What I was saying about implementation time....
RE01 Rice Brian T. EM2
Wed, 28 Oct 1998 18:01:00 -0800
>A message got posted to a ton of language newsgroups 2 days ago about a
>new project Microsoft Research Laboratories is starting in collaboration
>with the Oxford University Computing Laboratory. Here are some excerpts:
>> The goal is to develop a
>> new kind of environment for transformational programming
>> that permits software to be composed from a set of
>> independent design decisions or "intentions", using
>> domain-specific notations and optimization strategies. The
>> specific aim of the Oxford component of the work is to
>> design a meta-language for the environment, within which
>> domain-specific abstractions can be described, implemented
>> and reused.
>> (a) identifying suitable features of current meta languages
>> in compiler construction and automated theorem proving.
>> (b) designing and building a prototype implementation of a
>> suitable meta language.
>> (c) experimenting with the use of that meta language in
>> case studies.
>I'm recalling my original prediction that if you don't build Tunes in 15
>years, a large effort will form to meet the task and you'll be beaten to
I'm sorry, but I can't help but rebut.
I've looked at MRL's page and papers for this intention-based
programming before. They have quite a few concepts confused, and they
are sticking to traditional approaches to the system. In other words,
don't hold your breath waiting for them to mention "total reflection",
"open implementation" or non-standard human-computer interfaces. They
also apparently don't have the brains (and I'll assume easily that
Oxford doesn't as well) to come up with the kind of mathematical basis
for such a system to be even close to the capacity that Tunes will have
Remember the Fifth-Generation Project that Jecel mentioned?
Corporations and universities inherently induce conservatism into their
research teams' members. They simply won't see the full benefits of
this system. That itself will kill any possibility of Tunes-level
exploration. The only plausible explanation is that "deep down",
Microsoft's reign of terror was (premeditatively!) engineered so that
this "greater cause" could be implemented without the company going
under. This seems a total contradiction, considering the fact that the
project is being advertised to those interested in programming
languages, and that Microsoft is now seeing greater profits (through Q4
98) than ever due to sticking with its usual scheme.
Sure, I'd like to believe that Bill Gates had this planned from all
along, but I know that he doesn't have that much time in the day, and
neither is he that brilliant to come up with the idea or to hire someone
who would. Microsoft's entire frame of reference, remember, seems to
everyone to be the mighty dollar (or yen or deutschmark or ...). If
they're coming up with a Tunes-capable system, then the late 20th
century Western world has been duped by the greatest bluff of the entire
history of the world. Even if the case is so, what a psychotic group of
sadists to consciously build this destructive software empire which
wastes the lives of billions of person-hours, just to turn around and
play the angel at the last moment.