river-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dan Creswell <dan.cresw...@gmail.com>
Subject Re: build mechanisms
Date Sat, 15 Jan 2011 09:43:15 GMT
On 14 January 2011 23:29, Peter Firmstone <jini@zeus.net.au> wrote:

> It sounds like we need to re-discuss how we use svn for River's
> development.
>
> I'd like to wait until Patricia returns, so we've got everyone on board and
> we all have the opportunity to reconsider the development process.
>
> The following is my current understanding of River's development process:
>
> After I had been coding directly in the trunk, and we found there were
> issues in doing so for a team of developers, it was my understanding that
> experimental work could be performed in skunk, then when code had some peer
> review, was ready etc, it could be merged back into trunk.
>
> Minor changes or patches and bug fixes can be committed to trunk with peer
> review.
>
> Trunk is the stable branch used for release branches, in order to capture
> all the patches and bug fixes that are submitted to Jira issues etc.
>
> To cut the next release, I will branch the current trunk, I will rely on
> the JIRA system to generate the release notes, it won't be based on the
> previous release branch.
>
>
This would be the way I'd like to see things work after many, many bitter
experiences in development both opensource and commercial.

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message