guido@mcvax.UUCP (Guido van Rossum) (11/04/84)
In article <211@oliveb.UUCP> jerry@oliveb.UUCP (Jerry Aguirre) writes: > One can't write compilers, editors, linkers, or operating >systems with standard Pascal because it is so over protective. Huh? The very first Pascal compiler was written in Pascal (even in a RESTRICTED SUBSET, actually). So were most Pascal compilers since. I sould say that none of the types of system programs you mention need be hampered by Pascal's (not OVER) protection, except for operating systems (kernels, actually). -- Guido van Rossum, "Stamp Out BASIC" Committee, CWI, Amsterdam guido@mcvax.UUCP "The number of arguments is unimportant unless some of them are correct."
jerry@oliveb.UUCP (Jerry Aguirre) (11/06/84)
Yes, you can write a compiler or text editor in Pascal if you are
willing to work with fixed length strings. However if you want to be
able to handle large files with limited memory this can be
inefficient. Put the text out on disk you say? But remember standard
Pascal has no random I/O. To get somewhere in the file you have to
read the entire file up to that point. It seems to me that an editor
using fixed length records and seqential I/O is going to be real
ssssslllllooooowwwwww.
Please, no quotes about systems that handle text as fixed 80 column
records at 110 baud using paper tape. Or stories about cooking over
a camp fire and walking 15 miles to school (in a blizard naturaly).
Of course you can write editors and compilers (and probably an OS) in
Pascal. The language was invented for teaching people to do that sort
of thing. However the difference between a class project and a tool
that gets commercial use can be a big one.
Jerry Aguirre @ Olivetti ATC
{hplabs|fortune|idi|ihnp4|ios|tolerant|allegra|tymix}!oliveb!jerryarndt@ttds.UUCP (Arndt Jonasson) (11/11/84)
(Sorry if this gets posted twice, I just made a mistake)
Actually, there is an editor developed by a computer club here
in Stockholm, which was almost entirely written in Standard Pascal.
The editor (called AMIS (Anti-MISery)) is an (almost) subset of
the original EMACS, and has been quite a success. It was developed
on a DEC10 for the Tops-10 system, but has since then been ported to
VAX/VMS, Pr1mos, NORD-10 among others.
There is no question that AMIS is slower (perhaps even much slower)
than it would have been if implemented in some machinedependent
language, but compared to what editors are otherwise available on
the named machines, an AMIS, even a slow one, is quite acceptable.
{decvax,philabs}!mcvax!enea!ttds!arndthans@log-hb.UUCP (Hans Albertsson) (11/12/84)
[]
While AMIS is written in Pascal, it does NOT use any standard Pascal I/O
at all. Every I/O request is relegated to machine dependent I/O modules.
This is non-standard in itself, since it requires you to be able to
use separately compiled modules, and link to modules in other languages.
AMIS isn't terribly slow, either, beating any other EMACS-lookalike by a
small but comfortable margin on VMS at least.
--
Hans Albertsson, USENET/uucp: {decvax,philabs}!mcvax!enea!log-hb!hans
Real World: TeleLOGIC AB, Box 1001, S-14901 Nynashamn,SWEDENarndt@ttds.UUCP (Arndt Jonasson) (11/17/84)
Actually, there is an editor developed by a computer club here in Stockholm, which was almost entirely written in Standard Pascal. The editor (called AMIS (Anti-MISery)) is an (almost) subset of the original EMACS, and has been quite a success. It was developed on a DEC10 for the Tops-10 system, but has since then been ported to VAX/VMS, Pr1mos, NORD-10 among others. There is no question that AMIS is slower (perhaps even much slower) than it would have been if implemented in some machinedependent language, but compared to what editors are otherwise available on the named machines, an AMIS, even a slow one, is quite acceptable.
cottrell@nbs-vms.ARPA (01/30/85)
I assume you are berating me for not converting from C to Pascal. Tell me some good things about Pascal. I happen to think it is pathetic. Let's see, first there was Pascal, then Modula-1, then Modula-2. Looks like it took Wirth three tries to get it right (if in fact he did). The claim that "It was designed as a teaching language" doesn't cut it with me. It's a toy. If I were Blaise I'd be rolling over in my grave. */
g-frank@gumby.UUCP (01/30/85)
> I assume you are berating me for not converting from C to Pascal. Tell me some > good things about Pascal. I happen to think it is pathetic. Let's see, first > there was Pascal, then Modula-1, then Modula-2. Looks like it took Wirth three > tries to get it right (if in fact he did). The claim that "It was designed as a > teaching language" doesn't cut it with me. It's a toy. If I were Blaise I'd be > rolling over in my grave. > */ My goodness. Tell me some good things about Unix. First there was single user Unix. Then multi-user. Then System III. Then Version 6. Then Level 7. Then Berkeley. Then System V. Then Berkeley 4.x. Now those silly people at Bell Labs are talking about System 8! That must mean that Unix is a bad operating system, don't you think? -- Dan Frank "good news is just life's way of keeping you off balance."
cottrell@nbs-vms.ARPA (01/30/85)
/* i got bitten by the line eater, so i'll post this again... > > /* > > > > to paraphrase lyrics from sugar magnolia (grateful dead) > > > > C's got everything delightful > > C's got everything i need > > > > */ > > The more things change, the more they are the same. Try to convince a > dedicated FORTRAN hacker that it's time to move up to a (gasp!) structured > language like C, and see what kind of answer you get . . . > > > -- > Dan Frank > > "good news is just life's way of keeping you off balance." I assume you are berating me for not converting from C to Pascal. Tell me some good things about Pascal. I happen to think it is pathetic. Let's see, first there was Pascal, then Modula-1, then Modula-2. Looks like it took Wirth three tries to get it right (if in fact he did). The claim that "It was designed as a teaching language" doesn't cut it with me. It's a toy. If I were Blaise I'd be rolling over in my grave. */
g-frank@gumby.UUCP (01/31/85)
> /* i got bitten by the line eater, so i'll post this again... > > > > /* > > > > > > to paraphrase lyrics from sugar magnolia (grateful dead) > > > > > > C's got everything delightful > > > C's got everything i need > > > > > > */ > > > > The more things change, the more they are the same. Try to convince a > > dedicated FORTRAN hacker that it's time to move up to a (gasp!) structured > > language like C, and see what kind of answer you get . . . > > > > > > -- > > Dan Frank > > > > "good news is just life's way of keeping you off balance." > > I assume you are berating me for not converting from C to Pascal. Tell me some > good things about Pascal. I happen to think it is pathetic. Let's see, first > there was Pascal, then Modula-1, then Modula-2. Looks like it took Wirth three > tries to get it right (if in fact he did). The claim that "It was designed as a > teaching language" doesn't cut it with me. It's a toy. If I were Blaise I'd be > rolling over in my grave. > > */ I thought it was a silly thing to say when I replied to it (or the incomplete version) the first time. However, now that I know what you were replying to, I have a less flip answer: I am not berating you for not switching to Pascal. Pascal is a pretty old language, too. I AM berating people (I don't know if this includes you) who keep their minds firmly closed to new and better ways to solve problems, including newer programming languages, simply because they know one way to do something and don't want to learn another. To say, "X is all I need," is to say, "I have stopped growing." It's also unbearably smug (is there a net.smug?). -- Dan Frank "good news is just life's way of keeping you off balance."
brooks@lll-crg.ARPA (Eugene D. Brooks III) (01/31/85)
> I assume you are berating me for not converting from C to Pascal. Tell me some > good things about Pascal. I happen to think it is pathetic. Let's see, first > there was Pascal, then Modula-1, then Modula-2. Looks like it took Wirth three > tries to get it right (if in fact he did). The claim that "It was designed as a > teaching language" doesn't cut it with me. It's a toy. If I were Blaise I'd be > rolling over in my grave. > */ To add to your pain take a look in this months byte therin is an article on converting C to @#$?!. Its reprehensible!
larry@cci-bdc.UUCP (Larry DeLuca) (02/01/85)
> > My goodness. Tell me some good things about Unix. First there was single > user Unix. Then multi-user. Then System III. Then Version 6. Then Level 7. > Then Berkeley. Then System V. Then Berkeley 4.x. Now those silly people > at Bell Labs are talking about System 8! > > That must mean that Unix is a bad operating system, don't you think? > > > -- > Dan Frank > > "good news is just life's way of keeping you off balance." Not exactly...Unix was conceived virtually from the start as a multi-user operating system... the first implementations were written in Assembly language on the PDP-7... a rewrite in C followed and the stuff moved up to a PDP-11... The most recent tree is like this: Version 6 | PWB Unix (Programmer's Workbench) | Version 7 (Here comes the split between ATT and Berzerkely) +------------------------------+ (Bell) (Berkeley) | | System III UNIX 32/V | | 2 and 3 BSD (2 for PDP-11, 3 for VAX) Still System III | | 4.0BSD, followed by 4.1BSD and 4.2BSD...sometime after the release of 4.0BSD 2.8BSD was released, similarly for 4.1BSD and 2.9BSD, and support for PDP-11's was dropped by Berzerkeley before the release of 4.2BSD for the VAX (which is now available for many machines). | System V, R1 for the VAX | System V, R2 for the VAX and ATT Computers (i may have some release dates confused, but i believe this to be basically correct...)... larry... -- uucp: ..mit-eddie!cybvax0!cci-bdc!larry arpa: henrik@mit-mc.ARPA This mind intentionally left blank.
Doug Gwyn (VLD/VMB) <gwyn@Brl-Vld.ARPA> (02/01/85)
I bet Bell Labs was amused to see that Berkeley is supposed to have come up with UNIX 32/V. This topic seems to be in the wrong newsgroup.
ron@brl-tgr.ARPA (Ron Natalie <ron>) (02/01/85)
While Larry's described how Berkeley got into it, PWB was not a precursor to version 7. Feel free to flame at me if I make a mistake, but I think it was something like this: Early Research Versions | Version 5 - first version to see light of day | Version 6 - where I got my start | +---------------+--------------------+ PWB UNIX (Vers 1) | Semi-supported version (Troff and early version 7 compiler with RJE features released here) | | | VERSION 7 | | UNIX/TS and | PWB UNIX (Vers 2) | (included V7 kernel features) | | | | | | | | | | Research Version 8 SYSTEM III (WOW! I could have had a V8) Finally THE Unix | SYSTEM V Well, consider this standard | SYSTEM VR2 Well, consider it a moving target
guy@rlgvax.UUCP (Guy Harris) (02/02/85)
> While Larry's described how Berkeley got into it, PWB was not a precursor > to version 7. Some features from "late Version 6" made it into PWB (such minor improvements as the Standard I/O library, a modern C compiler that understands "long" and "unsigned", ...), and into Version 7, but weren't in vanilla Version 6; this may be where the idea that PWB was a precursor to V7 came about. ("late V6" here refers to V6 plus "troff/early version 7 compiler" plus Ken Thompson's "50 changes"). > Feel free to flame at me if I make a mistake, but I think it was > something like this: Pretty much accurate (although since the BSD stuff isn't in there the 4.1BSD influence on V8 isn't shown). Guy Harris {seismo,ihnp4,allegra}!rlgvax!guy
Doug Gwyn (VLD/VMB) <gwyn@Brl-Vld.ARPA> (02/02/85)
A workable commercial standard necessarily has to evolve in a controlled fashion.