Re: T&R of 2.4.2 this week.
I'll be T&Ring 2.4.2 this afternoon (eastern). On Apr 3, 2012, at 10:05 AM, Jim Jagielski wrote: > > I'd like to T&R 2.4.2 this week in hopes of doing the actual > release early next week... > > Comments? >
Re: [RE-VOTE #3] adoption of mod_combine subproject
On Apr 3, 2012, at 9:37 PM, William A. Rowe Jr. wrote: > > I will absolutely not shirk my own responsibility, which in this matter, is > neither the responsibility of a committer placing code at the ASF, an officer > acting under the direction of the BoD, nor a a director of the ASF. Which is > to say, my entire responsibility as a member of the project and the foundation > consisted of bringing the concern to the chair of the project and VP Legal, > and let you all have your fun. I'm done with this dialog. Cheers. > I will be honest: I have no idea what this whole debate is about. From the vote within the PMC, it's clear that the consensus is to fold the code in, that we are satisfied that we are covered, IP-wise, due to Graham's iCLA on file as well as other guarantees noted in the (long) thread regarding the code submission. So what is the problem?? That someone doesn't like the result of the vote and is hoping to have it overturned, somehow?? I also fail to see how this codebase is any different from other modules which we've folded in from "outside" like mod_proxy_html and Paul's various heartbeat/cluster stuff which came in with hardly any debate and certainly not dragging legal into it all...
2.4.2 Windows still beta.
Noticed (no discuss) that AccepFilter https none is still broken. No problem here, can use IIS or 2.2 in front to have SSL. -Original Message- From: Jim Jagielski Sent: Tuesday, April 03, 2012 4:05 PM Newsgroups: gmane.comp.apache.devel To: dev@httpd.apache.org Subject: T&R of 2.4.2 this week. I'd like to T&R 2.4.2 this week in hopes of doing the actual release early next week... Comments?
Re: 2.4.2 Windows still beta.
I'm not sure how to resolve this... there are only a handful of people who have ever hacked on the Windows stuff and they all seem pre-occupied w/ other issues. Yeah, it bothers me, but I don't know what can be done. At the very least we are no longer holding the rest of the 2.4.x community hostage (by delaying releases) because of Windows issues :/ We used to have access to Microsoft test environs, etc... maybe we need to see if we can still have that happen. I might ping Gianugo and see if we can get some MS support here. On Apr 4, 2012, at 10:02 AM, Steffen wrote: > Noticed (no discuss) that AccepFilter https none is still broken. > > No problem here, can use IIS or 2.2 in front to have SSL. > > -Original Message- From: Jim Jagielski Sent: Tuesday, April 03, 2012 > 4:05 PM Newsgroups: gmane.comp.apache.devel To: dev@httpd.apache.org Subject: > T&R of 2.4.2 this week. > > I'd like to T&R 2.4.2 this week in hopes of doing the actual > release early next week... > > Comments? >
Fwd: Re: [RE-VOTE #3] adoption of mod_combine subproject
Response to Jim's post sent to legal@, with PMC-specific notes; Let's put just legal@ questions to legal and keep the rest for this list, eh? If you remain confused, I suggest you actually read the bugzilla ticket to understand how Sam confused the issue. Sam effectively concludes, after spending a month suggesting that Roy's guidance was insufficent, that (while refusing to say as much), yup, Roy was right throughout. No big surprise. I will update both apr and httpd dev pages in the coming days to clarify and the issue shouldn't come up again. Returning to PMC business; Graham, please proceed as Sam's questions on the ticket have been asked and answered and he raises no further concern. The firehose contribution is already in core, now correctly considered and with no further IP steps needed. The policy module into modules/test/ is equally ready for you to import. Sorry we didn't reach agreement on combine. Original Message Subject: Re: [RE-VOTE #3] adoption of mod_combine subproject Date: Wed, 04 Apr 2012 10:13:39 -0500 From: William A. Rowe Jr. To: legal-disc...@apache.org On 4/4/2012 7:52 AM, Jim Jagielski wrote: > > On Apr 3, 2012, at 9:37 PM, William A. Rowe Jr. wrote: >> >> I will absolutely not shirk my own responsibility, which in this matter, is >> neither the responsibility of a committer placing code at the ASF, an officer >> acting under the direction of the BoD, nor a a director of the ASF. Which is >> to say, my entire responsibility as a member of the project and the >> foundation >> consisted of bringing the concern to the chair of the project and VP Legal, >> and let you all have your fun. I'm done with this dialog. Cheers. > > I will be honest: I have no idea what this whole debate is about. > From the vote within the PMC, it's clear that the consensus is > to fold the code in, that we are satisfied that we are covered, > IP-wise, due to Graham's iCLA on file as well as other guarantees > noted in the (long) thread regarding the code submission. > So what is the problem?? That someone doesn't like the result > of the vote and is hoping to have it overturned, somehow?? > > I also fail to see how this codebase is any different from other > modules which we've folded in from "outside" like mod_proxy_html > and Paul's various heartbeat/cluster stuff which came in with hardly > any debate and certainly not dragging legal into it all... In December I pointed out, owing to my own ignorance, that the submission needed to follow the incubator IP clearance process. You *agreed* with the comment somewhere along the way. Checking off that box appeared necessary to us both. There were other issues, nothing to do with legal@ which you are bringing up above, but that's all distraction from the purpose of this thread here at legal@ And then... Roy answered, No, Asked and Answered; we do not follow that *incubator* IP clearance process for new submissions authored by existing committers. The PMC simply accepts the code based on its own judgement that its own committers are correctly handling the IP concerns. And then... Of course, Sam said that legal provided no such guidance. I shoved this all back to legal-private and said "figure it out, then explain it to us". Sam suggests that revised legal IP clearance requirements, as a matter of policy, across top level projects, is sufficiently described by; http://mail-archives.apache.org/mod_mbox/www-legal-discuss/201204.mbox/%3C4F7C3832.4050509%40intertwingly.net%3E So, simply time for spring cleaning. I've asked Sam to document one specific legal policy and have gained this reference email post, and will ensure that neither the httpd.a.o or apr.a.o dev pages refer to other older, stale ip clearance policies. Certainly won't try to clean it up across the whole foundation, but at least everyone has this same email documentation to refer back to. In the meantime, on the referenced bug, there was no objection in 7 days to Graham's claim, so Graham will proceed and PMC considers this resolved and this thread is ended. As Roy argued in the first place, and Sam now finally concurs, the external IP Clearance process was never applicable to committers with icla's. One would suppose the IP Clearance process remains applicable only for multiparty or code-grant based submissions, just not for single party, author-contributed icla-based submissions. --- Begin Message --- On 04/03/2012 10:43 PM, William A. Rowe Jr. wrote: On 4/3/2012 9:33 PM, Sam Ruby wrote: The ASF goes through great pains to ensure that (for example) every commit results in an email to a mailing list that PMC is expected to monitor. When mistakes happen (note I said when, not if), it is a failing not only of the committer but of the PMC. Every time somebody runs a RAT report and identifies a problem, that is a problem that was missed previously. Somebody committed that change. Every PMC member wasn't doing their job monitoring commits.
Re: 2.4.2 Windows still beta.
Hi, >> Noticed (no discuss) that AccepFilter https none is still broken. > At the very least we are no longer holding the rest of the > 2.4.x community hostage (by delaying releases) because of > Windows issues :/ I realy understand Steffens point. But 2.4.1 with that issue is already released. I think it is a good idea to release 2.4.2 even if that is not solved yet like done with 2.4.1. And I also think that Steffen means that it is not a show stopper for 2.4.2. The other stuff is running well on windows like on the other OSs. > We used to have access to Microsoft test environs, etc... maybe > we need to see if we can still have that happen. I might ping > Gianugo and see if we can get some MS support here. That would be great if you can get some help there. There will be windows tester if you offer a patch for testing :-) Cheers Mario
Re: setting up testing
Bother: I did not need to install any extra modules - and this has taken forever, because I kept running into problems trying to install all the extra modules. However, "Is this all there is?", and why do I not have a mod_perl.c - is that a "tester/developer" only module, or just no longer in the "all" list? Would you consider this a PASS for the tests? Thx. root@x104:[/data/prj/SVN/Apache-Test]t/TEST [warning] setting ulimit to allow core files ulimit -c unlimited; /usr/opt/perl5/bin/perl /data/prj/SVN/Apache-Test/t/TEST /opt/httpd/sbin/httpd -d /data/prj/SVN/Apache-Test/t -f /data/prj/SVN/Apache-Test/t/conf/httpd.conf -D APACHE2 -D PERL_USEITHREADS using Apache/2.5.0-dev (worker MPM) waiting 60 seconds for server to start: ... waiting 60 seconds for server to start: ok (waited 2 secs) server loopback:8529 started [ info] adding source lib /data/prj/SVN/Apache-Test/../Apache-Test/lib to @INC t/alltest/all...skipped all skipped: testing all.t t/alltest2/all..skipped all skipped: testing more than one all.t t/bad_codingok t/cookies...ok t/importok t/log_watch.ok t/log_watch_for_broken_linesok t/more/all..skipped all skipped: cannot find module 'mod_perl.c' t/next_available_port...ok t/ping..ok t/redirect..ok t/request...ok t/sok...ok All tests successful, 3 tests skipped. Files=13, Tests=93, 24 wallclock secs ( 8.67 cusr + 6.69 csys = 15.36 CPU) [warning] server loopback:8529 shutdown [warning] port 8529 still in use... ..done On Thu, Mar 1, 2012 at 9:01 PM, Jeff Trawick wrote: > On Thu, Mar 1, 2012 at 2:50 PM, Michael Felt wrote: > > Want to get started on this. I read the links from > > http://httpd.apache.org/test/ and think I understand the flood > subproject. > > From reading the forums here recently I get the impression that more than > > flood is being used, or even something other than flood. > > > > Looking for assistance and suggestions. > > perl framework... > > get svn on your AIX box if you don't have it already > > check out /httpd/test/framework/trunk/ > > perl Makefile.PL -apxs /path/to/apxs > t/TEST > t/TEST -clean > > No it won't be fun because things will fail and there will be many > different reasons, few if any obvious ones :) > > You will likely need to install more Perl modules... Some old > instructions of mine: > >perl -MCPAN -e 'shell' >cpan>install Test::Harness >cpan>install URI >cpan>install LWP::Protocol::https >cpan>install HTTP::DAV >cpan>install Bundle::ApacheTest >cpan>install LWP::Simple > > (dunno if this is current or everything you need on your level of AIX; > this was for OpenSolaris a couple of years ago) >
Re: setting up testing
I found some info on a mod_perl, and have picked it up via svn - but am stuck - where does it expect to be? root@x104:[/data/prj/modperl]perl Makefile.PL MP_APXS=/opt/httpd/bin/apxs Reading Makefile.PL args from @ARGV MP_APXS = /opt/httpd/bin/apxs [ error] Can't find a sub-directory Apache-Test. Make sure that you are using a complete source distribution On Wed, Apr 4, 2012 at 10:13 PM, Michael Felt wrote: > Bother: I did not need to install any extra modules - and this has taken > forever, because I kept running into problems trying to install all the > extra modules. > > However, "Is this all there is?", and why do I not have a mod_perl.c - is > that a "tester/developer" only module, or just no longer in the "all" list? > > Would you consider this a PASS for the tests? > > Thx. > > root@x104:[/data/prj/SVN/Apache-Test]t/TEST > [warning] setting ulimit to allow core files > ulimit -c unlimited; /usr/opt/perl5/bin/perl > /data/prj/SVN/Apache-Test/t/TEST > /opt/httpd/sbin/httpd -d /data/prj/SVN/Apache-Test/t -f > /data/prj/SVN/Apache-Test/t/conf/httpd.conf -D APACHE2 -D PERL_USEITHREADS > using Apache/2.5.0-dev (worker MPM) > > waiting 60 seconds for server to start: ... > waiting 60 seconds for server to start: ok (waited 2 secs) > server loopback:8529 started > [ info] adding source lib /data/prj/SVN/Apache-Test/../Apache-Test/lib > to @INC > t/alltest/all...skipped > all skipped: testing all.t > t/alltest2/all..skipped > all skipped: testing more than one all.t > t/bad_codingok > > t/cookies...ok > > t/importok > > t/log_watch.ok > > t/log_watch_for_broken_linesok > > t/more/all..skipped > all skipped: cannot find module 'mod_perl.c' > t/next_available_port...ok > > t/ping..ok > > t/redirect..ok > > t/request...ok > > t/sok...ok > > All tests successful, 3 tests skipped. > Files=13, Tests=93, 24 wallclock secs ( 8.67 cusr + 6.69 csys = 15.36 CPU) > [warning] server loopback:8529 shutdown > [warning] port 8529 still in use... > ..done > > > > On Thu, Mar 1, 2012 at 9:01 PM, Jeff Trawick wrote: > >> On Thu, Mar 1, 2012 at 2:50 PM, Michael Felt wrote: >> > Want to get started on this. I read the links from >> > http://httpd.apache.org/test/ and think I understand the flood >> subproject. >> > From reading the forums here recently I get the impression that more >> than >> > flood is being used, or even something other than flood. >> > >> > Looking for assistance and suggestions. >> >> perl framework... >> >> get svn on your AIX box if you don't have it already >> >> check out /httpd/test/framework/trunk/ >> >> perl Makefile.PL -apxs /path/to/apxs >> t/TEST >> t/TEST -clean >> >> No it won't be fun because things will fail and there will be many >> different reasons, few if any obvious ones :) >> >> You will likely need to install more Perl modules... Some old >> instructions of mine: >> >>perl -MCPAN -e 'shell' >>cpan>install Test::Harness >>cpan>install URI >>cpan>install LWP::Protocol::https >>cpan>install HTTP::DAV >>cpan>install Bundle::ApacheTest >>cpan>install LWP::Simple >> >> (dunno if this is current or everything you need on your level of AIX; >> this was for OpenSolaris a couple of years ago) >> > >
Re: setting up testing
On 05 Apr 2012, at 12:03 AM, Michael Felt wrote: > I found some info on a mod_perl, and have picked it up via svn - but am stuck > - where does it expect to be? > root@x104:[/data/prj/modperl]perl Makefile.PL MP_APXS=/opt/httpd/bin/apxs > Reading Makefile.PL args from @ARGV >MP_APXS = /opt/httpd/bin/apxs > [ error] Can't find a sub-directory Apache-Test. Make sure that you are > using a complete source distribution Mod_perl is a completely separate project, it would need to be packaged separately. Regards, Graham -- smime.p7s Description: S/MIME cryptographic signature
Re: setting up testing
On Wed, Apr 4, 2012 at 4:13 PM, Michael Felt wrote: > Bother: I did not need to install any extra modules - and this has taken > forever, because I kept running into problems trying to install all the > extra modules. > What modules? > However, "Is this all there is?", and why do I not have a mod_perl.c - is > that a "tester/developer" only module, or just no longer in the "all" list? mod_perl is separate, but your test results imply the framework thinks you built almost no modules and thus ran the absolute bare minimum of tests. Most shared modules would be built and loaded in the default configuation, and there is a developer-mode flag to enable all the modules compiled. The test suite onl y tests modules enabled in your configuration. It finds them when you configure it (perl Makefile.PL...) > > Would you consider this a PASS for the tests? Yes, with an asterisk that seemingly most modules were skipped so most tests were not run. > > Files=13, Tests=93, 24 wallclock secs ( 8.67 cusr + 6.69 csys = 15.36 CPU)
Re: Apache httpd 2.4.x on Windows
Good news! Gianugo and Claudio have offered to see how they can help us in resolving the SSL issues w/ httpd 2.4.x... As you can see, Claudio is the main guru who may be able to provide the most insight. I've CC'ed both on this Email. Claudio, if you like, you may wish to subscribe to the dev@httpd.apache.org mailing list. On Apr 4, 2012, at 1:57 PM, Gianugo Rabellino wrote: > Jim, meet Claudio Caldato. Claudio is a fellow Italian and we work in the > same team. He's the one to blame for making node.js sing on Windows and he'd > be happy to start investigating how we can help httpd as well. Let me know if > I can be of any further help. > > -- > Gianugo Rabellino > Sr. Director, Open Source Communities, Microsoft Corp. > Mobile: +1 (425) 786 8646 - Twitter: @gianugo > > -Original Message- > From: Jim Jagielski [mailto:j...@apache.org] > Sent: Wednesday, April 04, 2012 7:25 AM > To: Gianugo Rabellino > Subject: Apache httpd 2.4.x on Windows > > Gianugo, > > I hope this email finds you well... I missed seeing you again at POSSCON this > year. > > We (the web server PMC) have an issue with the latest revs of Apache (2.4.x) > on Windows, such that we can't get SSL to work well with the Windows > networking stack... as such, even though it's GA for Unix and Unix-like OSs, > Windows is still considered beta. We'd like to address that :) > > So I was wondering if there's a team in MS we could ping to get some better > insights on what might be going on... we want to ensure that Windows is a > fully supported OS for httpd. > > Thanks! > > > > >
Re: mod_dav_fs does not check for return value on stream_close
On 15 Mar 2012, at 3:56 PM, Brian J. France wrote:
> Could somebody review the patch below for 2.2, 2.4, and trunk?
>
> A better error message could be sent, but I am more worried about how the
> return will effect the code after it.
>
> I am thinking the file needs to be removed either via a apr_file_remove call
> or:
>
> apr_pool_cleanup_kill(stream->p, stream, tmpfile_cleanup);
>
> call, but I don't know the code well enough to know which is right and
> 2.4/trunk adds even more complexity compared to 2.2.x.
>
> Thoughts/Comments?
>
> Thanks,
>
> Brian
>
> - I am still getting more details why close is failing, but for some reason
> it is happening enough to cause issues. (responding 200, but no file)
>
> -
>
> 2.2:
>
> Index: modules/dav/fs/repos.c
> ===
> --- modules/dav/fs/repos.c(revision 1300964)
> +++ modules/dav/fs/repos.c(working copy)
> @@ -881,6 +881,10 @@
> {
> apr_file_close(stream->f);
Would be above not be
status = apr_file_close(stream->f);
>
> +if (status != APR_SUCCESS) {
> + return dav_new_error(stream->p, MAP_IO2HTTP(status), 0, "There was a
> problem closing the stream");
> +}
> +
> if (!commit) {
> if (apr_file_remove(stream->pathname, stream->p) != APR_SUCCESS) {
> /* ### use a better description? */
>
>
> 2.24 and trunk:
>
>
> Index: modules/dav/fs/repos.c
> ===
> --- modules/dav/fs/repos.c(revision 1300964)
> +++ modules/dav/fs/repos.c(working copy)
> @@ -970,6 +970,10 @@
>
> apr_file_close(stream->f);
Same with this one.
>
> +if (status != APR_SUCCESS) {
> + return dav_new_error(stream->p, MAP_IO2HTTP(status), 0, "There was a
> problem closing the stream");
> +}
> +
> if (!commit) {
> if (stream->temppath) {
> apr_pool_cleanup_run(stream->p, stream, tmpfile_cleanup);
>
>
Regards,
Graham
--
smime.p7s
Description: S/MIME cryptographic signature
Re: mod_dav_fs does not check for return value on stream_close
Yes, you are correct. Looks like my merge from work code to httpd svn didn't
fully work.
We have been running this patch for a week or so with no issues.
Brian
On Apr 4, 2012, at 6:24 PM, Graham Leggett wrote:
> On 15 Mar 2012, at 3:56 PM, Brian J. France wrote:
>
>> Could somebody review the patch below for 2.2, 2.4, and trunk?
>>
>> A better error message could be sent, but I am more worried about how the
>> return will effect the code after it.
>>
>> I am thinking the file needs to be removed either via a apr_file_remove call
>> or:
>>
>> apr_pool_cleanup_kill(stream->p, stream, tmpfile_cleanup);
>>
>> call, but I don't know the code well enough to know which is right and
>> 2.4/trunk adds even more complexity compared to 2.2.x.
>>
>> Thoughts/Comments?
>>
>> Thanks,
>>
>> Brian
>>
>> - I am still getting more details why close is failing, but for some reason
>> it is happening enough to cause issues. (responding 200, but no file)
>>
>> -
>>
>> 2.2:
>>
>> Index: modules/dav/fs/repos.c
>> ===
>> --- modules/dav/fs/repos.c (revision 1300964)
>> +++ modules/dav/fs/repos.c (working copy)
>> @@ -881,6 +881,10 @@
>> {
>>apr_file_close(stream->f);
>
> Would be above not be
>
> status = apr_file_close(stream->f);
>
>>
>> +if (status != APR_SUCCESS) {
>> +return dav_new_error(stream->p, MAP_IO2HTTP(status), 0, "There was a
>> problem closing the stream");
>> +}
>> +
>>if (!commit) {
>>if (apr_file_remove(stream->pathname, stream->p) != APR_SUCCESS) {
>>/* ### use a better description? */
>>
>>
>> 2.24 and trunk:
>>
>>
>> Index: modules/dav/fs/repos.c
>> ===
>> --- modules/dav/fs/repos.c (revision 1300964)
>> +++ modules/dav/fs/repos.c (working copy)
>> @@ -970,6 +970,10 @@
>>
>>apr_file_close(stream->f);
>
> Same with this one.
>
>>
>> +if (status != APR_SUCCESS) {
>> +return dav_new_error(stream->p, MAP_IO2HTTP(status), 0, "There was a
>> problem closing the stream");
>> +}
>> +
>>if (!commit) {
>>if (stream->temppath) {
>>apr_pool_cleanup_run(stream->p, stream, tmpfile_cleanup);
>>
>>
>
> Regards,
> Graham
> --
>
Re: Apache httpd 2.4.x on Windows
Very cool :) Claudio, we can point you precisely to where httpd 2.4 goes off the rails. We instantiate a socket connection (accept() or AcceptEx() yet our presumptions about the state of the socket (block/nonblock timeouts etc) were not honored. There is a call, apr_os_socket_make() which APR makes certain assumptions about. Seems too fragile. Perhaps that _make() call needs to IoCtl at the socket to determine all of the 'anticipated' state conditions to be able to roll on with the correct assumptions. On 4/4/2012 5:13 PM, Jim Jagielski wrote: > Good news! Gianugo and Claudio have offered to see how they > can help us in resolving the SSL issues w/ httpd 2.4.x... > As you can see, Claudio is the main guru who may be able > to provide the most insight. > > I've CC'ed both on this Email. > > Claudio, if you like, you may wish to subscribe to the dev@httpd.apache.org > mailing list. > > On Apr 4, 2012, at 1:57 PM, Gianugo Rabellino wrote: > >> Jim, meet Claudio Caldato. Claudio is a fellow Italian and we work in the >> same team. He's the one to blame for making node.js sing on Windows and he'd >> be happy to start investigating how we can help httpd as well. Let me know >> if I can be of any further help. >> >> -- >> Gianugo Rabellino >> Sr. Director, Open Source Communities, Microsoft Corp. >> Mobile: +1 (425) 786 8646 - Twitter: @gianugo >> >> -Original Message- >> From: Jim Jagielski [mailto:j...@apache.org] >> Sent: Wednesday, April 04, 2012 7:25 AM >> To: Gianugo Rabellino >> Subject: Apache httpd 2.4.x on Windows >> >> Gianugo, >> >> I hope this email finds you well... I missed seeing you again at POSSCON >> this year. >> >> We (the web server PMC) have an issue with the latest revs of Apache (2.4.x) >> on Windows, such that we can't get SSL to work well with the Windows >> networking stack... as such, even though it's GA for Unix and Unix-like OSs, >> Windows is still considered beta. We'd like to address that :) >> >> So I was wondering if there's a team in MS we could ping to get some better >> insights on what might be going on... we want to ensure that Windows is a >> fully supported OS for httpd. >> >> Thanks! >> >> >> >> >> > >
