documentation build failing in hackage?

Sorry if this has been asked/discussed previously It seems that there are many packages in which the documentation is not generated. That is the visible effect since there is no indication if the build process failed or not. Could anyone give me some hints? I did a lazy look at the hackage2 documentation but I found nothing about that. Only that it is possible to upload the documentation if the build bot fails... -- Alberto.

On 14-10-05 05:40 AM, Alberto G. Corona wrote:
Sorry if this has been asked/discussed previously
It seems that there are many packages in which the documentation is not generated. That is the visible effect since there is no indication if the build process failed or not.
My recollection: One or two weeks ago, Hackage doc building was disabled because it was making the server really slow. (The same server had to serve you Hackage downloads. You don't want slow downloads.) It probably is still disabled.

On Sun, Oct 5, 2014 at 6:27 PM, Albert Y. C. Lai
My recollection: One or two weeks ago, Hackage doc building was disabled because it was making the server really slow. (The same server had to serve you Hackage downloads. You don't want slow downloads.)
Wasbn't Hackage emergency relocated to a new server about a week ago? Possibly something is still not quite set up right --- or it copied over the settings from the old including docs being disabled, and nobody remembered to re-enable them. -- brandon s allbery kf8nh sine nomine associates allbery.b@gmail.com ballbery@sinenomine.net unix, openafs, kerberos, infrastructure, xmonad http://sinenomine.net

On Sun, Oct 05, 2014 at 06:27:12PM -0400, Albert Y. C. Lai wrote:
On 14-10-05 05:40 AM, Alberto G. Corona wrote:
Sorry if this has been asked/discussed previously
It seems that there are many packages in which the documentation is not generated. That is the visible effect since there is no indication if the build process failed or not.
My recollection: One or two weeks ago, Hackage doc building was disabled because it was making the server really slow. (The same server had to serve you Hackage downloads. You don't want slow downloads.)
It probably is still disabled.
Wasn't there also an announcement of the possibility of uploading generated docs where Hackage fails? Indeed: - http://fuuzetsu.co.uk/blog/posts/2014-01-06-Fix-your-Hackage-documentation.h... - http://fuuzetsu.co.uk/blog/posts/2014-01-06-Hackage-documentation-v2.html Hopefully those two posts are still valid and useful. /M -- Magnus Therning OpenPGP: 0xAB4DFBA4 email: magnus@therning.org jabber: magnus@therning.org twitter: magthe http://therning.org/magnus Perl is another example of filling a tiny, short-term need, and then being a real problem in the longer term. -- Alan Kay

Thanks.
Looking at the log, using the above links, mine has not been processed
since it was uploaded at October 1 (there is no build log).
https://hackage.haskell.org/package/haste-perch-0.1.0.4/reports/
According with the logs, there are packages that are not build since the
end of September
For example:
http://hackage.haskell.org/package/haste-perch-0.1.0.4/reports/
It is not a question of documentation build, but the whole build process
2014-10-06 12:22 GMT+02:00 Magnus Therning
On Sun, Oct 05, 2014 at 06:27:12PM -0400, Albert Y. C. Lai wrote:
On 14-10-05 05:40 AM, Alberto G. Corona wrote:
Sorry if this has been asked/discussed previously
It seems that there are many packages in which the documentation is not generated. That is the visible effect since there is no indication if the build process failed or not.
My recollection: One or two weeks ago, Hackage doc building was disabled because it was making the server really slow. (The same server had to serve you Hackage downloads. You don't want slow downloads.)
It probably is still disabled.
Wasn't there also an announcement of the possibility of uploading generated docs where Hackage fails?
Indeed:
- http://fuuzetsu.co.uk/blog/posts/2014-01-06-Fix-your-Hackage-documentation.h... - http://fuuzetsu.co.uk/blog/posts/2014-01-06-Hackage-documentation-v2.html
Hopefully those two posts are still valid and useful.
/M
-- Magnus Therning OpenPGP: 0xAB4DFBA4 email: magnus@therning.org jabber: magnus@therning.org twitter: magthe http://therning.org/magnus
Perl is another example of filling a tiny, short-term need, and then being a real problem in the longer term. -- Alan Kay
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Alberto.

Niklas, This can be used to manually upload hackage docs in the meantime.
https://github.com/ekmett/lens/blob/master/scripts/hackage-docs.sh
On Sat, Oct 11, 2014 at 5:53 PM, Niklas Hambüchen
Is anything coming out of this?
Will builds come back? _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Cell: 1.630.740.8204

Will builds come back?
Looking at http://hackage.haskell.org/packages/recent I have the impression it works OK now (with some delay, as always). Are you still experiencing problems? For a package uploaded right now or in the preceding weeks?

In my case, the two packages are still not built. I looked at some others
which had documentation generated since then, for example:
https://hackage.haskell.org/package/haste-compiler-0.4.3
But the documentation seem not to be generated by hackage, since there is
no activity log: is empty:
https://hackage.haskell.org/package/haste-compiler-0.4.3/reports/
It may have been uploaded.
It is not a problem of documentation but the entire build process. That is
a big issue. Can we contribute in some way?. computing power should not be
a problem today....
2014-10-12 1:24 GMT+02:00 Mikolaj Konarski
Will builds come back?
Looking at
http://hackage.haskell.org/packages/recent
I have the impression it works OK now (with some delay, as always). Are you still experiencing problems? For a package uploaded right now or in the preceding weeks? _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Alberto.

Its a matter of manpower on the entirely volunteers infrastructure team.
Talk with them Ti figure out how you can help.
A few people are working on setting up some mitigation, but I hesitate to
name names until we've gotten the mitigations running live.
On Oct 11, 2014 7:38 PM, "Alberto G. Corona"
In my case, the two packages are still not built. I looked at some others which had documentation generated since then, for example:
https://hackage.haskell.org/package/haste-compiler-0.4.3
But the documentation seem not to be generated by hackage, since there is no activity log: is empty:
https://hackage.haskell.org/package/haste-compiler-0.4.3/reports/
It may have been uploaded.
It is not a problem of documentation but the entire build process. That is a big issue. Can we contribute in some way?. computing power should not be a problem today....
2014-10-12 1:24 GMT+02:00 Mikolaj Konarski
: Will builds come back?
Looking at
http://hackage.haskell.org/packages/recent
I have the impression it works OK now (with some delay, as always). Are you still experiencing problems? For a package uploaded right now or in the preceding weeks? _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Alberto.
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

On 10/12/2014 01:14 AM, Carter Schonwald wrote:
Its a matter of manpower on the entirely volunteers infrastructure team. Talk with them Ti figure out how you can help.
A few people are working on setting up some mitigation, but I hesitate to name names until we've gotten the mitigations running live.
The biggest problem is that there is no official statement as to why the builds stopped and an estimate of when they are coming back.
On Oct 11, 2014 7:38 PM, "Alberto G. Corona"
wrote: In my case, the two packages are still not built. I looked at some others which had documentation generated since then, for example:
https://hackage.haskell.org/package/haste-compiler-0.4.3
But the documentation seem not to be generated by hackage, since there is no activity log: is empty:
https://hackage.haskell.org/package/haste-compiler-0.4.3/reports/
It may have been uploaded.
It is not a problem of documentation but the entire build process. That is a big issue. Can we contribute in some way?. computing power should not be a problem today....
2014-10-12 1:24 GMT+02:00 Mikolaj Konarski
: Will builds come back?
Looking at
http://hackage.haskell.org/packages/recent
I have the impression it works OK now (with some delay, as always). Are you still experiencing problems? For a package uploaded right now or in the preceding weeks? _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Alberto.
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Mateusz K.

Indeed; this stuff is pretty much all done by volunteers so far as I know. So I'm not asking it to always work perfectly, but it would be nice to have a better way to find out that this stuff is turned off than just folklore and guessing. Perhaps in the future an announcement that explains the problem could be made? (Forgive me if one was made but I missed it) In the past, whether or not the documentation built for your package was a sort of cheap way to figure out if your package built for people other than you. Kind regards, Jon On Sun, Oct 12, 2014, at 04:50 AM, Mateusz Kowalczyk wrote:
On 10/12/2014 01:14 AM, Carter Schonwald wrote:
Its a matter of manpower on the entirely volunteers infrastructure team. Talk with them Ti figure out how you can help.
A few people are working on setting up some mitigation, but I hesitate to name names until we've gotten the mitigations running live.
The biggest problem is that there is no official statement as to why the builds stopped and an estimate of when they are coming back.
On Oct 11, 2014 7:38 PM, "Alberto G. Corona"
wrote: In my case, the two packages are still not built. I looked at some others which had documentation generated since then, for example:
https://hackage.haskell.org/package/haste-compiler-0.4.3
But the documentation seem not to be generated by hackage, since there is no activity log: is empty:
https://hackage.haskell.org/package/haste-compiler-0.4.3/reports/
It may have been uploaded.
It is not a problem of documentation but the entire build process. That is a big issue. Can we contribute in some way?. computing power should not be a problem today....
2014-10-12 1:24 GMT+02:00 Mikolaj Konarski
: Will builds come back?
Looking at
http://hackage.haskell.org/packages/recent
I have the impression it works OK now (with some delay, as always). Are you still experiencing problems? For a package uploaded right now or in the preceding weeks? _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Alberto.
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- Mateusz K. _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

Jon Sterling
Indeed; this stuff is pretty much all done by volunteers so far as I know. So I'm not asking it to always work perfectly, but it would be nice to have a better way to find out that this stuff is turned off than just folklore and guessing. Perhaps in the future an announcement that explains the problem could be made? (Forgive me if one was made but I missed it)
Perhaps it would make sense for those involved in the infrastructure effort to maintain a status page on the wiki for these sorts of things. It doesn't need to be much, just a quick sentence or two describing what they are currently working on. Cheers, - Ben

I was wrong and you were right about the halted haddock building. So I asked around and, just as Carter reports, I confirmed a few kind people are currently working on setting up new doc building bots and it doesn't look too bad. If you'd like to offer help to the volunteer infrastructure team, please send a message to the haskell-infrastructure list at http://community.galois.com/pipermail/haskell-infrastructure or keep pinging people on #haskell-infrastructure on freenode IRC. This is also where you can, with some luck, get information about the current status of the infrastructure. Major disruptions are usually reported on https://status.haskell.org/ but the page needs to be manually updated and, as we can see, it's not always accurate when partial disruptions take place. Cheers, Mikolaj

Indeed! And I'd like to emphasize that volunteering is really about jumping
in and doing legwork to help. This can be things like contributing to core
infrastructure, setting up a mirror. Or whatever.
A lot of the nice haskell tooling we have is a product of the hard work of
many (though relatively small pool of many) volunteers wanting to make
things better! Imagine how much nicer things can get with some deliberate
carefully aimed work by a few more folks!
On Oct 13, 2014 12:23 PM, "Mikolaj Konarski"
I was wrong and you were right about the halted haddock building.
So I asked around and, just as Carter reports, I confirmed a few kind people are currently working on setting up new doc building bots and it doesn't look too bad.
If you'd like to offer help to the volunteer infrastructure team, please send a message to the haskell-infrastructure list at http://community.galois.com/pipermail/haskell-infrastructure or keep pinging people on #haskell-infrastructure on freenode IRC.
This is also where you can, with some luck, get information about the current status of the infrastructure. Major disruptions are usually reported on https://status.haskell.org/ but the page needs to be manually updated and, as we can see, it's not always accurate when partial disruptions take place.
Cheers, Mikolaj _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
participants (11)
-
Albert Y. C. Lai
-
Alberto G. Corona
-
Ben Gamari
-
Brandon Allbery
-
Carter Schonwald
-
David Johnson
-
Jon Sterling
-
Magnus Therning
-
Mateusz Kowalczyk
-
Mikolaj Konarski
-
Niklas Hambüchen