Re: [platform-swt-dev] SWT port for Swing!

["Liyu Liu" <liyul@xxxxxxxxxxx>]

Just my 2c as some one new to this group.

It's interesting to have SWT over swing. However, IMHO, there are at least
three problems to be solved.

1) SWT becomes at least as slow as swing. (slow)
2) People find swing's MVC design more flexible and don't like the more
restrictive SWT interface. (less flexible)
3) You have to emulate native UI again rather than simple reused them. (it
may be ugly, and reinventing wheels)

Weigh against the small and questionable gain, --more portabitlity in theory
but in reality doesn't pan out very well. Actually SWT is proven to be more
portable than swing in real world and even better for native compilers like
gcj.

SWT is what AWTshould have been. So it is probably more useful to have
AWT/Swing over SWT so that all the work of swing can be reused and you can
freely host AWT/Swing components when flexibility is needed.

Also, it's needed to get SWT work in javabeans. Sun's implementation will
honor AWT/Swing only, but a good SWT UI component frame work is important
for the continuing prospering of SWT.


Luke

> Anyway, I think Swing is a really good idea for a port so that SWT
> applications could run without changing the code for the few
> non-supported platforms. Eventually developers could use some Swing
> specific code if they want (similar to OLE on Windows). Moreover, SWT
> programs would work right away whenever a new platform with new VM is
> out, while waiting for the real native port to be developed. There are
> of course many other possibilities brought by such a port...

^^^^^^^^^^^^^^^^^^^^^^^^
These effort are probably more laudable in the unlikely case we hit a hard
wall in doing native SWT.