Subject: Re: [RELEASE] 0.9.3 schedule ?
From: Martin Sevior (msevior@mccubbin.ph.unimelb.edu.au)
Date: Mon Aug 27 2001 - 20:58:56 CDT
On Mon, 27 Aug 2001, Randy Kramer wrote:
> Paul Rohr wrote:
> > 5. Then, of course, the big trick is just to make sure that we get
> > reasonably quick turnaround from the testers for each package. My belief is
> > that, so long as each tester knows far enough in advance to make their time
> > available, that shouldn't be too hard.
>
> I think I'm almost willing to volunteer to be a tester (probably with a
> binary for Win32, as I am not real confident about installing stuff on
> Linux), but I'd like to know more about how to test or what you expect
> from a tester.
>
> Is there a specific set of documents to test with, or a test procedure?
> Or do you pick a document (at random) and then, for example, try
> something with every option on the menu? Or does it depend on the
> "focus" of the release (the features to be emphasized)?
>
> How much time would be involved in testing a given "release candidate",
> and how often would "release candidates" be released?
>
> Would you expect bugs to be filed in bugzilla, or just a report issued
> back to the appropriate developer(s) (or list)?
>
> Other?
>
Yes! Actually Randy, what you've listed here shows that if you were a
tester, our binaries would geta lot more more testing than they currently
do before releases.
If you can think of a reasonable protocol to put tests through that will
not take a long time to fo through, feel free to post it here.
What has generally been done in the past is to test out the most recent
new features/bug fixed and check that they actually work as advertised. It
would be great to have a set of regressions to put abiword through
too. Joaquin's perl bindings would be a great help for this too.
Cheers
Martin
This archive was generated by hypermail 2b25 : Mon Aug 27 2001 - 20:59:08 CDT