You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@couchdb.apache.org by Dave Cottlehuber <dc...@jsonified.com> on 2013/03/31 15:29:22 UTC
Add OTP R16B support + upgrade Mochiweb to 2.4.2
Hi everybody,
I have a couple more tests to make on ensuring the timezone bug in
earlier OTP releases is correctly resolved, but the bulk of the work
is done. Unless somebody reports an issue, I'll include this branch in
master in the next week, probably cherry-picked.
[jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
[branch]: https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
A+
Dave
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Klaus Trainer <kl...@posteo.de>.
Btw., I just run the tests after I put tmp/lib onto a ramdisk, basically
the same results, except that the form_submit test now seems to pass
always (whereas without the ramdisk it only passes sometimes).
K
On Tue, 2013-04-02 at 10:33 +0200, Klaus Trainer wrote:
> On Mon, 2013-04-01 at 23:26 +0200, Dave Cottlehuber wrote:
> > If you run those by hand with an extra `-v` option, can you share the logs
> > somewhere?
>
> Here you are: https://www.friendpaste.com/19YnF5A2nZHea25LEShpAi
>
> >
> >
> > On 1 April 2013 00:33, Klaus Trainer <kl...@posteo.de> wrote:
> >
> > > On Sun, 2013-03-31 at 22:46 +0200, Jan Lehnardt wrote:
> > > > On Mar 31, 2013, at 22:43 , Klaus Trainer <kl...@posteo.de>
> > > wrote:
> > > >
> > > > > Dave, what tests do you have exactly failing?
> > > > >
> > > > > I got nine (consistently) failing tests. However, it might be possible
> > > > > that a few of them are timeout related, as I'm still working with
> > > > > spinning discs (do you remember those?) on my computer here. That's why
> > > > > I'd like to compare.
> > > >
> > > > Can you post your list? :)
> > >
> > > Sure, why not :)
> > >
> > > not ok 27 form_submit
> > > not ok 36 oauth
> > > not ok 39 purge
> > > not ok 46 replication
> > > not ok 47 replicator_db
> > > not ok 49 rev_stemming
> > > not ok 51 security_validation
> > > not ok 62 view_conflicts
> > > not ok 64 view_include_docs
> > >
> > > >
> > > >
> > > > >
> > > > > K
> > > > >
> > > > >
> > > > > On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
> > > > >> Hi everybody,
> > > > >>
> > > > >> I have a couple more tests to make on ensuring the timezone bug in
> > > > >> earlier OTP releases is correctly resolved, but the bulk of the work
> > > > >> is done. Unless somebody reports an issue, I'll include this branch in
> > > > >> master in the next week, probably cherry-picked.
> > > > >>
> > > > >> [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
> > > > >> [branch]:
> > > https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
> > > > >>
> > > > >> A+
> > > > >> Dave
> > > > >
> > > > >
> > > >
> > >
> > >
> > >
>
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Dave Cottlehuber <dc...@jsonified.com>.
Hi Benoit,
r16 has a fix for the TZ issue anyway, and I suspect mochiweb already has
something there. If it doesn't I'll send that upstream.
On 2 April 2013 11:41, Benoit Chesneau <bc...@gmail.com> wrote:
> not totally related but why not sending the TZ patch upstream btw? Is
> there anything against that?
>
> I'm worried we have to maintain our own versions for dependencies for
> a long time instead of pushing upstream our changes I reckon that for
> a limited time we have to maintain our own set of patches (upstream
> integration can be slower or don't want the patch as is - which may be
> a big warning for us -) but in that cad I suggest that we also put the
> patch as a diff in our own repo under a `patches/` folder so it can be
> integrated differently by vendors/distributors and in that case e
> those who are using a separate mochiweb install. Also it would help us
> to synchronize our changes with upstream.
>
> - benoît
>
>
> On Tue, Apr 2, 2013 at 10:33 AM, Klaus Trainer <kl...@posteo.de>
> wrote:
> > On Mon, 2013-04-01 at 23:26 +0200, Dave Cottlehuber wrote:
> >> If you run those by hand with an extra `-v` option, can you share the
> logs
> >> somewhere?
> >
> > Here you are: https://www.friendpaste.com/19YnF5A2nZHea25LEShpAi
> >
> >>
> >>
> >> On 1 April 2013 00:33, Klaus Trainer <kl...@posteo.de> wrote:
> >>
> >> > On Sun, 2013-03-31 at 22:46 +0200, Jan Lehnardt wrote:
> >> > > On Mar 31, 2013, at 22:43 , Klaus Trainer <kl...@posteo.de>
> >> > wrote:
> >> > >
> >> > > > Dave, what tests do you have exactly failing?
> >> > > >
> >> > > > I got nine (consistently) failing tests. However, it might be
> possible
> >> > > > that a few of them are timeout related, as I'm still working with
> >> > > > spinning discs (do you remember those?) on my computer here.
> That's why
> >> > > > I'd like to compare.
> >> > >
> >> > > Can you post your list? :)
> >> >
> >> > Sure, why not :)
> >> >
> >> > not ok 27 form_submit
> >> > not ok 36 oauth
> >> > not ok 39 purge
> >> > not ok 46 replication
> >> > not ok 47 replicator_db
> >> > not ok 49 rev_stemming
> >> > not ok 51 security_validation
> >> > not ok 62 view_conflicts
> >> > not ok 64 view_include_docs
> >> >
> >> > >
> >> > >
> >> > > >
> >> > > > K
> >> > > >
> >> > > >
> >> > > > On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
> >> > > >> Hi everybody,
> >> > > >>
> >> > > >> I have a couple more tests to make on ensuring the timezone bug
> in
> >> > > >> earlier OTP releases is correctly resolved, but the bulk of the
> work
> >> > > >> is done. Unless somebody reports an issue, I'll include this
> branch in
> >> > > >> master in the next week, probably cherry-picked.
> >> > > >>
> >> > > >> [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
> >> > > >> [branch]:
> >> >
> https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
> >> > > >>
> >> > > >> A+
> >> > > >> Dave
> >> > > >
> >> > > >
> >> > >
> >> >
> >> >
> >> >
> >
>
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Benoit Chesneau <bc...@gmail.com>.
not totally related but why not sending the TZ patch upstream btw? Is
there anything against that?
I'm worried we have to maintain our own versions for dependencies for
a long time instead of pushing upstream our changes I reckon that for
a limited time we have to maintain our own set of patches (upstream
integration can be slower or don't want the patch as is - which may be
a big warning for us -) but in that cad I suggest that we also put the
patch as a diff in our own repo under a `patches/` folder so it can be
integrated differently by vendors/distributors and in that case e
those who are using a separate mochiweb install. Also it would help us
to synchronize our changes with upstream.
- benoît
On Tue, Apr 2, 2013 at 10:33 AM, Klaus Trainer <kl...@posteo.de> wrote:
> On Mon, 2013-04-01 at 23:26 +0200, Dave Cottlehuber wrote:
>> If you run those by hand with an extra `-v` option, can you share the logs
>> somewhere?
>
> Here you are: https://www.friendpaste.com/19YnF5A2nZHea25LEShpAi
>
>>
>>
>> On 1 April 2013 00:33, Klaus Trainer <kl...@posteo.de> wrote:
>>
>> > On Sun, 2013-03-31 at 22:46 +0200, Jan Lehnardt wrote:
>> > > On Mar 31, 2013, at 22:43 , Klaus Trainer <kl...@posteo.de>
>> > wrote:
>> > >
>> > > > Dave, what tests do you have exactly failing?
>> > > >
>> > > > I got nine (consistently) failing tests. However, it might be possible
>> > > > that a few of them are timeout related, as I'm still working with
>> > > > spinning discs (do you remember those?) on my computer here. That's why
>> > > > I'd like to compare.
>> > >
>> > > Can you post your list? :)
>> >
>> > Sure, why not :)
>> >
>> > not ok 27 form_submit
>> > not ok 36 oauth
>> > not ok 39 purge
>> > not ok 46 replication
>> > not ok 47 replicator_db
>> > not ok 49 rev_stemming
>> > not ok 51 security_validation
>> > not ok 62 view_conflicts
>> > not ok 64 view_include_docs
>> >
>> > >
>> > >
>> > > >
>> > > > K
>> > > >
>> > > >
>> > > > On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
>> > > >> Hi everybody,
>> > > >>
>> > > >> I have a couple more tests to make on ensuring the timezone bug in
>> > > >> earlier OTP releases is correctly resolved, but the bulk of the work
>> > > >> is done. Unless somebody reports an issue, I'll include this branch in
>> > > >> master in the next week, probably cherry-picked.
>> > > >>
>> > > >> [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
>> > > >> [branch]:
>> > https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
>> > > >>
>> > > >> A+
>> > > >> Dave
>> > > >
>> > > >
>> > >
>> >
>> >
>> >
>
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Klaus Trainer <kl...@posteo.de>.
On Mon, 2013-04-01 at 23:26 +0200, Dave Cottlehuber wrote:
> If you run those by hand with an extra `-v` option, can you share the logs
> somewhere?
Here you are: https://www.friendpaste.com/19YnF5A2nZHea25LEShpAi
>
>
> On 1 April 2013 00:33, Klaus Trainer <kl...@posteo.de> wrote:
>
> > On Sun, 2013-03-31 at 22:46 +0200, Jan Lehnardt wrote:
> > > On Mar 31, 2013, at 22:43 , Klaus Trainer <kl...@posteo.de>
> > wrote:
> > >
> > > > Dave, what tests do you have exactly failing?
> > > >
> > > > I got nine (consistently) failing tests. However, it might be possible
> > > > that a few of them are timeout related, as I'm still working with
> > > > spinning discs (do you remember those?) on my computer here. That's why
> > > > I'd like to compare.
> > >
> > > Can you post your list? :)
> >
> > Sure, why not :)
> >
> > not ok 27 form_submit
> > not ok 36 oauth
> > not ok 39 purge
> > not ok 46 replication
> > not ok 47 replicator_db
> > not ok 49 rev_stemming
> > not ok 51 security_validation
> > not ok 62 view_conflicts
> > not ok 64 view_include_docs
> >
> > >
> > >
> > > >
> > > > K
> > > >
> > > >
> > > > On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
> > > >> Hi everybody,
> > > >>
> > > >> I have a couple more tests to make on ensuring the timezone bug in
> > > >> earlier OTP releases is correctly resolved, but the bulk of the work
> > > >> is done. Unless somebody reports an issue, I'll include this branch in
> > > >> master in the next week, probably cherry-picked.
> > > >>
> > > >> [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
> > > >> [branch]:
> > https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
> > > >>
> > > >> A+
> > > >> Dave
> > > >
> > > >
> > >
> >
> >
> >
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Dave Cottlehuber <dc...@jsonified.com>.
If you run those by hand with an extra `-v` option, can you share the logs
somewhere?
On 1 April 2013 00:33, Klaus Trainer <kl...@posteo.de> wrote:
> On Sun, 2013-03-31 at 22:46 +0200, Jan Lehnardt wrote:
> > On Mar 31, 2013, at 22:43 , Klaus Trainer <kl...@posteo.de>
> wrote:
> >
> > > Dave, what tests do you have exactly failing?
> > >
> > > I got nine (consistently) failing tests. However, it might be possible
> > > that a few of them are timeout related, as I'm still working with
> > > spinning discs (do you remember those?) on my computer here. That's why
> > > I'd like to compare.
> >
> > Can you post your list? :)
>
> Sure, why not :)
>
> not ok 27 form_submit
> not ok 36 oauth
> not ok 39 purge
> not ok 46 replication
> not ok 47 replicator_db
> not ok 49 rev_stemming
> not ok 51 security_validation
> not ok 62 view_conflicts
> not ok 64 view_include_docs
>
> >
> >
> > >
> > > K
> > >
> > >
> > > On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
> > >> Hi everybody,
> > >>
> > >> I have a couple more tests to make on ensuring the timezone bug in
> > >> earlier OTP releases is correctly resolved, but the bulk of the work
> > >> is done. Unless somebody reports an issue, I'll include this branch in
> > >> master in the next week, probably cherry-picked.
> > >>
> > >> [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
> > >> [branch]:
> https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
> > >>
> > >> A+
> > >> Dave
> > >
> > >
> >
>
>
>
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Klaus Trainer <kl...@posteo.de>.
On Sun, 2013-03-31 at 22:46 +0200, Jan Lehnardt wrote:
> On Mar 31, 2013, at 22:43 , Klaus Trainer <kl...@posteo.de> wrote:
>
> > Dave, what tests do you have exactly failing?
> >
> > I got nine (consistently) failing tests. However, it might be possible
> > that a few of them are timeout related, as I'm still working with
> > spinning discs (do you remember those?) on my computer here. That's why
> > I'd like to compare.
>
> Can you post your list? :)
Sure, why not :)
not ok 27 form_submit
not ok 36 oauth
not ok 39 purge
not ok 46 replication
not ok 47 replicator_db
not ok 49 rev_stemming
not ok 51 security_validation
not ok 62 view_conflicts
not ok 64 view_include_docs
>
>
> >
> > K
> >
> >
> > On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
> >> Hi everybody,
> >>
> >> I have a couple more tests to make on ensuring the timezone bug in
> >> earlier OTP releases is correctly resolved, but the bulk of the work
> >> is done. Unless somebody reports an issue, I'll include this branch in
> >> master in the next week, probably cherry-picked.
> >>
> >> [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
> >> [branch]: https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
> >>
> >> A+
> >> Dave
> >
> >
>
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Jan Lehnardt <ja...@apache.org>.
On Mar 31, 2013, at 22:43 , Klaus Trainer <kl...@posteo.de> wrote:
> Dave, what tests do you have exactly failing?
>
> I got nine (consistently) failing tests. However, it might be possible
> that a few of them are timeout related, as I'm still working with
> spinning discs (do you remember those?) on my computer here. That's why
> I'd like to compare.
Can you post your list? :)
>
> K
>
>
> On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
>> Hi everybody,
>>
>> I have a couple more tests to make on ensuring the timezone bug in
>> earlier OTP releases is correctly resolved, but the bulk of the work
>> is done. Unless somebody reports an issue, I'll include this branch in
>> master in the next week, probably cherry-picked.
>>
>> [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
>> [branch]: https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
>>
>> A+
>> Dave
>
>
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Klaus Trainer <kl...@posteo.de>.
That's totally strange and feels a bit embarrassing.
Here's what I enter in my shell in order to do the respective testing:
g fetch
g reset --hard origin/1696-update-mochiweb-2-4-2
./bootstrap && ./configure && make && make check
Actually, there are no untracked unignored files or any strange things
*that I'd know of*.
Now, when I clone a new copy of the repo to a different location and do
the same there, everything works fine, except
that ./test/etap/200-view-group-no-db-leaks.t keeps failing.
So, the good news is that I can confirm that
beside ./test/etap/200-view-group-no-db-leaks.t (which keeps failing)
everything seems to be fine. The bad news is that I still don't know
what's wrong in my other repo clone.
K
On Mon, 2013-04-01 at 23:20 +0200, Dave Cottlehuber wrote:
> Hey Klaus,
>
> There's nothing failing in make check for me on OSX, but I need to flick
> back to UTC +7 or whatever if was, and see if the TZ bug is still fixed.
> Bob did a patch for this originally and I think mochiweb has now done
> something similar, but I've not validated that.
>
> PS it's time to upgrade
> http://eightface.com/2012/11/worlds-oldest-working-digital-computer/
> http://laughingsquid.com/1951-harwell-dekatron-the-oldest-original-working-digital-computer/
>
>
> On 31 March 2013 22:43, Klaus Trainer <kl...@posteo.de> wrote:
>
> > Dave, what tests do you have exactly failing?
> >
> > I got nine (consistently) failing tests. However, it might be possible
> > that a few of them are timeout related, as I'm still working with
> > spinning discs (do you remember those?) on my computer here. That's why
> > I'd like to compare.
> >
> > K
> >
> >
> > On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
> > > Hi everybody,
> > >
> > > I have a couple more tests to make on ensuring the timezone bug in
> > > earlier OTP releases is correctly resolved, but the bulk of the work
> > > is done. Unless somebody reports an issue, I'll include this branch in
> > > master in the next week, probably cherry-picked.
> > >
> > > [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
> > > [branch]:
> > https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
> > >
> > > A+
> > > Dave
> >
> >
> >
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Dave Cottlehuber <dc...@jsonified.com>.
Hey Klaus,
There's nothing failing in make check for me on OSX, but I need to flick
back to UTC +7 or whatever if was, and see if the TZ bug is still fixed.
Bob did a patch for this originally and I think mochiweb has now done
something similar, but I've not validated that.
PS it's time to upgrade
http://eightface.com/2012/11/worlds-oldest-working-digital-computer/
http://laughingsquid.com/1951-harwell-dekatron-the-oldest-original-working-digital-computer/
On 31 March 2013 22:43, Klaus Trainer <kl...@posteo.de> wrote:
> Dave, what tests do you have exactly failing?
>
> I got nine (consistently) failing tests. However, it might be possible
> that a few of them are timeout related, as I'm still working with
> spinning discs (do you remember those?) on my computer here. That's why
> I'd like to compare.
>
> K
>
>
> On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
> > Hi everybody,
> >
> > I have a couple more tests to make on ensuring the timezone bug in
> > earlier OTP releases is correctly resolved, but the bulk of the work
> > is done. Unless somebody reports an issue, I'll include this branch in
> > master in the next week, probably cherry-picked.
> >
> > [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
> > [branch]:
> https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
> >
> > A+
> > Dave
>
>
>
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Klaus Trainer <kl...@posteo.de>.
Dave, what tests do you have exactly failing?
I got nine (consistently) failing tests. However, it might be possible
that a few of them are timeout related, as I'm still working with
spinning discs (do you remember those?) on my computer here. That's why
I'd like to compare.
K
On Sun, 2013-03-31 at 15:29 +0200, Dave Cottlehuber wrote:
> Hi everybody,
>
> I have a couple more tests to make on ensuring the timezone bug in
> earlier OTP releases is correctly resolved, but the bulk of the work
> is done. Unless somebody reports an issue, I'll include this branch in
> master in the next week, probably cherry-picked.
>
> [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
> [branch]: https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
>
> A+
> Dave
Re: Add OTP R16B support + upgrade Mochiweb to 2.4.2
Posted by Noah Slater <ns...@apache.org>.
Thanks Dave!
On 31 March 2013 14:29, Dave Cottlehuber <dc...@jsonified.com> wrote:
> Hi everybody,
>
> I have a couple more tests to make on ensuring the timezone bug in
> earlier OTP releases is correctly resolved, but the bulk of the work
> is done. Unless somebody reports an issue, I'll include this branch in
> master in the next week, probably cherry-picked.
>
> [jira]: https://issues.apache.org/jira/browse/COUCHDB-1696
> [branch]:
> https://git-wip-us.apache.org/repos/asf?p=couchdb.git;a=tree;h=refs/heads/1696-update-mochiweb-2-4-2;hb=1696-update-mochiweb-2-4-2
>
> A+
> Dave
>
--
NS