[gclist] Web hosts for the FAQ

Chris Reedy creedy@mitre.org
Tue, 5 Mar 1996 13:11:23 -0500


At  8:21 AM 3/5/96 -0600, Kaelin Colclasure wrote:
>Mike Spertus wrote:
>
>[...]
>> In general, I am a bit concerned about the emphasis of the FAQ. I don't
>>think that
>> the lack of detailed technical information on GC algorithms is really what
>>is holding
>> GC back. I think that the site should have a technical backbone but spend
>>most
>> of its effort orienting people to a GC mindset (without being fluff) and then
>> have pointers to the GC paper repository, etc.
>[...]
>
>I tend to disagree.  The people who're most likely to read the FAQ are
>those already considering either incorporating GC or using a GC'd language
>in their work/project/etc.  Yes, the FAQ should point out the advantages
>and address some of the myths-- but it should also provide a consolidated
>overview of the technology available (which I believe it is shaping up
>to do nicely).
>
I tend to disagree with your disagreement.  I think there are a lot of
people who might want to use GC but would say something like:

  "I saw a Lisp (or Smalltalk) program and it used GC and it behaved
unacceptably (too slow, paused too long)", or

  "It sounds nice but I don't have the time or skill to implement one,
where do I find the software to do it", or

  "I think it sounds like a good idea.  My boss thinks I'm crazy.  Give me
some ammunition to help me support my position."

I would like to see answers to these kinds of questions in the FAQ.

Plug:  I think you're going to see a lot more of this kind of question.  In
my job I am seeing a lot of organizations who want or are trying to do
business applications (you know, the kinds of things that used to be done
in COBOL) in C or C++.  Somebody talked earlier about difficulties of
handling multi-threading.  Most COBOL programmers have never had to deal
with any form of dynamic memory management.  About half of them will never
understand malloc and free, much less the complexities associated with
constructors and destructors in C++.  I keep hoping that good efficient
garbage collection will provide a way out for these people.

  Chris

This is an informal message and not an official Mitretek Systems position.
Dr. Christopher L. Reedy, Mail Stop Z667
Mitretek Systems, 7525 Colshire Drive, McLean, VA 22102-3481
Email: creedy@mitretek.org  Phone: (703) 883-7183  FAX: (703) 883-6991