ewhac@well.UUCP (Leo 'Bols Ewhac' Schwab) (01/09/87)
[ Purina Line Eater Chow ] Ever since the Big Changeover in the USENET, I've fallen behind. I finally managed to catch up on roughly 600 articles (thank God for 2400 baud) in this newsgroup. Now I'll have to catch up on rec.humor (is it worth it?). Juggler Demo: Concatenated all the .uue files, applied the posted fix, decoded it, uploaded it (remember, VAX to Amiga is uploading now; the Amiga is more powerful :-), and ran it. Totally blew away the saleslady at the local Amiga dealership. A zillion thanks for a great demo and the information with which to make it work. For masochists out there, it took roughly an hour to upload the movie.data file over Kermit at 2400 bps. Wecker's VT-100 emulator: Let me be the Nth to say it: it's a wonderful tool which I use ALL the time. However, it does suffer from some drawbacks which, if I ever get my programming energy back, I will address on my own. One of the drawbacks that irks me the most is being demonstrated now. You see, Wecker's program doesn't auto-wrap the cursor at the end of a line. Thus, you get the cursor stuck at the end of the line flickering away, and wonder what you are missing. This very long line demonstrates that limitation, and those of you trying to live with it are probably very irritated with me by now. I hope the mailer doesn't munch it. This problem was true in v2.2, anyway, and I don't know the escape code to cure it. I haven't tried 2.4 yet (but I will). The Workbench: I never use it. Mice are for people who can't type :-). Seriously, the drawbacks voiced in the Workbench are not the fault of the Workbench. The guy who wrote the Workbench did it "The Right Way" i.e. he settled on a scheme that would provide maximum flexibility for manipulation of Workbench information (okay, icons), and for presenting those icons to the user. At least, that's the way it appears to me. The real problem with the Workbench is addressed below. The CLI (and globbing, etc): Globbing is not a viable idea under AmigaDOS since directory searches are so slow. Doing a "list p #?.c" is already painful enough. Globbing is, in my humble opinion, the responsibility of the shell (or command interpreter, or whatever they're calling it this week). One of the more irritating "features" of the CLI is my inability to specify a long filename with a simple pattern. I can't say "ext*" to run the program "extremelylongprogramname". I have to type the whole thing out. If I make a mistrake, I gots to type it all over again. Thus globbing would be nice, but you'd probably have to listen to your drives grind an awful lot more. The Real Problem: This is the real problem with the CLI and The Workbench. Set the below line in 36 point italic boldface type: >> AmigaDOS SUCKS MAJOR HOSE << While the redundancy in the filesystem is nice, it is possible to have that redundancy without scattering your directory all over the disk. And has anyone done any studies on just how much time you really save using a hashed filename structure? With large directories, it's probably noticeable, but not everyone keeps 100 files in a single directory (Fish Disk #13 doesn't count). Not to mention the fact that the filesystem doesn't support links i.e. multiple names for files, one of the things I sort of like about UNIX (there's that operating system again!). BCPL (Bogus Compiler, Preposterous Language) is what the DOS is written in. Jim Goodnow ][ spoke at the December BADGE meeting. He has done extensive studies of the DOS (in an attempt to make his software run in a nice way (you don't honestly think fexec() uses AmigaDOS functions, do you?)). After listening to him speak for about an hour and a half, I came to two conclusions. 1. Jim is an incredibly dedicated, patient, and brilliant person who has gleaned more information out of the DOS by tracing the ROM code than anyone else. Someone should pay him to write a book on the subject (Mitchell Waite, are you listening?). 2. AmigaDOS is an absolute, unequivocal, inarguable -=> PIG! <=- Everything (and I mean EVERYTHING) is done the wrong way. The library function table is non-standard. The DOS caches the pointer to DOS library functions internally, so even if you do manage to patch the library table, you still lose. BCPL grows the stack in the wrong direction. BCPL uses register D0 as an address register (gaakk!!). The fact that it works at all is a miracle beyond comprehension. No wonder all the C-A reps praise MetaComCo for porting it over in three months. It should have taken three decades! It is a sad twist of fate that left us with the DOS we have today. I spoke to -=RJ Mical=- privately nearly a year ago on this subject. He related to me that they had come up internally with a spec for the DOS they wanted the system to run. But everyone within the company was already working on other projects. So they contracted out to another firm (which -=RJ=- would not name for legal/moral reasons) to write the DOS, asking, "Please have it ready by the time CES rolls around." After exhausting roughly half that time, the company came back and said, "We can't do it. Sorry." Sorry indeed! CES was but a few months away, and if they didn't have *some* kind of DOS on the machine, Commodore's financing could be terminated. Someone from MetaComCo managed to get wind of this situation, and told Amiga they already had a DOS similar to what they were looking for. It was either this or let the system effectively die. Amiga jumped, and AmigaDOS was born. So it really isn't Commodore-Amiga's fault. And if you think about it, it isn't even MetaComCo's fault. TRIPe-OS works on systems designed to use it, and works quite well, I imagine. Now, if we could just find out who this mysterious company was... The FIX: It's quite simple (sort of), and I see no other way around it. Re-write the DOS. From scratch. As per the original spec. Those of you out there who are still listening and agree with this postulate must also agree to the following one: Commodore-Amiga is not going to re-write the DOS. They have already committed themselves to it, and are doomed to support it and be compatible with it until The End of Time. Also, re-writing the DOS is a major effort, and I don't think C-A is in a position to afford such an expenditure, especially when the end result is that ALL existing software for the machine will suddenly break. However, you people out there don't have such a limitation. You could bring 4.3 BSD UNIX up on the thing and quite probably sell it. No one would expect the old stuff to work on it since it didn't come from Commodore. You could bring up CP/M-68K, or even a real version of TRIPe-OS (but then, why would you want to?). The point is that, if you want to make the DOS what it should be, you are unfortunately going to have to do it yourselves. In this regard, C-A has provided us all with enough documentation to pull this off. Just replace the entire DOS library, and away you go. We've got the trackdisk.device (which fortunately is an Exec device, so it works The Right Way), what more do you need to write a DOS? If a bunch of you decide to do this, a word of advice (from a flaming moron): Please try to coordinate your efforts, so we don't get 13 zillion copies of DOS's on seperate Fish Disks, all different. If you can, see if you can get a copy of the original specs for the DOS. C-A may still have a copy laying around in Los Gatos. Phew. Quite a mess, now that I look at it. Oh well, try to keep the flames turned down just a bit. Thanks for listening to me rant and rave. Final Note: From someone who has been reading and appreciating the content of this forum for over a year.... > > > T H A N K Y O U < < < Your information, code, examples, help, discussions, arguments, fights, flames, reviews, warnings, and anything else I've left out have been gleefully read by me. I've been helped, enlightened, and even amused at the content of this newsgroup. To the people at Commodore, The Software Distillery, Fred Fish, Matt Dillon, Mike Meyer, Ali Ozer, and hundreds of other people whose names I am embarrassed to admit I've forgotten, and to the system administrators and the backbone who carry this group, a hearty and thoroughly sincere THANK YOU from this one voice in the wilderness. _-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_ ________ ___ Leo L. Schwab \ /___--__ The Guy in The Cape ___ ___ /\ ---##\ ihnp4!ptsfa!well!ewhac / X \_____ | __ _---)) ..or.. / /_\-- -----+==____\ // \ _ well ---\ ___ ( o---+------------------O/ \/ \ dual ----> !unicom!ewhac \ / ___ \_ (`o ) hplabs -/ ("AE-wack") ____ \___/ \_/ Recumbent Bikes: "Work FOR? I don't work FOR The _O_n_l_y Way To Fly! anybody! I'm just having fun."
mwm@eris.BERKELEY.EDU (Mike Meyer) (01/10/87)
One comment, and one comment only: How about somebody trying to pick up where TLM left off (assuming they are dead?), and doing an OS/9 port? The Amiga is the kind of hardware that OS/9 Level I was designed to run on, and it runs well on it. They've already got windowing packages and other good things for it. That would seem to make a lot more sense than doing it from scratch. <mike
wagner@utcs.UUCP (01/10/87)
In article <2146@jade.BERKELEY.EDU> mwm@eris.BERKELEY.EDU (Mike Meyer) writes: >One comment, and one comment only: > >How about somebody trying to pick up where TLM left off (assuming they >are dead?), and doing an OS/9 port? > >The Amiga is the kind of hardware that OS/9 Level I was designed to >run on, and it runs well on it. They've already got windowing packages >and other good things for it. That would seem to make a lot more sense >than doing it from scratch. > > <mike I've thought about it. Unfortunately, one needs a fair amount of money to license the sources from Micro-Ware. I might do it if I weren't going off to Europe for a year. By the time I come back, someone else will undoubtedly have done it.
carroll@unirot.UUCP (mark carroll) (01/13/87)
In article <1987Jan10.114601.29610@utcs.uucp> wagner@utcs.UUCP (Michael Wagner) writes: >> >>How about somebody trying to pick up where TLM left off >> and doing an OS/9 port for the Amiga > Unfortunately, one needs a fair amount of money >to license the sources from Micro-Ware. How much do they charge to liscence the sources? And also, any hackers in the central Jersey area whod be interested in giving me a hand with this? I think Id like to give it a try, but I just dont have the time to do it all by myself. If you can, please reply to carroll@ru-aim.arpa ( ..!rutgers!ru-aim.arpa, I think) Mark Carroll /unirot!carroll ..rutgers \carroll@ru-aim.arpa or Carroll@aim.rutgers.edu As for disclaimers, my boss doesnt even know the net exists, much less that Im here.
mende@aramis.UUCP (01/13/87)
In article <288@unirot.UUCP>, carroll@unirot.UUCP (mark carroll) writes: > [--re a port of OS/9 to the amiga--] > > How much do they charge to liscence the sources? And also, any > hackers in the central Jersey area whod be interested in giving me a > hand with this? I think Id like to give it a try, but I just dont > have the time to do it all by myself. > If you can, please reply to carroll@ru-aim.arpa > (..!rutgers!ru-aim.arpa, I think) > Mark Carroll Mark, OS/9 is a nice operation system, but I have a suggestion that you (and many other amigaites will like. I suggest porting minix to the amiga. I already have the book on order and plan to purchace the software. Think about what what minix has to offer. I includes almost full AT&T look-alike system/utilities/compilers. If anyone out there loves unix like I have come to (I a reformed VMS weniie), then they should be jumping for joy about the prospest of minix on the amiga. I will take a look at the minix system, but I am not sure that I know enough about the internals of the amiga to take on a OS calaber project. If anyone else out there wants to take it on ... GO FOR IT! It's about time that the amiga gets a real multitaksing/multiuser (Please ... No AMIGAdos flames!!) for a real multitasking system. Bob Mende ps... Mark ... the path to ru-aim via uucp is...!rutgers!aim!carroll. -- {Both Reality and this message are figments of my imagination} ARPA: mende@rutgers.rutgers.edu Snail: RPO 4888 CN 5063 UUCP: {anywhere}!rutgers!mende New Brunswick NJ 08903
cc1@locus.ucla.edu (Michael Gersten) (01/14/87)
One other thing for anyone who is re-writting AmigaDos:
Be sure that any program that currently does OpenLibrary("dos.library"),
and calls the routine at offset X to get a routine that does Y
will still be able to do this latter. I.e. Any program that does not
mess with internal data structures should work properly without re-compiling.
And please put an exec() routine in.
Views expressed here may not be those of the Computer Club, UCLA, or
anyone in their left OR right mind. And that's the name o' that tune.steve@wlbreng1.UUCP (01/15/87)
So suppose someone does revisit the issue of porting OS9 to the amiga.
Other than having a nice O/S, what would we gain? There's only three or
so antiquated programs for OS9 and a quantity of uninteresting user group
utilities and BASIC programs for the neophite.
I think it makes sense ONLY if the addition of GKS to OS9 opens up
a CAD/CAM opportunity. However, GKS is, by many, thought to be, er, ah,
rather poor in terms of throughput.
So what say ye?
Regards,
Steve Childress
(818) 706-5247 (days)
{trwrb, scgvaxd, ihnp4, voder, jplgodo} !wlbr!wlbreng1!stevecsc@watmath.UUCP (01/28/87)
In article <2325@well.UUCP> ewhac@well.UUCP (Leo 'Bols Ewhac' Schwab) writes: > It is a sad twist of fate that left us with the DOS we have today. >I spoke to -=RJ Mical=- privately nearly a year ago on this subject. He >related to me that they had come up internally with a spec for the DOS they >wanted the system to run. But everyone within the company was already >working on other projects. So they contracted out to another firm (which >-=RJ=- would not name for legal/moral reasons) to write the DOS, asking, >"Please have it ready by the time CES rolls around." (etc...) Would some version of this spec be available? I'm not up to writing OS's (yet), but I'd still like to contemplate it. It would also be interesting to compare the spec'd OS with current AmigaDOS. (Did it ask for multitasking?) Is there someone I can get a photocopy from, or is it a secret internal document? Paying $x (for small x) for the copier's time is acceptable. Of course, if it's around in electronic form, that's even better. -Colin Plumb (c/o csc@watmath) Zippy says: I would like to urinate in an OVULAR, porcelain pool --