survey: Minor Commits during Release Cycle

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

survey: Minor Commits during Release Cycle

Erik Uzureau-3
Hey folks, I came to an interesting situation this morning while working on a
little cleanup of the tile section.

I found that there was a little error in the JSDOC for OpenLayers.Tile.WFS,
and so I made a patch, then made a ticket[1].

During normal development, coders with commit access do not need approval
for making changes to trunk that involve: tests, examples, or documentation.

Since we are in a release cycle, however, I reasoned the following:

If I just commit this directly to trunk, then it will never get
brought up to the
2.4 branch, because there is no ticket for it. In my mind, the way the release
process is done is by surveying the tickets marked 'pullup', and applying
thier patches.

Schuyler brought up that instead of the above procedure, it would be easier
to just make the commit directly to trunk. Then when it's time to make a new
RC candidate, the release manager should be in charge of looking through
the svn logs since the last RC and pulling up any relevant changesets.

The former seems like too much work to him, the latter to me :-)

Thoughts? Revelations?

[1] http://trac.openlayers.org/ticket/644
_______________________________________________
Dev mailing list
[hidden email]
http://openlayers.org/mailman/listinfo/dev
Reply | Threaded
Open this post in threaded view
|

Re: survey: Minor Commits during Release Cycle

Tim Schaub-2
Erik Uzureau wrote:

> If I just commit this directly to trunk, then it will never get
> brought up to the
> 2.4 branch, because there is no ticket for it. In my mind, the way the release
> process is done is by surveying the tickets marked 'pullup', and applying
> thier patches.
>
> Schuyler brought up that instead of the above procedure, it would be easier
> to just make the commit directly to trunk. Then when it's time to make a new
> RC candidate, the release manager should be in charge of looking through
> the svn logs since the last RC and pulling up any relevant changesets.
>
> The former seems like too much work to him, the latter to me :-)
>
> Thoughts? Revelations?
>

The latter relies on svn being in charge, the former relies on people
(making tickets and spelling things like 'pullup' correctly).

Doesn't say anything about which is more work, but it suggests that the
latter is more reliable to me.

Tim

_______________________________________________
Dev mailing list
[hidden email]
http://openlayers.org/mailman/listinfo/dev