Path: utzoo!attcan!uunet!pdn!tscs!tct!chip From: chip@tct.uucp (Chip Salzenberg) Newsgroups: comp.arch Subject: Re: Compiler Costs Message-ID: <2698D786.5310@tct.uucp> Date: 9 Jul 90 19:14:14 GMT References: <1797@apctrc.UUCP> <373@e2big.mko.dec.com> <2319@l.cc.purdue.edu> Organization: ComDev/TCT, Sarasota, FL Lines: 35 According to cik@l.cc.purdue.edu (Herman Rubin): >There are two problems with the present assembler. First, there is the >gosh-awful syntax and arbitrariness of them. Assembler syntax isn't an architectural issue. For example: The Z-80 has two widely used assembler syntaxes, the extended 8080 and the Zilog. Both of them are, to me, difficult to read. Compare, however, the smal2z language which was bundled with an old Intel C compiler. Using a bit of lex and yacc in front of the normal assembler, the smal2z translator accepted C syntax for Z-80 operations: A = B; /* LD A,B */ B = *HL++; /* LD B,(HL) | INC HL */ do { /* label: */ *IX++ = 0; /* LD (IX),0 | INC IX */ } while (--B); /* DJNZ label */ My point: Assembler syntax isn't an architectural issue. It's just an arbitrary way to present the architecture to programmers. >There is also the attitude in the computer field that the user need not take >these things into account, that the compiler is a great genius and should do >it all for you. The attitude I have observed is more like: "I'd rather let the computer do the grunt work for me, than do it myself." Micromanagement produces microsuccess. I prefer macrosuccess. If you don't like HLLs, don't use them. But don't heckle those of us who would use HLLs as a valuable tool -- even our primary tool. -- Chip Salzenberg at ComDev/TCT ,