Re: [GSWHackers] Re: OGo/GNUstep cooperation Re: Re[2]: Frameworks integration

[Manuel Guesdon]

Hi Helge,

On Tue, 2 Mar 2004 12:54:37 +0100 Helge Hess <helge.hess@opengroupware.org> 
wrote:
 >| Manuel, I'm sorry to see that you have no intentions to do work on a 
 >| solution.

Probably a misunderstood. Remember, since before the begining of gsweb, I was 
open to 
cooperation to avoid doing same things twice. But, honestly, begining 
cooperation message 
with "Yet with OGo we now do have a high quality and proven WebObjects clone  
and gnustep-web 
is somewhat obsolete for practical purposes" is very strange. Cultural 
difference ? :-(




 >| So, all the other GSWHackers, I would like to see whether we could work 
 >| together on a single project instead of doing things twice. Of course 
 >| this doesn't have to be a hard move from gnustep-web to SOPE, eg we 
 >| could start out to port some applications to work with both application 
 >| servers (shouldn't be difficult if gnustep-web is really WO like) and 
 >| then do some better evaluation whats SOPE gives or not.

May be that one of the key problem. I agreed to your previous post:
<<
 >| So while I'm open to any suggestions, it might be worth to fork a new 
 >| project focusing on a WO clone. 
>>

And started to think about CVS organisation,... (see  my message of this 
morning "Re: CVS hosting issues (was: Re:
Frameworks integration)")

Now you're talking again about "move from gnustep-web to SOPE". So I'm still 
not sure about what you have in mind.
I have no fixed idea on this, I'd just like  to have a more clear idea on what, 
how, what will be cvs hierarchy, 
who will be maintainer(s), who will 'own' code.... As I've said, I trust some 
people for being maintainers of gnustep
because I've seen what they do and how they do. But don't ask me to trust you 
right now without elements. In my opinion,
co-operation is a process which includes step by step progress and mutual 
respect, not something that could 
come one day from the sky.. I know we have/had different opinions about some 
things (licensing, copyright 
assignement, ...) so I hope you understand that I (and I'm not the only one) 
don't want to start running without thinking
about furture of all this. Just to avoid another misunderstood, I have nothing 
against different opinions, I have
nothing against working with people who have different opinions but I want to 
be relatively sure that there will 
be no blocking points which would impact projects before spending too much 
timedoing things.
And I see no urgency on this: gnustep is a ten year old project and gsweb a 5 
years old one,  so we can take 
some days to think about goals, details, organization, copyrights,...

For exemple, there's an xml framework in gnustep base, used by gsdoc, gsweb, 
... So if we replace it by skyrix one,
where to put it, how to handle dependencies,... It's just an exemple, once 
again I have nothing against this.

I'm open to cooperation but I don't want to jump through the window _before_ 
thinking about next steps. I want to get 
an idea about mid and long term before making choice and starting to modify 
things. 



 >| The primary goal is pretty clear and well stated before: streamline 
 >| resources on a single application server instead of doing two. Don't do 
 >| things twice.

Once again, I agree on this but that's not all. There's different ways to 
achieve this (dropping gsweb and moving to
SOPE, dropping gsweb and moving to SOPE with including gsweb enhancements, 
droping SOPE and moving to 
gsweb, droping SOPE and moving to gsweb with including SOPE enhancements, 
creating a new project and working 
on what and how to put best of SOPE and gsweb in it,....)


 >| > I'm not the only one to use gsweb and I just say that people who
 >| > have developped applications based on gsweb probably don't want to see 
 >| > it just
 >| > dropped;
 >| > like you won't just drop NGObjWeb from OOg to replace it by a something
 >| > else just next week.
 >| 
 >| GSWHackers:
 >| a) No one was suggesting to drop gsweb from one second to the other nor
 >|     was anyone suggesting an API change. So obviously all the 
 >| applications
 >|     written by gstep-web users must run first and obviously this 
 >| shouldn't
 >|     be too hard.

Great.  So let's take some time to think about all this, after solving 
organisation, cvs,... questions.


 >| Manuels application (eCommStep) seems to be proprietary software not 
 >| available in sourcecode form. 

Right.


Manuel
-- 
______________________________________________________________________
Manuel Guesdon - ORANGE CONCEPT <mguesdon@orange-concept.com>
14 rue Jean-Baptiste Clement  -  93200 Saint-Denis  -  France
Tel: +33 1 4940 0997  -  Fax: +33 1 4940 0998