Re: Release notes

[Riccardo]

Hello,

On Tuesday, July 5, 2005, at 04:08 AM, Adam Fedor wrote:

> I updated the Documentation/ReleaseNotes.gsdoc for base and gui for the 
> next release. Please check it and add or change anything I forgot.
>
> I still plan on making a release in 1 or 2 weeks.

I'd like to see some progress on the FreeBSD/x86 front before you make a 
release (it worked in the past release but current CVS is broken since 
some weeks, I filed a bug). Why ? because releases are "references" for 
many aspects. Features for example. Application XYZ works with release 
N.M. But essentially because releases are what people are most likely to 
use and often only releases make their way into distributions.

I think it would be a good idea for the future to try to deliver a 
working release on most (better, all) platforms gnustep runs on. 
ALthough this kind of quality control is probably difficult to achieve 
and probably overkill (as well as the definition of "works" is fuzzy) 
I'd like to make the following points:

- it should compile and run usably (no major showstoppers) on Linux (x86 
of course, ppc...) on FreeBSD/x86 and on NetBSD (x86, ppc, sparc). These 
being the major free OSes gnustep is likely to be used on.
- in a second place there would be Solaris, OpenBSD and possibly Windows.

Windows and Solaris (especially now that there is OpenSolaris and 
possible upcoming distributions like SchilliX) could open us some more 
professional and more enterprise market.

We can argue about making a "must be working" OS list and I would avoid 
that, criteria and personal preferences would clash. Most widespread OS? 
Windows! and it is an important one, recent talks from Gregory (but 
ironically I had a private talk with a developer the week before and 
told that gregory) confirm that. Linux and *BSD  are closer to us being 
Free Software and we are after all GNU... we should work on Hurd :)

Also the entity of the "showstopper" can be argued. Some poeple here use 
only gui, so glitches there are impportant, other poeple use more -base 
and the web/database stuff.
But everyone wil agree that the failure to build -base is bad! :)

I wrote a long mail and I wonder if someone read it up to this line, but 
the main concept is that we are pretending that gnustep is getting more 
mature and usable (which, in my opinion, it is indeed). SO as we 
approach "1.0" releases (see gorm) we should also try to act accordingly.

Thank you for your patience,
  R