Prev: TMA Assembler?
Next: pshufb
From: T.M. Sommers on 7 Nov 2006 00:23 Frank Kotler wrote: > > As I recall, when I installed gtk(+ ?), I had to install four or five > other libraries, in the correct order, first. I'm not familiar with > wxWidgets - didn't it used to be called wxWindows? I think I downloaded > that once - 7 or 8 Megs, IIRC. My reaction to this stuff is, "That looks > like a truck tire. I wanted a racing bike tire. I'm going to have to > reinvent the wheel!" The thing is, after you reinvent the wheel, you either have your very own truck tire, or you have a bike tire of very limited functionality, and you have spent ages developing it. On the other hand, if you had used existing libraries, you would have a pile of applications already built. -- Thomas M. Sommers -- tms(a)nj.net -- AB2SB
From: Herbert Kleebauer on 7 Nov 2006 04:11 Betov wrote: > Herbert Kleebauer <klee(a)unibwm.de> ?crivait news:454F6D79.3B2DF851 > > Assembler is only good for learning and understanding but > > not for programming applications > > You mean: "What i call an Assembler is ..." I never know whether I should take you seriously. Imagine you have an idea for a new application and hire a programmer to implement it. I suppose, you tell him what this program should do, how the interface looks like and maybe a coarse structure of the program. But I don't think you will tell him at CPU instruction level what the program should do. To be able judge the quality of the code produced by the programmer, you should have the knowledge to do the programing yourself. Now, suppose the hired programmer isn't a human but a computer system. Why does this completely change your behavior? Now you tell the system not what you want but how you want it at the CPU level. Why not use a HLL and tell it what you want and let the system decide how to do it at CPU level. But again you must be able to do it yourself at CPU level to be able to ask the compiler only for jobs which can be efficient translated to CPU level and also to judge the code generated by the compiler. And therefore you have to learn assembly programming and not for using assembly to write applications. > * First, you write an Assembler that cannot be used for > seriously developing Asm Apps. No. First you think about for what an assembler is useable. And then you write an assembler which is optimized for this purpose. > * Then, you say that doing so is using the wrong tool. It's not the wrong tool, it's the wrong language. > In some counter way, you are right: Using your Assembler for > "programming applications" would be absurd, as long as it has > no other purpose but "for learning and understanding". It's like speaking to very small children. People use a very different language (vocabulary) when speaking to babies than when speaking to grown up people. Now you can say: "I always have only spoken to babies and I don't know any other way to speak. I have adopted a few words (IF, ELSE) I heard from adults, but that has to be enough." There is nothing wrong with this, but if you tell other people, that they also should use the baby language because it's so simple and everybody can understand it, then this is wrong. I also don't speak (and don't like) OO languages, but I don't say, people shouldn't use OO languages and program in normal C, because this is much simpler.
From: Herbert Kleebauer on 7 Nov 2006 04:35 "T.M. Sommers" wrote: > Frank Kotler wrote: > > > > As I recall, when I installed gtk(+ ?), I had to install four or five > > other libraries, in the correct order, first. I'm not familiar with > > wxWidgets - didn't it used to be called wxWindows? I think I downloaded > > that once - 7 or 8 Megs, IIRC. My reaction to this stuff is, "That looks > > like a truck tire. I wanted a racing bike tire. I'm going to have to > > reinvent the wheel!" > > The thing is, after you reinvent the wheel, you either have your > very own truck tire, or you have a bike tire of very limited > functionality, and you have spent ages developing it. But now you understand how a tire is build, you know the weakness and strongness of a tire, you know about the problems which arise when you drive to fast in a curve or if the temperature is to high or to low. And that's the purpose of building your own tire and not to use this tire on your car. > On the > other hand, if you had used existing libraries, you would have a > pile of applications already built. That's the amateur approach, a professional has to understand the advantages and disadvantages of the tools he uses.
From: Betov on 7 Nov 2006 05:03 Herbert Kleebauer <klee(a)unibwm.de> ?crivait news:45504DD6.9D8E1554 @unibwm.de: > First you think about for what an assembler is useable. And > then you write an assembler which is optimized for this purpose. Nothing but what i mean. Exactly. * First you build yourself a concept of what Assembly is. * Then, you develop an Assembler, accordingly to your view. * Then, you say that Assembly is not the proper thing for developments. The snake eating its own tail. My own view is: Whatever development HLL (i don't talk about script engines), it always ends with a Tool calling for Asm Snippets, in a forced Syntax. The "forced Syntax" is a thing does not even deserve any comment (so hatable "containment"). The very end result of it, is useless complexity. If there had not been _money concerns_ at the beginning of this absurd history, no HLL, on earth, would ever have existed. The only thing that would exist would be collections of Snippets, in Asm to be called from Asm, in Asm Syntax, as long as there is absolutely no technical reason for CALLing from HLL rather than from Assembly. Betov. < http://rosasm.org >
From: T.M. Sommers on 7 Nov 2006 11:49
Herbert Kleebauer wrote: > "T.M. Sommers" wrote: >>Frank Kotler wrote: >> >>>As I recall, when I installed gtk(+ ?), I had to install four or five >>>other libraries, in the correct order, first. I'm not familiar with >>>wxWidgets - didn't it used to be called wxWindows? I think I downloaded >>>that once - 7 or 8 Megs, IIRC. My reaction to this stuff is, "That looks >>>like a truck tire. I wanted a racing bike tire. I'm going to have to >>>reinvent the wheel!" >> >>The thing is, after you reinvent the wheel, you either have your >>very own truck tire, or you have a bike tire of very limited >>functionality, and you have spent ages developing it. > > But now you understand how a tire is build, you know the weakness and > strongness of a tire, you know about the problems which arise when > you drive to fast in a curve or if the temperature is to high or to low. But you, not being a tire expert, have probably built a tire with different characteristics from those built by tire engineers with years of training and experience. > And that's the purpose of building your own tire and not to use this > tire on your car. That's a good thing, because if you used it you would likely kill yourself. >>On the >>other hand, if you had used existing libraries, you would have a >>pile of applications already built. > > That's the amateur approach, Hardly. Let's both bid on a contract, with you including the extra time and cost to build everything yourself, and me using existing libraries. Who do you think will get the contract? > a professional has to understand the > advantages and disadvantages of the tools he uses. You can do that without rebuilding those tools. Do carpenters make their own saws and hammers? If you really believed in your approach, your first step would be to reinvent mathematics. Then you would reinvent solid-state physics. Then electronics. Then you would design and build your own hardware. Then you would create your own operating system. Then, and only then, you could start reinventing X. Meanwhile, the rest of us will be getting some useful work done. -- Thomas M. Sommers -- tms(a)nj.net -- AB2SB |