[mac-lisp-ide] Where /is/ everyone?

[Duncan Rose]

I'm still here too; it's been a little slow with the McCLIM Cocoa back 
end over the last couple of weeks, but hopefully things are moving 
again now. Currently the McCLIM Listener isn't really usable - this is 
my priority at the moment.

Whilst not going into the same amount of detail as Mikel, here's what 
*needs* to be done, and what I'd like to do over the coming weeks:-

  + "Core" functionality needed to get the Listener working
      * scrolling (I'm working on this now) [1]
      * pixmap support
      * menus
      * flipping ink
      * keyboard events (currently only ASCII input works)

  + Other stuff to complete the back end
      * a couple of primitive drawing methods aren't implemented yet 
(ellipse + circle)
      * bringing it up to date with the latest McCLIM sources (Andy 
Heffner added scroll wheel support into McCLIM which should be easy to 
add into the Cocoa back end)
      * LOTS of tidying up (I'm still learning, and I think it shows!)
      * Native pane implementation - replace the standard McCLIM L&F 
with pretty Aqua as default in the Cocoa back end

Once this is done I think it will be ready to checkin to the McCLIM 
repository. Depending on how long all this takes, I might push for 
committing it after the Listener runs but before everything else is 
done.

Other than this, I want to work on finishing off McCLIM (it's not 100% 
complete as per the spec), including writing a comprehensive test 
suite. I'd like to see an CLIM editor, inspector, stepper and debugger 
(I don't think I'm alone in this) and also a documentation authoring / 
viewing tool would be nice. I've found some things in CLIM to be 
difficult to understand and whilst the code is always going to be the 
best documentation, some written documentation would be nice too.

As far as the editor goes, there appear to be a couple of options. The 
first is just to use Emacs, but I'd like something that is CL (and 
integrated better into the overall environment - which also needs 
implementing[2]). Other alternatives include Hemlock, or waiting to see 
what others create and stealing that ;-)  Another alternative that 
sounds interesting (to me) but which might not be practical would be to 
port the Dylan Deuce editor into CL + CLIM (there are many similarities 
between DUIM + CLIM, but also many differences. Whether it's possible 
to work around the differences or not remains to be seen).

Since I'm working in OpenMCL, using the OpenMCL inspector as a base for 
a McCLIM inspector has some appeal. Obviously the UI elements would 
need reworking (this may well hold true for all these tools of course; 
if an MVC approach is followed to their development there's no reason 
the same basic application code can't be reused with different UIs 
atop).

As to a stepper and debugger, I don't really have any ideas how to even 
start on these kinds of tools (and I suspect OpenMCL would need some 
hacking to permit highly-functional implementations too - but until I 
have time to sit down and trawl through its internals (that would be 
sometime in 2030 I think...) I'm just making wild guesses).

-Duncan

[1] In case you're wondering, the "scrolling problem of the day" is 
that McCLIM seems to perform scrolling by altering the scrollee's 
transformation and region. The functions that do this in CLX 
(clx:drawable-x etc.) appear to cause the server redraw the mirror 
automatically - not so in Cocoa, and using 'dispatch-repaint' in CLIM 
leads to infinite looping. Now I'm wishing I'd done the drawing in the 
Cocoa back end differently so I could just send a Cocoa 
"setNeedsDisplay". I guess more kludges are needed ;-)

[2] It's possible (likely?) this will be "in" McCLIM already by the 
time I get around to it...


On Tuesday, March 23, 2004, at 04:16 PM, Brian Mastenbrook wrote:

> Greetings all,
>
> This list has been dead for a while, so I'm wondering where everyone 
> is? As I recall people were working on a few cool things: Clotho, 
> McCLIM/Cocoa, and SBCL/PPC callbacks for Carbon. I'd love to know how 
> people have been doing on these!
>
> Brian
> --
> Brian Mastenbrook
> bmastenb at cs.indiana.edu
> http://cs.indiana.edu/~bmastenb/
>
>
> _______________________________________________
> mac-lisp-ide site list
> mac-lisp-ide at common-lisp.net
> http://common-lisp.net/mailman/listinfo/mac-lisp-ide
>