[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Release schedule
From: |
David Ayers |
Subject: |
Re: Release schedule |
Date: |
Tue, 01 Apr 2003 09:17:54 +0200 |
User-agent: |
Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3b) Gecko/20030210 |
Hi,
Collectively, we seem to be striving to 3 levels of compatiblity:
OpenStep Specification
OPENSTEP 4.2 implementation
current Cocoa implementation
enriched by some very useful GNUstep specific extensions.
I personally believe the OpenStep Specification is more of an academic
issue as there is probably *very* few people if any intersted in
maintaining code that will be portable between GNUstep and strict
OpenStep implementations. (Please correct me if I'm wrong.)
I still believe there is a viable amount of support for compatiblity of
the OPENSTEP 4.2 implementation as there are still
a.) real OPENSTEP Enterprise implementations out there, that
continuously consider GNUstep as a potential strategy to provide a
future for thier project.
b.) real ObjC WebObjects implementations, with the same goal (the API of
WO has already slightly advanced to a Cocoa like API though.)
It would also supply a stable baseline that developers implementing for
GNUstep exclusively can rely on.
The potential support from the Cocoa communtiy and the possiblitly of
writing portable code for GNUstep and Cocoa makes, maintaining a decent
level of compatiblilty to the current Cocoa API too attractive to be
ignored.
But when tracking Cocoa, we should not try to be more restrictive just
because Cocoa is being more restrictive We should also not add features
that simply don't map into the GNUstep environment.
I think the decision and it reasoning of whether or not to add a given
feature set or not, should be promanently advertised and remain
debatable at certain 'fixed' intervals.
Any feature added, lastly depends on someone contributing it. Yet we
also know that the integration may break or complicate existing code.
And this is the point where developers need mechanisms, guidelines and
discussion on how or whether to integrate these features. Maybe this
hasn't been happening in the right context. Maybe the GNUstep Project
needs some type of convention / forum to come by these decisions and
document the reasoning publicly, but keeping it debatable (maybe after
each major release).
Cheers,
Dave