Migrate to CommitTicketUpdater

[Amar Takhar]

The old script was ancient and outdated. I've now killed it off and moved to the internal system described here:

​https://trac.edgewall.org/wiki/CommitTicketUpdater

This should handle all scenarios if it does not please let me know. I will leave this ticket open for a week or so.

[Chris Johns]

Should ​https://devel.rtems.org/wiki/Developer/Git/Committers be updated with something about the triggers, even a link to Trac doco?

[Amar Takhar]

Replying to Chris Johns:

Should ​https://devel.rtems.org/wiki/Developer/Git/Committers be updated with something about the triggers, even a link to Trac doco?

Great idea how is this? wiki:Developer/Git/Committers#TicketUpdates

[Chris Johns]

Git pushes are not updating Trac up as report by Sebastian in an email..

​https://git.rtems.org/rtems_waf/commit/?id=1b896361d302aeda0145af90972aea863e28898f

It didn't update the ticket:

​https://devel.rtems.org/ticket/3551

It doesn't show up in the Git view in Trac:

​https://devel.rtems.org/log/rtems_waf

The common git hook /data/support/git-support/hooks/post-receive-3 for Trac is commented out:

stat -x /data/support/git-support/hooks/post-receive-3 | grep Modify
Modify: Sun Oct 21 00:09:04 2018

I suggest we manually update all tickets until this is resolved. I am not sure why this has been done.

[Amar Takhar]

Replying to Chris Johns:

I suggest we manually update all tickets until this is resolved. I am not sure why this has been done.

It was done due to the numerous complaints about tickets not updating especially with multiple commits.

The way we were doing has been unsupported for over 4 years now it's broken and hacky this change was inevitable I'll look into it I can manually force trac to update the tickets as required so we won't lose anything.

[Amar Takhar]

I forced an update and changed how I was doing the update we'll see if it works this way.

[Chris Johns]

Replying to Amar Takhar:

I forced an update and changed how I was doing the update we'll see if it works this way.

Could you please explain the change and is there anything we need to do, I am confused if we still need to manually handle ticket updates?

[Sebastian Huber]

Thanks for fixing this. The automatic update worked with this commit:

[05a5366469ee4ee7776f70dceea631ea57a63b19/rtems-docs]

[Sebastian Huber]

I checked in four commits at once:

​https://git.rtems.org/rtems/log/?qt=range&q=01595a4f321ad890270a3bc3e21c98ba51163562..24b58072ff1ac5a1691374aa139807bf866bf01a

Only the last commit updated the ticket system. This one for example
didn't update the corresponding ticket:

​https://git.rtems.org/rtems/commit/?id=4b801acc2003073c50b381435896a29f29e51a3c

​https://devel.rtems.org/ticket/2514#comment:41

[Amar Takhar]

While working on Buildbot I discovered what was causing this issue. The changeset updating in trac wasn't properly iterating over every commit in a push just the parent.

I have changed it now hopefully it should catch everything now. Let me know if it doesn't thanks.

[Sebastian Huber]

I committed this to rtems-docs:

​https://git.rtems.org/rtems-docs/commit/?id=e49e4056c7d4b7fa35881cfc3522c6e038940e84

I received this from Git:

git push origin e49e4056c7d4b7fa35881cfc3522c6e038940e84:master
Counting objects: 4, done.
Delta compression using up to 12 threads.
Compressing objects: 100% (4/4), done.
Writing objects: 100% (4/4), 481 bytes | 481.00 KiB/s, done.
Total 4 (delta 3), reused 0 (delta 0)
remote: 0: init
remote: 1: mail vc@rtems.org
remote: NoSuchChangeset: No changeset e49e4056c7d4b7fa35881cfc3522c6e038940e84 in the repository
remote: 4: IRC
remote: fatal: bad object e49e4056c7d4b7fa35881cfc3522c6e038940e84
remote: irc notice: Invalid number of commits: 0
remote: 5: Buildbot
To ssh://dispatch.rtems.org/data/git/rtems-docs.git
   1a704cd..e49e405  e49e4056c7d4b7fa35881cfc3522c6e038940e84 -> master

The corresponding tickets were not updated, e.g. #3584.

[Amar Takhar]

I reworked a lot of these scripts to make it much easier to modify going forward as a few more will be added. I left some old variables in one of them that were statically set to 'rtems'.

This should fix this problem please let me know if it does not.

[Amar Takhar]

I have not heard any complaints and multiple references should work now as the bug was extremely clear. If it's still an issue please re-open.