[josm-dev] Fwd: start of gwtosm the google webtoolkit port of josm

[Ævar Arnfjörð Bjarmason]

On Wed, May 26, 2010 at 20:36, Frederik Ramm <frederik at remote.org> wrote:
> jamesmikedupont at googlemail.com wrote:
>> gwt is a java to javascript compiler that promises to allow us to port
>> parts of the josm to an very thin client,
>> i have the absolute basics (just points) being displayed now, check it out.
>
> Forgive my being thick but I still don't see the point. Surely
> Javascript that has been auto-generated from JOSM Java code will never
> be able to match a proper editor devised in Javascript from the ground
> up, and there have been impressive steps in that direction (chrschmidt's
> OpenLayers based feature editor comes to mind).
>
> Is there really any hope that you will arrive at a anything remotely
> usable by modifying JOSM so that it is a better basis for
> auto-generating Javascript code? This sounds very much like wishful
> thinking to me. I mean you're free to try of course, I'm sure it is an
> interesting experiment ;-)

People are writing large applications in GWT. JavaScript is
dynamically sent over the wire as needed for the client. I don't have
any experience with it, but JavaScript engines are getting very fast
these days.

This also has the advantage that you don't need to write a completely
new editor in JavaScript, you can just use existing Java code (with
some changes).

Anyway, we'll see how it turns out.

> If one wants more than that, one can run JOSM as an applet in the browser.

I think the point is that there's Java backend code and a native
browser UI using HTML5+CSS like Canavas and SVG, not just an ugly
stand-alone SWING interface.

> And Avar, if one wanted some kind of more or less clean engine/library
> as a basis for other projects, JOSM would be the worst choice in my
> eyes. Much better to start with either the JOSM-NG codebase or the
> existing Java osmlib that has been built for Traveling Salesman!

Maybe. I'm not familiar with it. But it does seem to do a lot of
things right, and it's well maintained and complete. You could do a
lot worse than using its primatives and well tested upload/download
code for a new client.

Isn't josm-ng dead-ish at this point? I thought it was just a one-man
project whose ideas are mostly in the main josm by now, maybe I'm
wrong.

Anyway, we'll see how this GWT experiment turns out.