Login | Register
My pages Projects Community openCollabNet

Discussions > dev > [GEF]RE: Separation of GEF and Argo

gef
Discussion topic

Hide all messages in topic

All messages in topic

RE: [GEF]RE: Separation of GEF and Argo

Reply

Author Micheal J <open dot zone at virgin dot net>
Full name Micheal J <open dot zone at virgin dot net>
Date 2000-08-01 10:56:03 PDT
Message GEdiF - Graphical Editing Framework?

==> www.gedif.org??

Just a suggestion...


Micheal

> -----Original Message-----
> From: Jason Elliot Robbins [mailto:jrobbins at collab dot net]
> Sent: 01 August 2000 15:54
> To: dev at gef dot tigris dot org
> Subject: Re: [GEF]RE: Separation of GEF and Argo
>
>
> >
> >Perhaps the solution is to acquire gef.org and argo[uml].org for the
> >project?. I would have thought Jason or an Apache-style organisation
> >(perhaps called "Argonauts") would be the natural custodian of
> the project.
>
> I own argouml.org, it forwards to argouml.tigris.org.
>
> gef.org is surely _way_ expensive.
>
> jason!
>

Re: [GEF]RE: Separation of GEF and Argo

Reply

Author jrobbins9
Full name Jason Robbins
Date 2000-08-01 07:53:48 PDT
Message >
>Perhaps the solution is to acquire gef.org and argo[uml].org for the
>project?. I would have thought Jason or an Apache-style organisation
>(perhaps called "Argonauts") would be the natural custodian of the project.

I own argouml.org, it forwards to argouml.tigris.org.

gef.org is surely _way_ expensive.

jason!

RE: [GEF]RE: Separation of GEF and Argo

Reply

Author Micheal J <open dot zone at virgin dot net>
Full name Micheal J <open dot zone at virgin dot net>
Date 2000-08-01 04:17:46 PDT
Message > > It is my understanding that having package names reflect the
> reverse order
> > of the URL of the developing organization was designed to
> prevent package
> > name collisions. If the .tigris is striped from the package
> names can't an
> > organization which owns the name gef.org publish a org.gef package that
> > conflicts with your proposed 'org.gef' package name?
> >
>
> I agree with you. But GEF as well as Argo are not developed by tigris.org.
> This is only the place where the developing community places the
> code. Don't
> get me wrong. I appreciate the work of tigris and the people at
> Collab.net.
> But they don't own GEF, they simply host the project. Therefore I
> don't think,
> that their name should be reflected by the package names. If someday an
> organization called gef.org will exist, hopefully it will be the
> owner of GEF
> and coordinate the further developing. To put it the other way round. if
> someday there will be an organization owning and controlling GEF (e.g.
> something like Apache), its name should be part of the package names.
>
> Jason: I think this kind of discussion will raise up for the package
> restructuring of argo as well. Maybe we should find a "generic"
> solution that
> fits for both, GEF and Argo.

Perhaps the solution is to acquire gef.org and argo[uml].org for the
project?. I would have thought Jason or an Apache-style organisation
(perhaps called "Argonauts") would be the natural custodian of the project.

Micheal

Re: [GEF]RE: Separation of GEF and Argo

Reply

Author Thorsten Sturm <1sturm at informatik dot uni-hamburg dot de>
Full name Thorsten Sturm <1sturm at informatik dot uni-hamburg dot de>
Date 2000-08-01 01:52:22 PDT
Message Ted,

sorry for my late response to your message. I had a little accident last
weekend and therefore no time to read my mails.



> Do you have any feel at all for how much the version of GEF in the GEF cvs
> repository and the version of GEF currently used in ArgoUML differ?
>

I'm currently investigating this. From what I saw until now, there have been
some efforts on changing the code in order to force the separation of GEF. But
that doesn't look to difficult to integrate with the Argo version (Edwin,
Jason: maybe you can give me a hint on where to look at and what to be careful
about). The Argo version has gone through some changes so that I suggest to
integrate the GEF version into the Argo version and to take this as the new
codebase for further GEF development.

> It is my understanding that having package names reflect the reverse order
> of the URL of the developing organization was designed to prevent package
> name collisions. If the .tigris is striped from the package names can't an
> organization which owns the name gef.org publish a org.gef package that
> conflicts with your proposed 'org.gef' package name?
>

I agree with you. But GEF as well as Argo are not developed by tigris.org.
This is only the place where the developing community places the code. Don't
get me wrong. I appreciate the work of tigris and the people at Collab.net.
But they don't own GEF, they simply host the project. Therefore I don't think,
that their name should be reflected by the package names. If someday an
organization called gef.org will exist, hopefully it will be the owner of GEF
and coordinate the further developing. To put it the other way round. if
someday there will be an organization owning and controlling GEF (e.g.
something like Apache), its name should be part of the package names.

Jason: I think this kind of discussion will raise up for the package
restructuring of argo as well. Maybe we should find a "generic" solution that
fits for both, GEF and Argo.

Thorsten

[GEF]RE: Separation of GEF and Argo

Reply

Author Ted Kosan <tkosan at shawnee dot edu>
Full name Ted Kosan <tkosan at shawnee dot edu>
Date 2000-07-29 23:51:58 PDT
Message Thorsten,

>So I suggest to put some efforts on completing the separation of Argo
>and GEF so that all of the GEF code left in Argo can be removed.

Do you have any feel at all for how much the version of GEF in the GEF cvs
repository and the version of GEF currently used in ArgoUML differ?

>Making the separation of GEF and Argo work seems to be a good time to
>restructure the packages. Based on the suggestion Jason made a while
>ago, I would suggest the following structure (I replaced org.tigris by
>simply org, because I don't think that the location or organization
>supporting the project should be represented in the package names):

It is my understanding that having package names reflect the reverse order
of the URL of the developing organization was designed to prevent package
name collisions. If the .tigris is striped from the package names can't an
organization which owns the name gef.org publish a org.gef package that
conflicts with your proposed 'org.gef' package name?

Ted
Messages per page: