shebs@apple.com (Stan Shebs) (05/09/89)
I was talking with somebody here and repeated the line about ANDF as a form of
UNCOL holy grail, but couldn't remember how or where I had picked up that
belief. So the question is: what are the original sources and experimental
work to support the assertion that something like ANDF is either impractical,
limited to a small set of architectures, or limited to a small set of
languages?
stan shebs
shebs@apple.com
[ANDF is practically a restatement of the UNCOL problem -- a common
intermediate language that sits between the parser and code generator so as to
reduce the problem of compiling N languages on M machines from an N*M to an
N+M problem. Seems to me that the best evidence that UNCOLs are impractical is
the fact that people have been working on them since about 1956 without any
notable success. I'll try and write up some of the reasons I think UNCOLs
fail in a separate note, and encourage other opinions. -John]
--
Send compilers articles to compilers@ima.isc.com or, perhaps, Levine@YALE.EDU
Plausible paths are { decvax | harvard | yale | bbn}!ima
Please send responses to the originator of the message -- I cannot forward
mail accidentally sent back to compilers. Meta-mail to ima!compilers-request