You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@flex.apache.org by Carlos Rovira <ca...@codeoscopic.com> on 2012/10/18 13:31:52 UTC
Is Apache model killing this project?
Hi,
I want to share some thoughts that I have since August... maybe
exposing here could make us debate over the problems and see what is
happening with this project. Maybe Apache Flex is going forward, but
my sense is it's not. Maybe committers could share what they see.
My overall feeling is that Flex is a complex project and contribute
need lots of things on board since you, as a contributor, need to be
comfortable and have time to invest.
The actual model in Apache is very old and is proving to be a failure
in actual days. The fact that we don't have a valid workflow based on
the voted option some months ago with Git, Github and nvie Git
branching model is one of the things that is making this project fail.
People is moving from Flex to HTML5 and this migration seems not to
stop although what they could get in that world is a real pain
compared with Flex development workflows. People staying are the teams
and developers that comes from IT sector and wants an OOP robust
framework and technology with lots of commodities like AMF,
Annotations and so on. Things that still exist in Flex nowadays.
Tickets voted in JIRA by community out there still are not closed
successfully (Maven deployment,...), maybe because actual committers
have another concept of what Flex should be and invest the time in
other things, and people that could enter the project to make that
happen couldn't enter since the Apache model makes it very hard (and
that people ends investing their time in other things that have more
possibilities).
In August my coworker (Pepe Barragan) and I was trying to contribute
to mustella, but the apache's workflow (and our own work) make us fall
in discouragement. Finally we are unwilling to participate. As we sure
there are more people. Which is not good for this project.
I'm afraid that or we make some movement and recognized that this
project is somewhat in bad state or we will end in Apache Flex 4.8
I'm following this list and things seems stuck since august.
Another point is that apache Flex installer was reject many times so
people ends passing, what again is very bad, and making developers
burn out to reach RC7...
The last problem I see is the heritage of adobe libraries and legal
issues added that make all things go very slow or even fail to get it
done. We are trying to save a great technology and if we are
discussing lots of legal licenses we soon will be with a dead horse
but with all documentation in order. If those legal problems (i.e:
maven related) does not cause a problem for Adobe until now...less now
than they want to quit from this technology. If things start to go ok,
legal problems can be solved in that moment.
So, after share all this issues... is Apache the best option for Flex?
What do you think about it?
Could we think other options to make this work in places like github
or bitbucket and the return results here?
Note: My intention is make us think a bit in what is happening. Maybe
I'm wrong and my perception is not shared here. This would be good and
will mean the project is really going on, so go forward! :). If that
is not the case, and you think things are not evolving as expected, is
the time to express yourself and say what you think. Only with this
kind of exercise we could know more about what is really happen to
Apache Flex.
Thanks in advance for your thought on this topic.
Best,
--
Carlos Rovira
Re: Is Apache model killing this project?
Posted by Erik de Bruin <er...@ixsoftware.nl>.
Carlos,
> The actual model in Apache is very old and is proving to be a failure
> in actual days. The fact that we don't have a valid workflow based on
> the voted option some months ago with Git, Github and nvie Git
> branching model is one of the things that is making this project fail.
Well, not having Git doesn't mean there isn't a valid workflow, on the
contrary. SVN is currently functioning fine, while volunteers like
yourself are helping to set up a valid Git workflow.
> Tickets voted in JIRA by community out there still are not closed
> successfully (Maven deployment,...), maybe because actual committers
> have another concept of what Flex should be and invest the time in
> other things, and people that could enter the project to make that
> happen couldn't enter since the Apache model makes it very hard (and
> that people ends investing their time in other things that have more
> possibilities).
Just to be clear, committers are NOT the only contributors to the
project. Everyone can/should help out. Find a JIRA ticket with a lot
of votes and create a patch to resolve it. Committers will notice and
if the patch checks out include it in the development branch of the
source.
> In August my coworker (Pepe Barragan) and I was trying to contribute
> to mustella, but the apache's workflow (and our own work) make us fall
> in discouragement. Finally we are unwilling to participate. As we sure
> there are more people. Which is not good for this project.
IIRC you were trying to work on this using the temporary Git solution,
which is indeed rather involved and prone to confusion and errors. I
do know your efforts were much appreciated and certainly didn't go
unnoticed. Maybe you would find the process a bit less discouraging if
you (for the time being, while waiting for an integrated Git solution)
worked in SVN?
> Another point is that apache Flex installer was reject many times so
> people ends passing, what again is very bad, and making developers
> burn out to reach RC7...
This I do know about ;-) I was the developer you're talking about, and
my "burn out" email was indeed born in frustration, but it was
intended (much like your email is, I assume) to get people to care
again... And they did. The vote is nearly over, and there are
apparently enough +1's to release the Installer. This means that
instead of remaining stagnant since August, the project is ready for
it's second release and moving forward!
> The last problem I see is the heritage of adobe libraries and legal
> issues added that make all things go very slow or even fail to get it
> done. We are trying to save a great technology and if we are
> discussing lots of legal licenses we soon will be with a dead horse
> but with all documentation in order. If those legal problems (i.e:
> maven related) does not cause a problem for Adobe until now...less now
> than they want to quit from this technology. If things start to go ok,
> legal problems can be solved in that moment.
These issues are frustrating (7! release candidates for the
Installer), but needed. They are also a one time issues, once a valid
solution has been found/agreed upon, we don't have to worry about them
anymore and can focus on moving forward.
> So, after share all this issues... is Apache the best option for Flex?
> What do you think about it?
> Could we think other options to make this work in places like github
> or bitbucket and the return results here?
While you're always free to fork the project (I think?), moving to a
different type of version control won't make the legal (license etc.)
issues go away. I think an organisation like Apache (which provides
much more to the project than only version control) is exactly the
right place for a code base like this to grow and thrive.
Yes, it takes effort to get a like this project underway, and it is
very frustrating at times, but it is very much worth it. Please don't
give up on the project, we're only just getting started!
EdB
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: Is Apache model killing this project?
Posted by Omar Gonzalez <om...@gmail.com>.
On Friday, October 19, 2012, Greg Reddin wrote:
>
> > How much progress do we have to continue to make in order to appease your
> > worries?
>
> Sorry, I'm not the one with worries. This thread started with someone
> suggesting that the Apache model was inhibiting progress for this
> project. I don't think that's the case at all so I was trying to
> respond to that particular suggestion.
>
> Greg
>
I only brought that stuff up cause you said "I believe some of that fear
has verified. My belief was *and **still is* that we'd be better off
getting stuff done in Subversion", I was disputing that statement as I
believe lots of progress has continued to be made using SVN while we get
Git transitioned.
-omar
Re: Is Apache model killing this project?
Posted by Greg Reddin <gr...@gmail.com>.
On Fri, Oct 19, 2012 at 11:19 AM, Omar Gonzalez
<om...@gmail.com> wrote:
> I disagree entirely. No progress has stopped during this "transitional
> phase". People are still committing to SVN. People are using an updated
> branching strategy. And we are still actively trying to find a person
> proficient in Python to take us that last mile. I don't believe any of your
> fears have been verified, otherwise we wouldn't have gotten things done
> like Peter's (and other non-committer's) work on Mustella and getting
> 35000+ tests to pass.
Thanks for pointing out that there are more things going on than I've
been keeping up with :-) Sorry for implying that progress had stopped.
I was responding to Carlos' suggestion that the drawn-out move to Git
was a progress-impeder. And I was responding by pointing out that we
said it probably would be a progress impeder.
> How much progress do we have to continue to make in order to appease your
> worries?
Sorry, I'm not the one with worries. This thread started with someone
suggesting that the Apache model was inhibiting progress for this
project. I don't think that's the case at all so I was trying to
respond to that particular suggestion.
Greg
Re: Is Apache model killing this project?
Posted by Omar Gonzalez <om...@gmail.com>.
> Now, to your actual point about Git, etc. It's worth noting that all
> the mentors cautioned against that move precisely because it would be
> difficult and time-consuming to get it implemented. Our fear was that
> it would further stall the progress of the project because we'd be
> focusing on SCM and branching strategies rather than working on the
> code. I believe some of that fear has verified.
I disagree entirely. No progress has stopped during this "transitional
phase". People are still committing to SVN. People are using an updated
branching strategy. And we are still actively trying to find a person
proficient in Python to take us that last mile. I don't believe any of your
fears have been verified, otherwise we wouldn't have gotten things done
like Peter's (and other non-committer's) work on Mustella and getting
35000+ tests to pass.
> My belief was and
> still is that we'd be better off getting stuff done in Subversion
> right now than trying to figure out how make Git work for us. I'm
> doing that exact very thing at work right now and we are making
> progress on our project.
>
> There's more but it will have to wait. I'm out of time.
>
> Greg
>
I've moved a couple of teams from SVN to Git, its not that big a deal. Its
more griping than technical problems. Anyhow, nobody has ever stopped
getting things done in Subversion so I don't understand your apprehension
toward Git. The decision we made has had very little impact on forward
progress, from my perspective. The Apache Flex Installer Om/Erik are about
to release, on RC7 now, all done in Subversion. The 35000+ Mustella tests,
all updated in Subversion, ...using our "new" branching strategy.
How much progress do we have to continue to make in order to appease your
worries?
-omar
Re: Is Apache model killing this project?
Posted by Greg Reddin <gr...@gmail.com>.
I apologize if my points below are invalidated by other responses. I
haven't had time to read the whole thread. At any rate...
On Thu, Oct 18, 2012 at 6:31 AM, Carlos Rovira
<ca...@codeoscopic.com> wrote:
> The actual model in Apache is very old and is proving to be a failure
> in actual days. The fact that we don't have a valid workflow based on
> the voted option some months ago with Git, Github and nvie Git
> branching model is one of the things that is making this project fail.
I will take issue with that entire paragraph. The statement about the
apache model is just FUD until you give evidence to support how it is
a failure in today's world. There are actually many projects that are
thriving in the Apache model and the stability of that model is one
thing that attracts organizations who are looking for an open source
home to host projects they want to donate.
It's true that the apache model has been around for quite a while and
probably needs to evolve for the new decade it is living in. But
evolving that model into the current environment is difficult without
breaking some of the things that are extremely important conceptually
to that model. It's a long, frustrating, arduous process, but for
those of us who think there's value in the model as well as the
projects it supports, it's worth the effort.
Now, to your actual point about Git, etc. It's worth noting that all
the mentors cautioned against that move precisely because it would be
difficult and time-consuming to get it implemented. Our fear was that
it would further stall the progress of the project because we'd be
focusing on SCM and branching strategies rather than working on the
code. I believe some of that fear has verified. My belief was and
still is that we'd be better off getting stuff done in Subversion
right now than trying to figure out how make Git work for us. I'm
doing that exact very thing at work right now and we are making
progress on our project.
There's more but it will have to wait. I'm out of time.
Greg
Re: Is Apache model killing this project?
Posted by Dasa Paddock <dp...@esri.com>.
The "develop" branch is getting updates at:
https://github.com/apache/flex-sdk/tree/develop
On Oct 18, 2012, at 2:52 PM, Carlos Rovira <ca...@codeoscopic.com> wrote:
> AFAIK, apache flex git mirror was not updating from svn. It was stuck
> in some older commit... Do you know if that was fixed?
>
> 2012/10/18 Erik de Bruin <er...@ixsoftware.nl>:
>>> Yes, I think we all could be working together more easy in github and
>>> from time to time upgrade SVN.
>>
>> Well, if you feel so strongly about this, want to contribute and don't
>> see any way to work in SVN while Apache Git is being set up, why not
>> work in github and from time to time submit patches to JIRA?
>>
>> EdB
>>
>>
>>
>> --
>> Ix Multimedia Software
>>
>> Jan Luykenstraat 27
>> 3521 VB Utrecht
>>
>> T. 06-51952295
>> I. www.ixsoftware.nl
>
>
>
> --
> Carlos Rovira
> Director de Tecnología
> M: +34 607 22 60 05
> F: +34 912 35 57 77
> http://www.codeoscopic.com
> http://www.directwriter.es
> http://www.avant2.es
>
Re: Is Apache model killing this project?
Posted by Carlos Rovira <ca...@codeoscopic.com>.
AFAIK, apache flex git mirror was not updating from svn. It was stuck
in some older commit... Do you know if that was fixed?
2012/10/18 Erik de Bruin <er...@ixsoftware.nl>:
>> Yes, I think we all could be working together more easy in github and
>> from time to time upgrade SVN.
>
> Well, if you feel so strongly about this, want to contribute and don't
> see any way to work in SVN while Apache Git is being set up, why not
> work in github and from time to time submit patches to JIRA?
>
> EdB
>
>
>
> --
> Ix Multimedia Software
>
> Jan Luykenstraat 27
> 3521 VB Utrecht
>
> T. 06-51952295
> I. www.ixsoftware.nl
--
Carlos Rovira
Director de Tecnología
M: +34 607 22 60 05
F: +34 912 35 57 77
http://www.codeoscopic.com
http://www.directwriter.es
http://www.avant2.es
Re: Is Apache model killing this project?
Posted by Bertrand Delacretaz <bd...@apache.org>.
On Thu, Oct 18, 2012 at 6:40 PM, Erik de Bruin <er...@ixsoftware.nl> wrote:
>> Yes, I think we all could be working together more easy in github and
>> from time to time upgrade SVN.
>
> Well, if you feel so strongly about this, want to contribute and don't
> see any way to work in SVN while Apache Git is being set up, why not
> work in github and from time to time submit patches to JIRA?...
+1, if someone's eager to write code that makes Flex better, just fork
it somewhere and improve it.
Once that's done, show your code to this community, and if it's
actually better it shouldn't be hard to convince them to make you a
committer and accept your contributions.
-Bertrand
Re: Is Apache model killing this project?
Posted by Erik de Bruin <er...@ixsoftware.nl>.
> Yes, I think we all could be working together more easy in github and
> from time to time upgrade SVN.
Well, if you feel so strongly about this, want to contribute and don't
see any way to work in SVN while Apache Git is being set up, why not
work in github and from time to time submit patches to JIRA?
EdB
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: Is Apache model killing this project?
Posted by Stefan Horochovec <st...@gmail.com>.
Hello
About code contribuition:
I don´t believe the GIT and Maven will solve our problems and neither will
facilitate the contribution of people in the project.
We can contribute through SVN + JIRA (Diffs) without any problems.
About Apache Model
All we are adapting to the Apache model development, I believe that this
"delay" in results is normal in this first releases.
Once the entire process is clear to all who are contributing to the project
more actively these initial difficulties (such as Apache Flex Installer)
will not occur and the process will become faster
Regards
Stefan Horochovec
2012/10/18 Carlos Rovira <ca...@codeoscopic.com>
> Yes, I think we all could be working together more easy in github and
> from time to time upgrade SVN.
>
> 2012/10/18 <ni...@perfectedz.com>:
> > A small notice. I think what Carlos means when he mentions GitHub is not
> the
> > version system (git) but actually moving the project from apache to
> github.
> > Carlos correct me if I misunderstood.
> >
>
>
>
> --
> Carlos Rovira
> Director de Tecnología
> M: +34 607 22 60 05
> F: +34 912 35 57 77
> CODEOSCOPIC S.A.
> Avd. del General Perón, 32
> Planta 10, Puertas P-Q
> 28020 Madrid
>
Re: Is Apache model killing this project?
Posted by Carlos Rovira <ca...@codeoscopic.com>.
Yes, I think we all could be working together more easy in github and
from time to time upgrade SVN.
2012/10/18 <ni...@perfectedz.com>:
> A small notice. I think what Carlos means when he mentions GitHub is not the
> version system (git) but actually moving the project from apache to github.
> Carlos correct me if I misunderstood.
>
--
Carlos Rovira
Director de Tecnología
M: +34 607 22 60 05
F: +34 912 35 57 77
CODEOSCOPIC S.A.
Avd. del General Perón, 32
Planta 10, Puertas P-Q
28020 Madrid
Re: Is Apache model killing this project?
Posted by ni...@perfectedz.com.
A small notice. I think what Carlos means when he mentions GitHub is
not the version system (git) but actually moving the project from apache
to github. Carlos correct me if I misunderstood.
Other than that I think more conversations like these are vital. They
provide more clarity and show that there are people that care for the
project.
Re: Is Apache model killing this project?
Posted by Carol Frampton <cf...@adobe.com>.
On 10/18/12 11 :02AM, "Carlos Rovira" <ca...@codeoscopic.com>
wrote:
>
>Right now I don't know what people are doing...What is doing Alex?
>Carol? Peter? Gordon? Michael? Om? Omar? You?
If you watch the flex-commits list it is pretty clear what many of those
people have been doing recently. I realize it does not tell you what we
plan to do.
Carol
>
Re: Is Apache model killing this project?
Posted by Omar Gonzalez <om...@gmail.com>.
On Thu, Oct 18, 2012 at 8:41 AM, Erik de Bruin <er...@ixsoftware.nl> wrote:
> > Talking about the sensation of "Ghost Town": I think it could be
> > removed if people investing time could talk more about what is doing,
> > contributing to the overall feeling that the project is moving
> > forward.
>
> Well, all I can say is that besides the list, where most people talk
> about what they are doing or planning, you can look at the commit logs
> in the SVN repository and see exactly who's contributing what. Also
> JIRA is a good resource to keep tabs on the current state of things.
>
> Nobody's making work lists or things like that, but if you're willing
> to contribute something like that so others with your reservations can
> see what's going on, that would be great!
>
> Thanks,
>
> EdB
>
>
>
> --
> Ix Multimedia Software
>
> Jan Luykenstraat 27
> 3521 VB Utrecht
>
> T. 06-51952295
> I. www.ixsoftware.nl
>
Or you can subscribe to flex-commits@incubator.apache.org
-omar
Re: Is Apache model killing this project?
Posted by Erik de Bruin <er...@ixsoftware.nl>.
> Talking about the sensation of "Ghost Town": I think it could be
> removed if people investing time could talk more about what is doing,
> contributing to the overall feeling that the project is moving
> forward.
Well, all I can say is that besides the list, where most people talk
about what they are doing or planning, you can look at the commit logs
in the SVN repository and see exactly who's contributing what. Also
JIRA is a good resource to keep tabs on the current state of things.
Nobody's making work lists or things like that, but if you're willing
to contribute something like that so others with your reservations can
see what's going on, that would be great!
Thanks,
EdB
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: Is Apache model killing this project?
Posted by Omar Gonzalez <om...@gmail.com>.
On Thu, Oct 18, 2012 at 8:02 AM, Carlos Rovira <
carlos.rovira@codeoscopic.com> wrote:
> Erik,
>
> in the first point I think I was misunderstood. What I can see is that
> I can't help in INFRA things due to the fact that Apache organization
> is not prepared and I can't access the level to put work there. I only
> can contribute with patches to the flex project code (as I did). But
> as months pass my vision is that Apache is not interested in support
> Git nor Github and individuals can't make it happen without a level of
> commitment that would be prohibitive. You would need to dedicate your
> life to make this happen and even in that situation you can't ensure
> that would happen in the end...
>
This is a little bit overly dramatic. Apache is interested in Git, this is
why they are asking for volunteers from projects that wish to move to it
full time. The volunteer efforts do not involve you to "dedicate your life"
to make it happen. All they're asking for is someone with knowledge of
Python to help fix JIRA issues in their already existing Git admin written
in Python. I would have volunteered myself, but I am not proficient in
Python.
>
> Talking about the sensation of "Ghost Town": I think it could be
> removed if people investing time could talk more about what is doing,
> contributing to the overall feeling that the project is moving
> forward.
>
> This could make people (knowing what things are rolling) trying to
> contribute to that efforts.
>
> Right now I don't know what people are doing...What is doing Alex?
> Carol? Peter? Gordon? Michael? Om? Omar? You? and so on.... if people
> could see some page where actual efforts were listed, people could
> adhere to the cause...but what is currently needed? what plans are
> driven actualy? plans that really have opportunities to reach a decent
> level of completion? There's so much missinformation (I say this from
> the point of view of someone that is following the project so close).
> i.e. I recently know that almost 100% of mustella are working, there
> was no announcement here...or maybe I miss that mail
>
>From following the threads there are two things that seem to me like they
are the things that are the main focus.
1.) JIRA tickets.
2.) Mustella tests
The problem with #1 is that we can not verify that changes people make
don't break the framework because all the Mustella tests don't work.
Therefore, Alex and Peter have been concentrating their efforts on making
all the Mustella tests pass, which I forget which of you were also helping
with. This step is vital to being able to move forward with both JIRA fixes
and new development efforts.
I didn't need a webpage to tell me that, I just follow the threads.
As far as I can remember regarding Gordon, he is only working on the
project 1 day out of the week, concentrating on Falcon.
Personally, I've been busy with a big project at work and getting ready for
my wedding next week. I expect my participation in the project to increase
once I'm back from my honeymoon. At that point I'll be able to start
learning me some Python and get this Git shit over the hump if no other
volunteers come up by then. We aren't far off from moving it to Git
read/write repo, we just need to come up with that support guy for Apache
INFRA.
Omar Gonzalez
s9tpepper@apache.org
Apache Flex PPMC
Re: Is Apache model killing this project?
Posted by Carlos Rovira <ca...@codeoscopic.com>.
Erik,
in the first point I think I was misunderstood. What I can see is that
I can't help in INFRA things due to the fact that Apache organization
is not prepared and I can't access the level to put work there. I only
can contribute with patches to the flex project code (as I did). But
as months pass my vision is that Apache is not interested in support
Git nor Github and individuals can't make it happen without a level of
commitment that would be prohibitive. You would need to dedicate your
life to make this happen and even in that situation you can't ensure
that would happen in the end...
Talking about the sensation of "Ghost Town": I think it could be
removed if people investing time could talk more about what is doing,
contributing to the overall feeling that the project is moving
forward.
This could make people (knowing what things are rolling) trying to
contribute to that efforts.
Right now I don't know what people are doing...What is doing Alex?
Carol? Peter? Gordon? Michael? Om? Omar? You? and so on.... if people
could see some page where actual efforts were listed, people could
adhere to the cause...but what is currently needed? what plans are
driven actualy? plans that really have opportunities to reach a decent
level of completion? There's so much missinformation (I say this from
the point of view of someone that is following the project so close).
i.e. I recently know that almost 100% of mustella are working, there
was no announcement here...or maybe I miss that mail
2012/10/18 Erik de Bruin <er...@ixsoftware.nl>:
> Carlos,
>
> You make some valid points, but also some that come from frustration, not fact:
>
>> * Apache is prepared only to accept patches. People like me can't help
>> nothing with infrastructure or GIT things until we don't reach the
>> committer status.
>
> Why can't you help? Sure, you can't help on your own until you become
> a committer, but I'm sure you can find a committer to work with
> getting your ideas implemented. I for one would certainly be willing
> to help out. As an example: when I started working on the Installer, I
> could only submit patches as well. So I started working with a
> committer (Om) and after some startup issues, we got a lot of work
> done together. And guess what? The PPMC noticed and decided to make me
> a committer as well... There is a lesson in there somewhere ;-)
>
>> * This project does not work like others because is half million files
>> with lots of knowledge to know to get to be productive. Other projects
>> barrier is lower since it only requieres a know a technology
>> (i.e:java) and the goals are more moderated.
>
> Small steps, take one JIRA ticket and work from there. It's
> unrealistic to expect you (anyone!) can come to a codebase as large as
> this one and expect to implement structural changes right away. It is
> also unrealistic to expect a project that is just underway to
> immediately start action of that magnitude. First of all, the basis
> has to be solid, hence the work on Mustella before anything is done
> to the framework itself. You were helping with that, which was very
> much appreciated!
>
>> * Here we have frameworks, compilers, and so on...without organization
>> is impossible (is what Michael states that he alone can't do nothing).
>> While apache way is for individuals to contribute individual things.
>
> Individual can fix JIRA tickets, supply patches. Individuals working
> together can get bigger stuff done (Installer) and I'm sure groups
> working together can get all the major goals done. But it starts with
> individuals actively contributing and doing (some of) the grunt work.
>
>> * I'm with Michael that this is a ghost town since very few committers
>> from the initial setup are active. Others although love flex are in
>> actual wars with HTML5, so we shouldn't expect to see it come back.
>
> I disagree. Do you have number to support this? My gut feeling is that
> there are more people actively contributing code/patches now then at
> the start of the project.
>
>> * Regarding SVN. We were working with SVN (not GIT) while submitting
>> patches to mustella. Only we used github to share the patches/pull
>> request. But something that could be more agile was a pain and make us
>> to throw the towel....that is very - very - very - bad...can't express
>> what kind of damage make to this project such procedure...the real
>> thing is that we talked to come back as Git and Github was setup...but
>> taking into account that Apache infra needs some python guy that's a
>> dead lock.
>
> JIRA is much more suited to take SVN patches and keeps the process
> documented at the same time. Once the Git infrastructure is in place,
> that may change, I'm not familiar with the Git workflow yet. But for
> now, work with what works and contribute to the development of the
> code. The rest is tools and tools are interchangeable. Choose the tool
> that works and get work done ;-)
>
>> Erik, I'm a bit of frustrating, but in the end I'm solving my problems
>> at my company level, so can't live with this situation. But is a shame
>> that the apache community and the apache way is making this project a
>> ghost town (IMHO), while if we work in a for in github this could be
>> really going up.
>
> I disagree with (and feel a little hurt by) the ghost town remarks.
> Using a different version control system will not change anything if
> people are not willing to contribute. And if you're willing to
> contribute, having to use a different version control system shouldn't
> be a hurdle. We're all developers, aren't we, used to working with a
> variety of tools?
>
>> One thing is for sure...or this trend change, or flex will be dead in
>> few months.
>
> Well, I would love it if you would be the one to break this trend you
> see, and re-start your contributions to Mustella (or any aspect of the
> project). As I said, I would love to work with you to make the process
> of contributing your patches as smooth as possible.
>
> EdB
>
>
>
> --
> Ix Multimedia Software
>
> Jan Luykenstraat 27
> 3521 VB Utrecht
>
> T. 06-51952295
> I. www.ixsoftware.nl
--
Carlos Rovira
Director de Tecnología
M: +34 607 22 60 05
F: +34 912 35 57 77
CODEOSCOPIC S.A.
Avd. del General Perón, 32
Planta 10, Puertas P-Q
28020 Madrid
Re: Is Apache model killing this project?
Posted by Carlos Rovira <ca...@codeoscopic.com>.
Michael,
what about to document in the wiki concrete efforts for the compiler?
This would be your vision, and people could be interested in your
proposal and end making team with you...
2012/10/18 Michael Schmalle <ap...@teotigraphix.com>:
> Erik,
>
> I want to clarify what *I* meant by ghost town.
>
> We are all on a team, in the circle of life we all have roles we play and
> positions we excel at. The ghost town I am speaking of is the push for
> forward progress at the same time Om and yourself are making the current SDK
> available as an Apache release.
>
> Honestly, I cannot do what you or Om are doing, I know nothing about keys,
> release protocols, the mustella etc. I know about components, compilers and
> documenters.
>
> I have tried creating a circle so I can feel like there are a couple others
> thinking about "what is next" for the project beyond what you, Om and others
> are working hard on at the moment.
>
> The fact is, to move forward with the compiler, there needs to be a circle
> created where more than just casual observation is happening.
>
> For me, this is the first step of contribution I am offering. I feel I have
> offered a lot to this project already and want to continue. I am just laying
> out some of my opinions that I feel are prerequisites to start this huge
> migration that is inevitable to compete with new technologies.
>
> So let me repeat, there is a lot I cannot do on this project, there is a
> little that I could be really good at.
>
>
> Anyway, I probably sound self centered, but I really am. I also know what it
> takes for innovation, communication is the only way mankind has evolved.
>
> And yes, this has been a good conversation... I only wish it was about the
> compiler API and unit tests. ;-)
>
>
> Mike
>
>
> --
> Michael Schmalle - Teoti Graphix, LLC
> http://www.teotigraphix.com
> http://blog.teotigraphix.com
>
--
Carlos Rovira
Director de Tecnología
M: +34 607 22 60 05
F: +34 912 35 57 77
CODEOSCOPIC S.A.
Avd. del General Perón, 32
Planta 10, Puertas P-Q
28020 Madrid
Re: Is Apache model killing this project?
Posted by Erik de Bruin <er...@ixsoftware.nl>.
> And yes, this has been a good conversation... I only wish it was about the
> compiler API and unit tests. ;-)
Well, here's a thought about unit tests(just to round off this
conversation): apart from establishing the framework, it is my
understanding you write the actual tests one at a time... That sounds
like something someone like yourself would be able to get going.
I really hope you will reconsider your contribution, I honestly think
that once someone (yes, it takes ONE) gets the ball rolling, others
are sure to follow.
EdB
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: Is Apache model killing this project?
Posted by Michael Schmalle <ap...@teotigraphix.com>.
Erik,
I want to clarify what *I* meant by ghost town.
We are all on a team, in the circle of life we all have roles we play
and positions we excel at. The ghost town I am speaking of is the push
for forward progress at the same time Om and yourself are making the
current SDK available as an Apache release.
Honestly, I cannot do what you or Om are doing, I know nothing about
keys, release protocols, the mustella etc. I know about components,
compilers and documenters.
I have tried creating a circle so I can feel like there are a couple
others thinking about "what is next" for the project beyond what you,
Om and others are working hard on at the moment.
The fact is, to move forward with the compiler, there needs to be a
circle created where more than just casual observation is happening.
For me, this is the first step of contribution I am offering. I feel I
have offered a lot to this project already and want to continue. I am
just laying out some of my opinions that I feel are prerequisites to
start this huge migration that is inevitable to compete with new
technologies.
So let me repeat, there is a lot I cannot do on this project, there is
a little that I could be really good at.
Anyway, I probably sound self centered, but I really am. I also know
what it takes for innovation, communication is the only way mankind
has evolved.
And yes, this has been a good conversation... I only wish it was about
the compiler API and unit tests. ;-)
Mike
--
Michael Schmalle - Teoti Graphix, LLC
http://www.teotigraphix.com
http://blog.teotigraphix.com
Re: Is Apache model killing this project?
Posted by Erik de Bruin <er...@ixsoftware.nl>.
Carlos,
You make some valid points, but also some that come from frustration, not fact:
> * Apache is prepared only to accept patches. People like me can't help
> nothing with infrastructure or GIT things until we don't reach the
> committer status.
Why can't you help? Sure, you can't help on your own until you become
a committer, but I'm sure you can find a committer to work with
getting your ideas implemented. I for one would certainly be willing
to help out. As an example: when I started working on the Installer, I
could only submit patches as well. So I started working with a
committer (Om) and after some startup issues, we got a lot of work
done together. And guess what? The PPMC noticed and decided to make me
a committer as well... There is a lesson in there somewhere ;-)
> * This project does not work like others because is half million files
> with lots of knowledge to know to get to be productive. Other projects
> barrier is lower since it only requieres a know a technology
> (i.e:java) and the goals are more moderated.
Small steps, take one JIRA ticket and work from there. It's
unrealistic to expect you (anyone!) can come to a codebase as large as
this one and expect to implement structural changes right away. It is
also unrealistic to expect a project that is just underway to
immediately start action of that magnitude. First of all, the basis
has to be solid, hence the work on Mustella before anything is done
to the framework itself. You were helping with that, which was very
much appreciated!
> * Here we have frameworks, compilers, and so on...without organization
> is impossible (is what Michael states that he alone can't do nothing).
> While apache way is for individuals to contribute individual things.
Individual can fix JIRA tickets, supply patches. Individuals working
together can get bigger stuff done (Installer) and I'm sure groups
working together can get all the major goals done. But it starts with
individuals actively contributing and doing (some of) the grunt work.
> * I'm with Michael that this is a ghost town since very few committers
> from the initial setup are active. Others although love flex are in
> actual wars with HTML5, so we shouldn't expect to see it come back.
I disagree. Do you have number to support this? My gut feeling is that
there are more people actively contributing code/patches now then at
the start of the project.
> * Regarding SVN. We were working with SVN (not GIT) while submitting
> patches to mustella. Only we used github to share the patches/pull
> request. But something that could be more agile was a pain and make us
> to throw the towel....that is very - very - very - bad...can't express
> what kind of damage make to this project such procedure...the real
> thing is that we talked to come back as Git and Github was setup...but
> taking into account that Apache infra needs some python guy that's a
> dead lock.
JIRA is much more suited to take SVN patches and keeps the process
documented at the same time. Once the Git infrastructure is in place,
that may change, I'm not familiar with the Git workflow yet. But for
now, work with what works and contribute to the development of the
code. The rest is tools and tools are interchangeable. Choose the tool
that works and get work done ;-)
> Erik, I'm a bit of frustrating, but in the end I'm solving my problems
> at my company level, so can't live with this situation. But is a shame
> that the apache community and the apache way is making this project a
> ghost town (IMHO), while if we work in a for in github this could be
> really going up.
I disagree with (and feel a little hurt by) the ghost town remarks.
Using a different version control system will not change anything if
people are not willing to contribute. And if you're willing to
contribute, having to use a different version control system shouldn't
be a hurdle. We're all developers, aren't we, used to working with a
variety of tools?
> One thing is for sure...or this trend change, or flex will be dead in
> few months.
Well, I would love it if you would be the one to break this trend you
see, and re-start your contributions to Mustella (or any aspect of the
project). As I said, I would love to work with you to make the process
of contributing your patches as smooth as possible.
EdB
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: Is Apache model killing this project?
Posted by Carlos Rovira <ca...@codeoscopic.com>.
Some points to take into account:
* Apache is prepared only to accept patches. People like me can't help
nothing with infrastructure or GIT things until we don't reach the
committer status.
* This project does not work like others because is half million files
with lots of knowledge to know to get to be productive. Other projects
barrier is lower since it only requieres a know a technology
(i.e:java) and the goals are more moderated.
* Here we have frameworks, compilers, and so on...without organization
is impossible (is what Michael states that he alone can't do nothing).
While apache way is for individuals to contribute individual things.
* I'm with Michael that this is a ghost town since very few committers
from the initial setup are active. Others although love flex are in
actual wars with HTML5, so we shouldn't expect to see it come back.
* Regarding SVN. We were working with SVN (not GIT) while submitting
patches to mustella. Only we used github to share the patches/pull
request. But something that could be more agile was a pain and make us
to throw the towel....that is very - very - very - bad...can't express
what kind of damage make to this project such procedure...the real
thing is that we talked to come back as Git and Github was setup...but
taking into account that Apache infra needs some python guy that's a
dead lock.
Erik, I'm a bit of frustrating, but in the end I'm solving my problems
at my company level, so can't live with this situation. But is a shame
that the apache community and the apache way is making this project a
ghost town (IMHO), while if we work in a for in github this could be
really going up.
One thing is for sure...or this trend change, or flex will be dead in
few months.
2012/10/18 Michael Schmalle <ap...@teotigraphix.com>:
> Erik,
>
> Please man, I mean no disrespect in any of these emails to you it's a
> frustration I have been holding onto for about 10 months now.
>
> The github reference was in relation to supporting a small, manageable open
> source project in a github account. The point was a lot of those projects
> are small and manageable.
>
> The flex project is huge and all interdependent. Yes, Apache says no road
> maps but common, active contributors need to have conversations and create
> dialog about future implementations. Maybe not when creating higherlevel
> things but when we are talking about a compiler, I'm not just stepping into
> the wild unkown without a "little" backup here, meaning a couple other
> enthusiastic developers at my side. IE I'm not talking about making a
> TabNavigator implementation in my whiteboard here.
>
> So to answer, I have no knowledge of how to get GIT running on Apache.
>
> Mike
>
>
>
> Quoting Erik de Bruin <er...@ixsoftware.nl>:
>
>>> Supporting some github project is definitely where I could "if you don't
>>> like it, change it."
>>
>>
>> Well, that is a good place to start! Apache Infra has indicated that
>> they would help the project move to Git if the project 'donated' a few
>> volunteers to help with the move and to support the new workflow after
>> that. I'm sure the project would much appreciate it if you could be
>> one of those volunteers.
>>
>> Thank you,
>>
>> EdB
>>
>>
>>
>> --
>> Ix Multimedia Software
>>
>> Jan Luykenstraat 27
>> 3521 VB Utrecht
>>
>> T. 06-51952295
>> I. www.ixsoftware.nl
>>
>
> --
> Michael Schmalle - Teoti Graphix, LLC
> http://www.teotigraphix.com
> http://blog.teotigraphix.com
>
--
Carlos Rovira
Director de Tecnología
M: +34 607 22 60 05
F: +34 912 35 57 77
CODEOSCOPIC S.A.
Avd. del General Perón, 32
Planta 10, Puertas P-Q
28020 Madrid
Re: Is Apache model killing this project?
Posted by Martin Heidegger <mh...@leichtgewicht.at>.
Hello Eric, Carlos and list,
Personal and technological problems have kept me from being active an i think this is a good point to share my thoughts and experiences after a half a year off this list:
I work at a company now that still has a reasonable amount of "Flash" work left. However I have hardly touched a line of As 3 in the past months. And I don't regret it. Yes: browsers and html are messy(still, a lot) and having to know things like Date.now() breaks on IE is annoying but there are so many things that are just nice in the html world that I don't want to switch back to Fl(ex/ash).
1) I love the command line and there is none easily combined with flex. I would need to run(red) tamarin or some other- horribly complex tool to get to a decent cmd line. Nodejs handles compilation well.
2) happy language hacking. JavaScript is awesome: why? You ask. Well: because I don't need to write it. I write in CoffeeScript, Dart, Haxe etc. they contain a lot of interesting language design choices that animate me to try things out.
3) deployment. I started with ant, struggled with maven, learned with gradle and mastered with build scripts and have decided: cake rocks. I have never been able to write such detailed, easy to setup scripts as I gave with CoffeeScript and nodejs. What a difference to how damn unnecessary hard it is to deploy with flex...
4) I don't want to write unit tests for software others wrote. I love to see tests run. It's satisfying. What is not satisfying is writing tests for other persons code just to makes sure they do what I think they do. The (not existing) testing system of flex drove me to the leave. Having mustella doesn't make me want to come back.
5) Aim. What do I need flex for? The flash apps developed at this company will not get better with flex, everyone uses their own libraries and flex just doesn't play we'll together with other kids on the block. I can't use a part without feeling its all unnecessarily intertwined. When I developed my first open source library years ago it was already a buck load of work to separate it properly and it had a fraction of the size of what flex is now. The reason I point at this is because once its separated I could easily change focus or intent of the parts and was able to talk about them independently. I could plan release in small cycles or change the direction of the modules easily. Today Aim of flex seems too much to preserve what's here in contrast to make the world awesome. I want make the world awesome... I don't think I can do that by supporting flex.
To me open source is about things I can be passionate about. I can't be passionate about flex anymore. It's slow, uncomfortable and fixed on flash at the moment.
Yours
Martin Heidegger
Re: Is Apache model killing this project?
Posted by Erik de Bruin <er...@ixsoftware.nl>.
> I understand this as well but, if I cannot at least get a decent
> conversation from the project's dev list, why even bother spending time on a
> wiki or a website that these devs probably are not even viewing?
I think this is a pretty decent conversation ;-)
EdB
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: Is Apache model killing this project?
Posted by Michael Schmalle <ap...@teotigraphix.com>.
Right, I have a thick head, thus some of my comments.
Quoting Erik de Bruin <er...@ixsoftware.nl>:
> Mike,
>
> I'm taking none of this personal, I'm Dutch and have a very thick
> skin, so don't worry about that :-)
>
> The only thing I'm trying to achieve with my comments is to get you to
> contribute something to the project, besides constructive criticism
> and excellent ideas. You have to offer so much, why not share a
> little? As you correctly assert, no one can complete Falcon on their
> own, not even Gordon (well, given enough time... ;-)), so, like a long
> journey, maybe start with one small step?
Yes, I understand this unfortunately I am at a turning point here. I
am like any other business man that has to weigh the options of my
next step and decisions.
I have tried to start some dialog in the past couple of weeks about
documentation, the compiler etc. As Roland stated, the only feedback I
got was from him and we have talked about this stuff personally, so he
doesn't count. :)
I already risked my business on Adobe last year, that failed. I had to
pick up the pieces of last November and reinvent myself. Choosing to
commit myself to this project means seeing or in my case feeling
support, I'm not going to start something until I can get a decent
dialog with other developers. I don't care what Apache says about do
it yourself, if conversations cannot even be started on a dev list,
it's hopeless form my view and I'm not going to spend my free time
committing code to a project that I cannot even get decent feedback
from a dev list.
> If you think that there needs to be discussion about the direction of
> the project and it the goals the community would like to work towards,
> please start one. You can use this list, but better would be to use
> the Wiki or the website itself, so there is a better record of the
> arguments.
I understand this as well but, if I cannot at least get a decent
conversation from the project's dev list, why even bother spending
time on a wiki or a website that these devs probably are not even
viewing?
Mike
> EdB
>
>
>
> On Thu, Oct 18, 2012 at 2:54 PM, Michael Schmalle
> <ap...@teotigraphix.com> wrote:
>> Erik,
>>
>> Please man, I mean no disrespect in any of these emails to you it's a
>> frustration I have been holding onto for about 10 months now.
>>
>> The github reference was in relation to supporting a small, manageable open
>> source project in a github account. The point was a lot of those projects
>> are small and manageable.
>>
>> The flex project is huge and all interdependent. Yes, Apache says no road
>> maps but common, active contributors need to have conversations and create
>> dialog about future implementations. Maybe not when creating higherlevel
>> things but when we are talking about a compiler, I'm not just stepping into
>> the wild unkown without a "little" backup here, meaning a couple other
>> enthusiastic developers at my side. IE I'm not talking about making a
>> TabNavigator implementation in my whiteboard here.
>>
>> So to answer, I have no knowledge of how to get GIT running on Apache.
>>
>> Mike
>>
>>
>>
>> Quoting Erik de Bruin <er...@ixsoftware.nl>:
>>
>>>> Supporting some github project is definitely where I could "if you don't
>>>> like it, change it."
>>>
>>>
>>> Well, that is a good place to start! Apache Infra has indicated that
>>> they would help the project move to Git if the project 'donated' a few
>>> volunteers to help with the move and to support the new workflow after
>>> that. I'm sure the project would much appreciate it if you could be
>>> one of those volunteers.
>>>
>>> Thank you,
>>>
>>> EdB
>>>
>>>
>>>
>>> --
>>> Ix Multimedia Software
>>>
>>> Jan Luykenstraat 27
>>> 3521 VB Utrecht
>>>
>>> T. 06-51952295
>>> I. www.ixsoftware.nl
>>>
>>
>> --
>> Michael Schmalle - Teoti Graphix, LLC
>> http://www.teotigraphix.com
>> http://blog.teotigraphix.com
>>
>
>
>
> --
> Ix Multimedia Software
>
> Jan Luykenstraat 27
> 3521 VB Utrecht
>
> T. 06-51952295
> I. www.ixsoftware.nl
>
--
Michael Schmalle - Teoti Graphix, LLC
http://www.teotigraphix.com
http://blog.teotigraphix.com
Re: Is Apache model killing this project?
Posted by Erik de Bruin <er...@ixsoftware.nl>.
Mike,
I'm taking none of this personal, I'm Dutch and have a very thick
skin, so don't worry about that :-)
The only thing I'm trying to achieve with my comments is to get you to
contribute something to the project, besides constructive criticism
and excellent ideas. You have to offer so much, why not share a
little? As you correctly assert, no one can complete Falcon on their
own, not even Gordon (well, given enough time... ;-)), so, like a long
journey, maybe start with one small step?
If you think that there needs to be discussion about the direction of
the project and it the goals the community would like to work towards,
please start one. You can use this list, but better would be to use
the Wiki or the website itself, so there is a better record of the
arguments.
EdB
On Thu, Oct 18, 2012 at 2:54 PM, Michael Schmalle
<ap...@teotigraphix.com> wrote:
> Erik,
>
> Please man, I mean no disrespect in any of these emails to you it's a
> frustration I have been holding onto for about 10 months now.
>
> The github reference was in relation to supporting a small, manageable open
> source project in a github account. The point was a lot of those projects
> are small and manageable.
>
> The flex project is huge and all interdependent. Yes, Apache says no road
> maps but common, active contributors need to have conversations and create
> dialog about future implementations. Maybe not when creating higherlevel
> things but when we are talking about a compiler, I'm not just stepping into
> the wild unkown without a "little" backup here, meaning a couple other
> enthusiastic developers at my side. IE I'm not talking about making a
> TabNavigator implementation in my whiteboard here.
>
> So to answer, I have no knowledge of how to get GIT running on Apache.
>
> Mike
>
>
>
> Quoting Erik de Bruin <er...@ixsoftware.nl>:
>
>>> Supporting some github project is definitely where I could "if you don't
>>> like it, change it."
>>
>>
>> Well, that is a good place to start! Apache Infra has indicated that
>> they would help the project move to Git if the project 'donated' a few
>> volunteers to help with the move and to support the new workflow after
>> that. I'm sure the project would much appreciate it if you could be
>> one of those volunteers.
>>
>> Thank you,
>>
>> EdB
>>
>>
>>
>> --
>> Ix Multimedia Software
>>
>> Jan Luykenstraat 27
>> 3521 VB Utrecht
>>
>> T. 06-51952295
>> I. www.ixsoftware.nl
>>
>
> --
> Michael Schmalle - Teoti Graphix, LLC
> http://www.teotigraphix.com
> http://blog.teotigraphix.com
>
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: Is Apache model killing this project?
Posted by Michael Schmalle <ap...@teotigraphix.com>.
Erik,
Please man, I mean no disrespect in any of these emails to you it's a
frustration I have been holding onto for about 10 months now.
The github reference was in relation to supporting a small, manageable
open source project in a github account. The point was a lot of those
projects are small and manageable.
The flex project is huge and all interdependent. Yes, Apache says no
road maps but common, active contributors need to have conversations
and create dialog about future implementations. Maybe not when
creating higherlevel things but when we are talking about a compiler,
I'm not just stepping into the wild unkown without a "little" backup
here, meaning a couple other enthusiastic developers at my side. IE
I'm not talking about making a TabNavigator implementation in my
whiteboard here.
So to answer, I have no knowledge of how to get GIT running on Apache.
Mike
Quoting Erik de Bruin <er...@ixsoftware.nl>:
>> Supporting some github project is definitely where I could "if you don't
>> like it, change it."
>
> Well, that is a good place to start! Apache Infra has indicated that
> they would help the project move to Git if the project 'donated' a few
> volunteers to help with the move and to support the new workflow after
> that. I'm sure the project would much appreciate it if you could be
> one of those volunteers.
>
> Thank you,
>
> EdB
>
>
>
> --
> Ix Multimedia Software
>
> Jan Luykenstraat 27
> 3521 VB Utrecht
>
> T. 06-51952295
> I. www.ixsoftware.nl
>
--
Michael Schmalle - Teoti Graphix, LLC
http://www.teotigraphix.com
http://blog.teotigraphix.com
Re: Is Apache model killing this project?
Posted by Erik de Bruin <er...@ixsoftware.nl>.
> Supporting some github project is definitely where I could "if you don't
> like it, change it."
Well, that is a good place to start! Apache Infra has indicated that
they would help the project move to Git if the project 'donated' a few
volunteers to help with the move and to support the new workflow after
that. I'm sure the project would much appreciate it if you could be
one of those volunteers.
Thank you,
EdB
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: Is Apache model killing this project?
Posted by Michael Schmalle <ap...@teotigraphix.com>.
Quoting Erik de Bruin <er...@ixsoftware.nl>:
>> Carlos, I don't think the Apache model is to blame, I think it's the project
>> to blame, we have no real contributors that are active!
>
> Myself, and I'm sure the other active contributors also, take offense.
> We insist we're very real, not figments of someone's imagination ;-)
>
> I'm sorry you feel frustrated, but with Open Source I think the mantra
> is: "if you don't like it, change it."
>
> EdB
>
Ok, lost in translation, you can't possibly think I meant there are no
real contributors. I definitely worded that wrong and I apologize if
you misinterpreted this.
What I won't apologize for is the seemingly naive attitude this
project is taking about it's future and current contributions.
I would debate this issue face to face, in any open source project
their are tiers of contribution that make the project successful. As
it stands, where are the discussions about the project's future, I'm
not talking about a road map here?
If you think I'm going to believe that the Apache way tells me to do
it myself and "if you don't like it, change it.". This project is
monolithic to change anything myself.
Please reread what I wrote about the compiler, I cannot do that myself
and to sink time into that without others support is ridiculous from
my perspective.
Supporting some github project is definitely where I could "if you
don't like it, change it."
Mike
--
Michael Schmalle - Teoti Graphix, LLC
http://www.teotigraphix.com
http://blog.teotigraphix.com
Re: Is Apache model killing this project?
Posted by Erik de Bruin <er...@ixsoftware.nl>.
> Carlos, I don't think the Apache model is to blame, I think it's the project
> to blame, we have no real contributors that are active!
Myself, and I'm sure the other active contributors also, take offense.
We insist we're very real, not figments of someone's imagination ;-)
I'm sorry you feel frustrated, but with Open Source I think the mantra
is: "if you don't like it, change it."
EdB
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: Is Apache model killing this project?
Posted by Michael Schmalle <ap...@teotigraphix.com>.
I'm not going to rant, see my last two messages.
Being a comitter here is my "location". I have no interest in what
Flex was or is right now, I think its dead and unfortunate that so
many large companies are still tied to the current code base and
seeing nothing move forward with the project.
When I first got into this project I was looking at what it could
become, how it could be reinvented by the many developers that used it.
What I see now, the wagon loaded up onto JS/HTML5 and took most that
might have had any innovative interests.
I have so many years invested into ActionScript that I have tried
everything to keep my interest in something that could look like a
phoenix rising from the ashes. It feels like the horse is dead and the
momentum gone. In poetic words, you could see the high-water mark
where it all looked hopeful.
Flash Gaming is the last horizon for Adobe and Flash, so they pulled
out, put all there money there and are just trying to brush the last
bit of Flex off them.
It's nice of them to give us a 3/4 implemented compiler, very
thoughtful but the fact is, I see no interest from the ghost town of
Apache Flex that would give me any confidence of putting my time into
furthering a compiler that seems like it's going to just end up a
relic of time.
Do you people know how much man power it will take to get this
compiler going past the point of what Gordon could contribute? ... You
don't want the answer.
MXML doesn't even work right now...
This project desperately needs some tie with the future technology or
I can't bear just seeing it rot like this.
Carlos, I don't think the Apache model is to blame, I think it's the
project to blame, we have no real contributors that are active!
Mike
Quoting Carlos Rovira <ca...@codeoscopic.com>:
> Hi,
>
> I want to share some thoughts that I have since August... maybe
> exposing here could make us debate over the problems and see what is
> happening with this project. Maybe Apache Flex is going forward, but
> my sense is it's not. Maybe committers could share what they see.
>
> My overall feeling is that Flex is a complex project and contribute
> need lots of things on board since you, as a contributor, need to be
> comfortable and have time to invest.
>
> The actual model in Apache is very old and is proving to be a failure
> in actual days. The fact that we don't have a valid workflow based on
> the voted option some months ago with Git, Github and nvie Git
> branching model is one of the things that is making this project fail.
>
> People is moving from Flex to HTML5 and this migration seems not to
> stop although what they could get in that world is a real pain
> compared with Flex development workflows. People staying are the teams
> and developers that comes from IT sector and wants an OOP robust
> framework and technology with lots of commodities like AMF,
> Annotations and so on. Things that still exist in Flex nowadays.
>
> Tickets voted in JIRA by community out there still are not closed
> successfully (Maven deployment,...), maybe because actual committers
> have another concept of what Flex should be and invest the time in
> other things, and people that could enter the project to make that
> happen couldn't enter since the Apache model makes it very hard (and
> that people ends investing their time in other things that have more
> possibilities).
>
> In August my coworker (Pepe Barragan) and I was trying to contribute
> to mustella, but the apache's workflow (and our own work) make us fall
> in discouragement. Finally we are unwilling to participate. As we sure
> there are more people. Which is not good for this project.
>
> I'm afraid that or we make some movement and recognized that this
> project is somewhat in bad state or we will end in Apache Flex 4.8
>
> I'm following this list and things seems stuck since august.
>
> Another point is that apache Flex installer was reject many times so
> people ends passing, what again is very bad, and making developers
> burn out to reach RC7...
>
> The last problem I see is the heritage of adobe libraries and legal
> issues added that make all things go very slow or even fail to get it
> done. We are trying to save a great technology and if we are
> discussing lots of legal licenses we soon will be with a dead horse
> but with all documentation in order. If those legal problems (i.e:
> maven related) does not cause a problem for Adobe until now...less now
> than they want to quit from this technology. If things start to go ok,
> legal problems can be solved in that moment.
>
> So, after share all this issues... is Apache the best option for Flex?
> What do you think about it?
> Could we think other options to make this work in places like github
> or bitbucket and the return results here?
>
> Note: My intention is make us think a bit in what is happening. Maybe
> I'm wrong and my perception is not shared here. This would be good and
> will mean the project is really going on, so go forward! :). If that
> is not the case, and you think things are not evolving as expected, is
> the time to express yourself and say what you think. Only with this
> kind of exercise we could know more about what is really happen to
> Apache Flex.
>
> Thanks in advance for your thought on this topic.
>
> Best,
>
> --
> Carlos Rovira
>
--
Michael Schmalle - Teoti Graphix, LLC
http://www.teotigraphix.com
http://blog.teotigraphix.com