Return to the
comp.compilers page.
Search the
comp.compilers archives again.
| From: | phr@netcom.com (Paul Rubin) |
| Newsgroups: | comp.lang.perl.misc,comp.lang.misc,comp.compilers |
| Date: | 30 Oct 1996 13:18:06 -0500 |
| Organization: | NETCOM On-line Communication Services (408 261-4700 guest) |
| References: | <5437ev$30u@shell1.aimnet.com> <545mqn$qul@picasso.op.net> 96-10-099 |
| Keywords: | design, history |
>Sorry to keep blabbing about this; it's off-topic for most of these
>groups. But it's so interesting I can't shut up. The implementation
>notes for SNOBOL4, version 3, say that associative arrays are *not*
>implemented as hash tables, as you might expect they would be.
>[they were an array of pairs of pointers to hashed strings]
SPITBOL, a fast implementation of Snobol4 for the S/360, implemented
associative arrays as hash tables. There was very little difference
except Spitbol was a lot faster than BTL SNOBOL4.
--
Return to the
comp.compilers page.
Search the
comp.compilers archives again.