Re: Release philosophy (was: Re: Miscellaneous Things)

[Eric McDonald <mcdonald@phy.cmich.edu>]

Hi Lincoln, Stan,

On Thu, 4 Dec 2003, Lincoln Peters wrote:

> On Thu, 2003-12-04 at 18:04, Stan Shebs wrote:
> > The most important way to get visibility is to "release early and often".
> > 7.5 doesn't have to be perfect, because only 5% of the audience will
> > take note of the announce anyway; but 10% will notice the 7.5.1
> > message, having been primed by the 7.5 release, and so forth. Since
> > there hasn't been a release in a long time, most people will assume
> > Xconq is a dead project and not even look to see what's up with it.
> 
> How about if, as soon as all of the serious bugs such as instant crashes
> are fixed, we put out a release candidate (7.5rc1)?  Then we can wait
> for anyone to report additional bugs, and we fix them, put out a new
> release candidate (7.5rc2), etc.?  Then when a release candidate has
> been out for a while (maybe a week or two) with no new bug reports, we
> release 7.5.

I feel the same can be accomplished by packaging known good CVS 
snapshots. Perhaps the only thing missing from the picture is to 
announce the availability of the these snapshots.

I personally think we should release 7.5 when it's ready. Maybe 
the documentation doesn't need to be brought entirely up to date 
before a release. But I don't think there should be any 
regressions from 7.4.1 as far as "make check" is concerned. And we 
are still getting closer to having much more intelligent transport 
behavior; I would want to wait until that improvement is in and 
working properl; it would be a boost to AI play, IMO.

> seems to work very nicely for them.  And if nothing else, it becomes
> clear to anyone who's paying attention that Xconq is *not* a dead
> project.

Perhaps it would not appear dead anymore, but it could earn a 
reputation as being quite buggy and needing lots of work.

Eric