You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tuscany.apache.org by Oscar Castaneda <os...@gmail.com> on 2008/08/28 17:50:56 UTC
[GSoC] Final student update
Dear Mentors,
Since the last update I sent Google has released an updated version of the
Android SDK. Unfortunately for my GSoC project such release came too late.
On a more positive note, I will continue helping Adriano with the porting
efforts. We have begun testing the new SDK with the Android trunk and can
hopefully see Tuscany running on Android soon.
The last tests performed as part of my GSoC project were to remove all
@Remotable annotations and to test calculator-android with a componentType
side file. Both of these initiatives failed because of the fact the previous
version of the Android SDK still choked when encountering any code with
annotations. As a new SDK was released I stopped further testing. Since then
I have updated the JIRA ticket I created, with the reduced set of modules
and changes that Adriano made to SCA code. Additionally, I updated my
project wiki to summarize all of the work performed throughout the project.
I have now started testing the new SDK and currently am running into
problems with trying to include a core library from a desktop
virtual machine into an Android application, this being
"javax/xml/namespace/QName$1.class". I will continue these efforts outside
the scope of GSoC.
Lastly, I would like to thank the Tuscany community; I received help from
several committers throughout the project, and of course, from my mentor
Adriano. I thank you all for your patience and continued support. Despite
the fact that the project wasn't a complete success I definitely learned a
lot and was able to explore my interests. I am now continuing my studies in
Computer Science and look forward to continue learning from and
collaborating on this project.
--
best,
-oscar
Oscar Castañeda
Android port patches, was: A branch for the android port, was: [GSoC]
Final student update
Posted by Jean-Sebastien Delfino <js...@apache.org>.
Adriano Crestani wrote:
> Hi Jean,
>
> We weren't able to run successfully the calculator sample on the android
> platform yet. I copied the modules from the r643746 to the
> android-sandbox and started to modify and identify which parts of sca
> code will need to be modified to get sca runtime running on android
> platform. I think if we identify all the necessary modifications on
> r643746, they will be probably the same on the latest revision. So, in
> my opinion we could keep working on this revision, avoinding any problem
> while trying to "update" it to the latest revision. I really think it's
> up to Oscar. Lets wait for his response.
>
> My modifications would not be compatible with the code on the android
> branch, I restarted my work on getting it working on a most recent
> revision. Lets wait for Oscar's patch and I will check which
> modifications I did could be applied to the branch code.
>
OK, I'm happy with r643746 or any other revision you prefer, switching
should be easy anyway.
If that helps, I have pretty powerful diff/patch tools on my Linux
system that can help us identify all the necessary mofications to
r643746. I can try to create the following diff patches:
a) the current android branch vs r643746
b) Oscar's changes vs r643746 (unless Oscar already has that)
c) a 3-way diff/patch of the above.
I can try that after work this evening.
--
Jean-Sebastien
Re: A branch for the android port, was: [GSoC] Final student update
Posted by Adriano Crestani <ad...@apache.org>.
Hi Jean,
We weren't able to run successfully the calculator sample on the android
platform yet. I copied the modules from the r643746 to the android-sandbox
and started to modify and identify which parts of sca code will need to be
modified to get sca runtime running on android platform. I think if we
identify all the necessary modifications on r643746, they will be probably
the same on the latest revision. So, in my opinion we could keep working on
this revision, avoinding any problem while trying to "update" it to the
latest revision. I really think it's up to Oscar. Lets wait for his
response.
My modifications would not be compatible with the code on the android
branch, I restarted my work on getting it working on a most recent revision.
Lets wait for Oscar's patch and I will check which modifications I did could
be applied to the branch code.
Regards,
Adriano Crestani
On Wed, Sep 10, 2008 at 10:17 AM, Jean-Sebastien Delfino <
jsdelfino@apache.org> wrote:
> Adriano Crestani wrote:
>
>> Hi Oscar,
>>
>> Why you tried to commit? You will not be able to do it. I think you were
>> probably wanting to apply a patch on your trunk or something like that, was
>> it?
>>
>> Regards,
>> Adriano Crestani
>>
>> On Tue, Sep 9, 2008 at 2:09 PM, Oscar Castaneda <
>> oscar.castaneda.gsoc@gmail.com <ma...@gmail.com>>
>> wrote:
>>
>> Hi Guys,
>>
>> Sorry for the delay in responding, I went on vacation for a few days
>> before starting school again. About JIRA 2440, I updated it with a
>> .tar.gz file because when I deleted modules from my trunk and did an
>> 'svn commit' I was asked for credentials for the Tuscany svn
>> repository. Probably I should set it to do things locally, but
>> didn't do so because was in a hurry to provide a last update for GSoC.
>>
>> The .tar.gz file contains the set of reduced modules with the
>> changes Adriano had done to the code, disabling API code and such.
>> In other words, it's a subset of revision 643746 (the one in which
>> Adriano was originally working on). Please let me know if you would
>> like me to make a patch out of it instead of deleting the modules in
>> the trunk. The advantage of having this reduced set of modules is
>> that they are sufficient to run calculator2, and as such provide a
>> lightweight Tuscany runtime for Android.
>>
>>
> Oscar, only committers can commit that's why you weren't able to do it :)
> However svn rm is just local and does not require you to commit anything.
>
> When you say r643746, I'm assuming that you're talking about r643746 of
> trunk?
>
> If you provide a patch with the changes from r643746 and the list of
> modules to not pick from trunk I can help apply that to the sca-android
> branch.
>
> I think we should merge these changes with other changes you have Adriano,
> on top of r643746.
>
> Do you guys think that it's worth trying to apply the patches to a more
> recent revision, as r643746 is from April?
>
> Does that make sense? Please let me know if I've mis-understood anything.
> Thanks!
> --
> Jean-Sebastien
>
Re: A branch for the android port, was: [GSoC] Final student update
Posted by Jean-Sebastien Delfino <js...@apache.org>.
Adriano Crestani wrote:
> Hi Oscar,
>
> Why you tried to commit? You will not be able to do it. I think you were
> probably wanting to apply a patch on your trunk or something like that,
> was it?
>
> Regards,
> Adriano Crestani
>
> On Tue, Sep 9, 2008 at 2:09 PM, Oscar Castaneda
> <oscar.castaneda.gsoc@gmail.com <ma...@gmail.com>>
> wrote:
>
> Hi Guys,
>
> Sorry for the delay in responding, I went on vacation for a few days
> before starting school again. About JIRA 2440, I updated it with a
> .tar.gz file because when I deleted modules from my trunk and did an
> 'svn commit' I was asked for credentials for the Tuscany svn
> repository. Probably I should set it to do things locally, but
> didn't do so because was in a hurry to provide a last update for GSoC.
>
> The .tar.gz file contains the set of reduced modules with the
> changes Adriano had done to the code, disabling API code and such.
> In other words, it's a subset of revision 643746 (the one in which
> Adriano was originally working on). Please let me know if you would
> like me to make a patch out of it instead of deleting the modules in
> the trunk. The advantage of having this reduced set of modules is
> that they are sufficient to run calculator2, and as such provide a
> lightweight Tuscany runtime for Android.
>
Oscar, only committers can commit that's why you weren't able to do it
:) However svn rm is just local and does not require you to commit anything.
When you say r643746, I'm assuming that you're talking about r643746 of
trunk?
If you provide a patch with the changes from r643746 and the list of
modules to not pick from trunk I can help apply that to the sca-android
branch.
I think we should merge these changes with other changes you have
Adriano, on top of r643746.
Do you guys think that it's worth trying to apply the patches to a more
recent revision, as r643746 is from April?
Does that make sense? Please let me know if I've mis-understood
anything. Thanks!
--
Jean-Sebastien
Re: A branch for the android port, was: [GSoC] Final student update
Posted by Adriano Crestani <ad...@apache.org>.
Hi Oscar,
Why you tried to commit? You will not be able to do it. I think you were
probably wanting to apply a patch on your trunk or something like that, was
it?
Regards,
Adriano Crestani
On Tue, Sep 9, 2008 at 2:09 PM, Oscar Castaneda <
oscar.castaneda.gsoc@gmail.com> wrote:
> Hi Guys,
>
> Sorry for the delay in responding, I went on vacation for a few days before
> starting school again. About JIRA 2440, I updated it with a .tar.gz file
> because when I deleted modules from my trunk and did an 'svn commit' I was
> asked for credentials for the Tuscany svn repository. Probably I should set
> it to do things locally, but didn't do so because was in a hurry to provide
> a last update for GSoC.
>
> The .tar.gz file contains the set of reduced modules with the changes
> Adriano had done to the code, disabling API code and such. In other words,
> it's a subset of revision 643746 (the one in which Adriano was originally
> working on). Please let me know if you would like me to make a patch out of
> it instead of deleting the modules in the trunk. The advantage of having
> this reduced set of modules is that they are sufficient to run calculator2,
> and as such provide a lightweight Tuscany runtime for Android.
>
>
> On Tue, Sep 9, 2008 at 10:17 PM, Jean-Sebastien Delfino <
> jsdelfino@gmail.com> wrote:
>
>> Jean-Sebastien Delfino wrote:
>>
>>> Adriano Crestani wrote:
>>>
>>>> Hi Jean,
>>>>
>>>> Thanks for helping us to get the poms working. I'm actually waiting a
>>>> response from Oscar about the JIRA 2440 to apply his last modifications. He
>>>> was working on the poms too, I'm not sure, but the patch he will provide to
>>>> us might contain some poms modifications.
>>>>
>>>> What about wait some couple of days to check if he replies with a
>>>> patch?! : )
>>>>
>>>>
>>> OK great, I'm going to wait for his patch before making changes to the
>>> existing poms to avoid causing any merge conflicts.
>>>
>>> Any patches should be easy to apply to the branch as it's just a straight
>>> URL to URL svn copy of sandbox/mobile-android.
>>>
>>> So, for now I'll try to build the modules by hand and will just add a new
>>> distribution module which will use the Maven assembly plugin to assemble the
>>> pieces in a Zip.
>>>
>>>
>> I've moved the modules under a modules directory and created distribution,
>> samples, and itest directories as having the same structure as trunk in the
>> branch will help sync up with it.
>>
>> Still no changes to the modules poms so any patches will still apply
>> without conflicts.
>>
>>
>> Hope this helps.
>> --
>> Jean-Sebastien
>>
>
>
>
> --
> best,
> -oscar
>
> Oscar Castañeda
>
Re: A branch for the android port, was: [GSoC] Final student update
Posted by Oscar Castaneda <os...@gmail.com>.
Hi Guys,
Sorry for the delay in responding, I went on vacation for a few days before
starting school again. About JIRA 2440, I updated it with a .tar.gz file
because when I deleted modules from my trunk and did an 'svn commit' I was
asked for credentials for the Tuscany svn repository. Probably I should set
it to do things locally, but didn't do so because was in a hurry to provide
a last update for GSoC.
The .tar.gz file contains the set of reduced modules with the changes
Adriano had done to the code, disabling API code and such. In other words,
it's a subset of revision 643746 (the one in which Adriano was originally
working on). Please let me know if you would like me to make a patch out of
it instead of deleting the modules in the trunk. The advantage of having
this reduced set of modules is that they are sufficient to run calculator2,
and as such provide a lightweight Tuscany runtime for Android.
On Tue, Sep 9, 2008 at 10:17 PM, Jean-Sebastien Delfino <jsdelfino@gmail.com
> wrote:
> Jean-Sebastien Delfino wrote:
>
>> Adriano Crestani wrote:
>>
>>> Hi Jean,
>>>
>>> Thanks for helping us to get the poms working. I'm actually waiting a
>>> response from Oscar about the JIRA 2440 to apply his last modifications. He
>>> was working on the poms too, I'm not sure, but the patch he will provide to
>>> us might contain some poms modifications.
>>>
>>> What about wait some couple of days to check if he replies with a patch?!
>>> : )
>>>
>>>
>> OK great, I'm going to wait for his patch before making changes to the
>> existing poms to avoid causing any merge conflicts.
>>
>> Any patches should be easy to apply to the branch as it's just a straight
>> URL to URL svn copy of sandbox/mobile-android.
>>
>> So, for now I'll try to build the modules by hand and will just add a new
>> distribution module which will use the Maven assembly plugin to assemble the
>> pieces in a Zip.
>>
>>
> I've moved the modules under a modules directory and created distribution,
> samples, and itest directories as having the same structure as trunk in the
> branch will help sync up with it.
>
> Still no changes to the modules poms so any patches will still apply
> without conflicts.
>
>
> Hope this helps.
> --
> Jean-Sebastien
>
--
best,
-oscar
Oscar Castañeda
Re: A branch for the android port, was: [GSoC] Final student update
Posted by Jean-Sebastien Delfino <js...@gmail.com>.
Jean-Sebastien Delfino wrote:
> Adriano Crestani wrote:
>> Hi Jean,
>>
>> Thanks for helping us to get the poms working. I'm actually waiting a
>> response from Oscar about the JIRA 2440 to apply his last
>> modifications. He was working on the poms too, I'm not sure, but the
>> patch he will provide to us might contain some poms modifications.
>>
>> What about wait some couple of days to check if he replies with a
>> patch?! : )
>>
>
> OK great, I'm going to wait for his patch before making changes to the
> existing poms to avoid causing any merge conflicts.
>
> Any patches should be easy to apply to the branch as it's just a
> straight URL to URL svn copy of sandbox/mobile-android.
>
> So, for now I'll try to build the modules by hand and will just add a
> new distribution module which will use the Maven assembly plugin to
> assemble the pieces in a Zip.
>
I've moved the modules under a modules directory and created
distribution, samples, and itest directories as having the same
structure as trunk in the branch will help sync up with it.
Still no changes to the modules poms so any patches will still apply
without conflicts.
Hope this helps.
--
Jean-Sebastien
Re: A branch for the android port, was: [GSoC] Final student update
Posted by Jean-Sebastien Delfino <js...@gmail.com>.
Adriano Crestani wrote:
> Hi Jean,
>
> Thanks for helping us to get the poms working. I'm actually waiting a
> response from Oscar about the JIRA 2440 to apply his last modifications.
> He was working on the poms too, I'm not sure, but the patch he will
> provide to us might contain some poms modifications.
>
> What about wait some couple of days to check if he replies with a
> patch?! : )
>
OK great, I'm going to wait for his patch before making changes to the
existing poms to avoid causing any merge conflicts.
Any patches should be easy to apply to the branch as it's just a
straight URL to URL svn copy of sandbox/mobile-android.
So, for now I'll try to build the modules by hand and will just add a
new distribution module which will use the Maven assembly plugin to
assemble the pieces in a Zip.
--
Jean-Sebastien
Re: A branch for the android port, was: [GSoC] Final student update
Posted by Adriano Crestani <ad...@apache.org>.
Hi Jean,
Thanks for helping us to get the poms working. I'm actually waiting a
response from Oscar about the JIRA 2440 to apply his last modifications. He
was working on the poms too, I'm not sure, but the patch he will provide to
us might contain some poms modifications.
What about wait some couple of days to check if he replies with a patch?! :
)
Regards,
Adriano Crestani
On Mon, Sep 8, 2008 at 7:50 PM, Jean-Sebastien Delfino <jsdelfino@apache.org
> wrote:
> Mike Edwards wrote:
>
>> Jean-Sebastien,
>>
>> I think that it is a good idea to get the Android stuff more mainstream.
>>
>> One question that I have is about making the build of the Android branch
>> run on Continuum. I assume that the idea is that the build can take place
>> there along with some testing that assures that the build works on the
>> Android platform - is that right?
>>
>> Yours, Mike.
>>
>>
> At least verify that the code compiles and run enough unit tests to verify
> its health.
>
> I've just created the android port branch [1]. Next I'll try to help put
> together a multi-module pom, write poms for the modules that don't have poms
> and get the existing poms working.
>
> [1] http://svn.apache.org/repos/asf/tuscany/branches/sca-android/
>
> Hope this helps.
> --
> Jean-Sebastien
>
Re: A branch for the android port, was: [GSoC] Final student update
Posted by Jean-Sebastien Delfino <js...@apache.org>.
Mike Edwards wrote:
> Jean-Sebastien,
>
> I think that it is a good idea to get the Android stuff more mainstream.
>
> One question that I have is about making the build of the Android branch
> run on Continuum. I assume that the idea is that the build can take
> place there along with some testing that assures that the build works on
> the Android platform - is that right?
>
> Yours, Mike.
>
At least verify that the code compiles and run enough unit tests to
verify its health.
I've just created the android port branch [1]. Next I'll try to help put
together a multi-module pom, write poms for the modules that don't have
poms and get the existing poms working.
[1] http://svn.apache.org/repos/asf/tuscany/branches/sca-android/
Hope this helps.
--
Jean-Sebastien
Re: A branch for the android port, was: [GSoC] Final student update
Posted by Mike Edwards <mi...@gmail.com>.
Jean-Sebastien,
I think that it is a good idea to get the Android stuff more mainstream.
One question that I have is about making the build of the Android branch run on Continuum. I assume
that the idea is that the build can take place there along with some testing that assures that the
build works on the Android platform - is that right?
Yours, Mike.
Jean-Sebastien Delfino wrote:
> Jean-Sebastien Delfino wrote:
> ...
>> Thanks Oscar, I think the Android port looks very promising, you guys
>> have done a fantastic job this summer and I hope you can continue to
>> work on it.
>>
>> BTW the code is in a sandbox now, I think it would be great to move it
>> to trunk (not in the main daily build I guess, but still under a
>> folder in trunk) or maybe in a branch like branches/android or
>> something like that, as it'll make clear to all that the code is alive
>> and that people in the community can work on it.
> ...
>
> No comments or objections so I'd like to do the following:
>
> 1) copy the modules in sandbox to a branches/android branch
>
> 2) contribute additional poms for the modules that don't have ones
>
> 3) bring up an end to end build, and maybe try to have it run on
> continuum too?
>
> 4) also later try to compare the modules that are both in trunk and the
> port and see if they can benefit from recent trunk fixes.
>
> I'm hoping that will help the community see the nice job that you've
> done with the port and enable people to build it, try it and come on
> board to help develop it.
>
> I'll start with 1/2/3 in the next days.
A branch for the android port, was: [GSoC] Final student update
Posted by Jean-Sebastien Delfino <js...@gmail.com>.
Jean-Sebastien Delfino wrote:
...
> Thanks Oscar, I think the Android port looks very promising, you guys
> have done a fantastic job this summer and I hope you can continue to
> work on it.
>
> BTW the code is in a sandbox now, I think it would be great to move it
> to trunk (not in the main daily build I guess, but still under a folder
> in trunk) or maybe in a branch like branches/android or something like
> that, as it'll make clear to all that the code is alive and that people
> in the community can work on it.
...
No comments or objections so I'd like to do the following:
1) copy the modules in sandbox to a branches/android branch
2) contribute additional poms for the modules that don't have ones
3) bring up an end to end build, and maybe try to have it run on
continuum too?
4) also later try to compare the modules that are both in trunk and the
port and see if they can benefit from recent trunk fixes.
I'm hoping that will help the community see the nice job that you've
done with the port and enable people to build it, try it and come on
board to help develop it.
I'll start with 1/2/3 in the next days.
--
Jean-Sebastien
Re: [GSoC] Final student update
Posted by Jean-Sebastien Delfino <js...@apache.org>.
Simon Laws wrote:
>
>
> On Thu, Aug 28, 2008 at 4:50 PM, Oscar Castaneda
> <oscar.castaneda.gsoc@gmail.com <ma...@gmail.com>>
> wrote:
>
> Dear Mentors,
>
> Since the last update I sent Google has released an updated version
> of the Android SDK. Unfortunately for my GSoC project such release
> came too late. On a more positive note, I will continue helping
> Adriano with the porting efforts. We have begun testing the new SDK
> with the Android trunk and can hopefully see Tuscany running on
> Android soon.
>
> The last tests performed as part of my GSoC project were to remove
> all @Remotable annotations and to test calculator-android with a
> componentType side file. Both of these initiatives failed because of
> the fact the previous version of the Android SDK still choked when
> encountering any code with annotations. As a new SDK was released I
> stopped further testing. Since then I have updated the JIRA ticket I
> created, with the reduced set of modules and changes that Adriano
> made to SCA code. Additionally, I updated my project wiki to
> summarize all of the work performed throughout the project.
>
> I have now started testing the new SDK and currently am running into
> problems with trying to include a core library from a desktop
> virtual machine into an Android application, this being
> "javax/xml/namespace/QName$1.class". I will continue these efforts
> outside the scope of GSoC.
>
> Lastly, I would like to thank the Tuscany community; I received help
> from several committers throughout the project, and of course, from
> my mentor Adriano. I thank you all for your patience and continued
> support. Despite the fact that the project wasn't a complete success
> I definitely learned a lot and was able to explore my interests. I
> am now continuing my studies in Computer Science and look forward to
> continue learning from and collaborating on this project.
>
> --
> best,
> -oscar
>
> Oscar Castañeda
>
>
> Thanks for the update Oscar. Was great having you involved in Tuscany
> this summer and good to hear you're thinking of sticking around.
>
> Simon
Thanks Oscar, I think the Android port looks very promising, you guys
have done a fantastic job this summer and I hope you can continue to
work on it.
BTW the code is in a sandbox now, I think it would be great to move it
to trunk (not in the main daily build I guess, but still under a folder
in trunk) or maybe in a branch like branches/android or something like
that, as it'll make clear to all that the code is alive and that people
in the community can work on it.
Also you said "I definitely learned a lot and was able to explore my
interests", that sounds like a great success to me! :)
I'm really happy to hear that you're going to stick around too, it's
really good to have you on the project!
--
Jean-Sebastien
Re: [GSoC] Final student update
Posted by Simon Laws <si...@googlemail.com>.
On Thu, Aug 28, 2008 at 4:50 PM, Oscar Castaneda <
oscar.castaneda.gsoc@gmail.com> wrote:
> Dear Mentors,
>
> Since the last update I sent Google has released an updated version of the
> Android SDK. Unfortunately for my GSoC project such release came too late.
> On a more positive note, I will continue helping Adriano with the porting
> efforts. We have begun testing the new SDK with the Android trunk and can
> hopefully see Tuscany running on Android soon.
>
> The last tests performed as part of my GSoC project were to remove all
> @Remotable annotations and to test calculator-android with a componentType
> side file. Both of these initiatives failed because of the fact the previous
> version of the Android SDK still choked when encountering any code with
> annotations. As a new SDK was released I stopped further testing. Since then
> I have updated the JIRA ticket I created, with the reduced set of modules
> and changes that Adriano made to SCA code. Additionally, I updated my
> project wiki to summarize all of the work performed throughout the project.
>
> I have now started testing the new SDK and currently am running into
> problems with trying to include a core library from a desktop
> virtual machine into an Android application, this being
> "javax/xml/namespace/QName$1.class". I will continue these efforts outside
> the scope of GSoC.
>
> Lastly, I would like to thank the Tuscany community; I received help from
> several committers throughout the project, and of course, from my mentor
> Adriano. I thank you all for your patience and continued support. Despite
> the fact that the project wasn't a complete success I definitely learned a
> lot and was able to explore my interests. I am now continuing my studies in
> Computer Science and look forward to continue learning from and
> collaborating on this project.
>
> --
> best,
> -oscar
>
> Oscar Castañeda
>
Thanks for the update Oscar. Was great having you involved in Tuscany this
summer and good to hear you're thinking of sticking around.
Simon