Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [gef-dev] GEF3 resurrection

Hi Serge,

the simple solution for allowing you to bring your changes into GEF Legacy is to separate GEF Legacy into another Eclipse Project. But you or someone else would have to take over responsibility.

Best regards,
Matthias

On Fri, Feb 18, 2022 at 12:20 PM Serge Rider <serge@xxxxxxxxx> wrote:
Hi Matthias, Alexander, Lars,

Honestly, I'd prefer not to be "in charge" (in any form) for GEF legacy.
Because I am not the best expert in the Eclipse platform ecosystem as a whole, I never participated in GEF development in the past.
I am not even an Eclipse committer yet, but I could try to become one.

I (together with DBeaver team) would be happy to propose PRs with bug fixes and improvements.
I would be happy to participate in discussions about GEF developments and, hopefully, participate in GEF3/GEF5 transition process.

However, I doubt that we can bring features of GEF5 into GEF3 (at least major features).
I think GEF3 API should remain the same for backward compatibility as there are hundreds of existing GEF3-based solutions. But it should be improved.
This is the main idea of my letter - return GEF3 to life to be able to fix bugs and provide minor UI or API improvements. We don't want to turn GEF3 into GEF5.

Another idea is that we could try to add support of SWT-based UI rednering into GEF5. This will solve conceptual problems with GEF3->GEF5 migration.

Briefly: the main thing we need now is a possibility to commit into GEF3 legacy repository and return GEF3 into release cycle.

Best,
Serge


On Thu, Feb 17, 2022 at 3:29 PM Matthias Wienand <matthias.wienand@xxxxxxxxx> wrote:
Hey,

it is nice to see some action here :-)

@Lars: Yes, I think that is the way to go. Thank you for offering your support with the uncomfortable stuff ;-) Maybe we need Serge and Alois to bring in a few contributions first, before we do an election, but I would also be fine with skipping directly to an election TBH.

@Serge: It would be nice if you could try to answer Alexander's question "Do you plan to actively contribute to the GEF project as a whole, or do you just feel obliged to GEF-legacy?". I am also very interested in the answer.
Of course, I would like to see you try to bring features from GEF 5 into an improved GEF 3 (the idea you touch on in the presentation, IIUC?).

Best regards,
Matthias


On Thu, Feb 17, 2022 at 10:19 AM Pierre-Charles David <pierre-charles.david@xxxxxxx> wrote:
Le 16/02/2022 à 17:11, Serge Rider a écrit :
> Hi Team!
>
> I am from the DBeaver development team (https://dbeaver.io). We
> develop the universal database management tool based on the Eclipse
> RCP platform.
> Besides other Eclipse RCP extensions, we heavily rely on legacy
> GEF/draw2d.
>
> I was a speaker on several recent Eclipse Cons, the last session was
> about legacy GEF adoption:
> https://www.eclipsecon.org/2021/sessions/diagrams-eclipse-rcp-back-future
> https://www.youtube.com/watch?v=3ZIUB2XFDLE
>
> What we want is to resurrect GEF3 maintenance and include it back to
> the standard Eclipse RCP lifecycle.
> We have several PRs we'd like to merge in the codebase (currently we
> maintain all fixes in our forked repository on GitHub). We also want
> to propose legacy API improvements (keeping  API backward
> compatibility) and several features.


Hi.


This is great news, thanks for the initiative Serge! (Thanks for DBeaver
btw, I use it almost every day and it's a great tool).

As the maintainer of GMF Runtime, I'd like to remind that *a lot* of
projects depend on GEF Legacy (via GMF Runtime or not). In the SimRel
alone there's Graphiti, Sirius (and Capella, even if it's not in the
SimRel), Papyrus, Ecore Tools, and probably others.

Of course improvements would be welcome, and we (in GMF and Sirius) may
able to propose some fixes/improvements we've currently had to make in
our own copies of GEF classes like you, but care must be taken not to
break the many projects which depend on the current behavior (sometimes
in subtle ways).

Regards,
Pierre-Charles David

_______________________________________________
gef-dev mailing list
gef-dev@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/gef-dev
_______________________________________________
gef-dev mailing list
gef-dev@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/gef-dev
_______________________________________________
gef-dev mailing list
gef-dev@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/gef-dev

Back to the top