No More Object LISP

[lynch@aristotle.ils.nwu.edu (Richard Lynch)]

>
>Here is an excerpt from a recent posting of Bill St.Clair's in response
>to following question:
>>   
>>   Does the system and window interface implementation now use CLOS rather
>>   than (or as well as) ObjectLisp?
>>   
>>Everything is converted to CLOS.  ObjectLisp is no more.
>>
>>Bill St. Clair
>>bill@cambridge.apple.com
>>
>
>I suppose I'm doomed in this but I want to enter one last plea
>that some conversion functions or some sort of support (perhaps
>on top of clos) be given for object lisp. I have a VERY large
>program built on top of it and I want to be able to use 2.0.
>Do other people feel this need? I assume so since considerable
>interface work has likely been done with object lisp.
>  HOW ABOUT THOSE WHO AGREE LETTING APPLE KNOW, perhaps if
>there is sufficient concern it will be seen as a legitimate
>need.
>  One final comment, the reason I stuck with object lisp instead
>of PCL is that a long while ago when plans to upgrade to CLOS
>were announced it was said that object lisp would likely be
>supported on top of CLOS. The recommendation was that if  you needed
>speed in the short term (e.g., me) stick with object lisp rather
>than going immediately to aPCL version of CLOS on top of the then
>current lisp which would be slower.  It was pointed out that in 
>the long term the cost would be that object lisp would become
>slower than the full implementation of CLOS.  I chose the first
>route for speed figuring I could accept the slower running later
>of object lisp because I would presumably have a faster hardware
>platform by then.  Now I find I have, so to speak, no software
>leg to stand on.
>    BTW, is there a strong reason why such conversion code or
>support cannot be provided, does it turn out to be extremely
>hard to do?
>
>       Lucian
>
>
Ditto.  I was reassured (maybe 8 months ago?) that there would be
conversion functions, but they would not be "perfect" converters.  I HOPE
this hasn't changed (unless they got perfect :-)).  It would REALLY be a
pain to start all over again and rewrite by hand the extensive libraries of
code we have developed here at work and I have at home (yes, I purchased a
copy of MACL for myself...actually BEFORE I came to work here).
I suspect it WILL be (was?) fairly difficult to write a conversion
function.  Object LISP makes no distinction between classes and instances,
and CLOS is very rigid about that.

"TANSTAAFL" Rich lynch@aristotle.ils.nwu.edu