peter@sugar.hackercorp.com (Peter da Silva) (11/21/90)
> >Assembler is a minus ? What are you getting at ? Assembler programming > >certainly is the best and most taxing way of programming for programmers. Most taxing, yes. Best, NO! I just downloaded a program into my Amiga 3000. It came with source and the readme file went on and on about how great it was that this program was in assembler and how cool his programming style was and how mad he was that his boss was making him program in Aztec C, so he was doing this in his spare time... Crashes most beautifully on the 3000. See, there are these differences between the 68000 and the 68030. You can't see them from C, but in assembly... Coding in assembly, other then for runtimes and in critical routines, is just a waste of time. Even if your assembly code is 10 times as fast as compiled code (unlikely) your program is spending most of its time in the Amiga runtime waiting for events... If a program is written in assembly, I tend to put it back on the shelf. > C has only one good use for real programmers, and that is that it is portable. And it's easier to code, and you can spend your time writing code for the user instead of for yourself. > So all you lowclass programmers who > can only program in one language (especially if you can only code in C) Who? Where? I write programs in the language best suited for the job, and the highest level language that allows for enough efficiency is that language. When needed I code in assembly (for any of a dozen CPUs), C, Fortrash, PL/M, Forth, any of half a dozen interpreted languages, and even COBOL. Programming in a lower level language out of machismo is just plain stupid, and is the source of much of the bad software out there. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
cedman@golem.ps.uci.edu (Carl Edman) (11/22/90)
In article <7104@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: > >Assembler is a minus ? What are you getting at ? Assembler programming > >certainly is the best and most taxing way of programming for programmers. Most taxing, yes. Best, NO! I just downloaded a program into my Amiga 3000. It came with source and the readme file went on and on about how great it was that this program was in assembler and how cool his programming style was and how mad he was that his boss was making him program in Aztec C, so he was doing this in his spare time... Crashes most beautifully on the 3000. See, there are these differences between the 68000 and the 68030. You can't see them from C, but in assembly... So what ? There is one more person who can't program well enough to know what (very few, really) things he may not do to keep his program portable. Assembler is , par se, just as binary-portable as C (among Amigas, that is). Maybe your dislike for assembler programming goes back to the fact that there is a correlation between unclean programming methods and assembler programming. That means that people who programm mainly in assembler are (on the whole) somewhat more likely to do stupid tricks. But the fact remains that you CAN program just as portably in assembler and that you can write a nightmare of incompatiblity in C (or even pascal or basic, if you want to). Coding in assembly, other then for runtimes and in critical routines, is just a waste of time. Even if your assembly code is 10 times as fast as compiled code (unlikely) your program is spending most of its time in the Amiga runtime waiting for events... That is not correct. Many applications require a great deal of internal computation. But even if it was true in every case: What do you care ? The judgement whether writing assembler for a definitive product is a waste of time or not, is a judgement of the programmer. You , as a user, really shouldn't care about how long it took the programmer to write this(that is his problem), but merely about how good a program it is. And an important part in this is speed and size. And here assembler can beat C handily. If a program is written in assembly, I tend to put it back on the shelf. Again you fail to give a reason for this. If you say: When a program is incompatible/doesn't multitask/a.s.o. then I don't use it, fine ! But really what point of criciticism is it against a real existing program in what language it was writen. When I use a program, I don't care in what language it was writen, but merely how it preforms. And assembler has at least a chance of being better than ANY high-level language. Carl Edman Theorectical Physicist,N.:A physicist whose | Send mail existence is postulated, to make the numbers | to balance but who is never actually observed | cedman@golem.ps.uci.edu in the laboratory. | edmanc@uciph0.ps.uci.edu
peter@sugar.hackercorp.com (Peter da Silva) (11/25/90)
[ I described a wonderful crash-prone assembler program ] In article <CEDMAN.90Nov21183628@lynx.ps.uci.edu>, cedman@golem.ps.uci.edu (Carl Edman) writes: > So what ? There is one more person who can't program well enough to > know what (very few, really) things he may not do to keep his program > portable. [etc...] You just don't get it, do you? The lower-level a program is, the more the individual programmer is responsible for making it work, and handling all the special cases for all the different systems it might run on. All you saw was one particular example. OK, here's another: a mandelbrot program that used a tight inner loop to get the most speed on a 68000. If that loop was in C I could just recompile with the right flags and get FPU code and it'd run faster. Or the guy who wrote the program could have easily done an FPU and a non-FPU version. But as it is I'm out of luck. It takes longer to code in assembler, and that translates directly to less software being written, less testing, fewer features, and longer waits for updates. > You , as a user, really shouldn't care about how long it took the > programmer to write this(that is his problem), but merely about how > good a program it is. And the higher level the code it was written in, the better a program it is likely to be. Even C is too low level for my tastes, but it'll do until something better comes along... > If a program is written in assembly, I tend to put it back on the shelf. > Again you fail to give a reason for this. I've now given *two messages* full of reasons. That should be enough. > And assembler has at least a chance of being better than ANY > high-level language. But almost always the higher level program is a better program for the long haul. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
maniac@howlin.cs.unlv.edu (Eric J. Schwertfeger) (11/26/90)
In article <CEDMAN.90Nov21183628@lynx.ps.uci.edu>, cedman@golem.ps.uci.edu (Carl Edman) writes:
)
) And assembler has at least a chance of being better than ANY
) high-level language.
)
) Carl Edman
)
)
I couldn't agree more! I`m working on a resident
notepad/free-form-database, that since it stays in memory,
has to be pretty small. I started programming it in C, using
as much native-amiga library calls as possible to keep the
size of the code down. at 25K, it was just managing the
dynamic string algorithms I use, opening windows, and
displaying text in the windows. I decided to recode the
program in Assembly, and it dropped to 14K, and is almost
fully functional.
--
Eric J. Schwertfeger, maniac@jimi.cs.unlv.edu
dvljhg@cs.umu.se (J|rgen Holmberg) (11/26/90)
Why are you even discussing this? Assembler is potentially faster, buggier and certainly takes longer to write. So what? If I want a very fast program and are less concerned with developing time I use assembler. If I need to write a program fast that does something without bugs I might even use Basic :-) All languages out there has advantages and disadvantages, we use the one that is best suited for what we want. If a product doesn't do what you want it is 99% sure that the reason for that is poor programming practice rather than the language involved. Programmers out there should try to get a working knowledge of many languages and use the one that meets their present needs. Just my 0.02 SKR worth ( heck, I don't even own $0.02 ;-) Jorgen -- ******************************************************************************* email dvljhg@cs.umu.se - other ways to communicate are a waste of time. Everything I say is always true, just apply it to the right reality. "Credo, quia absurdum est."