[Openmcl-devel] openmcl 1.0 vs. GUI

David Steuber david at david-steuber.com
Mon Aug 29 09:33:10 PDT 2005


On Aug 29, 2005, at 1:00 AM, Hamilton Link wrote:

> No feedback IS feedback.  It means that people don't think it's good 
> enough to bother with yet, and I know why I personally don't think 
> it's worth bothering with yet, and I've said why before: speed.  Oh, 
> and pretty often the last time I used it it was still throwing errors 
> subtly suggesting I should Save Often.

Just to provide another data point here, I use Emacs + SLIME with 
OpenMCL (and also with SBCL).  I haven't even tried the IDE yet.

For OS X applications, I've really only been playing, and not with 
Cocoa.  I gather I am one of few (if there are any others) who use 
OpenMCL's FFI to access the Carbon APIs.  The only difficulty I have 
that really annoys me (and I have no clue as to what a fix would be) is 
if something in the main thread (where the GUI must reside) signals an 
error to send me to the debugger.  SLDB will come up and then never go 
away.  So I've used ignore-errors as a bandaid in my callback function. 
  The condition is logged in the grand tradition of printf style 
debugging and the code plows on in the face of adversity.

I haven't touched Carbon in a while but I just went and rebuilt the 
last app I was working on to confirm that it still builds with the 
bleeding edge (CVS as of a few days ago).  It built.  I got a bundle.  
The bundle launched.

I'll do my best to find time to exercise the code in the bleeding edge 
cvs.  That is my default Lisp development environment (used via SLIME 
as I stated above).  PFD's ANSI testing package should do a good job of 
exercising code paths, but I haven't installed it to try it out.

I would actually like to see more of the slime features available via 
the slime-openmcl.lisp backend.  Nicer argument names for autodoc-mode 
would be nice too.  I think Bryan O'Connor was working along those 
lines in relation to getting better docstrings in the image.  I use the 
CLHS for CL package symbols, so better args would be my preference in 
priority over docstrings.  Obviously the CCL package could use good 
docstrings.

PS:  I just hit CVS again and then had to download a new bootstrap 
image (those are sure changing frequently these days!).  It looks like 
things are still working.  I'll report bugs if I find them.




More information about the Openmcl-devel mailing list