You are viewing a plain text version of this content. The canonical link for it is here.
Posted to xap-dev@incubator.apache.org by James Margaris <jm...@nexaweb.com> on 2007/03/21 22:38:16 UTC
Optmizations
Tonight I am going to check in some optimizations that Michael Mikhaylov
and I have been working on. (95% Michael, 5% me) These optimizations are
geared around reducing startup time. The main optimization involves some
changes to our use of dojo.connectEvent - calling it less frequently and
changing the connect code itself a bit as well.
Recently Michael Turyn checked in dojo 0.4, which also makes some
performance improvements as well as fixing the behavior of some
components. For example the tabbed pane response time is now a lot
snappier when clicking on various tabs, and some operations like
scrolling a large table inside of a tabbed pane are an order of
magnitude faster.
So hopefully people will start seeing a fairly significant speedup.
James Margaris
Wiki updated to help new developers
Posted by James Margaris <jm...@nexaweb.com>.
http://wiki.apache.org/incubator/xap/CodeTour
I added this page to help people find their way around the codebase a bit. It describes some basic ideas taken from the xap web page architecture section and translates them into actual file names so people can match the code against the concepts.
It is mostly geared towards developing UI components as that is something many people are probably interested and is also some core functionality.
James Margaris
RE: Optmizations
Posted by James Margaris <jm...@nexaweb.com>.
XAP-346 and XAP-347 are the optimization patches.
There isn't a lot of new code and the new code that there is is basically one method copy/pasted 14 times with a new name, so it looks like a set of minor bug fixes to me.
James M
-----Original Message-----
From: James Margaris [mailto:jmargaris@nexaweb.com]
Sent: Wednesday, March 21, 2007 8:58 PM
To: xap-dev@incubator.apache.org
Subject: RE: Optmizations
I did a little experimenting, you can see the result here:
https://issues.apache.org/jira/browse/XAP-345
If you go to "Manage Attachments" it says in red that the attachments are not intended for use and license is not granted. Whenever you upload an attachment there is a radio button you have to check that says:
"Grant license to ASF for inclusion in ASF works (as per the Apache Software License §5) Contributions intended for inclusion in ASF products (eg. patches, code) must be licensed to ASF under the terms of the Apache Software License. Other attachments (eg. log dumps, test cases) need not be. "
It is kind of annoying that it seems you can *only* see this info if you click on manage attachments, if you click on the attachment itself you you can't see it.
I will ping you when the new patch comes in. I'll look at some older patches and make sure that radio button is checked for people without CLAs.
James M
-----Original Message-----
From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of Martin Cooper
Sent: Wednesday, March 21, 2007 8:44 PM
To: xap-dev@incubator.apache.org
Subject: Re: Optmizations
On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
>
>
> Martin it can be very difficult to navigate the Apache policies
> sometimes.
I know. ;-(
I looked at http://www.apache.org/foundation/how-it-works.html#roles and
> it seems to me that Carlos and Michael M fall under "developers".
Yes. However, you shouldn't read into the description of "committer" that those are the _only_ people who need to sign a CLA. The description of "developer" should probably say that they too may have to sign a CLA, depending on their specific contributions.
If we
> need CLAs for Jira patches I guess we can do that but I have in the
> past checked in patches from people without CLAs.
Again, it's a judgement call. Sometimes it's obvious: a multi-megabyte code contribution is going to need a CLA (and probably more than that); a one-line fix is not. Other times, though, it's not as clear cut.
There's some checkbox thing in JIRA that I've heard about that explicitly grants rights or something. I don't know if that alleviates the legal concerns, and I'm not sure if it shows up for all projects, but it might be worth looking into.
The actual code changes here are fairly small, although I am not
> familiar with the exact scope it is stuff on the level of creating a
> regular expression once and saving it rather than creating a new one
> every time, or using "connectOnce" instead of "connect" to connect
> events. I don't believe there are any new files or substantial blocks
> of new code.
>
> I can hold off on comitting the changes. (I don't see that patch filed
> yet anyway) Maybe you could look at the patch after it is filed and
> help make the judgement call if it is signicant changes or simple bug fixes?
I can try to do that if I can find some time. It would help if you could ping me directly when the patch shows up.
I would like to get the perf improvements checked in without waiting for
> a CLA if that is possible because the performance changes makes a
> fairly large difference. It isn't a lot of code change but the effect
> is rather large.
Understood.
--
Martin Cooper
James M
>
>
>
> -----Original Message-----
> From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> Martin Cooper
> Sent: Wednesday, March 21, 2007 8:18 PM
> To: xap-dev@incubator.apache.org
> Subject: Re: Optmizations
>
> On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> >
> > He has submitted Jira patches in the past and this one will be
> > submitted as well.
> >
> > I was under the impression that Jira patches did not need CLAs filed.
> > Isn't the process supposed to be that people file patches and we
> > patch
>
> > them into the repository?
> >
> > Michael Mikhaylov has submitted patches for:
> >
> > XAP-314
> > XAP-333
> > XAP-318
>
>
> Interesting. I wonder if there is a problem with JIRA. I searched for
> his name in JIRA, and it came back with only XAP-314, which is the
> only one Google finds too. Looking at the other issues you list does
> show his name, so it's strange that a search doesn't find them.
>
> Carlos Sanchez recently submitted a Maven patch that I or someone else
> > will hopefully take a look at and commit soon. Have you heard of
> > Carlos Sanchez before? I haven't but I appreciate the work that he
> did.
>
>
> If it's the same Carlos Sanchez, I believe he is a committer on the
> Maven project.
>
> More up-front communication about what people are working on would be
> > good but I don't see why the lack of that would rule out patching in
> > the code that they wrote and submitted to Jira.
> >
> > If I have some fundamental misunderstanding of how this works please
> > let me know, but I thought that Jira contributors did not need CLAs
> filed.
>
>
> At some level, it becomes a judgement call, but the default should be
> that a CLA is required. To quote from the CLA section on this page:
>
> http://www.apache.org/licenses/
>
> "The ASF desires that all contributors of ideas, code, or
> documentation to the Apache projects complete, sign, and submit [a CLA]"
>
> Many projects tend to let this slide for minor bug fixes and patches
> (which is where the judgement call comes in), but for anything more
> than that, we should really be getting a CLA, for legal protection
> reasons. I brought it up here because you mentioned that 95% of what
> seemed to me likely to be a significant commit was coming from someone
> who has not previously file a CLA.
>
> --
> Martin Cooper
>
>
> James M
> >
> >
> >
> >
> >
> > -----Original Message-----
> > From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> > Martin Cooper
> > Sent: Wednesday, March 21, 2007 6:53 PM
> > To: xap-dev@incubator.apache.org
> > Subject: Re: Optmizations
> >
> > On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> > >
> > > Tonight I am going to check in some optimizations that Michael
> > > Mikhaylov and I have been working on. (95% Michael, 5% me)
> >
> >
> > Since 95% of this is coming from someone else, and he's not a
> > committer, we're going to need him to file a CLA first. Please have
> > him fax one in, and hold off on your commit until it has been
> > received
>
> > by the ASF secretary.
> >
> > Who is this Michael Mikhaylov? I haven't seen his name in relation
> > to XAP anywhere before. If he's working on XAP, he should be on the
> > lists, and we should be seeing the discussions of the work he's
> > doing right here on the xap-dev list.
> >
> > Remember, this is a community project, not a stealth development
> > with a public face.
> >
> > --
> > Martin Cooper
> >
> >
> > These optimizations are
> > > geared around reducing startup time. The main optimization
> > > involves some changes to our use of dojo.connectEvent - calling it
> > > less frequently and changing the connect code itself a bit as well.
> > >
> > > Recently Michael Turyn checked in dojo 0.4, which also makes some
> > > performance improvements as well as fixing the behavior of some
> > > components. For example the tabbed pane response time is now a lot
> > > snappier when clicking on various tabs, and some operations like
> > > scrolling a large table inside of a tabbed pane are an order of
> > > magnitude faster.
> > >
> > > So hopefully people will start seeing a fairly significant speedup.
> > >
> > > James Margaris
> > >
> >
>
JIRA organization
Posted by James Margaris <jm...@nexaweb.com>.
I'm separating out "widgets" into more specific categories because the
vast majority of issues are widget issues and it is annoying for 90% of
all bugs to be grouped in one category.
James M
RE: Optmizations
Posted by David Gennaco <dg...@nexaweb.com>.
Hello,
In the few bug fixes I submitted with attachments the appropriate button should have been checked, please let me know if I did not file them correctly.
Thank you,
Dave
-----Original Message-----
From: James Margaris [mailto:jmargaris@nexaweb.com]
Sent: Wednesday, March 21, 2007 8:58 PM
To: xap-dev@incubator.apache.org
Subject: RE: Optmizations
I did a little experimenting, you can see the result here:
https://issues.apache.org/jira/browse/XAP-345
If you go to "Manage Attachments" it says in red that the attachments are not intended for use and license is not granted. Whenever you upload an attachment there is a radio button you have to check that says:
"Grant license to ASF for inclusion in ASF works (as per the Apache Software License §5) Contributions intended for inclusion in ASF products (eg. patches, code) must be licensed to ASF under the terms of the Apache Software License. Other attachments (eg. log dumps, test cases) need not be. "
It is kind of annoying that it seems you can *only* see this info if you click on manage attachments, if you click on the attachment itself you you can't see it.
I will ping you when the new patch comes in. I'll look at some older patches and make sure that radio button is checked for people without CLAs.
James M
-----Original Message-----
From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of Martin Cooper
Sent: Wednesday, March 21, 2007 8:44 PM
To: xap-dev@incubator.apache.org
Subject: Re: Optmizations
On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
>
>
> Martin it can be very difficult to navigate the Apache policies
> sometimes.
I know. ;-(
I looked at http://www.apache.org/foundation/how-it-works.html#roles and
> it seems to me that Carlos and Michael M fall under "developers".
Yes. However, you shouldn't read into the description of "committer" that those are the _only_ people who need to sign a CLA. The description of "developer" should probably say that they too may have to sign a CLA, depending on their specific contributions.
If we
> need CLAs for Jira patches I guess we can do that but I have in the
> past checked in patches from people without CLAs.
Again, it's a judgement call. Sometimes it's obvious: a multi-megabyte code contribution is going to need a CLA (and probably more than that); a one-line fix is not. Other times, though, it's not as clear cut.
There's some checkbox thing in JIRA that I've heard about that explicitly grants rights or something. I don't know if that alleviates the legal concerns, and I'm not sure if it shows up for all projects, but it might be worth looking into.
The actual code changes here are fairly small, although I am not
> familiar with the exact scope it is stuff on the level of creating a
> regular expression once and saving it rather than creating a new one
> every time, or using "connectOnce" instead of "connect" to connect
> events. I don't believe there are any new files or substantial blocks
> of new code.
>
> I can hold off on comitting the changes. (I don't see that patch filed
> yet anyway) Maybe you could look at the patch after it is filed and
> help make the judgement call if it is signicant changes or simple bug fixes?
I can try to do that if I can find some time. It would help if you could ping me directly when the patch shows up.
I would like to get the perf improvements checked in without waiting for
> a CLA if that is possible because the performance changes makes a
> fairly large difference. It isn't a lot of code change but the effect
> is rather large.
Understood.
--
Martin Cooper
James M
>
>
>
> -----Original Message-----
> From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> Martin Cooper
> Sent: Wednesday, March 21, 2007 8:18 PM
> To: xap-dev@incubator.apache.org
> Subject: Re: Optmizations
>
> On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> >
> > He has submitted Jira patches in the past and this one will be
> > submitted as well.
> >
> > I was under the impression that Jira patches did not need CLAs filed.
> > Isn't the process supposed to be that people file patches and we
> > patch
>
> > them into the repository?
> >
> > Michael Mikhaylov has submitted patches for:
> >
> > XAP-314
> > XAP-333
> > XAP-318
>
>
> Interesting. I wonder if there is a problem with JIRA. I searched for
> his name in JIRA, and it came back with only XAP-314, which is the
> only one Google finds too. Looking at the other issues you list does
> show his name, so it's strange that a search doesn't find them.
>
> Carlos Sanchez recently submitted a Maven patch that I or someone else
> > will hopefully take a look at and commit soon. Have you heard of
> > Carlos Sanchez before? I haven't but I appreciate the work that he
> did.
>
>
> If it's the same Carlos Sanchez, I believe he is a committer on the
> Maven project.
>
> More up-front communication about what people are working on would be
> > good but I don't see why the lack of that would rule out patching in
> > the code that they wrote and submitted to Jira.
> >
> > If I have some fundamental misunderstanding of how this works please
> > let me know, but I thought that Jira contributors did not need CLAs
> filed.
>
>
> At some level, it becomes a judgement call, but the default should be
> that a CLA is required. To quote from the CLA section on this page:
>
> http://www.apache.org/licenses/
>
> "The ASF desires that all contributors of ideas, code, or
> documentation to the Apache projects complete, sign, and submit [a CLA]"
>
> Many projects tend to let this slide for minor bug fixes and patches
> (which is where the judgement call comes in), but for anything more
> than that, we should really be getting a CLA, for legal protection
> reasons. I brought it up here because you mentioned that 95% of what
> seemed to me likely to be a significant commit was coming from someone
> who has not previously file a CLA.
>
> --
> Martin Cooper
>
>
> James M
> >
> >
> >
> >
> >
> > -----Original Message-----
> > From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> > Martin Cooper
> > Sent: Wednesday, March 21, 2007 6:53 PM
> > To: xap-dev@incubator.apache.org
> > Subject: Re: Optmizations
> >
> > On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> > >
> > > Tonight I am going to check in some optimizations that Michael
> > > Mikhaylov and I have been working on. (95% Michael, 5% me)
> >
> >
> > Since 95% of this is coming from someone else, and he's not a
> > committer, we're going to need him to file a CLA first. Please have
> > him fax one in, and hold off on your commit until it has been
> > received
>
> > by the ASF secretary.
> >
> > Who is this Michael Mikhaylov? I haven't seen his name in relation
> > to XAP anywhere before. If he's working on XAP, he should be on the
> > lists, and we should be seeing the discussions of the work he's
> > doing right here on the xap-dev list.
> >
> > Remember, this is a community project, not a stealth development
> > with a public face.
> >
> > --
> > Martin Cooper
> >
> >
> > These optimizations are
> > > geared around reducing startup time. The main optimization
> > > involves some changes to our use of dojo.connectEvent - calling it
> > > less frequently and changing the connect code itself a bit as well.
> > >
> > > Recently Michael Turyn checked in dojo 0.4, which also makes some
> > > performance improvements as well as fixing the behavior of some
> > > components. For example the tabbed pane response time is now a lot
> > > snappier when clicking on various tabs, and some operations like
> > > scrolling a large table inside of a tabbed pane are an order of
> > > magnitude faster.
> > >
> > > So hopefully people will start seeing a fairly significant speedup.
> > >
> > > James Margaris
> > >
> >
>
RE: Optmizations
Posted by James Margaris <jm...@nexaweb.com>.
I did a little experimenting, you can see the result here:
https://issues.apache.org/jira/browse/XAP-345
If you go to "Manage Attachments" it says in red that the attachments are not intended for use and license is not granted. Whenever you upload an attachment there is a radio button you have to check that says:
"Grant license to ASF for inclusion in ASF works (as per the Apache Software License §5)
Contributions intended for inclusion in ASF products (eg. patches, code) must be licensed to ASF under the terms of the Apache Software License. Other attachments (eg. log dumps, test cases) need not be. "
It is kind of annoying that it seems you can *only* see this info if you click on manage attachments, if you click on the attachment itself you you can't see it.
I will ping you when the new patch comes in. I'll look at some older patches and make sure that radio button is checked for people without CLAs.
James M
-----Original Message-----
From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of Martin Cooper
Sent: Wednesday, March 21, 2007 8:44 PM
To: xap-dev@incubator.apache.org
Subject: Re: Optmizations
On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
>
>
> Martin it can be very difficult to navigate the Apache policies
> sometimes.
I know. ;-(
I looked at http://www.apache.org/foundation/how-it-works.html#roles and
> it seems to me that Carlos and Michael M fall under "developers".
Yes. However, you shouldn't read into the description of "committer" that those are the _only_ people who need to sign a CLA. The description of "developer" should probably say that they too may have to sign a CLA, depending on their specific contributions.
If we
> need CLAs for Jira patches I guess we can do that but I have in the
> past checked in patches from people without CLAs.
Again, it's a judgement call. Sometimes it's obvious: a multi-megabyte code contribution is going to need a CLA (and probably more than that); a one-line fix is not. Other times, though, it's not as clear cut.
There's some checkbox thing in JIRA that I've heard about that explicitly grants rights or something. I don't know if that alleviates the legal concerns, and I'm not sure if it shows up for all projects, but it might be worth looking into.
The actual code changes here are fairly small, although I am not
> familiar with the exact scope it is stuff on the level of creating a
> regular expression once and saving it rather than creating a new one
> every time, or using "connectOnce" instead of "connect" to connect
> events. I don't believe there are any new files or substantial blocks
> of new code.
>
> I can hold off on comitting the changes. (I don't see that patch filed
> yet anyway) Maybe you could look at the patch after it is filed and
> help make the judgement call if it is signicant changes or simple bug fixes?
I can try to do that if I can find some time. It would help if you could ping me directly when the patch shows up.
I would like to get the perf improvements checked in without waiting for
> a CLA if that is possible because the performance changes makes a
> fairly large difference. It isn't a lot of code change but the effect
> is rather large.
Understood.
--
Martin Cooper
James M
>
>
>
> -----Original Message-----
> From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> Martin Cooper
> Sent: Wednesday, March 21, 2007 8:18 PM
> To: xap-dev@incubator.apache.org
> Subject: Re: Optmizations
>
> On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> >
> > He has submitted Jira patches in the past and this one will be
> > submitted as well.
> >
> > I was under the impression that Jira patches did not need CLAs filed.
> > Isn't the process supposed to be that people file patches and we
> > patch
>
> > them into the repository?
> >
> > Michael Mikhaylov has submitted patches for:
> >
> > XAP-314
> > XAP-333
> > XAP-318
>
>
> Interesting. I wonder if there is a problem with JIRA. I searched for
> his name in JIRA, and it came back with only XAP-314, which is the
> only one Google finds too. Looking at the other issues you list does
> show his name, so it's strange that a search doesn't find them.
>
> Carlos Sanchez recently submitted a Maven patch that I or someone else
> > will hopefully take a look at and commit soon. Have you heard of
> > Carlos Sanchez before? I haven't but I appreciate the work that he
> did.
>
>
> If it's the same Carlos Sanchez, I believe he is a committer on the
> Maven project.
>
> More up-front communication about what people are working on would be
> > good but I don't see why the lack of that would rule out patching in
> > the code that they wrote and submitted to Jira.
> >
> > If I have some fundamental misunderstanding of how this works please
> > let me know, but I thought that Jira contributors did not need CLAs
> filed.
>
>
> At some level, it becomes a judgement call, but the default should be
> that a CLA is required. To quote from the CLA section on this page:
>
> http://www.apache.org/licenses/
>
> "The ASF desires that all contributors of ideas, code, or
> documentation to the Apache projects complete, sign, and submit [a CLA]"
>
> Many projects tend to let this slide for minor bug fixes and patches
> (which is where the judgement call comes in), but for anything more
> than that, we should really be getting a CLA, for legal protection
> reasons. I brought it up here because you mentioned that 95% of what
> seemed to me likely to be a significant commit was coming from someone
> who has not previously file a CLA.
>
> --
> Martin Cooper
>
>
> James M
> >
> >
> >
> >
> >
> > -----Original Message-----
> > From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> > Martin Cooper
> > Sent: Wednesday, March 21, 2007 6:53 PM
> > To: xap-dev@incubator.apache.org
> > Subject: Re: Optmizations
> >
> > On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> > >
> > > Tonight I am going to check in some optimizations that Michael
> > > Mikhaylov and I have been working on. (95% Michael, 5% me)
> >
> >
> > Since 95% of this is coming from someone else, and he's not a
> > committer, we're going to need him to file a CLA first. Please have
> > him fax one in, and hold off on your commit until it has been
> > received
>
> > by the ASF secretary.
> >
> > Who is this Michael Mikhaylov? I haven't seen his name in relation
> > to XAP anywhere before. If he's working on XAP, he should be on the
> > lists, and we should be seeing the discussions of the work he's
> > doing right here on the xap-dev list.
> >
> > Remember, this is a community project, not a stealth development
> > with a public face.
> >
> > --
> > Martin Cooper
> >
> >
> > These optimizations are
> > > geared around reducing startup time. The main optimization
> > > involves some changes to our use of dojo.connectEvent - calling it
> > > less frequently and changing the connect code itself a bit as well.
> > >
> > > Recently Michael Turyn checked in dojo 0.4, which also makes some
> > > performance improvements as well as fixing the behavior of some
> > > components. For example the tabbed pane response time is now a lot
> > > snappier when clicking on various tabs, and some operations like
> > > scrolling a large table inside of a tabbed pane are an order of
> > > magnitude faster.
> > >
> > > So hopefully people will start seeing a fairly significant speedup.
> > >
> > > James Margaris
> > >
> >
>
Re: Optmizations
Posted by Martin Cooper <ma...@apache.org>.
On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
>
>
> Martin it can be very difficult to navigate the Apache policies
> sometimes.
I know. ;-(
I looked at http://www.apache.org/foundation/how-it-works.html#roles and
> it seems to me that Carlos and Michael M fall under "developers".
Yes. However, you shouldn't read into the description of "committer" that
those are the _only_ people who need to sign a CLA. The description of
"developer" should probably say that they too may have to sign a CLA,
depending on their specific contributions.
If we
> need CLAs for Jira patches I guess we can do that but I have in the past
> checked in patches from people without CLAs.
Again, it's a judgement call. Sometimes it's obvious: a multi-megabyte code
contribution is going to need a CLA (and probably more than that); a
one-line fix is not. Other times, though, it's not as clear cut.
There's some checkbox thing in JIRA that I've heard about that explicitly
grants rights or something. I don't know if that alleviates the legal
concerns, and I'm not sure if it shows up for all projects, but it might be
worth looking into.
The actual code changes here are fairly small, although I am not
> familiar with the exact scope it is stuff on the level of creating a
> regular expression once and saving it rather than creating a new one
> every time, or using "connectOnce" instead of "connect" to connect
> events. I don't believe there are any new files or substantial blocks of
> new code.
>
> I can hold off on comitting the changes. (I don't see that patch filed
> yet anyway) Maybe you could look at the patch after it is filed and help
> make the judgement call if it is signicant changes or simple bug fixes?
I can try to do that if I can find some time. It would help if you could
ping me directly when the patch shows up.
I would like to get the perf improvements checked in without waiting for
> a CLA if that is possible because the performance changes makes a fairly
> large difference. It isn't a lot of code change but the effect is rather
> large.
Understood.
--
Martin Cooper
James M
>
>
>
> -----Original Message-----
> From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> Martin Cooper
> Sent: Wednesday, March 21, 2007 8:18 PM
> To: xap-dev@incubator.apache.org
> Subject: Re: Optmizations
>
> On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> >
> > He has submitted Jira patches in the past and this one will be
> > submitted as well.
> >
> > I was under the impression that Jira patches did not need CLAs filed.
> > Isn't the process supposed to be that people file patches and we patch
>
> > them into the repository?
> >
> > Michael Mikhaylov has submitted patches for:
> >
> > XAP-314
> > XAP-333
> > XAP-318
>
>
> Interesting. I wonder if there is a problem with JIRA. I searched for
> his name in JIRA, and it came back with only XAP-314, which is the only
> one Google finds too. Looking at the other issues you list does show his
> name, so it's strange that a search doesn't find them.
>
> Carlos Sanchez recently submitted a Maven patch that I or someone else
> > will hopefully take a look at and commit soon. Have you heard of
> > Carlos Sanchez before? I haven't but I appreciate the work that he
> did.
>
>
> If it's the same Carlos Sanchez, I believe he is a committer on the
> Maven project.
>
> More up-front communication about what people are working on would be
> > good but I don't see why the lack of that would rule out patching in
> > the code that they wrote and submitted to Jira.
> >
> > If I have some fundamental misunderstanding of how this works please
> > let me know, but I thought that Jira contributors did not need CLAs
> filed.
>
>
> At some level, it becomes a judgement call, but the default should be
> that a CLA is required. To quote from the CLA section on this page:
>
> http://www.apache.org/licenses/
>
> "The ASF desires that all contributors of ideas, code, or documentation
> to the Apache projects complete, sign, and submit [a CLA]"
>
> Many projects tend to let this slide for minor bug fixes and patches
> (which is where the judgement call comes in), but for anything more than
> that, we should really be getting a CLA, for legal protection reasons. I
> brought it up here because you mentioned that 95% of what seemed to me
> likely to be a significant commit was coming from someone who has not
> previously file a CLA.
>
> --
> Martin Cooper
>
>
> James M
> >
> >
> >
> >
> >
> > -----Original Message-----
> > From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> > Martin Cooper
> > Sent: Wednesday, March 21, 2007 6:53 PM
> > To: xap-dev@incubator.apache.org
> > Subject: Re: Optmizations
> >
> > On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> > >
> > > Tonight I am going to check in some optimizations that Michael
> > > Mikhaylov and I have been working on. (95% Michael, 5% me)
> >
> >
> > Since 95% of this is coming from someone else, and he's not a
> > committer, we're going to need him to file a CLA first. Please have
> > him fax one in, and hold off on your commit until it has been received
>
> > by the ASF secretary.
> >
> > Who is this Michael Mikhaylov? I haven't seen his name in relation to
> > XAP anywhere before. If he's working on XAP, he should be on the
> > lists, and we should be seeing the discussions of the work he's doing
> > right here on the xap-dev list.
> >
> > Remember, this is a community project, not a stealth development with
> > a public face.
> >
> > --
> > Martin Cooper
> >
> >
> > These optimizations are
> > > geared around reducing startup time. The main optimization involves
> > > some changes to our use of dojo.connectEvent - calling it less
> > > frequently and changing the connect code itself a bit as well.
> > >
> > > Recently Michael Turyn checked in dojo 0.4, which also makes some
> > > performance improvements as well as fixing the behavior of some
> > > components. For example the tabbed pane response time is now a lot
> > > snappier when clicking on various tabs, and some operations like
> > > scrolling a large table inside of a tabbed pane are an order of
> > > magnitude faster.
> > >
> > > So hopefully people will start seeing a fairly significant speedup.
> > >
> > > James Margaris
> > >
> >
>
RE: Optmizations
Posted by James Margaris <jm...@nexaweb.com>.
Martin it can be very difficult to navigate the Apache policies
sometimes.
I looked at http://www.apache.org/foundation/how-it-works.html#roles and
it seems to me that Carlos and Michael M fall under "developers". If we
need CLAs for Jira patches I guess we can do that but I have in the past
checked in patches from people without CLAs.
The actual code changes here are fairly small, although I am not
familiar with the exact scope it is stuff on the level of creating a
regular expression once and saving it rather than creating a new one
every time, or using "connectOnce" instead of "connect" to connect
events. I don't believe there are any new files or substantial blocks of
new code.
I can hold off on comitting the changes. (I don't see that patch filed
yet anyway) Maybe you could look at the patch after it is filed and help
make the judgement call if it is signicant changes or simple bug fixes?
I would like to get the perf improvements checked in without waiting for
a CLA if that is possible because the performance changes makes a fairly
large difference. It isn't a lot of code change but the effect is rather
large.
James M
-----Original Message-----
From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
Martin Cooper
Sent: Wednesday, March 21, 2007 8:18 PM
To: xap-dev@incubator.apache.org
Subject: Re: Optmizations
On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
>
> He has submitted Jira patches in the past and this one will be
> submitted as well.
>
> I was under the impression that Jira patches did not need CLAs filed.
> Isn't the process supposed to be that people file patches and we patch
> them into the repository?
>
> Michael Mikhaylov has submitted patches for:
>
> XAP-314
> XAP-333
> XAP-318
Interesting. I wonder if there is a problem with JIRA. I searched for
his name in JIRA, and it came back with only XAP-314, which is the only
one Google finds too. Looking at the other issues you list does show his
name, so it's strange that a search doesn't find them.
Carlos Sanchez recently submitted a Maven patch that I or someone else
> will hopefully take a look at and commit soon. Have you heard of
> Carlos Sanchez before? I haven't but I appreciate the work that he
did.
If it's the same Carlos Sanchez, I believe he is a committer on the
Maven project.
More up-front communication about what people are working on would be
> good but I don't see why the lack of that would rule out patching in
> the code that they wrote and submitted to Jira.
>
> If I have some fundamental misunderstanding of how this works please
> let me know, but I thought that Jira contributors did not need CLAs
filed.
At some level, it becomes a judgement call, but the default should be
that a CLA is required. To quote from the CLA section on this page:
http://www.apache.org/licenses/
"The ASF desires that all contributors of ideas, code, or documentation
to the Apache projects complete, sign, and submit [a CLA]"
Many projects tend to let this slide for minor bug fixes and patches
(which is where the judgement call comes in), but for anything more than
that, we should really be getting a CLA, for legal protection reasons. I
brought it up here because you mentioned that 95% of what seemed to me
likely to be a significant commit was coming from someone who has not
previously file a CLA.
--
Martin Cooper
James M
>
>
>
>
>
> -----Original Message-----
> From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> Martin Cooper
> Sent: Wednesday, March 21, 2007 6:53 PM
> To: xap-dev@incubator.apache.org
> Subject: Re: Optmizations
>
> On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> >
> > Tonight I am going to check in some optimizations that Michael
> > Mikhaylov and I have been working on. (95% Michael, 5% me)
>
>
> Since 95% of this is coming from someone else, and he's not a
> committer, we're going to need him to file a CLA first. Please have
> him fax one in, and hold off on your commit until it has been received
> by the ASF secretary.
>
> Who is this Michael Mikhaylov? I haven't seen his name in relation to
> XAP anywhere before. If he's working on XAP, he should be on the
> lists, and we should be seeing the discussions of the work he's doing
> right here on the xap-dev list.
>
> Remember, this is a community project, not a stealth development with
> a public face.
>
> --
> Martin Cooper
>
>
> These optimizations are
> > geared around reducing startup time. The main optimization involves
> > some changes to our use of dojo.connectEvent - calling it less
> > frequently and changing the connect code itself a bit as well.
> >
> > Recently Michael Turyn checked in dojo 0.4, which also makes some
> > performance improvements as well as fixing the behavior of some
> > components. For example the tabbed pane response time is now a lot
> > snappier when clicking on various tabs, and some operations like
> > scrolling a large table inside of a tabbed pane are an order of
> > magnitude faster.
> >
> > So hopefully people will start seeing a fairly significant speedup.
> >
> > James Margaris
> >
>
Re: Optmizations
Posted by Martin Cooper <ma...@apache.org>.
On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
>
> He has submitted Jira patches in the past and this one will be submitted
> as well.
>
> I was under the impression that Jira patches did not need CLAs filed.
> Isn't the process supposed to be that people file patches and we patch
> them into the repository?
>
> Michael Mikhaylov has submitted patches for:
>
> XAP-314
> XAP-333
> XAP-318
Interesting. I wonder if there is a problem with JIRA. I searched for his
name in JIRA, and it came back with only XAP-314, which is the only one
Google finds too. Looking at the other issues you list does show his name,
so it's strange that a search doesn't find them.
Carlos Sanchez recently submitted a Maven patch that I or someone else
> will hopefully take a look at and commit soon. Have you heard of Carlos
> Sanchez before? I haven't but I appreciate the work that he did.
If it's the same Carlos Sanchez, I believe he is a committer on the Maven
project.
More up-front communication about what people are working on would be
> good but I don't see why the lack of that would rule out patching in the
> code that they wrote and submitted to Jira.
>
> If I have some fundamental misunderstanding of how this works please let
> me know, but I thought that Jira contributors did not need CLAs filed.
At some level, it becomes a judgement call, but the default should be that a
CLA is required. To quote from the CLA section on this page:
http://www.apache.org/licenses/
"The ASF desires that all contributors of ideas, code, or documentation to
the Apache projects complete, sign, and submit [a CLA]"
Many projects tend to let this slide for minor bug fixes and patches (which
is where the judgement call comes in), but for anything more than that, we
should really be getting a CLA, for legal protection reasons. I brought it
up here because you mentioned that 95% of what seemed to me likely to be a
significant commit was coming from someone who has not previously file a
CLA.
--
Martin Cooper
James M
>
>
>
>
>
> -----Original Message-----
> From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
> Martin Cooper
> Sent: Wednesday, March 21, 2007 6:53 PM
> To: xap-dev@incubator.apache.org
> Subject: Re: Optmizations
>
> On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
> >
> > Tonight I am going to check in some optimizations that Michael
> > Mikhaylov and I have been working on. (95% Michael, 5% me)
>
>
> Since 95% of this is coming from someone else, and he's not a committer,
> we're going to need him to file a CLA first. Please have him fax one in,
> and hold off on your commit until it has been received by the ASF
> secretary.
>
> Who is this Michael Mikhaylov? I haven't seen his name in relation to
> XAP anywhere before. If he's working on XAP, he should be on the lists,
> and we should be seeing the discussions of the work he's doing right
> here on the xap-dev list.
>
> Remember, this is a community project, not a stealth development with a
> public face.
>
> --
> Martin Cooper
>
>
> These optimizations are
> > geared around reducing startup time. The main optimization involves
> > some changes to our use of dojo.connectEvent - calling it less
> > frequently and changing the connect code itself a bit as well.
> >
> > Recently Michael Turyn checked in dojo 0.4, which also makes some
> > performance improvements as well as fixing the behavior of some
> > components. For example the tabbed pane response time is now a lot
> > snappier when clicking on various tabs, and some operations like
> > scrolling a large table inside of a tabbed pane are an order of
> > magnitude faster.
> >
> > So hopefully people will start seeing a fairly significant speedup.
> >
> > James Margaris
> >
>
RE: Optmizations
Posted by James Margaris <jm...@nexaweb.com>.
He has submitted Jira patches in the past and this one will be submitted
as well.
I was under the impression that Jira patches did not need CLAs filed.
Isn't the process supposed to be that people file patches and we patch
them into the repository?
Michael Mikhaylov has submitted patches for:
XAP-314
XAP-333
XAP-318
Carlos Sanchez recently submitted a Maven patch that I or someone else
will hopefully take a look at and commit soon. Have you heard of Carlos
Sanchez before? I haven't but I appreciate the work that he did.
More up-front communication about what people are working on would be
good but I don't see why the lack of that would rule out patching in the
code that they wrote and submitted to Jira.
If I have some fundamental misunderstanding of how this works please let
me know, but I thought that Jira contributors did not need CLAs filed.
James M
-----Original Message-----
From: mfncooper@gmail.com [mailto:mfncooper@gmail.com] On Behalf Of
Martin Cooper
Sent: Wednesday, March 21, 2007 6:53 PM
To: xap-dev@incubator.apache.org
Subject: Re: Optmizations
On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
>
> Tonight I am going to check in some optimizations that Michael
> Mikhaylov and I have been working on. (95% Michael, 5% me)
Since 95% of this is coming from someone else, and he's not a committer,
we're going to need him to file a CLA first. Please have him fax one in,
and hold off on your commit until it has been received by the ASF
secretary.
Who is this Michael Mikhaylov? I haven't seen his name in relation to
XAP anywhere before. If he's working on XAP, he should be on the lists,
and we should be seeing the discussions of the work he's doing right
here on the xap-dev list.
Remember, this is a community project, not a stealth development with a
public face.
--
Martin Cooper
These optimizations are
> geared around reducing startup time. The main optimization involves
> some changes to our use of dojo.connectEvent - calling it less
> frequently and changing the connect code itself a bit as well.
>
> Recently Michael Turyn checked in dojo 0.4, which also makes some
> performance improvements as well as fixing the behavior of some
> components. For example the tabbed pane response time is now a lot
> snappier when clicking on various tabs, and some operations like
> scrolling a large table inside of a tabbed pane are an order of
> magnitude faster.
>
> So hopefully people will start seeing a fairly significant speedup.
>
> James Margaris
>
Re: Optmizations
Posted by Martin Cooper <ma...@apache.org>.
On 3/21/07, James Margaris <jm...@nexaweb.com> wrote:
>
> Tonight I am going to check in some optimizations that Michael Mikhaylov
> and I have been working on. (95% Michael, 5% me)
Since 95% of this is coming from someone else, and he's not a committer,
we're going to need him to file a CLA first. Please have him fax one in, and
hold off on your commit until it has been received by the ASF secretary.
Who is this Michael Mikhaylov? I haven't seen his name in relation to XAP
anywhere before. If he's working on XAP, he should be on the lists, and we
should be seeing the discussions of the work he's doing right here on the
xap-dev list.
Remember, this is a community project, not a stealth development with a
public face.
--
Martin Cooper
These optimizations are
> geared around reducing startup time. The main optimization involves some
> changes to our use of dojo.connectEvent - calling it less frequently and
> changing the connect code itself a bit as well.
>
> Recently Michael Turyn checked in dojo 0.4, which also makes some
> performance improvements as well as fixing the behavior of some
> components. For example the tabbed pane response time is now a lot
> snappier when clicking on various tabs, and some operations like
> scrolling a large table inside of a tabbed pane are an order of
> magnitude faster.
>
> So hopefully people will start seeing a fairly significant speedup.
>
> James Margaris
>