Please build git-annex with webapp enabled

I'm sorry about the confusion here. Last cycle git-annex got caught up in a very complex transition to GHC 7.6 and we only barely managed to keep it in the release at all; sadly I didn't notice this bug or I might have tried to fix it. We weren't explicitly disabling the webapp, and I can't quite work out why it vanished, but -DWITH_WEBAPP is indeed missing from the preprocessor flags ...

Anyway, I've uploaded 4.20130501ubuntu1 to saucy (it still needs a handful of control file modifications for GHC 7.6, which I'll forward if necessary once Debian gets that far) and my local build has the webapp enabled. I'll close this bug once I've checked that the official builds do too.

I've confirmed that the official build of 4.20130501ubuntu1 has the webapp enabled. It should be available in the saucy archive shortly.

Please enable it for Raring. Thanks.

I'm not sure I'll have time to figure out why it went missing (since as I say we weren't intentionally disabling it), but if somebody wants to work it out then I would be happy to apply a patch.

Please enable it for Raring. Thanks.

Christoph, it's not useful to simply repeat previous comments word-for-word. Please see my comment #4.

Colin, ok. I'm sorry. I'd love to help you but I have no idea how... :-(

I'm sorry, it looks like I changed the status to Fix Released (somehow).
Since I cannot undo it, can someone else please do it?
Sorry again.

Is there a way we can manually activate the webapp?

On Mon, Jun 10, 2013 at 02:32:48AM -0000, Robin Sheat wrote:
> Is there a way we can manually activate the webapp?

For me, if I build the package myself the webapp is enabled.

Here's a paste of what I said on IRC; I found the reason the package
isn't getting the webapp enabled, but I don't know *why* this is
happening. I'd like to get a handle on this first.

,----
| 09/06 20:04:14 <Laney> hmm
| 09/06 20:04:32 <Laney> so I looked at why we didn't get the git-annex
| webapp on raring
| 09/06 20:05:32 <Laney>
| https://launchpadlibrarian.net/142026098/buildlog_ubuntu-raring-i386.git-annex_3.20121112ubuntu3+laney1_UPLOADING.txt.gz
| shows that the dpkg-query call to figure out if libghc-yesod-dev is
| installed returns "purge ok installed"
| 09/06 20:06:24 <Laney> which means that we use the 'else' branch which
| doesn't build the webapp
| 09/06 20:06:47 <Laney> but why does that happen? I get "install ok
| installed" locally.
`----

--
Iain Lane [ <email address hidden> ]
Debian Developer [ <email address hidden> ]
Ubuntu Developer [ <email address hidden> ]

The first field ("purge") is the desired state, mainly there for
dselect. Normally it would match up after an apt-get install, and I
agree it's weird that it doesn't here, but I'd be inclined to say that
we should just ignore that field and only worry about the third
("installed").

On Mon, Jun 10, 2013 at 08:51:08AM -0000, Colin Watson wrote:
> The first field ("purge") is the desired state, mainly there for
> dselect. Normally it would match up after an apt-get install, and I
> agree it's weird that it doesn't here, but I'd be inclined to say that
> we should just ignore that field and only worry about the third
> ("installed").

Right, I know what it's for, but I didn't know if this is something that
the buildds are doing wrong (so a problem to be fixed there rather than
in packages, given that local sbuild behaves differently); seems curious
to me.

But it does seem right that packages should ignore the desired state and
only care about the current, indeed.

 affects ubuntu/raring/git-annex
 assignee laney
 status inprgress

--
Iain Lane [ <email address hidden> ]
Debian Developer [ <email address hidden> ]
Ubuntu Developer [ <email address hidden> ]

Uploaded to raring's queue for review.

FWIW, this isn't something the buildds are doing "wrong", though arguably "unnecessary" now that we don't purge packages at the end of builds.

sbuild always set the desired state of everything it installs to "purge", so it's a single call to remove them all again at the end.

Hello Greg, or anyone else affected,

Accepted git-annex into raring-proposed. The package will build now and be available at http://launchpad.net/ubuntu/+source/git-annex/3.20121112ubuntu4 in a few hours, and then in the -proposed repository.

Please help us by testing this new package. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Your feedback will aid us getting this update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug, mentioning the version of the package you tested, and change the tag from verification-needed to verification-done. If it does not fix the bug for you, please add a comment stating that, and change the tag to verification-failed. In either case, details of your testing will help us make a better decision.

Further information regarding the verification process can be found at https://wiki.ubuntu.com/QATeam/PerformingSRUVerification . Thank you in advance!

The package installs for me on Mint 15 (based on Raring 13.04) and git-annex assistant works as I would expect! Many thanks.

Verification needs to be done on Ubuntu, not a derivative.

And this is about the git-annex webapp, not just the assistant. Sorry to be pedantic, but the "git-annex assistant" command wasn't the problem, it was "git-annex webapp". (The webapp depends on the assistant)

On Mon, Jun 17, 2013 at 10:22:04AM -0000, Scott Kitterman wrote:
> Verification needs to be done on Ubuntu, not a derivative.

I don't know why my earlier email didn't show up*. That was just a
coincidence. I self verified and don't know how to change tags via the
email interface so visited the bug to do that.

But since there are obviously people watching this, why doesn't someone
else do it too? :-)

* Oh, yes I do. I mangled To:.

--
Iain Lane [ <email address hidden> ]
Debian Developer [ <email address hidden> ]
Ubuntu Developer [ <email address hidden> ]

Confirmed. git-annex webapp does what is expected (ie: start the webapp and sync files)

greg@xps:~/Downloads$ apt-cache policy git-annex
git-annex:
  Installed: 3.20121112ubuntu4
  Candidate: 3.20121112ubuntu4
  Version table:
 *** 3.20121112ubuntu4 0
        100 /var/lib/dpkg/status
(I downloaded the deb directly instead of enabling -proposed)

There are some oddities with git-annex (the webapp has some UI bugs: interface not responding, etc) but those should probably be another bug report.

The verification of this Stable Release Update has completed successfully and the package has now been released to -updates. Subsequently, the Ubuntu Stable Release Updates Team is being unsubscribed and will not receive messages about this bug report. In the event that you encounter a regression using the package from -updates please report a new bug using ubuntu-bug and tag the bug report regression-update so we can easily find any regresssions.

This bug was fixed in the package git-annex - 3.20121112ubuntu4

---------------
git-annex (3.20121112ubuntu4) raring; urgency=low

  * Check for the 'ok installed' substring in the dpkg-query output when
    determining whether Yesod is installed; this matches against the 'current'
    state, rather than the 'desired' and 'current' states that checking for
    'install ok installed' does. The latter is a problem on the buildds where
    the desired state is 'purge'. (LP: #1084693)
 -- Iain Lane <email address hidden> Mon, 10 Jun 2013 22:05:18 +0100