On Sunday 28 June 2009, Øyvind Harboe wrote:
> I'd hate the idea of having 10 host platforms with a few outstanding
> problems on each and try to time a release until "all" problems have
> been solved.

Me too.  Not that I think there is anything like such a
mess; right now, it seems the issue is entirely how to
get MS-Windows to behave.  There can be "dot" releases.


> Better to let the package maintainers decide what's best for
> a particular platform.

That undermines the notion of formal release points.

Not all package maintainers will be "on top of" such issues.
I don't think it *should* be their role to evaluate all the
unresolved bugs and make such decisions.

We certainly don't want to take away the ability of any
packager to decide that e.g. they'll push something out
based on some SVN tag.  In fact we can't, without shutting
down the open development process.  And most developers
would object to the SVN tree being very broken for long.

IMO we should be aiming to have regular release points
so that users can synchronize on them and know that, for
example, while 0.2.0 has bug X, it's all platforms fix
that in their 0.2.2 versions.  And also, so that developers
can organize their work around release points:  big changes
happen earlier in the cycle, later the focus is bugfixing.

- Dave

_______________________________________________
Openocd-development mailing list
Openocd-development@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/openocd-development

Reply via email to