Mass rebuild for f33-java11 side tag completed

[Jiri Vanek]

Hello!

toatal packages: 610
passed: 427
failed: 176

From the failures, there is 29 which passed in the copr before, and now are thus failing
from two
reasons - unrelated change, or non-intel64-arch failure. I will put this to FTBF bugs for
those 29
pacakges,


In Monday, or as other duties allows I will fill FTBFS bugs for failures  with straces and
reproduce
steps.
In default CC will be, me, Severin, decathrope and java-devel(a)lists.fedoraproject.org.

Note that during non-sidetag rebuild during fedora 33 branching in start of August,  we
can expect
some indirect dependencies to fail.

Thoughts?


    J.

-- 
Jiri Vanek
Senior QE engineer, OpenJDK QE lead, Mgr.
Red Hat Czech
jvanek(a)redhat.com    M: [phone]
_______________________________________________
devel-announce mailing list -- devel-announce(a)lists.fedoraproject.org
To unsubscribe send an email to devel-announce-leave(a)lists.fedoraproject.org
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/[e-mail]...

[fedora-java] Mass rebuild for f33-java11 side tag completed

[Fabio Valentini]

On Sat, Jul 11, 2020 at 10:26 AM Jiri Vanek <jvanek(a)redhat.com&gt; wrote:
>  Hello!
>
>  toatal packages: 610
>  passed: 427
>  failed: 176
>
>  From the failures, there is 29 which passed in the copr before, and now are thus failing
> from two
>  reasons - unrelated change, or non-intel64-arch failure. I will put this to FTBF bugs for
> those 29
>  pacakges,
>
>
>  In Monday, or as other duties allows I will fill FTBFS bugs for failures  with straces
> and reproduce
>  steps.
>  In default CC will be, me, Severin, decathrope and java-devel(a)lists.fedoraproject.org.
>
>  Note that during non-sidetag rebuild during fedora 33 branching in start of August,  we
> can expect
>  some indirect dependencies to fail.
>
>  Thoughts? 
Looks like the result of the mass rebuild isn't too bad after all.
If I have some time next week, I can try to solve some of the
remaining FTBFS issues.

Do you have plans for when to merge builds back into rawhide?
In my experience, the longer you wait with that, the more merge
conflicts there are going to be.

Fabio

[fedora-java] Mass rebuild for f33-java11 side tag completed

[Mat Booth]

On Sat, 11 Jul 2020 at 09:24, Jiri Vanek <jvanek(a)redhat.com&gt; wrote:
>  Hello!
>
>  toatal packages: 610
>  passed: 427
>  failed: 176
>
>  From the failures, there is 29 which passed in the copr before, and now 
are thus
failing from two
>  reasons - unrelated change, or non-intel64-arch failure. I will put
> this 
to FTBF bugs for those 29
>  pacakges,
>
>
>  In Monday, or as other duties allows I will fill FTBFS bugs for failures 
 with
straces and reproduce
>  steps.
>  In default CC will be, me, Severin, decathrope and
java-devel(a)lists.fedoraproject.org.
>  Note that during non-sidetag rebuild during fedora 33 branching in start 
of August,
 we can expect
>  some indirect dependencies to fail.
>
>  Thoughts?
>
>  
A cursory glance shows that some failed with network problems. E.g.
plexus-velocity failed with this:

DEBUG util.py:621:  Errors during downloading metadata for repository
'build':
DEBUG util.py:621:    - Curl error (18): Transferred a partial file for
http://kojipkgs-cache01.s390.fedoraproject.org/repos/f33-java11/1710873/s...
[transfer closed with 23617146 bytes remaining to read]

(Sigh, flakey s390x machines, we meet again.)

When I retriggered the build it succeeded:
https://koji.fedoraproject.org/koji/buildinfo?buildID=1540681

Might be worth a retry of other failures before filing FTBFS bugs.

[fedora-java] Re: Mass rebuild for f33-java11 side tag completed

[Jiri Vanek]

>  
>  A cursory glance shows that some failed with network problems. E.g. plexus-velocity
> failed with this:
>  
>  DEBUG util.py:621:  Errors during downloading metadata for repository 'build':
>  DEBUG util.py:621:    - Curl error (18): Transferred a partial file for
>
> http://kojipkgs-cache01.s390.fedoraproject.org/repos/f33-java11/1710873/s...
>  [transfer closed with 23617146 bytes remaining to read]
>  
>  (Sigh, flakey s390x machines, we meet again.)
>  
>  When I retriggered the build it succeeded:
> https://koji.fedoraproject.org/koji/buildinfo?buildID=1540681
>  
>  Might be worth a retry of other failures before filing FTBFS bugs. 
Thanx!

Will elaborate on this!
>  
>  _______________________________________________
>  devel mailing list -- devel(a)lists.fedoraproject.org
>  To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
>  Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>  List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
>  List Archives:
> https://lists.fedoraproject.org/archives/list/[e-mail]
>   

-- 
Jiri Vanek
Senior QE engineer, OpenJDK QE lead, Mgr.
Red Hat Czech
jvanek(a)redhat.com    M: [phone]

[fedora-java] Re: Mass rebuild for f33-java11 side tag completed

[Mat Booth]

On Mon, 13 Jul 2020 at 19:14, Jiri Vanek <jvanek(a)redhat.com&gt; wrote:
>  >
>  > A cursory glance shows that some failed with network problems. E.g.
plexus-velocity failed with this:
>  >
>  > DEBUG util.py:621:  Errors during downloading metadata for repository
'build':
>  > DEBUG util.py:621:    - Curl error (18): Transferred a partial
> file for
>  >
http://kojipkgs-cache01.s390.fedoraproject.org/repos/f33-java11/1710873/s...
>  > [transfer closed with 23617146 bytes remaining to read]
>  >
>  > (Sigh, flakey s390x machines, we meet again.)
>  >
>  > When I retriggered the build it succeeded:
https://koji.fedoraproject.org/koji/buildinfo?buildID=1540681
>  >
>  > Might be worth a retry of other failures before filing FTBFS bugs.
>
>  Thanx!
>
>  Will elaborate on this! 
Any update? Any thoughts on when you want to merge the f33-java11 side tag
back into rawhide?

[fedora-java] Re: Mass rebuild for f33-java11 side tag completed

[Mat Booth]

On Wed, 15 Jul 2020 at 16:54, Mat Booth <fedora(a)matbooth.co.uk&gt; wrote:
>  On Mon, 13 Jul 2020 at 19:14, Jiri Vanek <jvanek(a)redhat.com&gt; wrote:
> >  
>  >
> >  
>  > > A cursory glance shows
> that some failed with network problems. E.g. 
plexus-velocity failed with this:
>  >
> >  
>  > > DEBUG util.py:621: 
> Errors during downloading metadata for repository 
'build':
>  > > DEBUG util.py:621:    - Curl error (18): Transferred a
> partial file 
for
>  > >
http://kojipkgs-cache01.s390.fedoraproject.org/repos/f33-java11/1710873/s...
> > > [transfer closed with 23617146 bytes remaining to read]
>  > > 
> > > (Sigh, flakey s390x machines, we meet
again.)
>  > > 
> > > When I retriggered the build it
succeeded:
https://koji.fedoraproject.org/koji/buildinfo?buildID=1540681
>  > > 
> > > Might be worth a retry of other
failures before filing FTBFS bugs.
>
>  
> > Thanx!
>
>  
> > Will elaborate on this!
>  
> Any update? Any thoughts on when you want to merge the
f33-java11 side
tag back into rawhide?
>  
BTW There are some packages, e.g. built by ant with no sauce/target level
specified at all, that are built with Java 11-level bytecode.

This is bad because if there is a dependent package that requires Java 8
for some reason it won't work because the bytecode of one your dependencies
is too new and cannot be interpreted by Java 8. In these cases

I am fixing such occurrences in the ```f33-java11``` build target as I
encounter them -- just something to be aware of in case you see any
UnsupportedClassVersionErrors.

--
Mat Booth
http://fedoraproject.org/get-fedora

[fedora-java] Re: Mass rebuild for f33-java11 side tag completed

[Fabio Valentini]

On Wed, Jul 15, 2020 at 5:55 PM Mat Booth <fedora(a)matbooth.co.uk&gt; wrote:
>  Any update? Any thoughts on when you want to merge the f33-java11
> side tag back into rawhide? 
I second this, please don't wait two weeks [0] for merging the side
tag, it's only going to cause merge conflict problems with packages
that get rebuilt in rawhide until then.
Please just do it now, and we'll fix any Java 11 FTBFS issues in
rawhide directly.

Fabio

[0]: https://bugzilla.redhat.com/show_bug.cgi?id=1858082#c3

[Christian Heimes]

>  Hello!
>  
>  toatal packages: 610
>  passed: 427
>  failed: 176
>  
>  From the failures, there is 29 which passed in the copr before, and now are thus failing
>  from two
>  reasons - unrelated change, or non-intel64-arch failure. I will put this to FTBF bugs
> for
>  those 29
>  pacakges,
>  
>  
>  In Monday, or as other duties allows I will fill FTBFS bugs for failures  with straces
> and
>  reproduce
>  steps.
>  In default CC will be, me, Severin, decathrope and java-devel(a)lists.fedoraproject.org.
>  
>  Note that during non-sidetag rebuild during fedora 33 branching in start of August,  we
>  can expect
>  some indirect dependencies to fail.
>  
>  Thoughts? 
Hi Jiri,

how will a merge of the Java 11 side tag into F33 Rawhide affect Dogtag
PKI? Dogtag PKI (aka pki-core package) is mostly written on top of the
Tomcat stack. As far as I know the Dogtag team is still working on Java
11 support.

Dogtag is a core component of FreeIPA, which is a one of three major
features of Fedora Server collection (the other two are modularity and
Cockpit) [0]. FreeIPA is also part of Fedora's OpenQA effort. It's
likely that Java 11 update is going to break a core feature of Fedora
and QA gating for a considerable amount of packages.

Could you please hold of the merge of the side tag and work with the
Dogtag team to solve the issue?

By the way I see a rebuild of Tomcat in the side tag but there is no
pki-core build. Why is pki-core missing from the tag? Is it one of the
FTBFS packages?

$ koji list-tagged f33-java11 | grep tomcat
tomcat-9.0.36-2.fc33                      f33-java11            jvanek
tomcat-native-1.2.23-2.fc33               f33-java11            jvanek
tomcat-taglibs-parent-3-12.fc33           f33-java11            jvanek
tomcatjss-7.5.0-0.5.fc33                  f33-java11            jvanek
$ koji list-tagged f33-java11 | grep pki
<n/a>

Christian

[0] https://getfedora.org/en/server/

[fedora-java] Re: Mass rebuild for f33-java11 side tag completed

[Iñaki Ucar]

Could someone with superpowers cancel the following task for me? Thanks.

https://koji.fedoraproject.org/koji/taskinfo?taskID=47330918

On Fri, 17 Jul 2020 at 09:44, Fabio Valentini <decathorpe(a)gmail.com&gt; wrote:
>  On Wed, Jul 15, 2020 at 5:55 PM Mat Booth <fedora(a)matbooth.co.uk&gt; wrote:
>  > Any update? Any thoughts on when you want to merge the f33-java11 side tag back into
> rawhide?
>
>  I second this, please don't wait two weeks [0] for merging the side
>  tag, it's only going to cause merge conflict problems with packages
>  that get rebuilt in rawhide until then.
>  Please just do it now, and we'll fix any Java 11 FTBFS issues in
>  rawhide directly.
>
>  Fabio
>
>  [0]: https://bugzilla.redhat.com/show_bug.cgi?id=1858082#c3
>  _______________________________________________
>  devel mailing list -- devel(a)lists.fedoraproject.org
>  To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
>  Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
>  List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
>  List Archives:
> https://lists.fedoraproject.org/archives/list/[e-mail] 


-- 
Iñaki Úcar

[fedora-java] Re: Mass rebuild for f33-java11 side tag completed

[Dan Horák]

On Fri, 17 Jul 2020 10:42:45 +0200
Iñaki Ucar <iucar(a)fedoraproject.org&gt; wrote:

>  Could someone with superpowers cancel the following task for me?
>  Thanks. 
done, it has already failed on i686


		Dan

> https://koji.fedoraproject.org/koji/taskinfo?taskID=47330918
> 
> On Fri, 17 Jul 2020 at 09:44, Fabio Valentini <decathorpe(a)gmail.com&gt;
> wrote:
> >
> > On Wed, Jul 15, 2020 at 5:55 PM Mat Booth <fedora(a)matbooth.co.uk&gt;
> > wrote:
> > > Any update? Any thoughts on when you want to merge the f33-java11
> > > side tag back into rawhide?
> >
> > I second this, please don't wait two weeks [0] for merging the side
> > tag, it's only going to cause merge conflict problems with packages
> > that get rebuilt in rawhide until then.
> > Please just do it now, and we'll fix any Java 11 FTBFS issues in
> > rawhide directly.
> >
> > Fabio
> >
> > [0]: https://bugzilla.redhat.com/show_bug.cgi?id=1858082#c3
> > _______________________________________________
> > devel mailing list -- devel(a)lists.fedoraproject.org
> > To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
> > Fedora Code of Conduct:
> > https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List
> > Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> > List Archives:
> > https://lists.fedoraproject.org/archives/list/[e-mail]
> 
> 
> 
> -- 
> Iñaki Úcar
> _______________________________________________
> devel mailing list -- devel(a)lists.fedoraproject.org
> To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
> Fedora Code of Conduct:
> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List
> Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives:
> https://lists.fedoraproject.org/archives/list/[e-mail]

[fedora-java] Re: Mass rebuild for f33-java11 side tag completed

[Iñaki Ucar]

On Fri, 17 Jul 2020 at 11:19, Dan Horák <dan(a)danny.cz&gt; wrote:
>  On Fri, 17 Jul 2020 10:42:45 +0200
>  Iñaki Ucar <iucar(a)fedoraproject.org&gt; wrote:
>
>  > Could someone with superpowers cancel the following task for me?
>  > Thanks.
>
>  done, it has already failed on i686 
Yeah, that's why I wanted to cancel it. The task was launched without
the --fail-fast flag and it takes many hours to complete on s390x and
armv7hl. Thanks!

-- 
Iñaki Úcar

Hi!

Sorry., I had missed your email. Lets continue the discussion in the ticket
https://pagure.io/fesco/issue/2449 you had risen.

In the mass rebuilds, I was working with packages with selected java dependencies or
packaes I know about tat had to be included. dogtag pki have likely only transitive 
dependence and I had not included it manually. Sorry.

Please rebuild the package in  f33-java11. Acording ot ticket, you already  did. So
continue here.

[Mat Booth]

On Fri, 17 Jul 2020 at 09:34, Christian Heimes <cheimes(a)redhat.com&gt; wrote:
>  > Hello!
>  >
>  > toatal packages: 610
>  > passed: 427
>  > failed: 176
>  >
>  > From the failures, there is 29 which passed in the copr before, and now are thus
> failing
>  > from two
>  > reasons - unrelated change, or non-intel64-arch failure. I will put this to FTBF
> bugs for
>  > those 29
>  > pacakges,
>  >
>  >
>  > In Monday, or as other duties allows I will fill FTBFS bugs for failures  with
> straces and
>  > reproduce
>  > steps.
>  > In default CC will be, me, Severin, decathrope and
> java-devel(a)lists.fedoraproject.org.
>  >
>  > Note that during non-sidetag rebuild during fedora 33 branching in start of August, 
> we
>  > can expect
>  > some indirect dependencies to fail.
>  >
>  > Thoughts?
>
>  Hi Jiri,
>
>  how will a merge of the Java 11 side tag into F33 Rawhide affect Dogtag
>  PKI? Dogtag PKI (aka pki-core package) is mostly written on top of the
>  Tomcat stack. As far as I know the Dogtag team is still working on Java
>  11 support.
>
>  Dogtag is a core component of FreeIPA, which is a one of three major
>  features of Fedora Server collection (the other two are modularity and
>  Cockpit) [0]. FreeIPA is also part of Fedora's OpenQA effort. It's
>  likely that Java 11 update is going to break a core feature of Fedora
>  and QA gating for a considerable amount of packages.
>
>  Could you please hold of the merge of the side tag and work with the
>  Dogtag team to solve the issue?
>
>  By the way I see a rebuild of Tomcat in the side tag but there is no
>  pki-core build. Why is pki-core missing from the tag? Is it one of the
>  FTBFS packages?
>
>  $ koji list-tagged f33-java11 | grep tomcat
>  tomcat-9.0.36-2.fc33                      f33-java11            jvanek
>  tomcat-native-1.2.23-2.fc33               f33-java11            jvanek
>  tomcat-taglibs-parent-3-12.fc33           f33-java11            jvanek
>  tomcatjss-7.5.0-0.5.fc33                  f33-java11            jvanek
>  $ koji list-tagged f33-java11 | grep pki
>  <n/a>
>  
pki-core fails to build in the side tag with this error:

  bad class file:
/usr/share/java/tomcatjss.jar(org/apache/tomcat/util/net/jss/IPasswordStore.class)
    class file has wrong version 55.0, should be 52.0
    Please remove or make sure it appears in the correct subdirectory
of the classpath.

That means that tomcatjss has Java 11 bytecode and I guess pki-core is
trying to use Java 8.

A quick glance at tomcatjss package shows it is built by ant with no
sauce/target level specified at all, which means it is built with Java
11 bytecode by default.

You should fix this by rebuilding tomcatjss with Java 8 bytecode.
Here's an example where I fixed the same problem in another package:
https://src.fedoraproject.org/rpms/aopalliance/c/5fdc6912e39fe15fe55cedcc...

-- 
Mat Booth
http://fedoraproject.org/get-fedora

[Fabio Valentini]

On Mon, Jul 20, 2020 at 2:24 PM Mat Booth <fedora(a)matbooth.co.uk&gt; wrote:
>  On Fri, 17 Jul 2020 at 09:34, Christian Heimes <cheimes(a)redhat.com&gt; wrote:
>  >
>  > > Hello!
>  > >
>  > > toatal packages: 610
>  > > passed: 427
>  > > failed: 176
>  > >
>  > > From the failures, there is 29 which passed in the copr before, and now are
> thus failing
>  > > from two
>  > > reasons - unrelated change, or non-intel64-arch failure. I will put this to
> FTBF bugs for
>  > > those 29
>  > > pacakges,
>  > >
>  > >
>  > > In Monday, or as other duties allows I will fill FTBFS bugs for failures  with
> straces and
>  > > reproduce
>  > > steps.
>  > > In default CC will be, me, Severin, decathrope and
> java-devel(a)lists.fedoraproject.org.
>  > >
>  > > Note that during non-sidetag rebuild during fedora 33 branching in start of
> August,  we
>  > > can expect
>  > > some indirect dependencies to fail.
>  > >
>  > > Thoughts?
>  >
>  > Hi Jiri,
>  >
>  > how will a merge of the Java 11 side tag into F33 Rawhide affect Dogtag
>  > PKI? Dogtag PKI (aka pki-core package) is mostly written on top of the
>  > Tomcat stack. As far as I know the Dogtag team is still working on Java
>  > 11 support.
>  >
>  > Dogtag is a core component of FreeIPA, which is a one of three major
>  > features of Fedora Server collection (the other two are modularity and
>  > Cockpit) [0]. FreeIPA is also part of Fedora's OpenQA effort. It's
>  > likely that Java 11 update is going to break a core feature of Fedora
>  > and QA gating for a considerable amount of packages.
>  >
>  > Could you please hold of the merge of the side tag and work with the
>  > Dogtag team to solve the issue?
>  >
>  > By the way I see a rebuild of Tomcat in the side tag but there is no
>  > pki-core build. Why is pki-core missing from the tag? Is it one of the
>  > FTBFS packages?
>  >
>  > $ koji list-tagged f33-java11 | grep tomcat
>  > tomcat-9.0.36-2.fc33                      f33-java11            jvanek
>  > tomcat-native-1.2.23-2.fc33               f33-java11            jvanek
>  > tomcat-taglibs-parent-3-12.fc33           f33-java11            jvanek
>  > tomcatjss-7.5.0-0.5.fc33                  f33-java11            jvanek
>  > $ koji list-tagged f33-java11 | grep pki
>  > <n/a>
>  >
>
>  pki-core fails to build in the side tag with this error:
>
>    bad class file:
>  /usr/share/java/tomcatjss.jar(org/apache/tomcat/util/net/jss/IPasswordStore.class)
>      class file has wrong version 55.0, should be 52.0
>      Please remove or make sure it appears in the correct subdirectory
>  of the classpath.
>
>  That means that tomcatjss has Java 11 bytecode and I guess pki-core is
>  trying to use Java 8.
>
>  A quick glance at tomcatjss package shows it is built by ant with no
>  sauce/target level specified at all, which means it is built with Java
>  11 bytecode by default. 
(snip)

>  You should fix this by rebuilding tomcatjss with Java 8 bytecode.
>  Here's an example where I fixed the same problem in another package:
>  https://src.fedoraproject.org/rpms/aopalliance/c/5fdc6912e39fe15fe55cedcc...
Been there, tried that, it fails due to an unrelated tomcat update.
See: https://pagure.io/fesco/issue/2449#comment-666061

Fabio

> --
> Mat Booth
> http://fedoraproject.org/get-fedora
> _______________________________________________
> devel mailing list -- devel(a)lists.fedoraproject.org
> To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
> Fedora Code of Conduct:
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
> List Archives:
https://lists.fedoraproject.org/archives/list/[e-mail]

[Jiri Vanek]

Hi!

Side tag was merged,  java-11-openjdk now should be system JDK.
Don't hesitate to reach me if you encounter some misbehaving!

J.

On 7/11/20 10:24 AM, Jiri Vanek wrote:
>  Hello!
>  
>  toatal packages: 610
>  passed: 427
>  failed: 176
>  
>  From the failures, there is 29 which passed in the copr before, and now are thus failing
> from two
>  reasons - unrelated change, or non-intel64-arch failure. I will put this to FTBF bugs for
> those 29
>  pacakges,
>  
>  
>  In Monday, or as other duties allows I will fill FTBFS bugs for failures  with straces
> and reproduce
>  steps.
>  In default CC will be, me, Severin, decathrope and java-devel(a)lists.fedoraproject.org.
>  
>  Note that during non-sidetag rebuild during fedora 33 branching in start of August,  we
> can expect
>  some indirect dependencies to fail.
>  
>  Thoughts?
>  
>  
>      J.
>  
>  

-- 
Jiri Vanek
Senior QE engineer, OpenJDK QE lead, Mgr.
Red Hat Czech
jvanek(a)redhat.com    M: [phone]
_______________________________________________
devel-announce mailing list -- devel-announce(a)lists.fedoraproject.org
To unsubscribe send an email to devel-announce-leave(a)lists.fedoraproject.org
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/[e-mail]...

[fedora-java] Re: Mass rebuild for f33-java11 side tag completed

[Jiri Vanek]

> >
> > Any update? Any thoughts on when you want to merge the f33-java11 side tag back into
> rawhide? 
done
> >
>  
>  BTW There are some packages, e.g. built by ant with no sauce/target level specified at
> all, that are
>  built with Java 11-level bytecode.
>  
>  This is bad because if there is a dependent package that requires Java 8 for some reason
> it won't
>  work because the bytecode of one your dependencies is too new and cannot be interpreted
> by Java 8.
>  In these cases
>  
>  I am fixing such occurrences in the ```f33-java11``` build target as I encounter them --
> just
>  something to be aware of in case you see any UnsupportedClassVersionErrors. 
Mat, I had come to same conclusion, which had lead me into this:
https://src.fedoraproject.org/rpms/javapackages-tools/pull-request/3#comm... .
Please
contriute, or sugest next steps here. I would liek to have it java-packaging gudelines
change, and
self-contained f33 change, but it may be to late.

I see yo already track the Fabio's to-high bytecode issue,  but my proposal is to
prevent it in
future. However, it do not seem to be facing to much sympathies.


J.



-- 
Jiri Vanek
Senior QE engineer, OpenJDK QE lead, Mgr.
Red Hat Czech
jvanek(a)redhat.com    M: [phone]

[fedora-java] Re: Mass rebuild for f33-java11 side tag completed

[Fabio Valentini]

On Tue, Jul 21, 2020 at 11:25 AM Jiri Vanek <jvanek(a)redhat.com&gt; wrote:
>  >>
>  >> Any update? Any thoughts on when you want to merge the f33-java11 side tag back
> into rawhide?
>
>  done
>  >>
>  >
>  > BTW There are some packages, e.g. built by ant with no sauce/target level specified
> at all, that are
>  > built with Java 11-level bytecode.
>  >
>  > This is bad because if there is a dependent package that requires Java 8 for some
> reason it won't
>  > work because the bytecode of one your dependencies is too new and cannot be
> interpreted by Java 8.
>  > In these cases
>  >
>  > I am fixing such occurrences in the ```f33-java11``` build target as I encounter
> them -- just
>  > something to be aware of in case you see any UnsupportedClassVersionErrors.
>
>  Mat, I had come to same conclusion, which had lead me into this:
>  https://src.fedoraproject.org/rpms/javapackages-tools/pull-request/3#comm... .
> Please
>  contriute, or sugest next steps here. I would liek to have it java-packaging gudelines
> change, and
>  self-contained f33 change, but it may be to late.
>
>  I see yo already track the Fabio's to-high bytecode issue,  but my proposal is to
> prevent it in
>  future. However, it do not seem to be facing to much sympathies. 
Hi Jiri,

I don't think it's a bad idea - in fact, I would like to have some way
to enforce this distro-wide.
But it's definitely too late for fedora 33 in my opinion. The
submission deadline for Self-Contained F33 Changes is today, and I
would argue that changing the way that ~600 packages are built might
not be considered self-contained :)

And if we really want to do this (enforce javac -source / target
versions and allow for macros to override them), it would be good to
have support for this in both maven and ant-based projects. For maven,
xmvn would be an easy place to put this (something like it was there
in old xmvn versions), but I'm not sure about ant (since build targets
in build.xml files are named differently in every upstream project I
looked at).

Fabio