voted in, IMO, was to reduce
>>this
>> > >> overhead of posting RCs, opening and closing vote threads, etc. so
>>we
>> > >>can
>> > >> more efficiently achieve the goal of serving our customers and
>> > >>attracting
>> > >> some of them to becoming committers so we can have more people find
>> bugs
>> > >> sooner by working with the develop branch.
>> > >>
>> > >> Recently, I’ve spent several days on improving build and approval
>> > >>scripts
>> > >> so testing what is in development takes less time.  In theory, you
>>can
>> > >>now
>> > >> start up the approval script which will pull down the bits, answer
>>a
>> few
>> > >> questions, then go do something else for 5 to 25 minutes and then
>>come
>> > >> back and poke at it.  I would have rather spent that time on
>>features
>> > >>for
>> > >> our customers, but I gambled that this would help us get the
>>release
>> out
>> > >> sooner.  I’m not sure that paid off.
>> > >>
>> > >>
>> > >> I don’t have any other ideas on how to make it easier for those of
>>you
>> > >>who
>> > >> contribute in your spare time to stay up on the commits and bugs.
>>It
>> > >> should be ok to take any nightly build and run it through your
>>tests
>> and
>> > >> report your findings.  Ideally, you would be up to date on the
>>commits
>> > >>and
>> > >> bugs and other discussions to know whether what you find is a
>> duplicate
>> > >>or
>> > >> not, but at this point, I don’t care if you report a duplicate.  At
>> > >>least
>> > >> that means you verified a bunch of other code paths worked for
>>you.  I
>> > >> don’t know how other Apache projects with really active code bases
>>do
>> > >>it.
>> > >>
>> > >> It is certainly fine to be too busy to vote on a release.  I was
>> hoping
>> > >>to
>> > >> get more folks to poke at the bits before starting a vote because
>>it
>> > >>will
>> > >> be a waste of community energy to start a vote and then have a
>>bunch
>> of
>> > >> PMC voters jump in and start reporting important bugs.  But I think
>> the
>> > >> community has waited too long already, so I am going to start a
>>vote
>> > >>soon,
>> > >> and Peter and I will vote and maybe Josh and/or Harbs and we’ll be
>> good
>> > >>to
>> > >> go.  Hopefully any others jumping in late won’t find release
>>blockers
>> > >>and
>> > >> we’ll just make another release later.
>> > >>
>> > >>
>> > >> -Alex
>> > >
>> >
>> >
>>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

Hi Yishay,

If you have a few moments to write down some things that were hard or
confusing, it would be great.

For sure, I haven’t spent any time on using FB to develop components.
Once we finish the back port so all of our code can be written in AS and
parts cross-compiled to JS, then I’ll put more time into making FB work
better.

-Alex

On 11/8/15, 7:58 AM, "yishayw" <yi...@hotmail.com> wrote:

>Hi Alex,
>
>I've been following these threads (usually in silence) because I still
>hope
>to use FlexJS and contribute to it. Recently I've had very little time for
>this, as I started a new Flex (not JS) project which is taking up most of
>my
>computer time.
>
>The comments made by Harbs and others regarding the complexity of setting
>up
>a working environment pertain to me as well. The WIKI had some mistakes
>and
>incomplete instructions, which I usually had to spend some time to figure
>out. This forum has proven very helpful, and I will probably surface
>problems earlier on, next time I have time to put into this. That would
>also
>help give you and others a better understanding of the kind of things that
>set me back.
>
>I still have a mind to finish the Video components and contribute them. If
>someone else wants to do it, that would be cool too. I could post what
>I've
>got somewhere accessible.
>
>As a side note, the project I'm working on now is in the process of being
>migrated to JS (I think they're using AS3JS). If they'd started the
>migration after this release (now that Grids and layouts are working
>better)
>I think I would recommend using FlexJS instead. So even from my limited
>vantage point, I can see an immediate market for what you guys are doing.
>I
>hope you keep it up.
>
>Thanks,
>Yishay
>
>
>
>
>
>--
>View this message in context:
>http://apache-flex-development.2333347.n4.nabble.com/DISCUSS-Release-Apach
>e-FlexJS-0-5-0-tp49347p49912.html
>Sent from the Apache Flex Development mailing list archive at Nabble.com.

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[yishayw <yi...@hotmail.com>]

Hi Alex,

I've been following these threads (usually in silence) because I still hope
to use FlexJS and contribute to it. Recently I've had very little time for
this, as I started a new Flex (not JS) project which is taking up most of my
computer time.

The comments made by Harbs and others regarding the complexity of setting up
a working environment pertain to me as well. The WIKI had some mistakes and
incomplete instructions, which I usually had to spend some time to figure
out. This forum has proven very helpful, and I will probably surface
problems earlier on, next time I have time to put into this. That would also
help give you and others a better understanding of the kind of things that
set me back.

I still have a mind to finish the Video components and contribute them. If
someone else wants to do it, that would be cool too. I could post what I've
got somewhere accessible.

As a side note, the project I'm working on now is in the process of being
migrated to JS (I think they're using AS3JS). If they'd started the
migration after this release (now that Grids and layouts are working better)
I think I would recommend using FlexJS instead. So even from my limited
vantage point, I can see an immediate market for what you guys are doing. I
hope you keep it up.

Thanks,
Yishay





--
View this message in context: http://apache-flex-development.2333347.n4.nabble.com/DISCUSS-Release-Apache-FlexJS-0-5-0-tp49347p49912.html
Sent from the Apache Flex Development mailing list archive at Nabble.com.

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

The script is looking for gnu privacy app to verify the signature. Please install it.


Sent from my LG G3, an AT&T 4G LTE smartphone


------ Original message------

From: Harbs

Date: Sat, Nov 7, 2015 12:48 PM

To: dev@flex.apache.org;

Subject:Re: [DISCUSS] Release Apache FlexJS 0.5.0


I just tried running the approval script and it failed:
/Users/harbs/Documents/ApacheFlex/ApproveFlexJS/ApproveFlexJS.xml:236: The following error occurred while executing this line:
/Users/harbs/Documents/ApacheFlex/ApproveFlexJS/ApproveFlexJS.xml:240: Execute failed: java.io.IOException: Cannot run program "gpg": error=2, No such file or directory

When I run ant -e -f ApproveFlexJS.xml -Drelease.version=0.5.0, I do not get the error. It’s only when I add -Drc=1 That it fails.

Clues?

On Nov 6, 2015, at 11:38 PM, Alex Harui <ah...@adobe.com> wrote:

>
>
> On 11/6/15, 12:47 PM, "jude" <fl...@gmail.com> wrote:
>
>> I said this in another thread but since you asked I would say this would
>> be
>> the perfect workflow for me (for both Flex and FlexJS):
>>
>> 1. Provide URL to source on Github (the project would have to be on github
>> obviously)
>
> Our repos are already mirrored to GitHub.  See:
> https://github.com/apache/flex-sdk
>
>> 2. I create a local repository from that
>> 3. I open FB and import a default *existing project* that is in the local
>> repository*
>>  -- in my project i add a spark Button
>>  -- I then open spark.Button.as and make changes
>
> Try importing all the projects from frameworks/projects.  For Flex, you
> should be able to create a new project and make these projects
> dependencies.  For FlexJS, maybe you can create that default existing
> project for us.
>
>> 4. I compile the project and run it. I see the changes immediately
>> 5. I commit those changes locally and push them up to the main branch to
>> be
>> accepted or rejected.
>
> From the mirrors you can generate a pull request.
>
> For FlexJS, I think you can install FlexJS as a Flex SDK and use it as the
> SDK for your project and you should be ok.  Give it a try.  Let us know
> what is still missing.
>
>
>>
>> I put a lot of work in one of my AS3 / Flex libraries and it bothers me
>> that that work is not in the Flex SDK. It may not be as high quality as
>> some code out there but that can be fixed by other community members
>> reviewing it and making it better.
>>
>> We have AS3corelib, tons of AS3 libraries and components on github and
>> google code and maybe not all of them, but a majority of them would give
>> the Flex SDK an amazing API to do just about anything. And I believe if
>> they were in the main repository they would be getting more use and
>> improved upon daily.
>
> For better or worse, Apache generally has more rules than many GitHub-only
> projects.  Those rules slow us down, but in theory are the price to pay to
> make corporations feel comfortable using our software, and supposedly,
> that’s where the money is for our future.
>
> You are welcome to donate more of your code to Apache Flex and get it into
> the default SDK.  If you wrote the code before becoming a committer, you
> have to go through the Software Grant process again for significant code
> bases.  Small stuff you can probably just commit, although a discussion
> before doing so would be recommended.
>
> -Alex
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Harbs <ha...@gmail.com>]

I just tried running the approval script and it failed:
/Users/harbs/Documents/ApacheFlex/ApproveFlexJS/ApproveFlexJS.xml:236: The following error occurred while executing this line:
/Users/harbs/Documents/ApacheFlex/ApproveFlexJS/ApproveFlexJS.xml:240: Execute failed: java.io.IOException: Cannot run program "gpg": error=2, No such file or directory

When I run ant -e -f ApproveFlexJS.xml -Drelease.version=0.5.0, I do not get the error. It’s only when I add -Drc=1 That it fails.

Clues?

On Nov 6, 2015, at 11:38 PM, Alex Harui <ah...@adobe.com> wrote:

> 
> 
> On 11/6/15, 12:47 PM, "jude" <fl...@gmail.com> wrote:
> 
>> I said this in another thread but since you asked I would say this would
>> be
>> the perfect workflow for me (for both Flex and FlexJS):
>> 
>> 1. Provide URL to source on Github (the project would have to be on github
>> obviously)
> 
> Our repos are already mirrored to GitHub.  See:
> https://github.com/apache/flex-sdk
> 
>> 2. I create a local repository from that
>> 3. I open FB and import a default *existing project* that is in the local
>> repository*
>>  -- in my project i add a spark Button
>>  -- I then open spark.Button.as and make changes
> 
> Try importing all the projects from frameworks/projects.  For Flex, you
> should be able to create a new project and make these projects
> dependencies.  For FlexJS, maybe you can create that default existing
> project for us.
> 
>> 4. I compile the project and run it. I see the changes immediately
>> 5. I commit those changes locally and push them up to the main branch to
>> be
>> accepted or rejected.
> 
> From the mirrors you can generate a pull request.
> 
> For FlexJS, I think you can install FlexJS as a Flex SDK and use it as the
> SDK for your project and you should be ok.  Give it a try.  Let us know
> what is still missing.
> 
> 
>> 
>> I put a lot of work in one of my AS3 / Flex libraries and it bothers me
>> that that work is not in the Flex SDK. It may not be as high quality as
>> some code out there but that can be fixed by other community members
>> reviewing it and making it better.
>> 
>> We have AS3corelib, tons of AS3 libraries and components on github and
>> google code and maybe not all of them, but a majority of them would give
>> the Flex SDK an amazing API to do just about anything. And I believe if
>> they were in the main repository they would be getting more use and
>> improved upon daily.
> 
> For better or worse, Apache generally has more rules than many GitHub-only
> projects.  Those rules slow us down, but in theory are the price to pay to
> make corporations feel comfortable using our software, and supposedly,
> that’s where the money is for our future.
> 
> You are welcome to donate more of your code to Apache Flex and get it into
> the default SDK.  If you wrote the code before becoming a committer, you
> have to go through the Software Grant process again for significant code
> bases.  Small stuff you can probably just commit, although a discussion
> before doing so would be recommended.
> 
> -Alex
> 

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

On 11/6/15, 12:47 PM, "jude" <fl...@gmail.com> wrote:

>I said this in another thread but since you asked I would say this would
>be
>the perfect workflow for me (for both Flex and FlexJS):
>
>1. Provide URL to source on Github (the project would have to be on github
>obviously)

Our repos are already mirrored to GitHub.  See:
https://github.com/apache/flex-sdk

>2. I create a local repository from that
>3. I open FB and import a default *existing project* that is in the local
>repository*
>   -- in my project i add a spark Button
>   -- I then open spark.Button.as and make changes

Try importing all the projects from frameworks/projects.  For Flex, you
should be able to create a new project and make these projects
dependencies.  For FlexJS, maybe you can create that default existing
project for us.

>4. I compile the project and run it. I see the changes immediately
>5. I commit those changes locally and push them up to the main branch to
>be
>accepted or rejected.

From the mirrors you can generate a pull request.

For FlexJS, I think you can install FlexJS as a Flex SDK and use it as the
SDK for your project and you should be ok.  Give it a try.  Let us know
what is still missing.


>
>I put a lot of work in one of my AS3 / Flex libraries and it bothers me
>that that work is not in the Flex SDK. It may not be as high quality as
>some code out there but that can be fixed by other community members
>reviewing it and making it better.
>
>We have AS3corelib, tons of AS3 libraries and components on github and
>google code and maybe not all of them, but a majority of them would give
>the Flex SDK an amazing API to do just about anything. And I believe if
>they were in the main repository they would be getting more use and
>improved upon daily.

For better or worse, Apache generally has more rules than many GitHub-only
projects.  Those rules slow us down, but in theory are the price to pay to
make corporations feel comfortable using our software, and supposedly,
that’s where the money is for our future.

You are welcome to donate more of your code to Apache Flex and get it into
the default SDK.  If you wrote the code before becoming a committer, you
have to go through the Software Grant process again for significant code
bases.  Small stuff you can probably just commit, although a discussion
before doing so would be recommended.

-Alex

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[jude <fl...@gmail.com>]

I said this in another thread but since you asked I would say this would be
the perfect workflow for me (for both Flex and FlexJS):

1. Provide URL to source on Github (the project would have to be on github
obviously)
2. I create a local repository from that
3. I open FB and import a default *existing project* that is in the local
repository*
   -- in my project i add a spark Button
   -- I then open spark.Button.as and make changes
4. I compile the project and run it. I see the changes immediately
5. I commit those changes locally and push them up to the main branch to be
accepted or rejected.

*No environment variables to setup, no projects to setup, etc. It should be
bog standard 2 step process: import repo locally, begin work.

I put a lot of work in one of my AS3 / Flex libraries and it bothers me
that that work is not in the Flex SDK. It may not be as high quality as
some code out there but that can be fixed by other community members
reviewing it and making it better.

We have AS3corelib, tons of AS3 libraries and components on github and
google code and maybe not all of them, but a majority of them would give
the Flex SDK an amazing API to do just about anything. And I believe if
they were in the main repository they would be getting more use and
improved upon daily.

On Fri, Nov 6, 2015 at 9:52 AM, Alex Harui <ah...@adobe.com> wrote:

> Let me see if I can summarize these various “simplify the build threads
> here”.
>
> For me, I have to go by what I read on our mailing lists.  Maybe there is
> some silent faction waiting for simpler builds, but I am going to reward
> those who speak up by trying to work on their issues.  It doesn’t seem
> right to tell someone who reports a bug that it will be several weeks
> before I can fix it because I am worried about attracting people who we
> don’t know exists.  Go back through the archives.  I rarely see new
> potential committers complaining about how to get started writing FlexJS
> code, especially after the recent changes I made to try to make that
> simpler which was in response to the last flurry of “simplify the build”
> emails from several of our current committers.  Strangely, when I
> completed that work and asked folks to test it, I got no responses.  For
> example, I didn’t see any responses from folks who could be in the silent
> faction saying that it made it better, so that’s why I’m just not sure how
> important it is to make it simpler than it is now.
>
> Harbs had issues getting started and IIRC, mostly it was environment
> variables so maybe all of these emails boils down to the need to try to
> reduce the reliance on environment variables.  I’m definitely open to
> spending some of my time on that.
>
> I believe Maven is important, but I do not wish stop working on other
> stuff to learn Maven and do the heavy lifting on it.  I hope Chris will
> stick around to do it and that others can help them.  I will try to help
> with refactoring needed to make Maven work better.  Those changes usually
> help make sure we have the loose-coupling we need between modules.  I’m
> not that interested in refactoring the Flex SDK.   I am interested in the
> notion of slicing off any Flex SDK dependencies in the Falcon code base so
> we can Mavenize FlexJS without dealing with the Flex SDK first.  And I’m
> actually looking forward to seeing the Mavenizing of FlexJS.  Maybe I’ll
> learn more about Maven by watching those commits.
>
> I’m sure in the early days of FlexJS and Falcon, going away for a while
> and coming back probably didn’t work.  But is that still true?  And if so,
> can you just blow away your repos and re-grab flex-asjs and run “ant all”
> and get back up and running?
>
> And for folks reading this who are in the silent faction:  Please speak
> up.  Or write me off-list if there is something you have to say that you
> don’t want to say on these lists.  And tell me/us why you have been
> silent.  Is there something we are doing or should be doing to make it
> easier for you to speak up?
>
> So, my takeaway is that we should fork off two actionable threads.  One
> about eliminating environment variables for Flash/AIR, and the other about
> folder restructuring in flex-asjs to make Mavenizing easier?  Did I miss
> anything?
>
> Meanwhile, I am trying to upgrade the build scripts to fix issues found in
> the FalconJX packages so that we can get releases out faster in the future
> by catching issues sooner.
>
> If you are still having problems with this FlexJS release candidate,
> please provide more detail.
>
> Thanks,
> -Alex
>
> On 11/6/15, 1:08 AM, "Justin Mclean" <ju...@me.com> wrote:
>
> >HI,
> >
> >> Nobody is saying simpler builds wouldn’t help.  But how much?
> >
> >A lot I think. Developer are unable to contribute if they can’t compile
> >and test their changes.
> >
> >>  More than a more functional DataGrid?  Or a more Spark and/or MX-like
> >>component set
> >> that reduces migration time for your app?  Or being able to write your
> >>JS
> >> using AS?
> >
> >IMO you would more likely get people contribution things like that if it
> >was simpler to build.
> >
> >> To folks reading this who are interested in
> >> FlexJS: what is the most important two or three things we should be
> >> working on?
> >
> >IMO:
> >1. Simpler builds
> >2. More frequent releases (say every 3 or 4 months).
> >3. Compiling MXML
> >4. Working documented examples / how to port existing applications
> >
> >> The build script was improved in this release such that if you have a
> >> couple of known sets of circumstances, like not having any of the repos,
> >> that running “ant all” will get all the repos and build them in the
> >>right
> >> order.
> >
> >I meant to ask before does that use well known versions or does it just
> >grab the latest out of the develop (or another) branch?
> >
> >Thanks,
> >Justin
>
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

On 11/6/15, 12:36 PM, "omuppi1@gmail.com on behalf of OmPrakash Muppirala"
<omuppi1@gmail.com on behalf of bigosmallm@gmail.com> wrote:
>>
>> I’m sure in the early days of FlexJS and Falcon, going away for a while
>> and coming back probably didn’t work.  But is that still true?  And if
>>so,
>> can you just blow away your repos and re-grab flex-asjs and run “ant
>>all”
>> and get back up and running?
>>
>
>It is an option.  But, I have several partially done cool things that were
>stopped due to personal time issues, build issues, etc.  I would hate to
>lose all that.  But I will probably just end up doing what you suggested.

Are those changes in repos other than flex-asjs?  You can probably leave
your flex-asjs folder intact.  Just blow away all the sibling folders like
flex-falcon and flex-sdk and then run “ant all” from flex-asjs.

>
>
>>
>> And for folks reading this who are in the silent faction:  Please speak
>> up.  Or write me off-list if there is something you have to say that you
>> don’t want to say on these lists.  And tell me/us why you have been
>> silent.  Is there something we are doing or should be doing to make it
>> easier for you to speak up?
>>
>
>I don't think the silent faction are paying attention, to this thread, or
>even to this forum.  Ideally, we should make things easy to build, do a
>couple of simple tutorials and invite folks to try it out.  That would be
>a
>better approach to make them speak up.

How do we reach them to invite them?  Why aren’t they on our mailing lists
and what can we do get them to listen here?  I think we have to try to use
this channel as our main place for discussion.

Thanks,
-Alex

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[OmPrakash Muppirala <bi...@gmail.com>]

>
>   I hope Chris will
> stick around to do it and that others can help them.  I will try to help
> with refactoring needed to make Maven work better.  Those changes usually
> help make sure we have the loose-coupling we need between modules.



I think that is a good start.


>   I’m
> not that interested in refactoring the Flex SDK.   I am interested in the
> notion of slicing off any Flex SDK dependencies in the Falcon code base so
> we can Mavenize FlexJS without dealing with the Flex SDK first.  And I’m
> actually looking forward to seeing the Mavenizing of FlexJS.  Maybe I’ll
> learn more about Maven by watching those commits.
>

I am hoping to do the same.


>
> I’m sure in the early days of FlexJS and Falcon, going away for a while
> and coming back probably didn’t work.  But is that still true?  And if so,
> can you just blow away your repos and re-grab flex-asjs and run “ant all”
> and get back up and running?
>

It is an option.  But, I have several partially done cool things that were
stopped due to personal time issues, build issues, etc.  I would hate to
lose all that.  But I will probably just end up doing what you suggested.


>
> And for folks reading this who are in the silent faction:  Please speak
> up.  Or write me off-list if there is something you have to say that you
> don’t want to say on these lists.  And tell me/us why you have been
> silent.  Is there something we are doing or should be doing to make it
> easier for you to speak up?
>

I don't think the silent faction are paying attention, to this thread, or
even to this forum.  Ideally, we should make things easy to build, do a
couple of simple tutorials and invite folks to try it out.  That would be a
better approach to make them speak up.


>
> So, my takeaway is that we should fork off two actionable threads.  One
> about eliminating environment variables for Flash/AIR, and the other about
> folder restructuring in flex-asjs to make Mavenizing easier?  Did I miss
> anything?
>

As I understand things, I think that is pretty much it.  And oh, the work
needed from Adobe to mavenize the Flash runtime artifacts.

Thanks,
Om


>
> Meanwhile, I am trying to upgrade the build scripts to fix issues found in
> the FalconJX packages so that we can get releases out faster in the future
> by catching issues sooner.
>
> If you are still having problems with this FlexJS release candidate,
> please provide more detail.
>
> Thanks,
> -Alex
>
> On 11/6/15, 1:08 AM, "Justin Mclean" <ju...@me.com> wrote:
>
> >HI,
> >
> >> Nobody is saying simpler builds wouldn’t help.  But how much?
> >
> >A lot I think. Developer are unable to contribute if they can’t compile
> >and test their changes.
> >
> >>  More than a more functional DataGrid?  Or a more Spark and/or MX-like
> >>component set
> >> that reduces migration time for your app?  Or being able to write your
> >>JS
> >> using AS?
> >
> >IMO you would more likely get people contribution things like that if it
> >was simpler to build.
> >
> >> To folks reading this who are interested in
> >> FlexJS: what is the most important two or three things we should be
> >> working on?
> >
> >IMO:
> >1. Simpler builds
> >2. More frequent releases (say every 3 or 4 months).
> >3. Compiling MXML
> >4. Working documented examples / how to port existing applications
> >
> >> The build script was improved in this release such that if you have a
> >> couple of known sets of circumstances, like not having any of the repos,
> >> that running “ant all” will get all the repos and build them in the
> >>right
> >> order.
> >
> >I meant to ask before does that use well known versions or does it just
> >grab the latest out of the develop (or another) branch?
> >
> >Thanks,
> >Justin
>
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

Let me see if I can summarize these various “simplify the build threads
here”.

For me, I have to go by what I read on our mailing lists.  Maybe there is
some silent faction waiting for simpler builds, but I am going to reward
those who speak up by trying to work on their issues.  It doesn’t seem
right to tell someone who reports a bug that it will be several weeks
before I can fix it because I am worried about attracting people who we
don’t know exists.  Go back through the archives.  I rarely see new
potential committers complaining about how to get started writing FlexJS
code, especially after the recent changes I made to try to make that
simpler which was in response to the last flurry of “simplify the build”
emails from several of our current committers.  Strangely, when I
completed that work and asked folks to test it, I got no responses.  For
example, I didn’t see any responses from folks who could be in the silent
faction saying that it made it better, so that’s why I’m just not sure how
important it is to make it simpler than it is now.

Harbs had issues getting started and IIRC, mostly it was environment
variables so maybe all of these emails boils down to the need to try to
reduce the reliance on environment variables.  I’m definitely open to
spending some of my time on that.

I believe Maven is important, but I do not wish stop working on other
stuff to learn Maven and do the heavy lifting on it.  I hope Chris will
stick around to do it and that others can help them.  I will try to help
with refactoring needed to make Maven work better.  Those changes usually
help make sure we have the loose-coupling we need between modules.  I’m
not that interested in refactoring the Flex SDK.   I am interested in the
notion of slicing off any Flex SDK dependencies in the Falcon code base so
we can Mavenize FlexJS without dealing with the Flex SDK first.  And I’m
actually looking forward to seeing the Mavenizing of FlexJS.  Maybe I’ll
learn more about Maven by watching those commits.

I’m sure in the early days of FlexJS and Falcon, going away for a while
and coming back probably didn’t work.  But is that still true?  And if so,
can you just blow away your repos and re-grab flex-asjs and run “ant all”
and get back up and running?

And for folks reading this who are in the silent faction:  Please speak
up.  Or write me off-list if there is something you have to say that you
don’t want to say on these lists.  And tell me/us why you have been
silent.  Is there something we are doing or should be doing to make it
easier for you to speak up?

So, my takeaway is that we should fork off two actionable threads.  One
about eliminating environment variables for Flash/AIR, and the other about
folder restructuring in flex-asjs to make Mavenizing easier?  Did I miss
anything?

Meanwhile, I am trying to upgrade the build scripts to fix issues found in
the FalconJX packages so that we can get releases out faster in the future
by catching issues sooner.

If you are still having problems with this FlexJS release candidate,
please provide more detail.

Thanks,
-Alex

On 11/6/15, 1:08 AM, "Justin Mclean" <ju...@me.com> wrote:

>HI,
>
>> Nobody is saying simpler builds wouldn’t help.  But how much?
>
>A lot I think. Developer are unable to contribute if they can’t compile
>and test their changes.
>
>>  More than a more functional DataGrid?  Or a more Spark and/or MX-like
>>component set
>> that reduces migration time for your app?  Or being able to write your
>>JS
>> using AS?
>
>IMO you would more likely get people contribution things like that if it
>was simpler to build.
>
>> To folks reading this who are interested in
>> FlexJS: what is the most important two or three things we should be
>> working on?
>
>IMO:
>1. Simpler builds
>2. More frequent releases (say every 3 or 4 months).
>3. Compiling MXML
>4. Working documented examples / how to port existing applications
>
>> The build script was improved in this release such that if you have a
>> couple of known sets of circumstances, like not having any of the repos,
>> that running “ant all” will get all the repos and build them in the
>>right
>> order.
>
>I meant to ask before does that use well known versions or does it just
>grab the latest out of the develop (or another) branch?
>
>Thanks,
>Justin

Re: [FlexJS] Back port

[Alex Harui <ah...@adobe.com>]

FWIW, I wrote down some hopefully useful concepts about the back port here
[1]

[1] 
https://cwiki.apache.org/confluence/display/FLEX/FlexJS+Component+Source+Co
de+Patterns

On 11/9/15, 1:12 PM, "Harbs" <ha...@gmail.com> wrote:

>Well, I’d be interested in your findings.
>
>If operator definitions is possible .. will be easier than . operators.
>If we allow single dot operators, it will be difficult to differentiate
>between E4X syntax and function calls. I’d really like to drop the () on
>length and some of the other methods… Those are a course for very common
>errors.
>
>On Nov 9, 2015, at 9:34 PM, Alex Harui <ah...@adobe.com> wrote:
>
>> I think “..” is already recognized as a token.
>> 
>> Other code says “hey, that token is not allowed here”.  We have control
>> over where that is.
>> 
>> Of course, I could be mis-understanding.
>> 
>> I’m also going to do a bit of reading to make sure we can’t do E4x.
>>Yeah,
>> maybe you can invent something better, but at least E4x already has a
>>spec
>> and an implementation on the AS side.
>> 
>> -Alex
>> 
>> On 11/9/15, 11:31 AM, "Harbs" <ha...@gmail.com> wrote:
>> 
>>> Huh? For custom classes? How would that work?
>>> 
>>> On Nov 9, 2015, at 9:30 PM, Alex Harui <ah...@adobe.com> wrote:
>>> 
>>>> I think the compiler already knows about that.  It can convert that
>>>>to a
>>>> function call.
>>>> 
>>>> On 11/9/15, 11:17 AM, "Harbs" <ha...@gmail.com> wrote:
>>>> 
>>>>> It seems to me that it’s only possible if there’s some way to define
>>>>> dot
>>>>> and dotdot operators. I don’t know of a way to do that in
>>>>>ActionScript.
>>>>> 
>>>>> It will be possible to use mySVG.descendents(“path”) or
>>>>> mySVG.child(“path”) (which the dot and dotdot operators are really
>>>>> shorthand for)
>>>>> 
>>>>> If someone can think of  a way to define the operators, I’d be happy
>>>>>to
>>>>> try to add that…
>>>>> 
>>>>> On Nov 9, 2015, at 9:08 PM, OmPrakash Muppirala
>>>>><bi...@gmail.com>
>>>>> wrote:
>>>>> 
>>>>>> On Mon, Nov 9, 2015 at 11:00 AM, Harbs <ha...@gmail.com>
>>>>>>wrote:
>>>>>> 
>>>>>>> I’m not 100% clear on the status of the back-ports or the best way
>>>>>>>to
>>>>>>> go
>>>>>>> about it. I figured I’d wait until we’re done with the release and
>>>>>>> then
>>>>>>> coordinate with you and Peter.
>>>>>>> 
>>>>>>> In the meantime, I started work on some XML classes. I pretty much
>>>>>>> decided
>>>>>>> on departing a bit from classic E4X for two reasons: 1. It’s just
>>>>>>> about
>>>>>>> impossibly to reproduce the E4X behavior. 2. E4X was always awkward
>>>>>>> to
>>>>>>> use.
>>>>>>> I’m going to more or less copy the public API of XML, but my
>>>>>>>version
>>>>>>> of
>>>>>>> XMLList is going to work differently. Both XML and XMLList (I’m
>>>>>>> calling
>>>>>>> them XMLNode and XMLNodeList) will have filter functions which work
>>>>>>> similar
>>>>>>> to Array filter functions.
>>>>>> 
>>>>>> 
>>>>>> The only E4X expression I have used over and over is the ability to
>>>>>> get
>>>>>> an
>>>>>> array of child nodes using the double-dot expression.
>>>>>> 
>>>>>> For example:
>>>>>> 
>>>>>> var mySVG:XML = <svg>
>>>>>> <desc />
>>>>>> <rect />
>>>>>> <ellipse />
>>>>>> <path />
>>>>>> <path />
>>>>>> <path />
>>>>>> <ellipse />
>>>>>> <path />
>>>>>> <path />
>>>>>> </svg>;
>>>>>> 
>>>>>> We could write:
>>>>>> 
>>>>>> var paths:XMLList = mySVG..path;
>>>>>> 
>>>>>> I think if we can add support to just this (maybe a few more based
>>>>>>on
>>>>>> popular usage), we should be able to get existing Flex apps easily
>>>>>> ported.
>>>>>> Although, I am not sure how hard something like this would be to
>>>>>> implement.
>>>>>> 
>>>>>> Thoughts?
>>>>>> 
>>>>>> Thanks,
>>>>>> Om
>>>>>> 
>>>>>> 
>>>>>>> This should allow getting chunks of XML in a more sane method than
>>>>>>> using
>>>>>>> hard-to-read E4X expressions.
>>>>>>> 
>>>>>>> I’m hoping to have something commit-able on that front by the end
>>>>>>>of
>>>>>>> the
>>>>>>> week.
>>>>>>> 
>>>>>>> I’ll try to follow what you are doing with DragDrop and Effects so
>>>>>>>I
>>>>>>> get a
>>>>>>> better feel of how it’s done.
>>>>>>> 
>>>>>>> On Nov 9, 2015, at 7:36 PM, Alex Harui <ah...@adobe.com> wrote:
>>>>>>> 
>>>>>>>> Harbs,
>>>>>>>> 
>>>>>>>> Thanks for the commits for the back port.  Let us know if you
>>>>>>>> actually
>>>>>>>> start work on porting an existing SWC.  I’m going to wait another
>>>>>>>> day
>>>>>>>> before cutting the FalconJX RC2 in case Justin has more feedback
>>>>>>>>on
>>>>>>>> the
>>>>>>>> LICENSE.  So I am going to try to port dragdrop and effects.
>>>>>>>> 
>>>>>>>> -Alex
>>>>>>>> 
>>>>>>>> On 11/9/15, 12:12 AM, "Alex Harui" <ah...@adobe.com> wrote:
>>>>>>>> 
>>>>>>>>> 
>>>>>>>>> 
>>>>>>>>> On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
>>>>>>>>>> 
>>>>>>>>>>> Meanwhile, Peter and I are back porting the SWCs one-by-one in
>>>>>>>>>>> the
>>>>>>>>>>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to
>>>>>>>>>>> or
>>>>>>>>>>> anything, but you can look at Binding, Core, Collections and
>>>>>>>>>>>HTML
>>>>>>>>>>> as
>>>>>>>>>>> examples.
>>>>>>>>>> 
>>>>>>>>>> Color me confused, but I still see all 3 source folders after
>>>>>>>>>> switching
>>>>>>>>>> to the back port branch. It looks like you’re doing the
>>>>>>>>>>back-port
>>>>>>>>>> in
>>>>>>> the
>>>>>>>>>> as folder? The folder clean-up will happen after the port is
>>>>>>>>>> complete?
>>>>>>>>> 
>>>>>>>>> The asjs folders should be empty (more or less) in Binding, Core,
>>>>>>>>> and
>>>>>>>>> other completed SWCs.  We move all .as files to the as/src folder
>>>>>>>>> and
>>>>>>> fix
>>>>>>>>> up the compile-asjs-config.xml files.  We are keeping the js/src
>>>>>>>>> folders
>>>>>>>>> for now, then we’ll delete them once we feel good we’ve ported
>>>>>>>>>them
>>>>>>>>> correctly.
>>>>>>>>>> 
>>>>>>>>>>> I think Peter is working on Charts and Mobile.  Graphics builds
>>>>>>> cleanly
>>>>>>>>>>> but Peter needs to verify it.  So, if you want, let us know
>>>>>>>>>>>what
>>>>>>>>>>> SWC
>>>>>>>>>>> you
>>>>>>>>>>> want to try and see if you can see the patterns from the other
>>>>>>> projects
>>>>>>>>>>> and change logs.
>>>>>>>>>> 
>>>>>>>>>> I’ll spend some time studying this today. I just added compiler
>>>>>>> arguments
>>>>>>>>>> to the Flash Builder project files so AFB does not choke on the
>>>>>>> compiler
>>>>>>>>>> arguments. I’ll commit those changes shortly. FWIW, the only
>>>>>>>>>>real
>>>>>>>>>> error
>>>>>>>>>> I'm getting is in HTML IScrollingLayoutParent that ILayoutParent
>>>>>>>>>> is
>>>>>>>>>> not
>>>>>>>>>> found.
>>>>>>>>> 
>>>>>>>>> I have to stop for the night, but I don’t think
>>>>>>>>> IScrollingLayoutParent
>>>>>>>>> should be there anymore.  Check the logs in the develop branch to
>>>>>>> verify.
>>>>>>>>> 
>>>>>>>>> Thanks for helping.
>>>>>>>>> -Alex
>>>>>>>>> 
>>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>> 
>>>> 
>>> 
>> 
>

Re: [FlexJS] Back port

[Harbs <ha...@gmail.com>]

Well, I’d be interested in your findings.

If operator definitions is possible .. will be easier than . operators. If we allow single dot operators, it will be difficult to differentiate between E4X syntax and function calls. I’d really like to drop the () on length and some of the other methods… Those are a course for very common errors.

On Nov 9, 2015, at 9:34 PM, Alex Harui <ah...@adobe.com> wrote:

> I think “..” is already recognized as a token.
> 
> Other code says “hey, that token is not allowed here”.  We have control
> over where that is.
> 
> Of course, I could be mis-understanding.
> 
> I’m also going to do a bit of reading to make sure we can’t do E4x.  Yeah,
> maybe you can invent something better, but at least E4x already has a spec
> and an implementation on the AS side.
> 
> -Alex
> 
> On 11/9/15, 11:31 AM, "Harbs" <ha...@gmail.com> wrote:
> 
>> Huh? For custom classes? How would that work?
>> 
>> On Nov 9, 2015, at 9:30 PM, Alex Harui <ah...@adobe.com> wrote:
>> 
>>> I think the compiler already knows about that.  It can convert that to a
>>> function call.
>>> 
>>> On 11/9/15, 11:17 AM, "Harbs" <ha...@gmail.com> wrote:
>>> 
>>>> It seems to me that it’s only possible if there’s some way to define
>>>> dot
>>>> and dotdot operators. I don’t know of a way to do that in ActionScript.
>>>> 
>>>> It will be possible to use mySVG.descendents(“path”) or
>>>> mySVG.child(“path”) (which the dot and dotdot operators are really
>>>> shorthand for)
>>>> 
>>>> If someone can think of  a way to define the operators, I’d be happy to
>>>> try to add that…
>>>> 
>>>> On Nov 9, 2015, at 9:08 PM, OmPrakash Muppirala <bi...@gmail.com>
>>>> wrote:
>>>> 
>>>>> On Mon, Nov 9, 2015 at 11:00 AM, Harbs <ha...@gmail.com> wrote:
>>>>> 
>>>>>> I’m not 100% clear on the status of the back-ports or the best way to
>>>>>> go
>>>>>> about it. I figured I’d wait until we’re done with the release and
>>>>>> then
>>>>>> coordinate with you and Peter.
>>>>>> 
>>>>>> In the meantime, I started work on some XML classes. I pretty much
>>>>>> decided
>>>>>> on departing a bit from classic E4X for two reasons: 1. It’s just
>>>>>> about
>>>>>> impossibly to reproduce the E4X behavior. 2. E4X was always awkward
>>>>>> to
>>>>>> use.
>>>>>> I’m going to more or less copy the public API of XML, but my version
>>>>>> of
>>>>>> XMLList is going to work differently. Both XML and XMLList (I’m
>>>>>> calling
>>>>>> them XMLNode and XMLNodeList) will have filter functions which work
>>>>>> similar
>>>>>> to Array filter functions.
>>>>> 
>>>>> 
>>>>> The only E4X expression I have used over and over is the ability to
>>>>> get
>>>>> an
>>>>> array of child nodes using the double-dot expression.
>>>>> 
>>>>> For example:
>>>>> 
>>>>> var mySVG:XML = <svg>
>>>>> <desc />
>>>>> <rect />
>>>>> <ellipse />
>>>>> <path />
>>>>> <path />
>>>>> <path />
>>>>> <ellipse />
>>>>> <path />
>>>>> <path />
>>>>> </svg>;
>>>>> 
>>>>> We could write:
>>>>> 
>>>>> var paths:XMLList = mySVG..path;
>>>>> 
>>>>> I think if we can add support to just this (maybe a few more based on
>>>>> popular usage), we should be able to get existing Flex apps easily
>>>>> ported.
>>>>> Although, I am not sure how hard something like this would be to
>>>>> implement.
>>>>> 
>>>>> Thoughts?
>>>>> 
>>>>> Thanks,
>>>>> Om
>>>>> 
>>>>> 
>>>>>> This should allow getting chunks of XML in a more sane method than
>>>>>> using
>>>>>> hard-to-read E4X expressions.
>>>>>> 
>>>>>> I’m hoping to have something commit-able on that front by the end of
>>>>>> the
>>>>>> week.
>>>>>> 
>>>>>> I’ll try to follow what you are doing with DragDrop and Effects so I
>>>>>> get a
>>>>>> better feel of how it’s done.
>>>>>> 
>>>>>> On Nov 9, 2015, at 7:36 PM, Alex Harui <ah...@adobe.com> wrote:
>>>>>> 
>>>>>>> Harbs,
>>>>>>> 
>>>>>>> Thanks for the commits for the back port.  Let us know if you
>>>>>>> actually
>>>>>>> start work on porting an existing SWC.  I’m going to wait another
>>>>>>> day
>>>>>>> before cutting the FalconJX RC2 in case Justin has more feedback on
>>>>>>> the
>>>>>>> LICENSE.  So I am going to try to port dragdrop and effects.
>>>>>>> 
>>>>>>> -Alex
>>>>>>> 
>>>>>>> On 11/9/15, 12:12 AM, "Alex Harui" <ah...@adobe.com> wrote:
>>>>>>> 
>>>>>>>> 
>>>>>>>> 
>>>>>>>> On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
>>>>>>>>> 
>>>>>>>>>> Meanwhile, Peter and I are back porting the SWCs one-by-one in
>>>>>>>>>> the
>>>>>>>>>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to
>>>>>>>>>> or
>>>>>>>>>> anything, but you can look at Binding, Core, Collections and HTML
>>>>>>>>>> as
>>>>>>>>>> examples.
>>>>>>>>> 
>>>>>>>>> Color me confused, but I still see all 3 source folders after
>>>>>>>>> switching
>>>>>>>>> to the back port branch. It looks like you’re doing the back-port
>>>>>>>>> in
>>>>>> the
>>>>>>>>> as folder? The folder clean-up will happen after the port is
>>>>>>>>> complete?
>>>>>>>> 
>>>>>>>> The asjs folders should be empty (more or less) in Binding, Core,
>>>>>>>> and
>>>>>>>> other completed SWCs.  We move all .as files to the as/src folder
>>>>>>>> and
>>>>>> fix
>>>>>>>> up the compile-asjs-config.xml files.  We are keeping the js/src
>>>>>>>> folders
>>>>>>>> for now, then we’ll delete them once we feel good we’ve ported them
>>>>>>>> correctly.
>>>>>>>>> 
>>>>>>>>>> I think Peter is working on Charts and Mobile.  Graphics builds
>>>>>> cleanly
>>>>>>>>>> but Peter needs to verify it.  So, if you want, let us know what
>>>>>>>>>> SWC
>>>>>>>>>> you
>>>>>>>>>> want to try and see if you can see the patterns from the other
>>>>>> projects
>>>>>>>>>> and change logs.
>>>>>>>>> 
>>>>>>>>> I’ll spend some time studying this today. I just added compiler
>>>>>> arguments
>>>>>>>>> to the Flash Builder project files so AFB does not choke on the
>>>>>> compiler
>>>>>>>>> arguments. I’ll commit those changes shortly. FWIW, the only real
>>>>>>>>> error
>>>>>>>>> I'm getting is in HTML IScrollingLayoutParent that ILayoutParent
>>>>>>>>> is
>>>>>>>>> not
>>>>>>>>> found.
>>>>>>>> 
>>>>>>>> I have to stop for the night, but I don’t think
>>>>>>>> IScrollingLayoutParent
>>>>>>>> should be there anymore.  Check the logs in the develop branch to
>>>>>> verify.
>>>>>>>> 
>>>>>>>> Thanks for helping.
>>>>>>>> -Alex
>>>>>>>> 
>>>>>>> 
>>>>>> 
>>>>>> 
>>>> 
>>> 
>> 
> 

Re: [FlexJS] Back port

[Alex Harui <ah...@adobe.com>]

I think “..” is already recognized as a token.

Other code says “hey, that token is not allowed here”.  We have control
over where that is.

Of course, I could be mis-understanding.

I’m also going to do a bit of reading to make sure we can’t do E4x.  Yeah,
maybe you can invent something better, but at least E4x already has a spec
and an implementation on the AS side.

-Alex

On 11/9/15, 11:31 AM, "Harbs" <ha...@gmail.com> wrote:

>Huh? For custom classes? How would that work?
>
>On Nov 9, 2015, at 9:30 PM, Alex Harui <ah...@adobe.com> wrote:
>
>> I think the compiler already knows about that.  It can convert that to a
>> function call.
>> 
>> On 11/9/15, 11:17 AM, "Harbs" <ha...@gmail.com> wrote:
>> 
>>> It seems to me that it’s only possible if there’s some way to define
>>>dot
>>> and dotdot operators. I don’t know of a way to do that in ActionScript.
>>> 
>>> It will be possible to use mySVG.descendents(“path”) or
>>> mySVG.child(“path”) (which the dot and dotdot operators are really
>>> shorthand for)
>>> 
>>> If someone can think of  a way to define the operators, I’d be happy to
>>> try to add that…
>>> 
>>> On Nov 9, 2015, at 9:08 PM, OmPrakash Muppirala <bi...@gmail.com>
>>> wrote:
>>> 
>>>> On Mon, Nov 9, 2015 at 11:00 AM, Harbs <ha...@gmail.com> wrote:
>>>> 
>>>>> I’m not 100% clear on the status of the back-ports or the best way to
>>>>> go
>>>>> about it. I figured I’d wait until we’re done with the release and
>>>>>then
>>>>> coordinate with you and Peter.
>>>>> 
>>>>> In the meantime, I started work on some XML classes. I pretty much
>>>>> decided
>>>>> on departing a bit from classic E4X for two reasons: 1. It’s just
>>>>>about
>>>>> impossibly to reproduce the E4X behavior. 2. E4X was always awkward
>>>>>to
>>>>> use.
>>>>> I’m going to more or less copy the public API of XML, but my version
>>>>>of
>>>>> XMLList is going to work differently. Both XML and XMLList (I’m
>>>>>calling
>>>>> them XMLNode and XMLNodeList) will have filter functions which work
>>>>> similar
>>>>> to Array filter functions.
>>>> 
>>>> 
>>>> The only E4X expression I have used over and over is the ability to
>>>>get
>>>> an
>>>> array of child nodes using the double-dot expression.
>>>> 
>>>> For example:
>>>> 
>>>> var mySVG:XML = <svg>
>>>> <desc />
>>>> <rect />
>>>> <ellipse />
>>>> <path />
>>>> <path />
>>>> <path />
>>>> <ellipse />
>>>> <path />
>>>> <path />
>>>> </svg>;
>>>> 
>>>> We could write:
>>>> 
>>>> var paths:XMLList = mySVG..path;
>>>> 
>>>> I think if we can add support to just this (maybe a few more based on
>>>> popular usage), we should be able to get existing Flex apps easily
>>>> ported.
>>>> Although, I am not sure how hard something like this would be to
>>>> implement.
>>>> 
>>>> Thoughts?
>>>> 
>>>> Thanks,
>>>> Om
>>>> 
>>>> 
>>>>> This should allow getting chunks of XML in a more sane method than
>>>>> using
>>>>> hard-to-read E4X expressions.
>>>>> 
>>>>> I’m hoping to have something commit-able on that front by the end of
>>>>> the
>>>>> week.
>>>>> 
>>>>> I’ll try to follow what you are doing with DragDrop and Effects so I
>>>>> get a
>>>>> better feel of how it’s done.
>>>>> 
>>>>> On Nov 9, 2015, at 7:36 PM, Alex Harui <ah...@adobe.com> wrote:
>>>>> 
>>>>>> Harbs,
>>>>>> 
>>>>>> Thanks for the commits for the back port.  Let us know if you
>>>>>>actually
>>>>>> start work on porting an existing SWC.  I’m going to wait another
>>>>>>day
>>>>>> before cutting the FalconJX RC2 in case Justin has more feedback on
>>>>>> the
>>>>>> LICENSE.  So I am going to try to port dragdrop and effects.
>>>>>> 
>>>>>> -Alex
>>>>>> 
>>>>>> On 11/9/15, 12:12 AM, "Alex Harui" <ah...@adobe.com> wrote:
>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
>>>>>>>> 
>>>>>>>>> Meanwhile, Peter and I are back porting the SWCs one-by-one in
>>>>>>>>>the
>>>>>>>>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to
>>>>>>>>>or
>>>>>>>>> anything, but you can look at Binding, Core, Collections and HTML
>>>>>>>>> as
>>>>>>>>> examples.
>>>>>>>> 
>>>>>>>> Color me confused, but I still see all 3 source folders after
>>>>>>>> switching
>>>>>>>> to the back port branch. It looks like you’re doing the back-port
>>>>>>>>in
>>>>> the
>>>>>>>> as folder? The folder clean-up will happen after the port is
>>>>>>>> complete?
>>>>>>> 
>>>>>>> The asjs folders should be empty (more or less) in Binding, Core,
>>>>>>>and
>>>>>>> other completed SWCs.  We move all .as files to the as/src folder
>>>>>>>and
>>>>> fix
>>>>>>> up the compile-asjs-config.xml files.  We are keeping the js/src
>>>>>>> folders
>>>>>>> for now, then we’ll delete them once we feel good we’ve ported them
>>>>>>> correctly.
>>>>>>>> 
>>>>>>>>> I think Peter is working on Charts and Mobile.  Graphics builds
>>>>> cleanly
>>>>>>>>> but Peter needs to verify it.  So, if you want, let us know what
>>>>>>>>> SWC
>>>>>>>>> you
>>>>>>>>> want to try and see if you can see the patterns from the other
>>>>> projects
>>>>>>>>> and change logs.
>>>>>>>> 
>>>>>>>> I’ll spend some time studying this today. I just added compiler
>>>>> arguments
>>>>>>>> to the Flash Builder project files so AFB does not choke on the
>>>>> compiler
>>>>>>>> arguments. I’ll commit those changes shortly. FWIW, the only real
>>>>>>>> error
>>>>>>>> I'm getting is in HTML IScrollingLayoutParent that ILayoutParent
>>>>>>>>is
>>>>>>>> not
>>>>>>>> found.
>>>>>>> 
>>>>>>> I have to stop for the night, but I don’t think
>>>>>>> IScrollingLayoutParent
>>>>>>> should be there anymore.  Check the logs in the develop branch to
>>>>> verify.
>>>>>>> 
>>>>>>> Thanks for helping.
>>>>>>> -Alex
>>>>>>> 
>>>>>> 
>>>>> 
>>>>> 
>>> 
>> 
>

Re: [FlexJS] Back port

[Harbs <ha...@gmail.com>]

Huh? For custom classes? How would that work?

On Nov 9, 2015, at 9:30 PM, Alex Harui <ah...@adobe.com> wrote:

> I think the compiler already knows about that.  It can convert that to a
> function call.
> 
> On 11/9/15, 11:17 AM, "Harbs" <ha...@gmail.com> wrote:
> 
>> It seems to me that it’s only possible if there’s some way to define dot
>> and dotdot operators. I don’t know of a way to do that in ActionScript.
>> 
>> It will be possible to use mySVG.descendents(“path”) or
>> mySVG.child(“path”) (which the dot and dotdot operators are really
>> shorthand for)
>> 
>> If someone can think of  a way to define the operators, I’d be happy to
>> try to add that…
>> 
>> On Nov 9, 2015, at 9:08 PM, OmPrakash Muppirala <bi...@gmail.com>
>> wrote:
>> 
>>> On Mon, Nov 9, 2015 at 11:00 AM, Harbs <ha...@gmail.com> wrote:
>>> 
>>>> I’m not 100% clear on the status of the back-ports or the best way to
>>>> go
>>>> about it. I figured I’d wait until we’re done with the release and then
>>>> coordinate with you and Peter.
>>>> 
>>>> In the meantime, I started work on some XML classes. I pretty much
>>>> decided
>>>> on departing a bit from classic E4X for two reasons: 1. It’s just about
>>>> impossibly to reproduce the E4X behavior. 2. E4X was always awkward to
>>>> use.
>>>> I’m going to more or less copy the public API of XML, but my version of
>>>> XMLList is going to work differently. Both XML and XMLList (I’m calling
>>>> them XMLNode and XMLNodeList) will have filter functions which work
>>>> similar
>>>> to Array filter functions.
>>> 
>>> 
>>> The only E4X expression I have used over and over is the ability to get
>>> an
>>> array of child nodes using the double-dot expression.
>>> 
>>> For example:
>>> 
>>> var mySVG:XML = <svg>
>>> <desc />
>>> <rect />
>>> <ellipse />
>>> <path />
>>> <path />
>>> <path />
>>> <ellipse />
>>> <path />
>>> <path />
>>> </svg>;
>>> 
>>> We could write:
>>> 
>>> var paths:XMLList = mySVG..path;
>>> 
>>> I think if we can add support to just this (maybe a few more based on
>>> popular usage), we should be able to get existing Flex apps easily
>>> ported.
>>> Although, I am not sure how hard something like this would be to
>>> implement.
>>> 
>>> Thoughts?
>>> 
>>> Thanks,
>>> Om
>>> 
>>> 
>>>> This should allow getting chunks of XML in a more sane method than
>>>> using
>>>> hard-to-read E4X expressions.
>>>> 
>>>> I’m hoping to have something commit-able on that front by the end of
>>>> the
>>>> week.
>>>> 
>>>> I’ll try to follow what you are doing with DragDrop and Effects so I
>>>> get a
>>>> better feel of how it’s done.
>>>> 
>>>> On Nov 9, 2015, at 7:36 PM, Alex Harui <ah...@adobe.com> wrote:
>>>> 
>>>>> Harbs,
>>>>> 
>>>>> Thanks for the commits for the back port.  Let us know if you actually
>>>>> start work on porting an existing SWC.  I’m going to wait another day
>>>>> before cutting the FalconJX RC2 in case Justin has more feedback on
>>>>> the
>>>>> LICENSE.  So I am going to try to port dragdrop and effects.
>>>>> 
>>>>> -Alex
>>>>> 
>>>>> On 11/9/15, 12:12 AM, "Alex Harui" <ah...@adobe.com> wrote:
>>>>> 
>>>>>> 
>>>>>> 
>>>>>> On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
>>>>>>> 
>>>>>>>> Meanwhile, Peter and I are back porting the SWCs one-by-one in the
>>>>>>>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to or
>>>>>>>> anything, but you can look at Binding, Core, Collections and HTML
>>>>>>>> as
>>>>>>>> examples.
>>>>>>> 
>>>>>>> Color me confused, but I still see all 3 source folders after
>>>>>>> switching
>>>>>>> to the back port branch. It looks like you’re doing the back-port in
>>>> the
>>>>>>> as folder? The folder clean-up will happen after the port is
>>>>>>> complete?
>>>>>> 
>>>>>> The asjs folders should be empty (more or less) in Binding, Core, and
>>>>>> other completed SWCs.  We move all .as files to the as/src folder and
>>>> fix
>>>>>> up the compile-asjs-config.xml files.  We are keeping the js/src
>>>>>> folders
>>>>>> for now, then we’ll delete them once we feel good we’ve ported them
>>>>>> correctly.
>>>>>>> 
>>>>>>>> I think Peter is working on Charts and Mobile.  Graphics builds
>>>> cleanly
>>>>>>>> but Peter needs to verify it.  So, if you want, let us know what
>>>>>>>> SWC
>>>>>>>> you
>>>>>>>> want to try and see if you can see the patterns from the other
>>>> projects
>>>>>>>> and change logs.
>>>>>>> 
>>>>>>> I’ll spend some time studying this today. I just added compiler
>>>> arguments
>>>>>>> to the Flash Builder project files so AFB does not choke on the
>>>> compiler
>>>>>>> arguments. I’ll commit those changes shortly. FWIW, the only real
>>>>>>> error
>>>>>>> I'm getting is in HTML IScrollingLayoutParent that ILayoutParent is
>>>>>>> not
>>>>>>> found.
>>>>>> 
>>>>>> I have to stop for the night, but I don’t think
>>>>>> IScrollingLayoutParent
>>>>>> should be there anymore.  Check the logs in the develop branch to
>>>> verify.
>>>>>> 
>>>>>> Thanks for helping.
>>>>>> -Alex
>>>>>> 
>>>>> 
>>>> 
>>>> 
>> 
> 

Re: [FlexJS] Back port

[Alex Harui <ah...@adobe.com>]

I think the compiler already knows about that.  It can convert that to a
function call.

On 11/9/15, 11:17 AM, "Harbs" <ha...@gmail.com> wrote:

>It seems to me that it’s only possible if there’s some way to define dot
>and dotdot operators. I don’t know of a way to do that in ActionScript.
>
>It will be possible to use mySVG.descendents(“path”) or
>mySVG.child(“path”) (which the dot and dotdot operators are really
>shorthand for)
>
>If someone can think of  a way to define the operators, I’d be happy to
>try to add that…
>
>On Nov 9, 2015, at 9:08 PM, OmPrakash Muppirala <bi...@gmail.com>
>wrote:
>
>> On Mon, Nov 9, 2015 at 11:00 AM, Harbs <ha...@gmail.com> wrote:
>> 
>>> I’m not 100% clear on the status of the back-ports or the best way to
>>>go
>>> about it. I figured I’d wait until we’re done with the release and then
>>> coordinate with you and Peter.
>>> 
>>> In the meantime, I started work on some XML classes. I pretty much
>>>decided
>>> on departing a bit from classic E4X for two reasons: 1. It’s just about
>>> impossibly to reproduce the E4X behavior. 2. E4X was always awkward to
>>>use.
>>> I’m going to more or less copy the public API of XML, but my version of
>>> XMLList is going to work differently. Both XML and XMLList (I’m calling
>>> them XMLNode and XMLNodeList) will have filter functions which work
>>>similar
>>> to Array filter functions.
>> 
>> 
>> The only E4X expression I have used over and over is the ability to get
>>an
>> array of child nodes using the double-dot expression.
>> 
>> For example:
>> 
>> var mySVG:XML = <svg>
>>  <desc />
>>  <rect />
>>  <ellipse />
>>  <path />
>>  <path />
>>  <path />
>>  <ellipse />
>>  <path />
>>  <path />
>> </svg>;
>> 
>> We could write:
>> 
>> var paths:XMLList = mySVG..path;
>> 
>> I think if we can add support to just this (maybe a few more based on
>> popular usage), we should be able to get existing Flex apps easily
>>ported.
>> Although, I am not sure how hard something like this would be to
>>implement.
>> 
>> Thoughts?
>> 
>> Thanks,
>> Om
>> 
>> 
>>> This should allow getting chunks of XML in a more sane method than
>>>using
>>> hard-to-read E4X expressions.
>>> 
>>> I’m hoping to have something commit-able on that front by the end of
>>>the
>>> week.
>>> 
>>> I’ll try to follow what you are doing with DragDrop and Effects so I
>>>get a
>>> better feel of how it’s done.
>>> 
>>> On Nov 9, 2015, at 7:36 PM, Alex Harui <ah...@adobe.com> wrote:
>>> 
>>>> Harbs,
>>>> 
>>>> Thanks for the commits for the back port.  Let us know if you actually
>>>> start work on porting an existing SWC.  I’m going to wait another day
>>>> before cutting the FalconJX RC2 in case Justin has more feedback on
>>>>the
>>>> LICENSE.  So I am going to try to port dragdrop and effects.
>>>> 
>>>> -Alex
>>>> 
>>>> On 11/9/15, 12:12 AM, "Alex Harui" <ah...@adobe.com> wrote:
>>>> 
>>>>> 
>>>>> 
>>>>> On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
>>>>>> 
>>>>>>> Meanwhile, Peter and I are back porting the SWCs one-by-one in the
>>>>>>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to or
>>>>>>> anything, but you can look at Binding, Core, Collections and HTML
>>>>>>>as
>>>>>>> examples.
>>>>>> 
>>>>>> Color me confused, but I still see all 3 source folders after
>>>>>>switching
>>>>>> to the back port branch. It looks like you’re doing the back-port in
>>> the
>>>>>> as folder? The folder clean-up will happen after the port is
>>>>>>complete?
>>>>> 
>>>>> The asjs folders should be empty (more or less) in Binding, Core, and
>>>>> other completed SWCs.  We move all .as files to the as/src folder and
>>> fix
>>>>> up the compile-asjs-config.xml files.  We are keeping the js/src
>>>>>folders
>>>>> for now, then we’ll delete them once we feel good we’ve ported them
>>>>> correctly.
>>>>>> 
>>>>>>> I think Peter is working on Charts and Mobile.  Graphics builds
>>> cleanly
>>>>>>> but Peter needs to verify it.  So, if you want, let us know what
>>>>>>>SWC
>>>>>>> you
>>>>>>> want to try and see if you can see the patterns from the other
>>> projects
>>>>>>> and change logs.
>>>>>> 
>>>>>> I’ll spend some time studying this today. I just added compiler
>>> arguments
>>>>>> to the Flash Builder project files so AFB does not choke on the
>>> compiler
>>>>>> arguments. I’ll commit those changes shortly. FWIW, the only real
>>>>>>error
>>>>>> I'm getting is in HTML IScrollingLayoutParent that ILayoutParent is
>>>>>>not
>>>>>> found.
>>>>> 
>>>>> I have to stop for the night, but I don’t think
>>>>>IScrollingLayoutParent
>>>>> should be there anymore.  Check the logs in the develop branch to
>>> verify.
>>>>> 
>>>>> Thanks for helping.
>>>>> -Alex
>>>>> 
>>>> 
>>> 
>>> 
>

Re: [FlexJS] Back port

[Harbs <ha...@gmail.com>]

It seems to me that it’s only possible if there’s some way to define dot and dotdot operators. I don’t know of a way to do that in ActionScript.

It will be possible to use mySVG.descendents(“path”) or mySVG.child(“path”) (which the dot and dotdot operators are really shorthand for)

If someone can think of  a way to define the operators, I’d be happy to try to add that…

On Nov 9, 2015, at 9:08 PM, OmPrakash Muppirala <bi...@gmail.com> wrote:

> On Mon, Nov 9, 2015 at 11:00 AM, Harbs <ha...@gmail.com> wrote:
> 
>> I’m not 100% clear on the status of the back-ports or the best way to go
>> about it. I figured I’d wait until we’re done with the release and then
>> coordinate with you and Peter.
>> 
>> In the meantime, I started work on some XML classes. I pretty much decided
>> on departing a bit from classic E4X for two reasons: 1. It’s just about
>> impossibly to reproduce the E4X behavior. 2. E4X was always awkward to use.
>> I’m going to more or less copy the public API of XML, but my version of
>> XMLList is going to work differently. Both XML and XMLList (I’m calling
>> them XMLNode and XMLNodeList) will have filter functions which work similar
>> to Array filter functions.
> 
> 
> The only E4X expression I have used over and over is the ability to get an
> array of child nodes using the double-dot expression.
> 
> For example:
> 
> var mySVG:XML = <svg>
>  <desc />
>  <rect />
>  <ellipse />
>  <path />
>  <path />
>  <path />
>  <ellipse />
>  <path />
>  <path />
> </svg>;
> 
> We could write:
> 
> var paths:XMLList = mySVG..path;
> 
> I think if we can add support to just this (maybe a few more based on
> popular usage), we should be able to get existing Flex apps easily ported.
> Although, I am not sure how hard something like this would be to implement.
> 
> Thoughts?
> 
> Thanks,
> Om
> 
> 
>> This should allow getting chunks of XML in a more sane method than using
>> hard-to-read E4X expressions.
>> 
>> I’m hoping to have something commit-able on that front by the end of the
>> week.
>> 
>> I’ll try to follow what you are doing with DragDrop and Effects so I get a
>> better feel of how it’s done.
>> 
>> On Nov 9, 2015, at 7:36 PM, Alex Harui <ah...@adobe.com> wrote:
>> 
>>> Harbs,
>>> 
>>> Thanks for the commits for the back port.  Let us know if you actually
>>> start work on porting an existing SWC.  I’m going to wait another day
>>> before cutting the FalconJX RC2 in case Justin has more feedback on the
>>> LICENSE.  So I am going to try to port dragdrop and effects.
>>> 
>>> -Alex
>>> 
>>> On 11/9/15, 12:12 AM, "Alex Harui" <ah...@adobe.com> wrote:
>>> 
>>>> 
>>>> 
>>>> On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
>>>>> 
>>>>>> Meanwhile, Peter and I are back porting the SWCs one-by-one in the
>>>>>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to or
>>>>>> anything, but you can look at Binding, Core, Collections and HTML as
>>>>>> examples.
>>>>> 
>>>>> Color me confused, but I still see all 3 source folders after switching
>>>>> to the back port branch. It looks like you’re doing the back-port in
>> the
>>>>> as folder? The folder clean-up will happen after the port is complete?
>>>> 
>>>> The asjs folders should be empty (more or less) in Binding, Core, and
>>>> other completed SWCs.  We move all .as files to the as/src folder and
>> fix
>>>> up the compile-asjs-config.xml files.  We are keeping the js/src folders
>>>> for now, then we’ll delete them once we feel good we’ve ported them
>>>> correctly.
>>>>> 
>>>>>> I think Peter is working on Charts and Mobile.  Graphics builds
>> cleanly
>>>>>> but Peter needs to verify it.  So, if you want, let us know what SWC
>>>>>> you
>>>>>> want to try and see if you can see the patterns from the other
>> projects
>>>>>> and change logs.
>>>>> 
>>>>> I’ll spend some time studying this today. I just added compiler
>> arguments
>>>>> to the Flash Builder project files so AFB does not choke on the
>> compiler
>>>>> arguments. I’ll commit those changes shortly. FWIW, the only real error
>>>>> I'm getting is in HTML IScrollingLayoutParent that ILayoutParent is not
>>>>> found.
>>>> 
>>>> I have to stop for the night, but I don’t think IScrollingLayoutParent
>>>> should be there anymore.  Check the logs in the develop branch to
>> verify.
>>>> 
>>>> Thanks for helping.
>>>> -Alex
>>>> 
>>> 
>> 
>> 

Re: [FlexJS] Back port

[OmPrakash Muppirala <bi...@gmail.com>]

On Mon, Nov 9, 2015 at 11:00 AM, Harbs <ha...@gmail.com> wrote:

> I’m not 100% clear on the status of the back-ports or the best way to go
> about it. I figured I’d wait until we’re done with the release and then
> coordinate with you and Peter.
>
> In the meantime, I started work on some XML classes. I pretty much decided
> on departing a bit from classic E4X for two reasons: 1. It’s just about
> impossibly to reproduce the E4X behavior. 2. E4X was always awkward to use.
> I’m going to more or less copy the public API of XML, but my version of
> XMLList is going to work differently. Both XML and XMLList (I’m calling
> them XMLNode and XMLNodeList) will have filter functions which work similar
> to Array filter functions.


The only E4X expression I have used over and over is the ability to get an
array of child nodes using the double-dot expression.

For example:

var mySVG:XML = <svg>
  <desc />
  <rect />
  <ellipse />
  <path />
  <path />
  <path />
  <ellipse />
  <path />
  <path />
</svg>;

We could write:

var paths:XMLList = mySVG..path;

I think if we can add support to just this (maybe a few more based on
popular usage), we should be able to get existing Flex apps easily ported.
Although, I am not sure how hard something like this would be to implement.

Thoughts?

Thanks,
Om


> This should allow getting chunks of XML in a more sane method than using
> hard-to-read E4X expressions.
>
> I’m hoping to have something commit-able on that front by the end of the
> week.
>
> I’ll try to follow what you are doing with DragDrop and Effects so I get a
> better feel of how it’s done.
>
> On Nov 9, 2015, at 7:36 PM, Alex Harui <ah...@adobe.com> wrote:
>
> > Harbs,
> >
> > Thanks for the commits for the back port.  Let us know if you actually
> > start work on porting an existing SWC.  I’m going to wait another day
> > before cutting the FalconJX RC2 in case Justin has more feedback on the
> > LICENSE.  So I am going to try to port dragdrop and effects.
> >
> > -Alex
> >
> > On 11/9/15, 12:12 AM, "Alex Harui" <ah...@adobe.com> wrote:
> >
> >>
> >>
> >> On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
> >>>
> >>>> Meanwhile, Peter and I are back porting the SWCs one-by-one in the
> >>>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to or
> >>>> anything, but you can look at Binding, Core, Collections and HTML as
> >>>> examples.
> >>>
> >>> Color me confused, but I still see all 3 source folders after switching
> >>> to the back port branch. It looks like you’re doing the back-port in
> the
> >>> as folder? The folder clean-up will happen after the port is complete?
> >>
> >> The asjs folders should be empty (more or less) in Binding, Core, and
> >> other completed SWCs.  We move all .as files to the as/src folder and
> fix
> >> up the compile-asjs-config.xml files.  We are keeping the js/src folders
> >> for now, then we’ll delete them once we feel good we’ve ported them
> >> correctly.
> >>>
> >>>> I think Peter is working on Charts and Mobile.  Graphics builds
> cleanly
> >>>> but Peter needs to verify it.  So, if you want, let us know what SWC
> >>>> you
> >>>> want to try and see if you can see the patterns from the other
> projects
> >>>> and change logs.
> >>>
> >>> I’ll spend some time studying this today. I just added compiler
> arguments
> >>> to the Flash Builder project files so AFB does not choke on the
> compiler
> >>> arguments. I’ll commit those changes shortly. FWIW, the only real error
> >>> I'm getting is in HTML IScrollingLayoutParent that ILayoutParent is not
> >>> found.
> >>
> >> I have to stop for the night, but I don’t think IScrollingLayoutParent
> >> should be there anymore.  Check the logs in the develop branch to
> verify.
> >>
> >> Thanks for helping.
> >> -Alex
> >>
> >
>
>

Re: [FlexJS] Back port

[Harbs <ha...@gmail.com>]

I’m not 100% clear on the status of the back-ports or the best way to go about it. I figured I’d wait until we’re done with the release and then coordinate with you and Peter.

In the meantime, I started work on some XML classes. I pretty much decided on departing a bit from classic E4X for two reasons: 1. It’s just about impossibly to reproduce the E4X behavior. 2. E4X was always awkward to use. I’m going to more or less copy the public API of XML, but my version of XMLList is going to work differently. Both XML and XMLList (I’m calling them XMLNode and XMLNodeList) will have filter functions which work similar to Array filter functions. This should allow getting chunks of XML in a more sane method than using hard-to-read E4X expressions.

I’m hoping to have something commit-able on that front by the end of the week.

I’ll try to follow what you are doing with DragDrop and Effects so I get a better feel of how it’s done.

On Nov 9, 2015, at 7:36 PM, Alex Harui <ah...@adobe.com> wrote:

> Harbs,
> 
> Thanks for the commits for the back port.  Let us know if you actually
> start work on porting an existing SWC.  I’m going to wait another day
> before cutting the FalconJX RC2 in case Justin has more feedback on the
> LICENSE.  So I am going to try to port dragdrop and effects.
> 
> -Alex
> 
> On 11/9/15, 12:12 AM, "Alex Harui" <ah...@adobe.com> wrote:
> 
>> 
>> 
>> On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
>>> 
>>>> Meanwhile, Peter and I are back porting the SWCs one-by-one in the
>>>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to or
>>>> anything, but you can look at Binding, Core, Collections and HTML as
>>>> examples.
>>> 
>>> Color me confused, but I still see all 3 source folders after switching
>>> to the back port branch. It looks like you’re doing the back-port in the
>>> as folder? The folder clean-up will happen after the port is complete?
>> 
>> The asjs folders should be empty (more or less) in Binding, Core, and
>> other completed SWCs.  We move all .as files to the as/src folder and fix
>> up the compile-asjs-config.xml files.  We are keeping the js/src folders
>> for now, then we’ll delete them once we feel good we’ve ported them
>> correctly.
>>> 
>>>> I think Peter is working on Charts and Mobile.  Graphics builds cleanly
>>>> but Peter needs to verify it.  So, if you want, let us know what SWC
>>>> you
>>>> want to try and see if you can see the patterns from the other projects
>>>> and change logs.
>>> 
>>> I’ll spend some time studying this today. I just added compiler arguments
>>> to the Flash Builder project files so AFB does not choke on the compiler
>>> arguments. I’ll commit those changes shortly. FWIW, the only real error
>>> I'm getting is in HTML IScrollingLayoutParent that ILayoutParent is not
>>> found.
>> 
>> I have to stop for the night, but I don’t think IScrollingLayoutParent
>> should be there anymore.  Check the logs in the develop branch to verify.
>> 
>> Thanks for helping.
>> -Alex
>> 
> 

Re: [FlexJS] Back port

[Alex Harui <ah...@adobe.com>]

Harbs,

Thanks for the commits for the back port.  Let us know if you actually
start work on porting an existing SWC.  I’m going to wait another day
before cutting the FalconJX RC2 in case Justin has more feedback on the
LICENSE.  So I am going to try to port dragdrop and effects.

-Alex

On 11/9/15, 12:12 AM, "Alex Harui" <ah...@adobe.com> wrote:

>
>
>On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
>>
>>> Meanwhile, Peter and I are back porting the SWCs one-by-one in the
>>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to or
>>> anything, but you can look at Binding, Core, Collections and HTML as
>>> examples.
>>
>>Color me confused, but I still see all 3 source folders after switching
>>to the back port branch. It looks like you’re doing the back-port in the
>>as folder? The folder clean-up will happen after the port is complete?
>
>The asjs folders should be empty (more or less) in Binding, Core, and
>other completed SWCs.  We move all .as files to the as/src folder and fix
>up the compile-asjs-config.xml files.  We are keeping the js/src folders
>for now, then we’ll delete them once we feel good we’ve ported them
>correctly.
>>
>>> I think Peter is working on Charts and Mobile.  Graphics builds cleanly
>>> but Peter needs to verify it.  So, if you want, let us know what SWC
>>>you
>>> want to try and see if you can see the patterns from the other projects
>>> and change logs.
>>
>>I’ll spend some time studying this today. I just added compiler arguments
>>to the Flash Builder project files so AFB does not choke on the compiler
>>arguments. I’ll commit those changes shortly. FWIW, the only real error
>>I'm getting is in HTML IScrollingLayoutParent that ILayoutParent is not
>>found.
>
>I have to stop for the night, but I don’t think IScrollingLayoutParent
>should be there anymore.  Check the logs in the develop branch to verify.
>
>Thanks for helping.
>-Alex
>

Re: [FlexJS] Back port

[Alex Harui <ah...@adobe.com>]

On 11/8/15, 11:55 PM, "Harbs" <ha...@gmail.com> wrote:
>
>> Meanwhile, Peter and I are back porting the SWCs one-by-one in the
>> core_js_to_as branch of flex-asjs.  We haven’t written a how-to or
>> anything, but you can look at Binding, Core, Collections and HTML as
>> examples.
>
>Color me confused, but I still see all 3 source folders after switching
>to the back port branch. It looks like you’re doing the back-port in the
>as folder? The folder clean-up will happen after the port is complete?

The asjs folders should be empty (more or less) in Binding, Core, and
other completed SWCs.  We move all .as files to the as/src folder and fix
up the compile-asjs-config.xml files.  We are keeping the js/src folders
for now, then we’ll delete them once we feel good we’ve ported them
correctly.
>
>> I think Peter is working on Charts and Mobile.  Graphics builds cleanly
>> but Peter needs to verify it.  So, if you want, let us know what SWC you
>> want to try and see if you can see the patterns from the other projects
>> and change logs.
>
>I’ll spend some time studying this today. I just added compiler arguments
>to the Flash Builder project files so AFB does not choke on the compiler
>arguments. I’ll commit those changes shortly. FWIW, the only real error
>I'm getting is in HTML IScrollingLayoutParent that ILayoutParent is not
>found.

I have to stop for the night, but I don’t think IScrollingLayoutParent
should be there anymore.  Check the logs in the develop branch to verify.

Thanks for helping.
-Alex

Re: [FlexJS] Back port

[Harbs <ha...@gmail.com>]

On Nov 9, 2015, at 9:45 AM, Alex Harui <ah...@adobe.com> wrote:

> Well, top priority is to get FalconJX out so we can release it and FlexJS.

Sure. I meant top priority for me. I don’t know how to work on the compiler… ;-)

> Meanwhile, Peter and I are back porting the SWCs one-by-one in the
> core_js_to_as branch of flex-asjs.  We haven’t written a how-to or
> anything, but you can look at Binding, Core, Collections and HTML as
> examples.

Color me confused, but I still see all 3 source folders after switching to the back port branch. It looks like you’re doing the back-port in the as folder? The folder clean-up will happen after the port is complete?

> I think Peter is working on Charts and Mobile.  Graphics builds cleanly
> but Peter needs to verify it.  So, if you want, let us know what SWC you
> want to try and see if you can see the patterns from the other projects
> and change logs.

I’ll spend some time studying this today. I just added compiler arguments to the Flash Builder project files so AFB does not choke on the compiler arguments. I’ll commit those changes shortly. FWIW, the only real error I'm getting is in HTML IScrollingLayoutParent that ILayoutParent is not found.

> Thanks,
> -Alex  
> 
> On 11/8/15, 11:29 PM, "Harbs" <ha...@gmail.com> wrote:
> 
>> I did not realize that you were in middle of a back port. It seems to me
>> that this is top priority and I’d like to help.
>> 
>> Is there a to-do list somewhere of things that need to be done for the
>> back-port?
>> 
>> On Nov 9, 2015, at 3:14 AM, Alex Harui <ah...@adobe.com> wrote:
>> 
>>> This will change once we finish the back port.  There will be one source
>>> folder.
>> 
> 

Re: [FlexJS] Back port

[Alex Harui <ah...@adobe.com>]

Well, top priority is to get FalconJX out so we can release it and FlexJS.

Meanwhile, Peter and I are back porting the SWCs one-by-one in the
core_js_to_as branch of flex-asjs.  We haven’t written a how-to or
anything, but you can look at Binding, Core, Collections and HTML as
examples.

I think Peter is working on Charts and Mobile.  Graphics builds cleanly
but Peter needs to verify it.  So, if you want, let us know what SWC you
want to try and see if you can see the patterns from the other projects
and change logs.

Thanks,
-Alex  

On 11/8/15, 11:29 PM, "Harbs" <ha...@gmail.com> wrote:

>I did not realize that you were in middle of a back port. It seems to me
>that this is top priority and I’d like to help.
>
>Is there a to-do list somewhere of things that need to be done for the
>back-port?
>
>On Nov 9, 2015, at 3:14 AM, Alex Harui <ah...@adobe.com> wrote:
>
>> This will change once we finish the back port.  There will be one source
>> folder.
>

[FlexJS] Back port

[Harbs <ha...@gmail.com>]

I did not realize that you were in middle of a back port. It seems to me that this is top priority and I’d like to help.

Is there a to-do list somewhere of things that need to be done for the back-port?

On Nov 9, 2015, at 3:14 AM, Alex Harui <ah...@adobe.com> wrote:

> This will change once we finish the back port.  There will be one source
> folder.

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

On 11/8/15, 2:56 AM, "Harbs" <ha...@gmail.com> wrote:

>I was kind of expecting a way to link the source code to the SWCs. I sort
>of expect users to want to be able to jump into the source, so they can
>see how some part of the SDK is working. As a longer term goal, I think
>it’s a good one.

Agreed.  I think that you add the SWC projects to your app project.  I’ll
know more once I get back to it after finishing the back port.

>
>I’m fine adding each piece as a separate project for now, but I’m still
>having issues.
>
>There’s various missing swc files, but I can solve that by just removing
>them.
>
>There’s one error I’m not sure what to do with:
>CreateJS cannot find the src folder
>
>My biggest issue right now, is I do not understand the structure of the
>projects:
>
>There are 3 different src folders. We have as, js and asjs. I’m not sure
>how to deal with these three folders — especially in the context of
>debugging in an IDE.

This will change once we finish the back port.  There will be one source
folder.

>
>The js folder is pretty clear. It’s js code which is used by the Closure
>Compiler. Right? I’m assuming that this code bypasses FlaconJX completely.
>The as folder is for Flash specific code?
>What’s the asjs folder for? Is that for common code?

Yes.

>
>Let’s say I want to create a new XML class which is a proxy for Document
>on the javacript side and a proxy for XML on the flash side. How would I
>set that up? I’d ideally like to write the whole thing in ActionScript,
>and let FalconJX create the javascript code necessary for JS. I know Fred
>got conditional compiling working, and that seems like the most
>attractive way of coding this to me, but I’m not sure what the right way
>to set that up would be.

If you want, you can switch to the core_js_to_as branch and try using the
conditional compile there.  But the FB integration may not work as well
yet.

-Alex

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Harbs <ha...@gmail.com>]

I was kind of expecting a way to link the source code to the SWCs. I sort of expect users to want to be able to jump into the source, so they can see how some part of the SDK is working. As a longer term goal, I think it’s a good one.

I’m fine adding each piece as a separate project for now, but I’m still having issues.

There’s various missing swc files, but I can solve that by just removing them.

There’s one error I’m not sure what to do with:
CreateJS cannot find the src folder

My biggest issue right now, is I do not understand the structure of the projects:

There are 3 different src folders. We have as, js and asjs. I’m not sure how to deal with these three folders — especially in the context of debugging in an IDE.

The js folder is pretty clear. It’s js code which is used by the Closure Compiler. Right? I’m assuming that this code bypasses FlaconJX completely.
The as folder is for Flash specific code?
What’s the asjs folder for? Is that for common code?

Let’s say I want to create a new XML class which is a proxy for Document on the javacript side and a proxy for XML on the flash side. How would I set that up? I’d ideally like to write the whole thing in ActionScript, and let FalconJX create the javascript code necessary for JS. I know Fred got conditional compiling working, and that seems like the most attractive way of coding this to me, but I’m not sure what the right way to set that up would be.

On Nov 6, 2015, at 12:13 PM, OmPrakash Muppirala <bi...@gmail.com> wrote:

> For 2: I have checked in the Flash Builder project files for all the FlexJS
> framework swc projects.  Have you tried simply importing those Flex Library
> projects into FlashBuilder all at once?  What kind of errors are you seeing?

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Harbs <ha...@gmail.com>]

I’ll give this a go again on Sunday. Even if I do get this setup, the fact that I’m having so much trouble is probably an indicator that others who want to contribute will also.

Bottom line: The simpler we make it to contribute, the more likely we will get more contributors. The more contributors we have, the quicker FlexJS will mature and the more releases we can get out the door. (which should then be a self-feeding process)

I think that making things simple should be higher priority than adding features because if things are simpler, features should be added quicker by default with more contributors.

When I do get my IDE setup for working on the source, I will try to document the process on the wiki.

Harbs

On Nov 6, 2015, at 12:13 PM, OmPrakash Muppirala <bi...@gmail.com> wrote:

> On Fri, Nov 6, 2015 at 2:02 AM, Harbs <ha...@gmail.com> wrote:
> 
>> I agree with Justin here. I’ve been playing with FlexJS for a good few
>> weeks already and I still do not have a good setup for working on the SDK.
>> I would have been productive a long time ago already if we would have the
>> following:
>> 
>> 1. Simple building of the SDK. Environment variables are tough to get
>> right. Either Maven or some GUI which would help users set environment
>> variables would be very helpful.
>> 2. A way to make the SDK built from source IDE compatible.
>> 3. Source linking between the source and built SDK.
>> 
>> I think these three items will go a long way in adding contributors to the
>> SDK and hopefully get out quicker releases.
>> 
>> 
> I have been able to get 2 and 3 working for me. It's a bit tricky to setup
> and I am not sure how to share that setup with others.
> 
> For 2: I have checked in the Flash Builder project files for all the FlexJS
> framework swc projects.  Have you tried simply importing those Flex Library
> projects into FlashBuilder all at once?  What kind of errors are you seeing?
> 
> For 3: After selecting FlexJS as your SDK in Flash Builder, you can simply
> add the above mentioned Flex Library projects as 'project dependencies' to
> the FlexJS app project.  This way you can tweak the SDK code as you are
> building up your FlexJS app example.  Again, if you can try this setup and
> let me know what errors you are getting, I can try to help you with that.
> 
> Thanks,
> Om
> 
> 
>> On Nov 6, 2015, at 11:08 AM, Justin Mclean <ju...@me.com> wrote:
>> 
>>> HI,
>>> 
>>>> Nobody is saying simpler builds wouldn’t help.  But how much?
>>> 
>>> A lot I think. Developer are unable to contribute if they can’t compile
>> and test their changes.
>>> 
>>>> More than a more functional DataGrid?  Or a more Spark and/or MX-like
>> component set
>>>> that reduces migration time for your app?  Or being able to write your
>> JS
>>>> using AS?
>>> 
>>> IMO you would more likely get people contribution things like that if it
>> was simpler to build.
>>> 
>>>> To folks reading this who are interested in
>>>> FlexJS: what is the most important two or three things we should be
>>>> working on?
>>> 
>>> IMO:
>>> 1. Simpler builds
>>> 2. More frequent releases (say every 3 or 4 months).
>>> 3. Compiling MXML
>>> 4. Working documented examples / how to port existing applications
>>> 
>>>> The build script was improved in this release such that if you have a
>>>> couple of known sets of circumstances, like not having any of the repos,
>>>> that running “ant all” will get all the repos and build them in the
>> right
>>>> order.
>>> 
>>> I meant to ask before does that use well known versions or does it just
>> grab the latest out of the develop (or another) branch?
>>> 
>>> Thanks,
>>> Justin
>> 
>> 

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[OmPrakash Muppirala <bi...@gmail.com>]

On Fri, Nov 6, 2015 at 2:02 AM, Harbs <ha...@gmail.com> wrote:

> I agree with Justin here. I’ve been playing with FlexJS for a good few
> weeks already and I still do not have a good setup for working on the SDK.
> I would have been productive a long time ago already if we would have the
> following:
>
> 1. Simple building of the SDK. Environment variables are tough to get
> right. Either Maven or some GUI which would help users set environment
> variables would be very helpful.
> 2. A way to make the SDK built from source IDE compatible.
> 3. Source linking between the source and built SDK.
>
> I think these three items will go a long way in adding contributors to the
> SDK and hopefully get out quicker releases.
>
>
I have been able to get 2 and 3 working for me. It's a bit tricky to setup
and I am not sure how to share that setup with others.

For 2: I have checked in the Flash Builder project files for all the FlexJS
framework swc projects.  Have you tried simply importing those Flex Library
projects into FlashBuilder all at once?  What kind of errors are you seeing?

For 3: After selecting FlexJS as your SDK in Flash Builder, you can simply
add the above mentioned Flex Library projects as 'project dependencies' to
the FlexJS app project.  This way you can tweak the SDK code as you are
building up your FlexJS app example.  Again, if you can try this setup and
let me know what errors you are getting, I can try to help you with that.

Thanks,
Om


> On Nov 6, 2015, at 11:08 AM, Justin Mclean <ju...@me.com> wrote:
>
> > HI,
> >
> >> Nobody is saying simpler builds wouldn’t help.  But how much?
> >
> > A lot I think. Developer are unable to contribute if they can’t compile
> and test their changes.
> >
> >> More than a more functional DataGrid?  Or a more Spark and/or MX-like
> component set
> >> that reduces migration time for your app?  Or being able to write your
> JS
> >> using AS?
> >
> > IMO you would more likely get people contribution things like that if it
> was simpler to build.
> >
> >> To folks reading this who are interested in
> >> FlexJS: what is the most important two or three things we should be
> >> working on?
> >
> > IMO:
> > 1. Simpler builds
> > 2. More frequent releases (say every 3 or 4 months).
> > 3. Compiling MXML
> > 4. Working documented examples / how to port existing applications
> >
> >> The build script was improved in this release such that if you have a
> >> couple of known sets of circumstances, like not having any of the repos,
> >> that running “ant all” will get all the repos and build them in the
> right
> >> order.
> >
> > I meant to ask before does that use well known versions or does it just
> grab the latest out of the develop (or another) branch?
> >
> > Thanks,
> > Justin
>
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Harbs <ha...@gmail.com>]

I agree with Justin here. I’ve been playing with FlexJS for a good few weeks already and I still do not have a good setup for working on the SDK. I would have been productive a long time ago already if we would have the following:

1. Simple building of the SDK. Environment variables are tough to get right. Either Maven or some GUI which would help users set environment variables would be very helpful.
2. A way to make the SDK built from source IDE compatible.
3. Source linking between the source and built SDK.

I think these three items will go a long way in adding contributors to the SDK and hopefully get out quicker releases.

On Nov 6, 2015, at 11:08 AM, Justin Mclean <ju...@me.com> wrote:

> HI,
> 
>> Nobody is saying simpler builds wouldn’t help.  But how much?
> 
> A lot I think. Developer are unable to contribute if they can’t compile and test their changes.
> 
>> More than a more functional DataGrid?  Or a more Spark and/or MX-like component set
>> that reduces migration time for your app?  Or being able to write your JS
>> using AS?
> 
> IMO you would more likely get people contribution things like that if it was simpler to build.
> 
>> To folks reading this who are interested in
>> FlexJS: what is the most important two or three things we should be
>> working on?
> 
> IMO:
> 1. Simpler builds
> 2. More frequent releases (say every 3 or 4 months).
> 3. Compiling MXML
> 4. Working documented examples / how to port existing applications
> 
>> The build script was improved in this release such that if you have a
>> couple of known sets of circumstances, like not having any of the repos,
>> that running “ant all” will get all the repos and build them in the right
>> order.
> 
> I meant to ask before does that use well known versions or does it just grab the latest out of the develop (or another) branch?
> 
> Thanks,
> Justin

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Justin Mclean <ju...@me.com>]

HI,

> Nobody is saying simpler builds wouldn’t help.  But how much?

A lot I think. Developer are unable to contribute if they can’t compile and test their changes.

>  More than a more functional DataGrid?  Or a more Spark and/or MX-like component set
> that reduces migration time for your app?  Or being able to write your JS
> using AS?

IMO you would more likely get people contribution things like that if it was simpler to build.

> To folks reading this who are interested in
> FlexJS: what is the most important two or three things we should be
> working on?

IMO:
1. Simpler builds
2. More frequent releases (say every 3 or 4 months).
3. Compiling MXML
4. Working documented examples / how to port existing applications

> The build script was improved in this release such that if you have a
> couple of known sets of circumstances, like not having any of the repos,
> that running “ant all” will get all the repos and build them in the right
> order.

I meant to ask before does that use well known versions or does it just grab the latest out of the develop (or another) branch?

Thanks,
Justin

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

On 11/5/15, 11:30 PM, "Justin Mclean" <ju...@me.com> wrote:

>I think trying to have a simpler build would help a lot. This ties into
>what Chris has just posted as well.

Nobody is saying simpler builds wouldn’t help.  But how much?  More than a
more functional DataGrid?  Or a more Spark and/or MX-like component set
that reduces migration time for your app?  Or being able to write your JS
using AS?  And what does simpler mean?  Since Apache is a do-ocracy, will
you help us make it simpler?  To folks reading this who are interested in
FlexJS: what is the most important two or three things we should be
working on?

AIUI, Apache Open Office has thousands of users, most of whom can’t build
AOO from source.  They just want to use the product.  I think Flex is
similar.  Most of our users don’t care if they can build the release from
source.  You are correct that for both AOO and Flex they should care to
the extent that knowing the source is truly open and accounted for and
attracting new committers is what keeps a project going.  But my sense is
that they are mainly focused on getting their apps up and running.

The build script was improved in this release such that if you have a
couple of known sets of circumstances, like not having any of the repos,
that running “ant all” will get all the repos and build them in the right
order.  But for existing committers like you, who may have repos in
different folder structures with different names, it is much harder to
make “ant all” work.

Anyway, let’s get some concrete suggestions about what else we could do to
the build system to make it even simpler.  I’m off to bed for tonight.

Thanks,
-Alex

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Justin Mclean <ju...@me.com>]

Hi,

> Apache is not “people”.  Apache is the organization that we have certain
> obligations to, such as the source package you mention.  “People” is our
> customers and plenty of them are using the Installer and the binary
> packages.  Let’s focus on the customers.

In this context I would of said Apache is a community not an organisation. Apache project creates software for the common good. If the project can’t easily build from source code and/or only a few people can produce what's usable then that’s potentially a issue for the long term health of the project and is likely to hinderance growth. That’s why one of the requirements for checking a release is being able to build it.  I know it’s still alpha/beta so there going to be a few rough edges and hoops to jump through but I think trying to have a simpler build would help a lot. This ties into  what Chris has just posted as well.

Thanks,
Justin

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

I was just about to respond similarly.  My version is:

Apache is not “people”.  Apache is the organization that we have certain
obligations to, such as the source package you mention.  “People” is our
customers and plenty of them are using the Installer and the binary
packages.  Let’s focus on the customers.

-Alex

On 11/5/15, 10:29 PM, "Harbs" <ha...@gmail.com> wrote:

>Justin,
>
>While that’s technically true, the vast majority of users would want a
>convenience package.
>
>On Nov 6, 2015, at 8:23 AM, Justin Mclean <ju...@me.com> wrote:
>
>> Hi,
>> 
>>> People who are having trouble with building are trying to build from
>>>the
>>> source distribution and/or the Git repository. Things are a little more
>>> complicated that way, but the vast majority of people won't need to
>>>deal
>>> with that type of build.
>> 
>> The building from source is important as far as Apache is concerned. In
>>fact of the source package is considered the release [1][2]
>> 
>> Thanks,
>> Justin
>> 
>> 1. http://www.apache.org/dev/release.html#what
>> 2. 
>>http://www.apache.org/dev/release.html#what-must-every-release-contain
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Harbs <ha...@gmail.com>]

Justin,

While that’s technically true, the vast majority of users would want a convenience package.

On Nov 6, 2015, at 8:23 AM, Justin Mclean <ju...@me.com> wrote:

> Hi,
> 
>> People who are having trouble with building are trying to build from the
>> source distribution and/or the Git repository. Things are a little more
>> complicated that way, but the vast majority of people won't need to deal
>> with that type of build.
> 
> The building from source is important as far as Apache is concerned. In fact of the source package is considered the release [1][2]
> 
> Thanks,
> Justin
> 
> 1. http://www.apache.org/dev/release.html#what
> 2. http://www.apache.org/dev/release.html#what-must-every-release-contain

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Justin Mclean <ju...@me.com>]

Hi,

> People who are having trouble with building are trying to build from the
> source distribution and/or the Git repository. Things are a little more
> complicated that way, but the vast majority of people won't need to deal
> with that type of build.

The building from source is important as far as Apache is concerned. In fact of the source package is considered the release [1][2]

Thanks,
Justin

1. http://www.apache.org/dev/release.html#what
2. http://www.apache.org/dev/release.html#what-must-every-release-contain

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Josh Tynjala <jo...@gmail.com>]

As Alex said, most people who want to try out FlexJS 0.5.0 will simply use
the Apache Flex SDK Installer. It'll be super easy for them. In fact, you
can already install the nightlies and RC1 using the installer if you
right-click and choose Show Dev Builds. Once 0.5.0 is released, the build
will show up on the regular list, obviously.

People who are having trouble with building are trying to build from the
source distribution and/or the Git repository. Things are a little more
complicated that way, but the vast majority of people won't need to deal
with that type of build.

- Josh

On Thu, Nov 5, 2015 at 3:21 PM, jude <fl...@gmail.com> wrote:

> Alex,
> I'm going to go ahead and put together an app so that you can install
> FlexJS with one click and compile. I want it to be as simple as possible
> for anyone (new web developers). I've downloaded the build script and
> looked over it but there is some code in it for voting and release. I'm
> sure you will be repeating yourself again but here's my requirements:
>
> • button to validate FlexJS requirements (paths, etc)
> • button to set paths for the user
> • one button to download FlexJS
> • another button to run FlexJS on a project app (without an IDE)
>
> Are there targets to do each of these (or where should I start)? Since it's
> an AIR app I want to run each target independently. BTW I'm bundling ant
> with the app so that the user doesn't have to install anything extra. If
> someone has FB or IntelliJ installed does the install directory need to
> change? I'll put the project on github or post it to the group as soon as
> it's usable.
>
>
> On Thu, Nov 5, 2015 at 6:32 AM, Alex Harui <ah...@adobe.com> wrote:
>
> >
> >
> > On 11/5/15, 3:56 AM, "Harbs" <ha...@gmail.com> wrote:
> >
> > >I’m a bit confused about the release process.
> > >
> > >I thought we were creating release branches in git for each release to
> > >“freeze” the code, so we do not have a wildly moving target. It does not
> > >seem like that’s happening, so I’m not sure if I just misunderstood.
> >
> > You understand correctly.  I’m cheating right now because it is just more
> > work to set up release branches, and there isn’t a lot of non-critical
> > work going on the develop branch.  Peter and I are working on the back
> > port from JS to AS in a separate branch.  Only important fixes are being
> > pushed to develop.  If we had more folks contributing more often, then I
> > would have used a release branch.
> >
> > Even release branches have historically moved because folks don’t start
> > testing until late in the game and find important bugs.
> >
> > -Alex
> >
> > >
> > >Harbs
> > >
> > >On Nov 3, 2015, at 12:04 AM, Alex Harui <ah...@adobe.com> wrote:
> > >
> > >>
> > >>
> > >> On 10/30/15, 3:19 PM, "Justin Mclean" <ju...@classsoftware.com>
> wrote:
> > >>
> > >>> Hi,
> > >>>
> > >>>> Hmm, I was hoping more PMC folks would respond.  Remember that,
> > >>>> according
> > >>>> to the release process, the PMC folks planning to vote are supposed
> to
> > >>>> be
> > >>>> running tests now. In theory, the only new test to be run  after we
> > >>>> start
> > >>>> the vote is whether the PGP signature is valid.
> > >>>
> > >>> We’re continually trying to test a moving target which involves a
> > >>>greater
> > >>> time commitment that I currently have available. You’ve never quite
> > >>>sure
> > >>> if the version you testing is going to be the version in the release
> > >>> candidate and unless you very carefully follow all of the commits
> it’s
> > >>> not obvious what needs to be retested at two different time
> intervals.
> > >>
> > >> Hmm, pleading is working so maybe I’ll try guilt-tripping.
> > >>
> > >> Yes, the nightly builds are a moving target.  IMO, we all want to grow
> > >>the
> > >> community by attracting customers and hopefully convert a few to be
> > >> committers and the only way I know to do it is to keep making the code
> > >> better and releasing the best release we can in the most efficient
> > >>manner.
> > >> IMO, freezing a branch and not allowing important bug fixes that might
> > >> make a difference in whether someone becomes more active in our
> > >>community
> > >> doesn’t make sense.  Taking the time to build out an RC and post it
> and
> > >> start a vote thread in order to finally get some testing isn’t very
> > >> efficient either.
> > >>
> > >> Historically, when we produced an RC and immediately started a release
> > >> vote, bugs would be found and we’d cancel the RC and roll out another
> > >>one.
> > >> The goal of the release process we voted in, IMO, was to reduce this
> > >> overhead of posting RCs, opening and closing vote threads, etc. so we
> > >>can
> > >> more efficiently achieve the goal of serving our customers and
> > >>attracting
> > >> some of them to becoming committers so we can have more people find
> bugs
> > >> sooner by working with the develop branch.
> > >>
> > >> Recently, I’ve spent several days on improving build and approval
> > >>scripts
> > >> so testing what is in development takes less time.  In theory, you can
> > >>now
> > >> start up the approval script which will pull down the bits, answer a
> few
> > >> questions, then go do something else for 5 to 25 minutes and then come
> > >> back and poke at it.  I would have rather spent that time on features
> > >>for
> > >> our customers, but I gambled that this would help us get the release
> out
> > >> sooner.  I’m not sure that paid off.
> > >>
> > >>
> > >> I don’t have any other ideas on how to make it easier for those of you
> > >>who
> > >> contribute in your spare time to stay up on the commits and bugs.  It
> > >> should be ok to take any nightly build and run it through your tests
> and
> > >> report your findings.  Ideally, you would be up to date on the commits
> > >>and
> > >> bugs and other discussions to know whether what you find is a
> duplicate
> > >>or
> > >> not, but at this point, I don’t care if you report a duplicate.  At
> > >>least
> > >> that means you verified a bunch of other code paths worked for you.  I
> > >> don’t know how other Apache projects with really active code bases do
> > >>it.
> > >>
> > >> It is certainly fine to be too busy to vote on a release.  I was
> hoping
> > >>to
> > >> get more folks to poke at the bits before starting a vote because it
> > >>will
> > >> be a waste of community energy to start a vote and then have a bunch
> of
> > >> PMC voters jump in and start reporting important bugs.  But I think
> the
> > >> community has waited too long already, so I am going to start a vote
> > >>soon,
> > >> and Peter and I will vote and maybe Josh and/or Harbs and we’ll be
> good
> > >>to
> > >> go.  Hopefully any others jumping in late won’t find release blockers
> > >>and
> > >> we’ll just make another release later.
> > >>
> > >>
> > >> -Alex
> > >
> >
> >
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[jude <fl...@gmail.com>]

Alex,
I'm going to go ahead and put together an app so that you can install
FlexJS with one click and compile. I want it to be as simple as possible
for anyone (new web developers). I've downloaded the build script and
looked over it but there is some code in it for voting and release. I'm
sure you will be repeating yourself again but here's my requirements:

• button to validate FlexJS requirements (paths, etc)
• button to set paths for the user
• one button to download FlexJS
• another button to run FlexJS on a project app (without an IDE)

Are there targets to do each of these (or where should I start)? Since it's
an AIR app I want to run each target independently. BTW I'm bundling ant
with the app so that the user doesn't have to install anything extra. If
someone has FB or IntelliJ installed does the install directory need to
change? I'll put the project on github or post it to the group as soon as
it's usable.


On Thu, Nov 5, 2015 at 6:32 AM, Alex Harui <ah...@adobe.com> wrote:

>
>
> On 11/5/15, 3:56 AM, "Harbs" <ha...@gmail.com> wrote:
>
> >I’m a bit confused about the release process.
> >
> >I thought we were creating release branches in git for each release to
> >“freeze” the code, so we do not have a wildly moving target. It does not
> >seem like that’s happening, so I’m not sure if I just misunderstood.
>
> You understand correctly.  I’m cheating right now because it is just more
> work to set up release branches, and there isn’t a lot of non-critical
> work going on the develop branch.  Peter and I are working on the back
> port from JS to AS in a separate branch.  Only important fixes are being
> pushed to develop.  If we had more folks contributing more often, then I
> would have used a release branch.
>
> Even release branches have historically moved because folks don’t start
> testing until late in the game and find important bugs.
>
> -Alex
>
> >
> >Harbs
> >
> >On Nov 3, 2015, at 12:04 AM, Alex Harui <ah...@adobe.com> wrote:
> >
> >>
> >>
> >> On 10/30/15, 3:19 PM, "Justin Mclean" <ju...@classsoftware.com> wrote:
> >>
> >>> Hi,
> >>>
> >>>> Hmm, I was hoping more PMC folks would respond.  Remember that,
> >>>> according
> >>>> to the release process, the PMC folks planning to vote are supposed to
> >>>> be
> >>>> running tests now. In theory, the only new test to be run  after we
> >>>> start
> >>>> the vote is whether the PGP signature is valid.
> >>>
> >>> We’re continually trying to test a moving target which involves a
> >>>greater
> >>> time commitment that I currently have available. You’ve never quite
> >>>sure
> >>> if the version you testing is going to be the version in the release
> >>> candidate and unless you very carefully follow all of the commits it’s
> >>> not obvious what needs to be retested at two different time intervals.
> >>
> >> Hmm, pleading is working so maybe I’ll try guilt-tripping.
> >>
> >> Yes, the nightly builds are a moving target.  IMO, we all want to grow
> >>the
> >> community by attracting customers and hopefully convert a few to be
> >> committers and the only way I know to do it is to keep making the code
> >> better and releasing the best release we can in the most efficient
> >>manner.
> >> IMO, freezing a branch and not allowing important bug fixes that might
> >> make a difference in whether someone becomes more active in our
> >>community
> >> doesn’t make sense.  Taking the time to build out an RC and post it and
> >> start a vote thread in order to finally get some testing isn’t very
> >> efficient either.
> >>
> >> Historically, when we produced an RC and immediately started a release
> >> vote, bugs would be found and we’d cancel the RC and roll out another
> >>one.
> >> The goal of the release process we voted in, IMO, was to reduce this
> >> overhead of posting RCs, opening and closing vote threads, etc. so we
> >>can
> >> more efficiently achieve the goal of serving our customers and
> >>attracting
> >> some of them to becoming committers so we can have more people find bugs
> >> sooner by working with the develop branch.
> >>
> >> Recently, I’ve spent several days on improving build and approval
> >>scripts
> >> so testing what is in development takes less time.  In theory, you can
> >>now
> >> start up the approval script which will pull down the bits, answer a few
> >> questions, then go do something else for 5 to 25 minutes and then come
> >> back and poke at it.  I would have rather spent that time on features
> >>for
> >> our customers, but I gambled that this would help us get the release out
> >> sooner.  I’m not sure that paid off.
> >>
> >>
> >> I don’t have any other ideas on how to make it easier for those of you
> >>who
> >> contribute in your spare time to stay up on the commits and bugs.  It
> >> should be ok to take any nightly build and run it through your tests and
> >> report your findings.  Ideally, you would be up to date on the commits
> >>and
> >> bugs and other discussions to know whether what you find is a duplicate
> >>or
> >> not, but at this point, I don’t care if you report a duplicate.  At
> >>least
> >> that means you verified a bunch of other code paths worked for you.  I
> >> don’t know how other Apache projects with really active code bases do
> >>it.
> >>
> >> It is certainly fine to be too busy to vote on a release.  I was hoping
> >>to
> >> get more folks to poke at the bits before starting a vote because it
> >>will
> >> be a waste of community energy to start a vote and then have a bunch of
> >> PMC voters jump in and start reporting important bugs.  But I think the
> >> community has waited too long already, so I am going to start a vote
> >>soon,
> >> and Peter and I will vote and maybe Josh and/or Harbs and we’ll be good
> >>to
> >> go.  Hopefully any others jumping in late won’t find release blockers
> >>and
> >> we’ll just make another release later.
> >>
> >>
> >> -Alex
> >
>
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Alex Harui <ah...@adobe.com>]

On 11/5/15, 3:56 AM, "Harbs" <ha...@gmail.com> wrote:

>I’m a bit confused about the release process.
>
>I thought we were creating release branches in git for each release to
>“freeze” the code, so we do not have a wildly moving target. It does not
>seem like that’s happening, so I’m not sure if I just misunderstood.

You understand correctly.  I’m cheating right now because it is just more
work to set up release branches, and there isn’t a lot of non-critical
work going on the develop branch.  Peter and I are working on the back
port from JS to AS in a separate branch.  Only important fixes are being
pushed to develop.  If we had more folks contributing more often, then I
would have used a release branch.

Even release branches have historically moved because folks don’t start
testing until late in the game and find important bugs.

-Alex

>
>Harbs
>
>On Nov 3, 2015, at 12:04 AM, Alex Harui <ah...@adobe.com> wrote:
>
>> 
>> 
>> On 10/30/15, 3:19 PM, "Justin Mclean" <ju...@classsoftware.com> wrote:
>> 
>>> Hi,
>>> 
>>>> Hmm, I was hoping more PMC folks would respond.  Remember that,
>>>> according
>>>> to the release process, the PMC folks planning to vote are supposed to
>>>> be
>>>> running tests now. In theory, the only new test to be run  after we
>>>> start
>>>> the vote is whether the PGP signature is valid.
>>> 
>>> We’re continually trying to test a moving target which involves a
>>>greater
>>> time commitment that I currently have available. You’ve never quite
>>>sure
>>> if the version you testing is going to be the version in the release
>>> candidate and unless you very carefully follow all of the commits it’s
>>> not obvious what needs to be retested at two different time intervals.
>> 
>> Hmm, pleading is working so maybe I’ll try guilt-tripping.
>> 
>> Yes, the nightly builds are a moving target.  IMO, we all want to grow
>>the
>> community by attracting customers and hopefully convert a few to be
>> committers and the only way I know to do it is to keep making the code
>> better and releasing the best release we can in the most efficient
>>manner.
>> IMO, freezing a branch and not allowing important bug fixes that might
>> make a difference in whether someone becomes more active in our
>>community
>> doesn’t make sense.  Taking the time to build out an RC and post it and
>> start a vote thread in order to finally get some testing isn’t very
>> efficient either.
>> 
>> Historically, when we produced an RC and immediately started a release
>> vote, bugs would be found and we’d cancel the RC and roll out another
>>one.
>> The goal of the release process we voted in, IMO, was to reduce this
>> overhead of posting RCs, opening and closing vote threads, etc. so we
>>can
>> more efficiently achieve the goal of serving our customers and
>>attracting
>> some of them to becoming committers so we can have more people find bugs
>> sooner by working with the develop branch.
>> 
>> Recently, I’ve spent several days on improving build and approval
>>scripts
>> so testing what is in development takes less time.  In theory, you can
>>now
>> start up the approval script which will pull down the bits, answer a few
>> questions, then go do something else for 5 to 25 minutes and then come
>> back and poke at it.  I would have rather spent that time on features
>>for
>> our customers, but I gambled that this would help us get the release out
>> sooner.  I’m not sure that paid off.
>> 
>> 
>> I don’t have any other ideas on how to make it easier for those of you
>>who
>> contribute in your spare time to stay up on the commits and bugs.  It
>> should be ok to take any nightly build and run it through your tests and
>> report your findings.  Ideally, you would be up to date on the commits
>>and
>> bugs and other discussions to know whether what you find is a duplicate
>>or
>> not, but at this point, I don’t care if you report a duplicate.  At
>>least
>> that means you verified a bunch of other code paths worked for you.  I
>> don’t know how other Apache projects with really active code bases do
>>it.
>> 
>> It is certainly fine to be too busy to vote on a release.  I was hoping
>>to
>> get more folks to poke at the bits before starting a vote because it
>>will
>> be a waste of community energy to start a vote and then have a bunch of
>> PMC voters jump in and start reporting important bugs.  But I think the
>> community has waited too long already, so I am going to start a vote
>>soon,
>> and Peter and I will vote and maybe Josh and/or Harbs and we’ll be good
>>to
>> go.  Hopefully any others jumping in late won’t find release blockers
>>and
>> we’ll just make another release later.
>> 
>> 
>> -Alex
>

Re: [DISCUSS] Release Apache FlexJS 0.5.0

[Harbs <ha...@gmail.com>]

I’m a bit confused about the release process.

I thought we were creating release branches in git for each release to “freeze” the code, so we do not have a wildly moving target. It does not seem like that’s happening, so I’m not sure if I just misunderstood.

Harbs

On Nov 3, 2015, at 12:04 AM, Alex Harui <ah...@adobe.com> wrote:

> 
> 
> On 10/30/15, 3:19 PM, "Justin Mclean" <ju...@classsoftware.com> wrote:
> 
>> Hi,
>> 
>>> Hmm, I was hoping more PMC folks would respond.  Remember that,
>>> according
>>> to the release process, the PMC folks planning to vote are supposed to
>>> be
>>> running tests now. In theory, the only new test to be run  after we
>>> start
>>> the vote is whether the PGP signature is valid.
>> 
>> We’re continually trying to test a moving target which involves a greater
>> time commitment that I currently have available. You’ve never quite sure
>> if the version you testing is going to be the version in the release
>> candidate and unless you very carefully follow all of the commits it’s
>> not obvious what needs to be retested at two different time intervals.
> 
> Hmm, pleading is working so maybe I’ll try guilt-tripping.
> 
> Yes, the nightly builds are a moving target.  IMO, we all want to grow the
> community by attracting customers and hopefully convert a few to be
> committers and the only way I know to do it is to keep making the code
> better and releasing the best release we can in the most efficient manner.
> IMO, freezing a branch and not allowing important bug fixes that might
> make a difference in whether someone becomes more active in our community
> doesn’t make sense.  Taking the time to build out an RC and post it and
> start a vote thread in order to finally get some testing isn’t very
> efficient either.
> 
> Historically, when we produced an RC and immediately started a release
> vote, bugs would be found and we’d cancel the RC and roll out another one.
> The goal of the release process we voted in, IMO, was to reduce this
> overhead of posting RCs, opening and closing vote threads, etc. so we can
> more efficiently achieve the goal of serving our customers and attracting
> some of them to becoming committers so we can have more people find bugs
> sooner by working with the develop branch.
> 
> Recently, I’ve spent several days on improving build and approval scripts
> so testing what is in development takes less time.  In theory, you can now
> start up the approval script which will pull down the bits, answer a few
> questions, then go do something else for 5 to 25 minutes and then come
> back and poke at it.  I would have rather spent that time on features for
> our customers, but I gambled that this would help us get the release out
> sooner.  I’m not sure that paid off.
> 
> 
> I don’t have any other ideas on how to make it easier for those of you who
> contribute in your spare time to stay up on the commits and bugs.  It
> should be ok to take any nightly build and run it through your tests and
> report your findings.  Ideally, you would be up to date on the commits and
> bugs and other discussions to know whether what you find is a duplicate or
> not, but at this point, I don’t care if you report a duplicate.  At least
> that means you verified a bunch of other code paths worked for you.  I
> don’t know how other Apache projects with really active code bases do it.
> 
> It is certainly fine to be too busy to vote on a release.  I was hoping to
> get more folks to poke at the bits before starting a vote because it will
> be a waste of community energy to start a vote and then have a bunch of
> PMC voters jump in and start reporting important bugs.  But I think the
> community has waited too long already, so I am going to start a vote soon,
> and Peter and I will vote and maybe Josh and/or Harbs and we’ll be good to
> go.  Hopefully any others jumping in late won’t find release blockers and
> we’ll just make another release later.
> 
> 
> -Alex