Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control)

[Philip Mötteli]

Am 26.10.2003 um 16:19 schrieb Philippe C.D. Robert:
> On Sunday, October 26, 2003, at 10:38  Uhr, Philip Mötteli wrote:
> > > > CoreFoundation is almost ported. Apart from that, it's very little 
> > > > used from Cocoa developpers.
> > > I seem to recall seeing it used by a lot of frameworks and other 
> > > lower level stuff.
> >
> > 1. "Well, seem to recall" is not very convincing.
> > 2. If those guys knew, that they will be able to deploy with very 
> > little effort also on Windows, I think, they would carefully think 
> > about using CF.
> > 3. What is used from CF is essentially the CFMap. And we have an 
> > equivalent here in Gnustep. So porting this is not a problem.
> > 4. CoreFoundation is almost ported to Linux.
>
> CoreFoundation, CoreAudio, CoreGraphics, CoreMIDI, ... all of these 
> are created to be used by Mac OS X programmers

Yes. But the aim of the core, that's at least an Apple employee, who 
told me, is to serve as a base to other programming languages. So that 
not only ObjC is usable on MOSX. But if you use ObjC, you usually don't 
use Cores, when available.
The problem is alwys the same: When you use a library, that is not 
available on the other platform, you're in a mess. But it's not up to 
us to implement for every library on every platform a clone. No, the 
programmers will chose, what is available on their target platforms. 
And sometimes, they will even port something. I do agree, that this is 
not always easy. But we don't need all the programs. 60% of them would 
already be a huge progress. I just do not agree to say: "we gonna loose 
40% of the programs, so let them all throw away!"


> - there are no Cocoa or Carbon programmers anymore

Yes, there are. Give me some non legacy programs, that use Carbon. Very 
few! And the few, you gonna find, are usually people, who don't want to 
learn Cocoa. And the people, who really need something, that is really 
not available in Cocoa, try to implement as much as possible in Cocoa 
and just the strict minimum in Carbon.
Yes, those worlds are very separated!


> in these days, they are all just Mac OS X

No!


> programmers who will probably use whatever framework there is as long 
> as it offers what they need.

Of course, they don't take, what they don't need. But they also think 
what they need in the future. And some of them also need to make a 
living of it. So they need clients. And selling ten times more, means 
ten times more money. At least me, I wouldn't say no to that and would 
prefer to learn some Cocoa, then continuing with my classic Carbon.


> Wrt Carbon, Apple's standpoint is clearly that Cocoa and Carbon are 
> both first class citizens which even can/should/be mixed, also because 
> Carbon offers functionality which is not available via Cocoa. 
> Personally I do not like this strategy very much, but then it makes 
> sense from a business perspective.

No, this is pure marketing! The CEO before Jobs came back, even wanted 
to drop Classic Mac OS completely. But the big companies like Adobe 
didn't agree and threatened them to leave the Apple market. So Apple 
made the whole effort to bring out Carbon, so that these companies just 
had to (officially) change 10% in their code to make their software a 
MOSX citicen. Of course Apple never said that, publically, because 
Apple is living from their "die hards". Those people reject everything, 
that doesn't come from Apple. So they rejected MOSX, Cocoa, 
Objective-C, just plain simple everything. So Apple was forced, for 
marketing reasons, to say, that actually MOSX is nothing new, but a 
smoth upgrade from MOS. I mean, no really new technology. The same as 
before. We all know, that MOSX is a completely new OS and has strictly 
nothing to do with the old MOS (fortunately). You see the problems of 
Apple? Do you think, that they would ever have told the people, that 
Carbon has no future? They don't even say that from MOS-Classic. The 
official version is still, that Classic will continue to be developped 
– new features and everything. But every "non die hard" knows, that MOS 
Classic is just debugged and adapted to work smoothly with MOSX. Apart 
from that it's dead. But Apple never sais that. Please don't take 
Apples word everywhere literally. I think you, with a NeXTstep 
background, you perfectly know, where they should go and they usually 
do exactly that.


> > But the question was, if it is important to have companies on our 
> > side or not. M$ had that. Phil Roberts doubted, if that would help 
> > GS. He said, the hobbyists would be enough for GS to succed. And I 
> > said, if we have the companies on our side, we've almost won the 
> > battle.
>
> No, I never say that, I never said anything about Hobbyists.

That was your word. I didn't even know, that it exists in English.


> I said that the chance to find programmers in the free *nix world 
> (Linux, BSD, ... ) who would be willing to support GNUstep to create a 
> complete, stable end-user environment solution

And those are not hobbyists? (concerning their work for GS)


> is probably higher and more realistic than trying to port GS to 
> Windows in order to come up with a full blown,

Not full blown. Just finish the port to Windows. No ProjectBuilder and 
all that.


> cross-platform development solution which perfectly integrates into 
> the native host OS (and thus not being an end user environment!)

We don't need the perfect solution. We just need as much, that 
companies say: "it's almost complete for our needs. We could complete 
the rest with a reasonable investment."


Re
Phil