Just to be clear, I don't speak for the entire development team.
Everyone will bring a different view to the evaluations. Having said
that, my primary interest is in improving the ArgoEclipse user
experience. Bogdan did a great job on getting initial implementation
done two summers ago, but it hasn't really been improved since then
(largely my fault), and in some cases has degraded due to "bit rot"
(ie things that have changed in ArgoUML core that we haven't tracked).
My priorities would be:
1. develop an agreed set of priorities and plan (in other words, not
everything needs to be decided today)
2. stabilize and clean up what's there
- there are little things that have become broken like the
positioning of newly created dialogs (positioned half off the screen),
missing icons, etc
- implement some of the SWT/AWT tips & tricks from the article you
mentioned (although many of the basic ones are already done)
- implement headless PDE build (although I'll try to get this done
myself in the near future)
- update branding to match latest ArgoUML icons and splashscreen
3. implement missing functionality from standalone ArgoUML to achieve
functional parity
- Eclipse preferences page for ArgoUML prefs
- Add support for code generation/reverse engineering plugins
- Enable use of eUML model subsystem (Should just be a new launch
configuration)
- anything else (I think that's it though)
4. Improve integration with other Eclipse plugins & Eclipse environment
(This can be advanced in multiple dimensions: visual appearance,
functional integration, etc. These are just examples)
- Reverse engineering Java code from Eclipse project instead of
native o/s directory
- Implement context sensitive help
- Allow multiple diagrams to be open at one time (hard - probably
requires changes to ArgoUML core)
- Allow multiple models to be open at one time (harder)
- other stuff (Mylyn integration, integration with CDT, Python, etc
plugins, ... let your mind run wild)
The list gets more vague as it progresses, but the important things
are a) the top level bullets, not the detail and b) achieving an equal
or better user experience (ie goals 1-3).
If anyone's application gets accepted for ArgoEclipse and I'm the
mentor, this is very much what the priorities will look like. If
someone else is the mentor, they'll probably at least have the
successful applicant doing #1. :-)
Tom
On Mon, Apr 7, 2008 at 3:48 PM, Brian Hudson <[EMAIL PROTECTED]> wrote:
> " I know that this seems like 11th hour feedback in the context of GSoC,
> but we can work to fine tune your proposal."
>
> As far as submitting modifications to my application(s) through Google's
> system thats true =)
>
>
> It sounds like there are a handful of things that could be useful tasks to
> undertake as part of GSoC:
>
> - Improve the existing integration between ArgoUMLs Swing user interface,
> and Eclipse workbench. There are a few tips in this Eclipse Corner Artcile
> that could be incorporated pretty quickly:
> http://www.eclipse.org/articles/article.php?file=Article-Swing-SWT-Integration/index.html
>
> - Begin rewriting some of the ArgoUML dialogs in SWT
>
> - Tie ArgoUMLs editors to built in Eclipse functionality (global actions,
> jobs, etc)
>
> - Begin writing an ArgoUML editor based on GMF or GEF (though if its true
> that Tigris GEF is assumed in many places this may be pretty difficult)
>
> - Begin to create extension points for well defined systems: Models,
> Modules, Code Generators, Reverse Engineering, etc...
>
> - Examine EMFs reverse engineering capabilities
>
>
> Of course not all of these could be completed within the GSoC timeframe.
> I've got some experience with Eclipse RCP, Extension Registry, Eclipse GEF,
> SWT, etc so I think these are all realistic tasks for me.
>
> What would be of most use to ArgoUML?
>
>
>
>
> On Mon, Apr 7, 2008 at 3:03 PM, Tom Morris <[EMAIL PROTECTED]> wrote:
> >
> >
> >
> >
> > > I believe we are all agreed to move ArgoUML to the eclipse platform
> > > By this, do you mean the only way to run it will be *inside* Eclipse?
> >
> > Yes.
> >
> >
> > > Or do you mean that it will use (I think I'm getting this right…) the
> Eclipse
> > > "Rich Client Platform," which is basically a framework providing UI and
> > > functionality?
> >
> > Yes. :-)
> >
> > Basically when double-clicked your ArgoUML icon (the shiny new one), a
> > mini instance of Eclipse, bundled with the rest of ArgoUML, would get
> > launched. It would act just like the ArgoUML of today, except that it
> > would have context sensitive help, automatic software updates, and a
> > bunch of other functionality without us having to implement it
> > ourselves.
> >
> > I'll send some screen shots in a separate thread.
> >
> > Tom
> >
> > ---------------------------------------------------------------------
> >
> >
> >
> >
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
> >
> >
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
