amos@nsta.UUCP (Amos Shapir) (08/03/87)
In article <3664@well.UUCP> ewhac@well.UUCP (Leo Schwab) writes: > ... However, once you have the >basic algorithm down, your next step should logically be to translate that >algorithm into assembly by hand. Compilers can do a good job, but never as >good as a human. > > I point to accomplishments of the past. > [examples of small & efficient assembly programs] Have you seen the output of a *recent* compiler? Some of the modern compilers compile into code no human could keep track of. On the other hand, have you ever tried to maintain a programming project in assembly, or port it to a different architecture (I mean a *real* program, not something that can fit in 4k)? > Final note: This is not a personal attack. I simply feel that >assembly code should be used more often than it is currently. The increasing power of processors and the decreasing price of memory on one hand, and the growing number of different architectures (RISC etc.) on the other hand, supply both the incentive and the means for writing better HLL compilers. It is most probable that you'll see much less assembler programming in the coming years, not more. Just a small example of the trend: SUN are in the process of changing the processor architecture of their machines; do you suppose they could even think of doing that if most of their software were written in assembler? -- Amos Shapir (My other cpu is a NS32532) National Semiconductor (Israel) 6 Maskit st. P.O.B. 3007, Herzlia 46104, Israel Tel. +972 52 522261 amos%nsta@nsc.com @{hplabs,pyramid,sun,decwrl} 34 48 E / 32 10 N