Moving forward our frontend stack
Hello oVirt developers! I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1). This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1 Since GWT 2.6 features support for Java 7 it was possible to increment language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend). To facilitate emitting bytecode compatible with Java 7 also upgrade of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features even support for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the future. Most of the changes in the patch are due to upgrade of GWT-P - i.e. changing packages of TokenFormatter and PlaceRequest. Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much less risky. I have tested draft-compile, debug-mode and also tried to use the resulting application manually for some time. So far everything worked (surprisingly well!) and I have not detected any defects. Of course I invite anyone to test this patch on his own since it is and upgrade of our core infrastructure. That having said I think is comparatively simple and the benefits outweigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cycle. Reviews, comments and testing are very welcome :-) Best regards, Martin [1] http://gerrit.ovirt.org/#/c/32135/
Hey Martin! I've just reviewed your patch, looks good overall. Here are my thoughts: * Jenkins CI fails on patch due to missing dependencies, we need to provide following new dependencies in order to proceed with the upgrade: org.aspectj:aspectjweaver:1.8.2 org.aspectj:aspectjrt:1.8.2 com.google.gwt:gwt-user:2.6.1 com.google.gwt:gwt-dev:2.6.1 com.google.gwt:gwt-servlet:2.6.1 com.google.gwt:2.6.1 org.codehaus.mojo:gwt-maven-plugin:2.6.1 com.gwtplatform:gwtp-processors:1.3.1 com.gwtplatform:gwtp-mvp-client:1.3.1 com.google.gwt.inject:gin:2.1.2 Who will take charge of that? * patch itself looks quite harmless (not too risky) * consolidating Java source & target version across frontend and backend is nice! -> this means we could use Java 7 features also on the frontend (Java/GWT) side * TODO-GWT tags [1] proved to be helpful [2] -> reminder to all UI maintainers to use TODO-GWT tags whenever we have some GWT(P) workaround, so that the future upgrade will be safer w.r.t. existing code [1] https://www.mail-archive.com/devel@ovirt.org/msg00761.html [2] http://gerrit.ovirt.org/#/c/32135/1/frontend/webadmin/modules/webadmin/src/m... Regards, Vojtech PS: I went through GWT-Platform release notes, found an interesting new feature in recent GWTP release, worth investigating: https://github.com/ArcBees/GWTP/wiki/Release-Notes #346 : Map more than one name token to presenter ----- Original Message -----
From: "Martin Betak" <mbetak@redhat.com> To: devel@ovirt.org Cc: "Vojtech Szocs" <vszocs@redhat.com>, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 4:11:10 PM Subject: Moving forward our frontend stack
Hello oVirt developers!
I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1).
This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1
Since GWT 2.6 features support for Java 7 it was possible to increment language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend).
To facilitate emitting bytecode compatible with Java 7 also upgrade of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features even support for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the future.
Most of the changes in the patch are due to upgrade of GWT-P - i.e. changing packages of TokenFormatter and PlaceRequest.
Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much less risky.
I have tested draft-compile, debug-mode and also tried to use the resulting application manually for some time. So far everything worked (surprisingly well!) and I have not detected any defects. Of course I invite anyone to test this patch on his own since it is and upgrade of our core infrastructure.
That having said I think is comparatively simple and the benefits outweigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cycle.
Reviews, comments and testing are very welcome :-)
Best regards,
Martin
Does this mean we can use the diamond operator at common after the patch is merged? For example: HashSet<String> names = new HashSet<>(); Insead of HashSet<String> names = new HashSet<String>(); Yair ----- Original Message -----
From: "Vojtech Szocs" <vszocs@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: devel@ovirt.org Sent: Thursday, August 28, 2014 6:38:04 PM Subject: Re: [ovirt-devel] Moving forward our frontend stack
Hey Martin!
I've just reviewed your patch, looks good overall.
Here are my thoughts:
* Jenkins CI fails on patch due to missing dependencies, we need to provide following new dependencies in order to proceed with the upgrade:
org.aspectj:aspectjweaver:1.8.2 org.aspectj:aspectjrt:1.8.2 com.google.gwt:gwt-user:2.6.1 com.google.gwt:gwt-dev:2.6.1 com.google.gwt:gwt-servlet:2.6.1 com.google.gwt:2.6.1 org.codehaus.mojo:gwt-maven-plugin:2.6.1 com.gwtplatform:gwtp-processors:1.3.1 com.gwtplatform:gwtp-mvp-client:1.3.1 com.google.gwt.inject:gin:2.1.2
Who will take charge of that?
* patch itself looks quite harmless (not too risky)
* consolidating Java source & target version across frontend and backend is nice!
-> this means we could use Java 7 features also on the frontend (Java/GWT) side
* TODO-GWT tags [1] proved to be helpful [2]
-> reminder to all UI maintainers to use TODO-GWT tags whenever we have some GWT(P) workaround, so that the future upgrade will be safer w.r.t. existing code
[1] https://www.mail-archive.com/devel@ovirt.org/msg00761.html [2] http://gerrit.ovirt.org/#/c/32135/1/frontend/webadmin/modules/webadmin/src/m...
Regards, Vojtech
PS: I went through GWT-Platform release notes, found an interesting new feature in recent GWTP release, worth investigating:
https://github.com/ArcBees/GWTP/wiki/Release-Notes #346 : Map more than one name token to presenter
----- Original Message -----
From: "Martin Betak" <mbetak@redhat.com> To: devel@ovirt.org Cc: "Vojtech Szocs" <vszocs@redhat.com>, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 4:11:10 PM Subject: Moving forward our frontend stack
Hello oVirt developers!
I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1).
This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1
Since GWT 2.6 features support for Java 7 it was possible to increment language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend).
To facilitate emitting bytecode compatible with Java 7 also upgrade of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features even support for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the future.
Most of the changes in the patch are due to upgrade of GWT-P - i.e. changing packages of TokenFormatter and PlaceRequest.
Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much less risky.
I have tested draft-compile, debug-mode and also tried to use the resulting application manually for some time. So far everything worked (surprisingly well!) and I have not detected any defects. Of course I invite anyone to test this patch on his own since it is and upgrade of our core infrastructure.
That having said I think is comparatively simple and the benefits outweigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cycle.
Reviews, comments and testing are very welcome :-)
Best regards,
Martin
_______________________________________________ Devel mailing list Devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/devel
----- Original Message -----
From: "Yair Zaslavsky" <yzaslavs@redhat.com> To: "Vojtech Szocs" <vszocs@redhat.com> Cc: devel@ovirt.org Sent: Friday, August 29, 2014 1:58:58 AM Subject: Re: [ovirt-devel] Moving forward our frontend stack
Does this mean we can use the diamond operator at common after the patch is merged?
For example:
HashSet<String> names = new HashSet<>();
Insead of
HashSet<String> names = new HashSet<String>();
yes, we can! :)
Yair
----- Original Message -----
From: "Vojtech Szocs" <vszocs@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: devel@ovirt.org Sent: Thursday, August 28, 2014 6:38:04 PM Subject: Re: [ovirt-devel] Moving forward our frontend stack
Hey Martin!
I've just reviewed your patch, looks good overall.
Here are my thoughts:
* Jenkins CI fails on patch due to missing dependencies, we need to provide following new dependencies in order to proceed with the upgrade:
org.aspectj:aspectjweaver:1.8.2 org.aspectj:aspectjrt:1.8.2 com.google.gwt:gwt-user:2.6.1 com.google.gwt:gwt-dev:2.6.1 com.google.gwt:gwt-servlet:2.6.1 com.google.gwt:2.6.1 org.codehaus.mojo:gwt-maven-plugin:2.6.1 com.gwtplatform:gwtp-processors:1.3.1 com.gwtplatform:gwtp-mvp-client:1.3.1 com.google.gwt.inject:gin:2.1.2
Who will take charge of that?
* patch itself looks quite harmless (not too risky)
* consolidating Java source & target version across frontend and backend is nice!
-> this means we could use Java 7 features also on the frontend (Java/GWT) side
* TODO-GWT tags [1] proved to be helpful [2]
-> reminder to all UI maintainers to use TODO-GWT tags whenever we have some GWT(P) workaround, so that the future upgrade will be safer w.r.t. existing code
[1] https://www.mail-archive.com/devel@ovirt.org/msg00761.html [2] http://gerrit.ovirt.org/#/c/32135/1/frontend/webadmin/modules/webadmin/src/m...
Regards, Vojtech
PS: I went through GWT-Platform release notes, found an interesting new feature in recent GWTP release, worth investigating:
https://github.com/ArcBees/GWTP/wiki/Release-Notes #346 : Map more than one name token to presenter
----- Original Message -----
From: "Martin Betak" <mbetak@redhat.com> To: devel@ovirt.org Cc: "Vojtech Szocs" <vszocs@redhat.com>, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 4:11:10 PM Subject: Moving forward our frontend stack
Hello oVirt developers!
I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1).
This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1
Since GWT 2.6 features support for Java 7 it was possible to increment language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend).
To facilitate emitting bytecode compatible with Java 7 also upgrade of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features even support for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the future.
Most of the changes in the patch are due to upgrade of GWT-P - i.e. changing packages of TokenFormatter and PlaceRequest.
Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much less risky.
I have tested draft-compile, debug-mode and also tried to use the resulting application manually for some time. So far everything worked (surprisingly well!) and I have not detected any defects. Of course I invite anyone to test this patch on his own since it is and upgrade of our core infrastructure.
That having said I think is comparatively simple and the benefits outweigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cycle.
Reviews, comments and testing are very welcome :-)
Best regards,
Martin
_______________________________________________ Devel mailing list Devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/devel
_______________________________________________ Devel mailing list Devel@ovirt.org http://lists.ovirt.org/mailman/listinfo/devel
----- Original Message -----
From: "Vojtech Szocs" <vszocs@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: devel@ovirt.org, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 5:38:04 PM Subject: Re: Moving forward our frontend stack
Hey Martin!
I've just reviewed your patch, looks good overall.
Here are my thoughts:
* Jenkins CI fails on patch due to missing dependencies, we need to provide following new dependencies in order to proceed with the upgrade:
org.aspectj:aspectjweaver:1.8.2 org.aspectj:aspectjrt:1.8.2 com.google.gwt:gwt-user:2.6.1 com.google.gwt:gwt-dev:2.6.1 com.google.gwt:gwt-servlet:2.6.1 com.google.gwt:2.6.1 org.codehaus.mojo:gwt-maven-plugin:2.6.1 com.gwtplatform:gwtp-processors:1.3.1 com.gwtplatform:gwtp-mvp-client:1.3.1 com.google.gwt.inject:gin:2.1.2
Who will take charge of that?
I already spoke to David Caro about the jenkins failure of patch http://gerrit.ovirt.org/#/c/32135/. It seems the files for appropriate versions are there but the files are corrupted. Let us hope this can be resolved easier than Node.js :-) Best regards, Martin
* patch itself looks quite harmless (not too risky)
* consolidating Java source & target version across frontend and backend is nice!
-> this means we could use Java 7 features also on the frontend (Java/GWT) side
* TODO-GWT tags [1] proved to be helpful [2]
-> reminder to all UI maintainers to use TODO-GWT tags whenever we have some GWT(P) workaround, so that the future upgrade will be safer w.r.t. existing code
[1] https://www.mail-archive.com/devel@ovirt.org/msg00761.html [2] http://gerrit.ovirt.org/#/c/32135/1/frontend/webadmin/modules/webadmin/src/m...
Regards, Vojtech
PS: I went through GWT-Platform release notes, found an interesting new feature in recent GWTP release, worth investigating:
https://github.com/ArcBees/GWTP/wiki/Release-Notes #346 : Map more than one name token to presenter
----- Original Message -----
From: "Martin Betak" <mbetak@redhat.com> To: devel@ovirt.org Cc: "Vojtech Szocs" <vszocs@redhat.com>, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 4:11:10 PM Subject: Moving forward our frontend stack
Hello oVirt developers!
I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1).
This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1
Since GWT 2.6 features support for Java 7 it was possible to increment language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend).
To facilitate emitting bytecode compatible with Java 7 also upgrade of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features even support for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the future.
Most of the changes in the patch are due to upgrade of GWT-P - i.e. changing packages of TokenFormatter and PlaceRequest.
Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much less risky.
I have tested draft-compile, debug-mode and also tried to use the resulting application manually for some time. So far everything worked (surprisingly well!) and I have not detected any defects. Of course I invite anyone to test this patch on his own since it is and upgrade of our core infrastructure.
That having said I think is comparatively simple and the benefits outweigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cycle.
Reviews, comments and testing are very welcome :-)
Best regards,
Martin
--lCAWRPmW1mITcIfM Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On 08/29, Martin Betak wrote:
From: "Vojtech Szocs" <vszocs@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: devel@ovirt.org, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels= " <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Ver= nia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 5:38:04 PM Subject: Re: Moving forward our frontend stack =20 Hey Martin! =20 I've just reviewed your patch, looks good overall. =20 Here are my thoughts: =20 * Jenkins CI fails on patch due to missing dependencies, we need to provide following new dependencies in order to proceed with the upgrade: =20 org.aspectj:aspectjweaver:1.8.2 org.aspectj:aspectjrt:1.8.2 com.google.gwt:gwt-user:2.6.1 com.google.gwt:gwt-dev:2.6.1 com.google.gwt:gwt-servlet:2.6.1 com.google.gwt:2.6.1 org.codehaus.mojo:gwt-maven-plugin:2.6.1 com.gwtplatform:gwtp-processors:1.3.1 com.gwtplatform:gwtp-mvp-client:1.3.1 com.google.gwt.inject:gin:2.1.2 =20 Who will take charge of that? =20 I already spoke to David Caro about the jenkins failure of=20
----- Original Message ----- patch http://gerrit.ovirt.org/#/c/32135/. It seems the files for appropriate versions are there but the files are corrupted. =20 Let us hope this can be resolved easier than Node.js :-)
=20 Best regards, =20 Martin =20
=20 * patch itself looks quite harmless (not too risky) =20 * consolidating Java source & target version across frontend and backend is nice! =20 -> this means we could use Java 7 features also on the frontend (Java/GWT) side =20 * TODO-GWT tags [1] proved to be helpful [2] =20 -> reminder to all UI maintainers to use TODO-GWT tags whenever we have some GWT(P) workaround, so that the future upgrade will be safer w.r.t. existing code =20 [1] https://www.mail-archive.com/devel@ovirt.org/msg00761.html [2] http://gerrit.ovirt.org/#/c/32135/1/frontend/webadmin/modules/webadmin/= src/main/java/org/ovirt/engine/ui/webadmin/section/main/view/SearchPanelVie= w.java =20 Regards, Vojtech =20 PS: I went through GWT-Platform release notes, found an interesting new feature in recent GWTP release, worth investigating: =20 https://github.com/ArcBees/GWTP/wiki/Release-Notes #346 : Map more than one name token to presenter =20 =20 ----- Original Message -----
From: "Martin Betak" <mbetak@redhat.com> To: devel@ovirt.org Cc: "Vojtech Szocs" <vszocs@redhat.com>, "Einav Cohen" <ecohen@redhat= =2Ecom>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 4:11:10 PM Subject: Moving forward our frontend stack =20 Hello oVirt developers! =20 I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1). =20 This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1 =20 Since GWT 2.6 features support for Java 7 it was possible to increment language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend). =20 To facilitate emitting bytecode compatible with Java 7 also upgrade of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features even s= upport for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the fu= ture. =20 Most of the changes in the patch are due to upgrade of GWT-P - i.e. changing packages of TokenFormatter and PlaceRequest. =20 Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much less = risky. =20 I have tested draft-compile, debug-mode and also tried to use the res=
It seems that the pom on jboss repository has a checksum that does not match the ones there. http://repository.jboss.org/nexus/content/groups/public-jboss/com/gwtplat= form/gwtp/1.3.1/ If you get the gwtp-1.3.1.pom file, you get the hashed: MD5: 9acbb4e5088825d31e2c8307ae679243 SHA1: eb6e92c012926fb265cf4f359c3d5adca8c5c0f9 While the ones there are: MD5: c33a1cd1ffa88aa4cec42e03364b0e0c SHA1: ca6c0155356765fc23f0a557373193f013adfd6c Any of you have any contact with the guys that maintain that repo (or know about who can know about?) I seem to be unable to find any contant in the repo itself... Anyhow, I've disabled the checksum checking and just pass through, so it should not complain about unable to download it (it might fail due to the checksum, it should at least xd). ulting
application manually for some time. So far everything worked (surprisingly well!)= and I have not detected any defects. Of course I invite anyone to test this patch on= his own since it is and upgrade of our core infrastructure. =20 That having said I think is comparatively simple and the benefits out= weigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cyc= le. =20 Reviews, comments and testing are very welcome :-) =20 Best regards, =20 Martin =20 [1] http://gerrit.ovirt.org/#/c/32135/ =20 =20
--=20 David Caro Red Hat S.L. Continuous Integration Engineer - EMEA ENG Virtualization R&D Tel.: +420 532 294 605 Email: dcaro@redhat.com Web: www.redhat.com RHT Global #: 82-62605 --lCAWRPmW1mITcIfM Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJUAEbFAAoJEEBxx+HSYmnDU4YH/isNEML5u8LAkxH56INs4STV jYzan8beYhIq9cxBLIy7ViGCKmTM7W7JqL7SB11a8sNxVimPDD0RNrmT0StPlK/P +s7NFw0oUh7dFvJOv7prmtr7dBGI7tMPHdCyhzRQHy4mi7R6aSrZrSXst57v04Y4 7wH5QzEcSFeaXIHvx0krbItw30Gps2micEG4Szqvqll7a4UyYL15T6ouxv099liB 78J14dViAYtdyEUqeY+PnJ+/DxgiyBXGT6JlHrJKelP9j9w4Vu311vfBzhc8JlGI 9Kbh3WwSELml9OruiiIX0eWVQRzakWHBqSf1d7gfeRgnqRCcpCG44b6Wvyrifkg= =b8n4 -----END PGP SIGNATURE----- --lCAWRPmW1mITcIfM--
----- Original Message -----
From: "David Caro" <dcaroest@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: "Vojtech Szocs" <vszocs@redhat.com>, devel@ovirt.org, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Friday, August 29, 2014 11:24:21 AM Subject: Re: Moving forward our frontend stack
On 08/29, Martin Betak wrote:
----- Original Message -----
From: "Vojtech Szocs" <vszocs@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: devel@ovirt.org, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 5:38:04 PM Subject: Re: Moving forward our frontend stack
Hey Martin!
I've just reviewed your patch, looks good overall.
Here are my thoughts:
* Jenkins CI fails on patch due to missing dependencies, we need to provide following new dependencies in order to proceed with the upgrade:
org.aspectj:aspectjweaver:1.8.2 org.aspectj:aspectjrt:1.8.2 com.google.gwt:gwt-user:2.6.1 com.google.gwt:gwt-dev:2.6.1 com.google.gwt:gwt-servlet:2.6.1 com.google.gwt:2.6.1 org.codehaus.mojo:gwt-maven-plugin:2.6.1 com.gwtplatform:gwtp-processors:1.3.1 com.gwtplatform:gwtp-mvp-client:1.3.1 com.google.gwt.inject:gin:2.1.2
Who will take charge of that?
I already spoke to David Caro about the jenkins failure of patch http://gerrit.ovirt.org/#/c/32135/. It seems the files for appropriate versions are there but the files are corrupted.
Let us hope this can be resolved easier than Node.js :-)
It seems that the pom on jboss repository has a checksum that does not match the ones there.
http://repository.jboss.org/nexus/content/groups/public-jboss/com/gwtplatfor...
If you get the gwtp-1.3.1.pom file, you get the hashed:
MD5: 9acbb4e5088825d31e2c8307ae679243 SHA1: eb6e92c012926fb265cf4f359c3d5adca8c5c0f9
While the ones there are:
MD5: c33a1cd1ffa88aa4cec42e03364b0e0c SHA1: ca6c0155356765fc23f0a557373193f013adfd6c
Any of you have any contact with the guys that maintain that repo (or know about who can know about?)
I seem to be unable to find any contant in the repo itself...
Anyhow, I've disabled the checksum checking and just pass through, so it should not complain about unable to download it (it might fail due to the checksum, it should at least xd).
Sorry, I've got no idea who to contact about that repo :( I'd rather avoid disabling checksum checking for our builds, bad checksums should be fixed in that repo.
Best regards,
Martin
* patch itself looks quite harmless (not too risky)
* consolidating Java source & target version across frontend and backend is nice!
-> this means we could use Java 7 features also on the frontend (Java/GWT) side
* TODO-GWT tags [1] proved to be helpful [2]
-> reminder to all UI maintainers to use TODO-GWT tags whenever we have some GWT(P) workaround, so that the future upgrade will be safer w.r.t. existing code
[1] https://www.mail-archive.com/devel@ovirt.org/msg00761.html [2] http://gerrit.ovirt.org/#/c/32135/1/frontend/webadmin/modules/webadmin/src/m...
Regards, Vojtech
PS: I went through GWT-Platform release notes, found an interesting new feature in recent GWTP release, worth investigating:
https://github.com/ArcBees/GWTP/wiki/Release-Notes #346 : Map more than one name token to presenter
----- Original Message -----
From: "Martin Betak" <mbetak@redhat.com> To: devel@ovirt.org Cc: "Vojtech Szocs" <vszocs@redhat.com>, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 4:11:10 PM Subject: Moving forward our frontend stack
Hello oVirt developers!
I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1).
This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1
Since GWT 2.6 features support for Java 7 it was possible to increment language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend).
To facilitate emitting bytecode compatible with Java 7 also upgrade of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features even support for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the future.
Most of the changes in the patch are due to upgrade of GWT-P - i.e. changing packages of TokenFormatter and PlaceRequest.
Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much less risky.
I have tested draft-compile, debug-mode and also tried to use the resulting application manually for some time. So far everything worked (surprisingly well!) and I have not detected any defects. Of course I invite anyone to test this patch on his own since it is and upgrade of our core infrastructure.
That having said I think is comparatively simple and the benefits outweigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cycle.
Reviews, comments and testing are very welcome :-)
Best regards,
Martin
-- David Caro
Red Hat S.L. Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605 Email: dcaro@redhat.com Web: www.redhat.com RHT Global #: 82-62605
=20 =20 ----- Original Message -----
From: "David Caro" <dcaroest@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: "Vojtech Szocs" <vszocs@redhat.com>, devel@ovirt.org, "Einav Cohen"= <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jeli= nek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Friday, August 29, 2014 11:24:21 AM Subject: Re: Moving forward our frontend stack =20 On 08/29, Martin Betak wrote:
From: "Vojtech Szocs" <vszocs@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: devel@ovirt.org, "Einav Cohen" <ecohen@redhat.com>, "Alexander = Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 5:38:04 PM Subject: Re: Moving forward our frontend stack =20 Hey Martin! =20 I've just reviewed your patch, looks good overall. =20 Here are my thoughts: =20 * Jenkins CI fails on patch due to missing dependencies, we need to provide following new dependencies in order to proceed with the upgrade: =20 org.aspectj:aspectjweaver:1.8.2 org.aspectj:aspectjrt:1.8.2 com.google.gwt:gwt-user:2.6.1 com.google.gwt:gwt-dev:2.6.1 com.google.gwt:gwt-servlet:2.6.1 com.google.gwt:2.6.1 org.codehaus.mojo:gwt-maven-plugin:2.6.1 com.gwtplatform:gwtp-processors:1.3.1 com.gwtplatform:gwtp-mvp-client:1.3.1 com.google.gwt.inject:gin:2.1.2 =20 Who will take charge of that? =20 I already spoke to David Caro about the jenkins failure of
----- Original Message ----- patch http://gerrit.ovirt.org/#/c/32135/. It seems the files for appropriate versions are there but the files are corrupted. =20 Let us hope this can be resolved easier than Node.js :-) =20 It seems that the pom on jboss repository has a checksum that does not match the ones there. =20 http://repository.jboss.org/nexus/content/groups/public-jboss/com/gwt=
--3xoW37o/FfUZJwQG Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On 08/29, Vojtech Szocs wrote: platform/gwtp/1.3.1/
=20 If you get the gwtp-1.3.1.pom file, you get the hashed: =20 MD5: 9acbb4e5088825d31e2c8307ae679243 SHA1: eb6e92c012926fb265cf4f359c3d5adca8c5c0f9 =20 While the ones there are: =20 MD5: c33a1cd1ffa88aa4cec42e03364b0e0c SHA1: ca6c0155356765fc23f0a557373193f013adfd6c =20 Any of you have any contact with the guys that maintain that repo (or know about who can know about?) =20 I seem to be unable to find any contant in the repo itself... =20 Anyhow, I've disabled the checksum checking and just pass through, so it should not complain about unable to download it (it might fail due to the checksum, it should at least xd). =20 Sorry, I've got no idea who to contact about that repo :( =20 I'd rather avoid disabling checksum checking for our builds, bad checksums should be fixed in that repo.
I've opened a ticket to servicedesk, maybe they know. In the meantime, do you prefer failing all the jobs that use that artfact?
=20
=20 =20 =20
=20 Best regards, =20 Martin =20
=20 * patch itself looks quite harmless (not too risky) =20 * consolidating Java source & target version across frontend and backend is nice! =20 -> this means we could use Java 7 features also on the frontend (Java/GWT) side =20 * TODO-GWT tags [1] proved to be helpful [2] =20 -> reminder to all UI maintainers to use TODO-GWT tags whenever we have some GWT(P) workaround, so that the future upgrade will be safer w.r.t. existing code =20 [1] https://www.mail-archive.com/devel@ovirt.org/msg00761.html [2] http://gerrit.ovirt.org/#/c/32135/1/frontend/webadmin/modules/webad= min/src/main/java/org/ovirt/engine/ui/webadmin/section/main/view/SearchPane= lView.java =20 Regards, Vojtech =20 PS: I went through GWT-Platform release notes, found an interesting new feature in recent GWTP release, worth investigating: =20 https://github.com/ArcBees/GWTP/wiki/Release-Notes #346 : Map more than one name token to presenter =20 =20 ----- Original Message -----
From: "Martin Betak" <mbetak@redhat.com> To: devel@ovirt.org Cc: "Vojtech Szocs" <vszocs@redhat.com>, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 4:11:10 PM Subject: Moving forward our frontend stack =20 Hello oVirt developers! =20 I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1). =20 This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1 =20 Since GWT 2.6 features support for Java 7 it was possible to incr= ement language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend). =20 To facilitate emitting bytecode compatible with Java 7 also upgra= de of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features ev= en support for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the future. =20 Most of the changes in the patch are due to upgrade of GWT-P - i.= e. changing packages of TokenFormatter and PlaceRequest. =20 Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much l= ess risky. =20 I have tested draft-compile, debug-mode and also tried to use the resulting application manually for some time. So far everything worked (surprisingly we= ll!) and I have not detected any defects. Of course I invite anyone to test this patc= h on his own since it is and upgrade of our core infrastructure. =20 That having said I think is comparatively simple and the benefits outweigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cycle. =20 Reviews, comments and testing are very welcome :-) =20 Best regards, =20 Martin =20 [1] http://gerrit.ovirt.org/#/c/32135/ =20 =20 =20 -- David Caro =20 Red Hat S.L. Continuous Integration Engineer - EMEA ENG Virtualization R&D =20 Tel.: +420 532 294 605 Email: dcaro@redhat.com Web: www.redhat.com RHT Global #: 82-62605 =20
--=20 David Caro Red Hat S.L. Continuous Integration Engineer - EMEA ENG Virtualization R&D Tel.: +420 532 294 605 Email: dcaro@redhat.com Web: www.redhat.com RHT Global #: 82-62605 --3xoW37o/FfUZJwQG Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJUAI7jAAoJEEBxx+HSYmnDaxsH/0/1Cy4TED0I+kn3j3Vbk8D0 f13ZL4xbVDKbCT8rOC4VB8kbBAi/r2dCJJy7kX5KkHgyIShV7vriaDbTrNEp1sYY 1Pm80JW2waD2BiVBuOpS4TBgBZ6ijcq8YQkZk1xfpjg62zj8y0Zxovy63CnLa1wV nHsYMUC8QkD/KJXZbgQbF7ydYPe2pPxu1SA6LbHtFoof/fNQhK3itg+Zh3OBr/77 hfFIWDyDteID41KYpJ5BndUXAFXmrXUw77k3qiwZvmSegPZP3ftOsVuyyhhEJi9K dqy1BpFssHrGYev8yFjtyZUujGi9QCchdrSeRdTOyxeOgs1ZLso0D9dyReUDQ8k= =8aqt -----END PGP SIGNATURE----- --3xoW37o/FfUZJwQG--
----- Original Message -----
From: "David Caro" <dcaroest@redhat.com> To: "Vojtech Szocs" <vszocs@redhat.com> Cc: "Martin Betak" <mbetak@redhat.com>, devel@ovirt.org, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Friday, August 29, 2014 4:32:03 PM Subject: Re: Moving forward our frontend stack
On 08/29, Vojtech Szocs wrote:
----- Original Message -----
From: "David Caro" <dcaroest@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: "Vojtech Szocs" <vszocs@redhat.com>, devel@ovirt.org, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Friday, August 29, 2014 11:24:21 AM Subject: Re: Moving forward our frontend stack
On 08/29, Martin Betak wrote:
----- Original Message -----
From: "Vojtech Szocs" <vszocs@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: devel@ovirt.org, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 5:38:04 PM Subject: Re: Moving forward our frontend stack
Hey Martin!
I've just reviewed your patch, looks good overall.
Here are my thoughts:
* Jenkins CI fails on patch due to missing dependencies, we need to provide following new dependencies in order to proceed with the upgrade:
org.aspectj:aspectjweaver:1.8.2 org.aspectj:aspectjrt:1.8.2 com.google.gwt:gwt-user:2.6.1 com.google.gwt:gwt-dev:2.6.1 com.google.gwt:gwt-servlet:2.6.1 com.google.gwt:2.6.1 org.codehaus.mojo:gwt-maven-plugin:2.6.1 com.gwtplatform:gwtp-processors:1.3.1 com.gwtplatform:gwtp-mvp-client:1.3.1 com.google.gwt.inject:gin:2.1.2
Who will take charge of that?
I already spoke to David Caro about the jenkins failure of patch http://gerrit.ovirt.org/#/c/32135/. It seems the files for appropriate versions are there but the files are corrupted.
Let us hope this can be resolved easier than Node.js :-)
It seems that the pom on jboss repository has a checksum that does not match the ones there.
http://repository.jboss.org/nexus/content/groups/public-jboss/com/gwtplatfor...
If you get the gwtp-1.3.1.pom file, you get the hashed:
MD5: 9acbb4e5088825d31e2c8307ae679243 SHA1: eb6e92c012926fb265cf4f359c3d5adca8c5c0f9
While the ones there are:
MD5: c33a1cd1ffa88aa4cec42e03364b0e0c SHA1: ca6c0155356765fc23f0a557373193f013adfd6c
Any of you have any contact with the guys that maintain that repo (or know about who can know about?)
I seem to be unable to find any contant in the repo itself...
Anyhow, I've disabled the checksum checking and just pass through, so it should not complain about unable to download it (it might fail due to the checksum, it should at least xd).
Sorry, I've got no idea who to contact about that repo :(
I'd rather avoid disabling checksum checking for our builds, bad checksums should be fixed in that repo.
I've opened a ticket to servicedesk, maybe they know.
Thanks!
In the meantime, do you prefer failing all the jobs that use that artfact?
Well, I meant to avoid disabling checksum checking globally. If we can selectively disable checksum checking for specific Maven artifacts like "gwtp" etc. then we could do that.. Anyway, I feel like Martin's patch needs some more testing before it's merged, so this is not super-urgent I guess :)
Best regards,
Martin
* patch itself looks quite harmless (not too risky)
* consolidating Java source & target version across frontend and backend is nice!
-> this means we could use Java 7 features also on the frontend (Java/GWT) side
* TODO-GWT tags [1] proved to be helpful [2]
-> reminder to all UI maintainers to use TODO-GWT tags whenever we have some GWT(P) workaround, so that the future upgrade will be safer w.r.t. existing code
[1] https://www.mail-archive.com/devel@ovirt.org/msg00761.html [2] http://gerrit.ovirt.org/#/c/32135/1/frontend/webadmin/modules/webadmin/src/m...
Regards, Vojtech
PS: I went through GWT-Platform release notes, found an interesting new feature in recent GWTP release, worth investigating:
https://github.com/ArcBees/GWTP/wiki/Release-Notes #346 : Map more than one name token to presenter
----- Original Message -----
From: "Martin Betak" <mbetak@redhat.com> To: devel@ovirt.org Cc: "Vojtech Szocs" <vszocs@redhat.com>, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 4:11:10 PM Subject: Moving forward our frontend stack
Hello oVirt developers!
I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1).
This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1
Since GWT 2.6 features support for Java 7 it was possible to increment language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend).
To facilitate emitting bytecode compatible with Java 7 also upgrade of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features even support for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the future.
Most of the changes in the patch are due to upgrade of GWT-P - i.e. changing packages of TokenFormatter and PlaceRequest.
Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much less risky.
I have tested draft-compile, debug-mode and also tried to use the resulting application manually for some time. So far everything worked (surprisingly well!) and I have not detected any defects. Of course I invite anyone to test this patch on his own since it is and upgrade of our core infrastructure.
That having said I think is comparatively simple and the benefits outweigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cycle.
Reviews, comments and testing are very welcome :-)
Best regards,
Martin
-- David Caro
Red Hat S.L. Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605 Email: dcaro@redhat.com Web: www.redhat.com RHT Global #: 82-62605
-- David Caro
Red Hat S.L. Continuous Integration Engineer - EMEA ENG Virtualization R&D
Tel.: +420 532 294 605 Email: dcaro@redhat.com Web: www.redhat.com RHT Global #: 82-62605
Well I really think this is a very bad idea. I'm pretty sure I don't need to introduce the concept of checksums to anybody on this list and why this is done. You must decide whether you want the correct code or not. If you decide to disable it on one repo, what is the justification to not disable it on another/all repos/projects? just my 2 cents :) Am 29.08.2014 16:54, schrieb Vojtech Szocs:
Well, I meant to avoid disabling checksum checking globally. If we can selectively disable checksum checking for specific Maven artifacts like "gwtp" etc. then we could do that..
-- Mit freundlichen Grüßen / Regards Sven Kieske Systemadministrator Mittwald CM Service GmbH & Co. KG Königsberger Straße 6 32339 Espelkamp T: +49-5772-293-100 F: +49-5772-293-333 https://www.mittwald.de Geschäftsführer: Robert Meyer St.Nr.: 331/5721/1033, USt-IdNr.: DE814773217, HRA 6640, AG Bad Oeynhausen Komplementärin: Robert Meyer Verwaltungs GmbH, HRB 13260, AG Bad Oeynhausen
--FFoLq8A0u+X9iRU8 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On 08/29, Martin Betak wrote:
From: "Vojtech Szocs" <vszocs@redhat.com> To: "Martin Betak" <mbetak@redhat.com> Cc: devel@ovirt.org, "Einav Cohen" <ecohen@redhat.com>, "Alexander Wels= " <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Ver= nia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 5:38:04 PM Subject: Re: Moving forward our frontend stack =20 Hey Martin! =20 I've just reviewed your patch, looks good overall. =20 Here are my thoughts: =20 * Jenkins CI fails on patch due to missing dependencies, we need to provide following new dependencies in order to proceed with the upgrade: =20 org.aspectj:aspectjweaver:1.8.2 org.aspectj:aspectjrt:1.8.2 com.google.gwt:gwt-user:2.6.1 com.google.gwt:gwt-dev:2.6.1 com.google.gwt:gwt-servlet:2.6.1 com.google.gwt:2.6.1 org.codehaus.mojo:gwt-maven-plugin:2.6.1 com.gwtplatform:gwtp-processors:1.3.1 com.gwtplatform:gwtp-mvp-client:1.3.1 com.google.gwt.inject:gin:2.1.2 =20 Who will take charge of that? =20 I already spoke to David Caro about the jenkins failure of=20
----- Original Message ----- patch http://gerrit.ovirt.org/#/c/32135/. It seems the files for appropriate versions are there but the files are corrupted. =20 Let us hope this can be resolved easier than Node.js :-)
=20 Best regards, =20 Martin =20
=20 * patch itself looks quite harmless (not too risky) =20 * consolidating Java source & target version across frontend and backend is nice! =20 -> this means we could use Java 7 features also on the frontend (Java/GWT) side =20 * TODO-GWT tags [1] proved to be helpful [2] =20 -> reminder to all UI maintainers to use TODO-GWT tags whenever we have some GWT(P) workaround, so that the future upgrade will be safer w.r.t. existing code =20 [1] https://www.mail-archive.com/devel@ovirt.org/msg00761.html [2] http://gerrit.ovirt.org/#/c/32135/1/frontend/webadmin/modules/webadmin/= src/main/java/org/ovirt/engine/ui/webadmin/section/main/view/SearchPanelVie= w.java =20 Regards, Vojtech =20 PS: I went through GWT-Platform release notes, found an interesting new feature in recent GWTP release, worth investigating: =20 https://github.com/ArcBees/GWTP/wiki/Release-Notes #346 : Map more than one name token to presenter =20 =20 ----- Original Message -----
From: "Martin Betak" <mbetak@redhat.com> To: devel@ovirt.org Cc: "Vojtech Szocs" <vszocs@redhat.com>, "Einav Cohen" <ecohen@redhat= =2Ecom>, "Alexander Wels" <awels@redhat.com>, "Greg Sheremeta" <gshereme@redhat.com>, "Tomas Jelinek" <tjelinek@redhat.com>, "Lior Vernia" <lvernia@redhat.com>, "Daniel Erez" <derez@redhat.com> Sent: Thursday, August 28, 2014 4:11:10 PM Subject: Moving forward our frontend stack =20 Hello oVirt developers! =20 I have prepared patch [1] that upgrades our frontend stack to use GWT version 2.6.1 (from previous 2.5.1). =20 This patch also updates GIN to version 2.1.2 and GWT-P to 1.3.1 =20 Since GWT 2.6 features support for Java 7 it was possible to increment language levels of all projects stuck at Java 6 (common, compat, searchbackend and entire of frontend). =20 To facilitate emitting bytecode compatible with Java 7 also upgrade of AspectJ was necessary. This patch upgrades it to AspectJ 1.8 that features even s= upport for Java 8 which will save effort when upgrading to GWT 2.7/3.0 in the fu= ture. =20 Most of the changes in the patch are due to upgrade of GWT-P - i.e. changing packages of TokenFormatter and PlaceRequest. =20 Overall this patch is *MUCH* simpler than the previous http://gerrit.ovirt.org/#/c/16739/ which facilitated upgrade from 2.3 to 2.5.1, and hopefully much less = risky. =20 I have tested draft-compile, debug-mode and also tried to use the res=
Ok job passed: http://jenkins.ovirt.org/job/ovirt_engine_master_compile_checkstyle_gerri= t/32071 ulting
application manually for some time. So far everything worked (surprisingly well!)= and I have not detected any defects. Of course I invite anyone to test this patch on= his own since it is and upgrade of our core infrastructure. =20 That having said I think is comparatively simple and the benefits out= weigh the risks if this upgrade is done at the beginning of ovirt-3.6 development cyc= le. =20 Reviews, comments and testing are very welcome :-) =20 Best regards, =20 Martin =20 [1] http://gerrit.ovirt.org/#/c/32135/ =20 =20
--=20 David Caro Red Hat S.L. Continuous Integration Engineer - EMEA ENG Virtualization R&D Tel.: +420 532 294 605 Email: dcaro@redhat.com Web: www.redhat.com RHT Global #: 82-62605 --FFoLq8A0u+X9iRU8 Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJUAEjZAAoJEEBxx+HSYmnDR4IH/34lwKCRu+B3HhgQ+sBZuaGp JlK6W+ZBghZh7fgshb71K079FJ2I7KIlbO3OTI8YOpGk9jGklHQOkFNpzGllJR0F M6cmgCE3zw4o+teO7pWycFusyzl32FWDXPUWfflpv2GV+uAnoctyN8oEuRJHAvFl 6zqalHQGZJvGHJQfuS5LunNtWluDpofSQFyvqtK+/PdEzgHuXkBdD9MxwOnsGtOM kElLbc4tzwX4YXvJirlvPf5bR/+oteW/stZ4AGb4ixVaMlb1v+xrMv29aPyYb3PO S7QL0SoWmiumjq1vZ6yTPItXzci666kKTBg/OpVeKfnU7T2zXZKlK5ow+1v7ltI= =n+g1 -----END PGP SIGNATURE----- --FFoLq8A0u+X9iRU8--
participants (6)
-
David Caro -
Martin Betak -
Sven Kieske -
Tomas Jelinek -
Vojtech Szocs -
Yair Zaslavsky