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

[Helge Hess]

On Mar 4, 2004, at 10:53 PM, Manuel Guesdon wrote:
> I'd only like we (all people which want to) discuss about how to 
> cooperate, where to put things, dependencies, licencies problem (if 
> any), maintainers questions, etc... There's a 2 threads about part of 
> this: "CVS hosting issues  (was: Re: Frameworks integration)" and 
> "Frameworks integration".

OK.

If the issue of copyright assignment is not important for you, I 
suggest that we use a decentralized approach first. The goal isn't 
necessarily the setup of a common project, there still can be SOPE and 
gnustep-web projects which are just different forks of the same code 
basis. The primary goal is to share code, also sharing projects is two 
steps further and requires much more discussion.
Though having a shared project of course gives additional benefits.

Since gnustep-web is like SOPE LGPL, we should not have any license 
issues?

> Your proposition is only 7 days old and I don't think all GNUstep 
> contributors have given their opinion.
> Iknow you've developped a lot of things which are interesting but I 
> think
> we can save time if we think a little about all this before doing 
> things.

Yes, indeed I think we should put that to the rest for now to give 
people time to look into each others sources. Maybe we should wait some 
months and then discuss that again.

> Few days is nothing on a ten years old project, IMHO.

Of course, we have no hurry in any way. This decision has time (and 
needs time for implementation anyway).

> About "keeping to your codebase  whatever?": personnaly, one think I 
> don't want is to invest time studying SOPE or starting a merge and 
> after that, more or less quickly, find that there's a big problem

Well, as I mentioned cooperation means that both parties actually need 
to put work in something to reach a common goal. If you don't want to 
invest work, we are stuck.

> (organisation,

irrelevant for sharing code

> license,

Obvious

> project orientation,

irrelevant for sharing code

>  maintainer choices

irrelevant for sharing code

> or god knows what).

Hm? ;-)

> You may say it is LGPL projects  so one can fork if he doesn't agree 
> on something but, _in this kind of situation_, I think I'd prefer to 
> keep on  working like before, just to avoid spending time and because 
> I know better gsxml, gsweb stuff than Slyrix stuff.

This conflicts with your previous statement that you are interested in 
cooperation. A joint effort will imply more work in the short term, but 
more functionality/maturity in the midterm and much less work in the 
long term.
This was my proposal.

> I don't say this will append and I hope not, of course; I don't say 
> there will be problems; I don't say there will be a fork;I don't say I 
> won't work to study SOPE or merge. I'm only carefull.

A fork is not necessarily bad if there are two different goals. This 
doesn't imply that no code can be shared.

As an example, we use a stable fork of gstep-make for a long time in 
OGo packaging because API and directory setup stability is extremely 
important in our view, while the head/GNUstep version of gstep-make 
focused on enhancements and improvements even if this breaks 
compatibility between minor revisions.
So our fork focuses on different goal, yet shares code with gstep-make.

> There's nothing personnal against you, Helge,

You started to get VERY personal several times in your replies.

Helge
--
http://docs.opengroupware.org/Members/helge/
OpenGroupware.org