Re: What's involved with signing/testing releases?

On 1/18/07, Ed Hillmann <edhillmann@yahoo.com> wrote:
> I've been lurking on the dev list, and I've noticed how various people are signing and testing releases on various platforms. Is there much involved with this?

No, not really - most folks are just downloading the release,
compiling it, and trying it out. Each person comes up with their own
metrics as to how or why they choose to vote for a release. Most
committers tend to run some combinations of the regression test suites
to determine whether the release is suitable for public consumption.

> I typically build our Subversion on a Solaris 9 environment for our own use internally. If no one is already checking this environment, I'd be happy to help out with this (if someone can nudge me in the right direction). :)

If there's something broken with a release, we want to know about it
as soon as possible. So, everyone on this list should feel free to
test the releases in their own environment. The one thing we ask is
that downstream developers and packagers do not release anything with
this new code until we have completed our testing and posted it on the
official Subversion site.

However, what we do before we post it on the official Subversion site,
is that we require that at least three committers have tested the
release before it can be posted. Each committer indicates their
approval of the release by submitting their signatures for the release
package.

Hope this helps! -- justin

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Jan 19 06:09:24 2007