William "Chops" Westfield wrote: > On Sep 21, 2005, at 8:12 AM, Dave Tweed wrote: > > > But the worst part, the part they should be killed for, they > > > were running out of code space - so they DELETED THE COMMENTS!!! > > > > That used to be a trick used with very early tape-based microcomputer > > BASIC interpreters, which kept the entire program -- tokenized source > > along with comments -- in main memory. Deleting comments could free up > > significant amounts of space for additional code tokens. > > Tokenized, hell. Early basic interpreters didn't tokenize; you could > save space by using short forms of keywords, or omitting the "let" in > assignment statements (if that was allowed.) It's presumably still an > issue on BASIC51 micros. Note that omitting comments made your program > go FASTER, too! > > > AFAIK, this technique never applied to any other language > > It's a potential problem with any pure interpreter. Forth comes to > mind, though its disk access was so primitive that space not used > by comments was wasted anyway. Postscript would be a modern example. > A postscript program with no comments downloads to the printer faster, > too. I thought about Forth, but that was *never* implemented as a pure interpreter, except when executing a typed-in command line. The comments didn't get included in the memory image (dictionary), so they couldn't affect program size or performance. As you say, the primitive block storage system used for source code wasted disk space anyway. -- Dave Tweed -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist