On Fri, 25 Mar 2005, Paul Koning wrote: The biggest reason why I don't think that I should retire yet - I teach an assembly language class. They're LISP (actually "SCHEME") fanatics. They demo'd an example that "can not possibly be solved in ANY way but recursion". While they were keying it in in Scheme, I wrote out a non-recursive solution (with a 2D array) in C, BASIC, FORTRAN, and I got halfway through writing the COBOL form. -- Grumpy Ol' Fred cisin at xenosoft.com

On Fri, 25 Mar 2005, Fred Cisin wrote: That's because youre a real programmer and understand what a computer is. None of our (actually very smart and some even experienced) grad students actually know what a computer is. I don't think hardly anyone does. We had a little robotics project, a bunch of little controllers that take commands from a big host, and they wanted to write it all in Java, threading, tasks, etc. I showed the programmer how to do non-blocking parallel tasks (serial command string <header ID command args cksum> and a PWM task driving pneumatic actuators) using a simple one-byte state variable and a dispatch table. They'd absolutely never heard of such a thing. We did this in a PIC. The lightbulb went on in the programmer kid's head -- on an unrelated project he used the state variable thing in a c++ prog and got his program done faster and it ran faster than all the threaded junk with interprocess communications. Well duh.

On Sat, 26 Mar 2005 21:57:16 +0000 (GMT) ard at p850ug1.demon.co.uk (Tony Duell) wrote: I thought a computer was a man, or woman, who specialized in computing sums and other mathematical calculations. At least that's what my 1920 Funk & Wagnells says. Being more on-topic, there is a lot of historical Data Processing equipment out there that can not be considered 'a computer.' IBM had a thriving business in punched card equipment decades before the first stored-instruction digital computer. Is a 68HC11 microprocessor a 'computer' if I use it in multi-chip mode, with some sort of serial terminal arrangement that allows user input and programmability, but NOT a computer if I use it in single-chip mode and the on-chip serial port is just used to read logged values from the on-chip A/D? -Scott

Anyone have the docs on this DT1712-32DI-PG (Unibus) module by Data Translation? Apparently it's either a AD or DA or both, and it's supposed to be DEC compatible, but as to which DEC module it's compatible with...? /wai-sun

Roger Merchberger wrote: Jerome Fine replies: I was told the following definition of a computer about 45 years ago: SPEEDY MORON! I find that it still fits (GRIN)!! In fact, until successful AI is developed, it will continue to fit - with the emphasis being more on the SPEEDY as opposed to the MORON. Sincerely yours, Jerome Fine -- If you attempted to send a reply and the original e-mail address has been discontinued due a high volume of junk e-mail, then the semi-permanent e-mail address can be obtained by replacing the four characters preceding the 'at' with the four digits of the current year.

On Sun, 27 Mar 2005, Roger Merchberger wrote: "A computer terminal is not some clunky old television with a typewriter in fron of it. It is an interface where the mind and body can connect with the universe and move bits of it about." - Douglas Adams

Megan wrote: Which is presumably not actually very useful as a definition of a computer, since a cow would qualify. Eric

Chickens do that; they turn bugs into birdshit and eggs. For a stored program, automatic, digital, computer, the thing that distinguishes it from other human artifacts is this one simple-seeming thing: the machine that modifies itself. (I know that now a bunch of pointless threads about "well as I drive my car the engine wears, that's 'changing itself'" but include me out of that time-wasting nonsense.)

And, to a first, second, and even third approximation, with no errors. "I really hate this damn machine I wish that they would sell it It never does quite what I want But only what I tell it." We'd be ecstatic if practically anything else had error rates well under one in 1e9[%]. But most computers, if they failed one instruction in 1e9 - never mind one clock cycle in 1e9 - wouldn't last even long enough to boot; it's not only common but *expected* that they do exactly what they're told to do, to within ten, eleven, twelve nines of "always". It's really pretty impressive. [%] Given our population demographics, I hesitate to say "billion" no matter what number I mean by it. :) /~\ The ASCII der Mouse \ / Ribbon Campaign X Against HTML mouse at rodents.montreal.qc.ca / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

Problem is, AFIAK this list is not just about stored-program digital computers... Analogue machines are on-topic, so arguably are machines that can't change their own program. Quote from the preface of the TRS80 microcomputer technical reference handbook (for the model 1 of course). 'If you are a technical person, but have never played around with microprocessors, dive right in! A computer is nothing more than a fast idiot. Don't let the "idiot" intimidate you. All of the familiar logic is there; there just may be more of it than you're used to seeing in one place.' -tony

:-) as would we all... Just an example of how someone viewed it...

On Fri, Mar 25, 2005 at 03:03:54PM -0800, Fred Cisin wrote: If they claim any such bullshit then they obviously weren't paying attention during class. Any recursive algorithm can be implemented as an iterative solution. At worst by emulation of the recursion stack, but usually much better. Recursive algorithms are often more elegant and easier to implement. Unfortunately they tend to blow up the stack if one throws sufficiently large datasets at them. At which point one switches to a non-recursive implementation ;-) Regards, Alex. -- "Opportunity is missed by most people because it is dressed in overalls and looks like work." -- Thomas A. Edison

On Sat, 26 Mar 2005 00:03:54 +0100, Fred Cisin <cisin at xenosoft.com> wrote: ... I once programmed quicksort in RPG-II (because we did not have an assembler for the System 3, IBM wanted lots of money for it. -- -bv

William Donzelli wrote: I fully disagree; compilers output assembly. As long as there are compilers, there will be assembly. -- Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/ Want to help an ambitious games project? http://www.mobygames.com/ Or check out some trippy MindCandy at http://www.mindcandydvd.com/

Eric Smith wrote: Now REAL programming in machine code does require the real blinking lights... :) Ben alias woodelf

Jim wrote: I'm not sure I buy that either. Hundreds of years from now (or after the singularity), when most machines are designed by other machines with no human intervention other than stating the requirements, will there be anyone who understands the lowest level of how they work? Will they even be simple enough that it's possible to understand them the way we can understand a microprocessor today? But I agree with your point that assembly language is not dead, or even "nearly dead". Though it certainly seems to be much less used relative to other languages than it was even just ten years ago. Eric

Jim wrote: level", > and people will still understand and work with it. Possibly. But it likely won't be anything at all like assembly language. I wrote about the prevalence of assembly language programming: Jim wrote: Sure, it's opinion. I don't have solid numbers, and I don't know where to get them. But I'd be willing to wager real money that if well-researched figures on this subject existed, they would show that there are at least a thousand times more lines of code written worldwide per year in languages other than assembler, than there are in assembler. In fact, it wouldn't surprise me if the factor was over a million or even ten million. And I'm fairly sure that whatever the ratio is, it's larger than it was ten years ago, and still on an upward trend. I've been a programmer for 29 years now, professionally for more than 20 of that. Much of that has been in embedded systems, which in general are *more* likely to use assembly than software for general purpose computing. I even write a lot of assembly myself for microcontrollers. But even so, I *still* see that the industry as a whole appears to be using a very much lower percentage of assembly code than it was ten years ago. I understood assembly well back then, and still do now. Yet I write a much lower percentage of my code in assembly now than I did back then. Why? Because it's rarely necessary, and even though I'm fluent in assembly language, I still have better productivity in C or Ada. When profiling shows that an inner loop is a bottleneck, sure I'll consider recoding it in assembly. And when I'm writing timing-critical code to control hardware, which I've done plenty of (e.g., generate an NTSC composite color video signal in real time entirely in software), I'll do it in assembly. When I'm writing test routines for the PDP-1 restoration project, or for my own PDP-8, PDP-10, and PDP-11 systems, I do that in assembly. But for me, and for most programmers, those are the exceptions, not the rule. I never said that it wasn't still in use. But a program that uses 1% assembly code today probably would have been written using more like 5% assembly code ten years ago. Eric

At 12:38 AM 3/26/2005, Eric Smith wrote: Sure. Sellam's and Tony's heads will be preserved in jars and still talking. I saw it on TV in a show about the future. - John

<raises hand> That's a silly question to ask on this list, isn't it? Gordon.

der Mouse wrote: On the basis that posting the recipe would probably make those readers in the less politically-free parts of the world into terrist suspec's, I will keep quiet on that one. Heat iron ore with charcoal. The carbon reduces the oxygen, leaving you with fairly poor quality iron. This is called "carburisation". You can then batter that about until you get wrought iron. Gordon.

Tony Duell wrote: And always use a scratch monkey before testing any neural implants. :)

On Fri, 25 Mar 2005 21:58:52 -0600 Jim Leonard <trixter at oldskool.org> wrote: Furthermore, there are machines for which a stack-intensive environment like C just will not do the job. What comes to mind to me are some of the PIC and the tinier 6805 controllers, i.e. programming environments where you might have from 32-512 bytes of Read/Write memory for ALL your data needs. When you've got an app for a controller and the part can't cost more than fifty cents, you code it in Assembly Language. -Scott

Fred Cisin wrote: I think the problem is nobody DOC's assembly langauge programming or have a GUI that you can use with bare hardware. I think software has kept the same speed, because the value of Memory size VS memory speed has been mostly constant. Ben alias woodelf. PS. -- MICROS-1 must be origin of system crashes. :)

It was written... I do, darn near daily. Idiots. Jay