val@netw23.uucp (Oberon Kenobi) (06/08/91)
I sent the following letter to Commodore (by U.S. mail) in February.
Commodore hasn't replied yet, but I thought that this letter would be of
general interest to the network community:
--------------------------------------------------------------------------------
Commodore Business Machines
Research and Development
1200 Wilson Drive
West Chester, PA 19380
Dear Commodore,
I've been a Commodore Amiga (1000) owner since late 1985. I now
have another Amiga 1000, and will soon purchase another one. I
like the look-and-feel of AmigaDOS V2.0 (or whatever the Amiga's
operating system is called). Unfortunately, it won't run on the
Amiga 1000's because it requires 512K of Kickstart. Couldn't
you make it so that V2.0 will run on the Amiga 1000 by loading
it into some RAM location?
The current Amigas are wonderful, but there is insufficient
incentive for me to upgrade. I await the forthcomming Amigas
with 68040s, and faster 32-bit custom chips.
Thanks for multitasking. It is the best feature of the Amiga,
and I can't live without it. There are also features that I
hope to see in future versions of AmigaDOS:
- Resource tracking: When tasks terminate, they should return
all system resources back to the operating system even
under abnormal termination (i.e.: task deletion). This
most definately includes memory.
- Memory protection: One of the (occasional) problems with
programs on the Amiga is that if they are not perfectly
behaved, they crash the system. This isn't too bad most
of the time, but it happens EVERY time I try to
demonstrate it to someone who owns another computer
system.
- Virtual memory: It would be nice, on Amigas with MMUs, to
have virtual memory possible. Special consideration must
be made for memory that was allocated with the MEMF_PUBLIC
bit. There should also be a way to allocate
non-paged/swapped (but not necessarily public) memory.
- Networking: Standard networking layers should be
distributed with the operating system. The network should
be accessed either as a library or a "device". This
networking "device" should be independant of the phyical
medium (i.e.: serial port, parallel port, ethernet, token
ring, etc.) that is used. And multiple protocols (i.e.:
TCP/IP, DECnet, AMIGAnet (?)) should be able to share the
same physical device. (A cheap ethernet interface would
be nice.)
Apple supplies networking with their Macintosh and even
includes the physical port with which to do it.
Networking support should be built into the file system
(as DECnet on VMS is), and the programs should not have to
worry about routing. The ability to cluster and/or remote
mount file systems should be included (i.e.: VAXcluster,
NFS, RFS) with the networking software.
- Graphics: Support for varying sizes of graphics screens,
aspect ratios, and color depth should be included.
Different, and multiple simultaneous, display devices
should be supported. (The NeXT and Macintosh do this.)
This should include expanding the address space of the
custom chips to 8-megabytes. (As well as faster custom
chips.)
Overall, however, I am pleased with the Amiga line of computers and
still recommend them to all my friends. Unfortunately, many of them
want to buy NeXT computers (when they can afford them).
Eagerly awaiting your reply,
Val Kartchner
--------------------------------------------------------------------------------
I will comment on some areas of the above letter before I hear about it
from the NET.
I have an Amiga 1000. I am happy with my Amiga 1000. I was going to get
a third one (as stated above), but that deal fell through. The other Amigas,
including the Amiga 3000, aren't that much of an improvement to cause me to
want to upgrade. I want a 68040 AND 32-bit custom chips.
Networking is a biggie. I've heard, in one of the Amiga magazies, that
Commodore is developing the Standard Amiga Networking Architecture (SANA). I
think that they should, instead, call it Standard Amiga Networking Environment
(SANE). It can be marketed as "The computer for the SANE mind."
Speaking of networking, would any hardware types out there be willing to
create a network adaptor for which a SANE driver could be written. It should be
a bus type network. To make it work with all Amigas and get a reasonably fast
throughput, the adaptor could be made to fit on the back of the floppy drive
chain. Any takers?
Also, considering the price of SCO Xenix comared with the price of the
Amiga with System V release 4 (a "standard" Unix), Commodore could market it as
"Buy our SVR4 Unix, and we'll throw in the machine for free." This could sell
some more systems. (NeXT could make this same claim, but the Amiga has color.)
Most of the rest of the letter is self-explainatory. Some of the rest has
been done, but I want a Commodore standard that comes with the machine.
--
____ ____ ____ |=============================================================|
\ \/ \/ / | "vi? Because I don't have a real editor." -- U*x user |
\Weber State/ |=============================================================|
\University/ |Oberon| Internet: val@csulx.weber.edu, val@net23.weber.edu |
\___/\___/ |Kenobi| Usenet: uunet!viusys!uxui!unislc!bryce!netw23!val |
Computer Science |=============================================================|chrisl@cbmvax.commodore.com (Christian Ludwig - CATS) (06/12/91)
> Networking is a biggie. I've heard, in one of the Amiga magazies, that >Commodore is developing the Standard Amiga Networking Architecture (SANA). I >think that they should, instead, call it Standard Amiga Networking Environment >(SANE). It can be marketed as "The computer for the SANE mind." SANE is a trademark of Apple Computer that stands for "Standard Apple Numerics Environment" It is a toolkit that supports IEEE math on their platforms.
manes@vger.nsu.edu ((Mark D. Manes), Norfolk State University) (06/12/91)
In article <177458@netw23.uucp>, val@netw23.uucp (Oberon Kenobi) writes: > The current Amigas are wonderful, but there is insufficient > incentive for me to upgrade. I await the forthcomming Amigas > with 68040s, and faster 32-bit custom chips. Insufficient incentive? Sounds to me like you are techno-spoiled. :-) If Commodore had shipped the A3000 with a 68040 a year ago (which was impossible) you probably would have changed that to a 68050. Just a guess however. Not flamming... really! > > > Thanks for multitasking. It is the best feature of the Amiga, > and I can't live without it. There are also features that I > hope to see in future versions of AmigaDOS: > - Resource tracking: When tasks terminate, they should return > all system resources back to the operating system even > under abnormal termination (i.e.: task deletion). This > most definately includes memory. Though I would like to see resource tracking, it would really hurt the performance of the Amiga. If it were implemented, I would like it to be able to be turned off via preferences. > - Memory protection: One of the (occasional) problems with > programs on the Amiga is that if they are not perfectly > behaved, they crash the system. This isn't too bad most > of the time, but it happens EVERY time I try to > demonstrate it to someone who owns another computer > system. Talk about abandoning the A500/A1000/A2000! To do effective memory protection would require the MMU of a 68020/68030. To be honest however, the Amiga operating system is very stable (certainly 2.0 is!) and this feature would really be the most effective on a developers machine. I like the idea that someone from Commodore had, about creating a special version of AmigaDOS that did _not_ have the 1.3 work-arounds, and had enforcer, memmung and all of that jazz tied in. I think that would improve the stability of user programs significantly. > - Virtual memory: It would be nice, on Amigas with MMUs, to > have virtual memory possible. Special consideration must > be made for memory that was allocated with the MEMF_PUBLIC > bit. There should also be a way to allocate > non-paged/swapped (but not necessarily public) memory. With the price of memory dropping like dirt what would be the real advantage to this? The only things I can think of is handling large 24 bit graphics data and perhaps being able to run numerous multi-megabyte applications. I like the fact that the Amiga operating system is 'real time'. I would want to be able to turn this feature off if it were implemented. > - Networking: Standard networking layers should be > distributed with the operating system. The network should > be accessed either as a library or a "device". This > networking "device" should be independant of the phyical > medium (i.e.: serial port, parallel port, ethernet, token > ring, etc.) that is used. And multiple protocols (i.e.: > TCP/IP, DECnet, AMIGAnet (?)) should be able to share the > same physical device. (A cheap ethernet interface would > be nice.) Why should it be distributed with the operating system? Most users would have zero use for it. I believe the operating system should provide the hooks (file protection and record locking) but not the protocols. Leave something for the third parties to make money on! Those who need it, will buy it, why bloat the Amiga operating system? > Apple supplies networking with their Macintosh and even > includes the physical port with which to do it. > Networking support should be built into the file system > (as DECnet on VMS is), and the programs should not have to > worry about routing. The ability to cluster and/or remote > mount file systems should be included (i.e.: VAXcluster, > NFS, RFS) with the networking software. As of 2.0 networking support is built into the operating system at least at the filesystem level. I think that individual networking solutions should be provided by the third parties. I don't want to be in the situation where I am stuck with one networking protocol and networking device. I don't want to pay for networking hardware I don't intend to use. > > Overall, however, I am pleased with the Amiga line of computers and > still recommend them to all my friends. Unfortunately, many of them > want to buy NeXT computers (when they can afford them). > Your pleased and yet you do not wish to invest in any of the new technology? You stated that while the current amiga line is nice, you think the A1000 is suitable for your needs, and yet you ask for essentially a workstation. Commodore has not seen a nickel from you since 1985, and that is assuming you bought your A1000 new. To me, your arguments are a bit weak. You are saying to Commodore, please keep building new boxes and let me judge them. If I find them adequent I will drop a few bucks into them. While that is how most people buy, it is not a real incentive for Commodore to react positively to your letter. "Buy all you want, we will make more" applies to Commodore. :-) I think most people miss the boat when they talk about the Amiga. The miracle of the Amiga is power without the price. (Sorry Atari). Getting the most computer for the least amount of coins is what intriques me about Commodore and the Amiga. That should _never_ be forgotten by Commodore or by the users. You compared the Amiga to the NeXT and the Mac, and of course technically that is ok. Just remember the features that you have asked for would push the Amiga into the same price range as the NeXT and the high-end Mac's. Commodore is probably asking the question: Can we really afford to play in the workstation arena? High-end systems are wonderful, I certainly love my A3000, but at what point will the consumer say "at this price, I ought to buy a Sun!". Commodore has to be real careful and not let the perspective buyer ask this question. I thought you had a good letter basically, I just think that you should re-evaluate your opinion of the A3000. It has: - 32 bit memory - 2 megabytes of chip memory (why is 8 meg necessary?) - Ability to easily upgrade to a 68040 - AmigaDOS 2.0 - Built in de-interlacer - UNIX SVR4. On and on... Commodore built your dream machine. I believe action is stronger than words. Commodore answered your letter with the A3000 and the A3000T. Now, before June 31st comes, go "Power-up!". > > Eagerly awaiting your reply, > > > > Val Kartchner > -- > ____ ____ ____ |=============================================================| > \ \/ \/ / | "vi? Because I don't have a real editor." -- U*x user | > \Weber State/ |=============================================================| > \University/ |Oberon| Internet: val@csulx.weber.edu, val@net23.weber.edu | > \___/\___/ |Kenobi| Usenet: uunet!viusys!uxui!unislc!bryce!netw23!val | > Computer Science |=============================================================| -mark= +--------+ ================================================== | \/ | Mark D. Manes "The Most lopsided deal since ..." | /\ \/ | manes@vger.nsu.edu | / | (804) 683-2532 "Make up your own mind! - AMIGA" +--------+ ================================================== "I protest Captain! I am not a merry man!" - Lt. Worf
thad@public.BTR.COM (Thaddeus P. Floryan) (06/12/91)
In article <22335@cbmvax.commodore.com> chrisl@cbmvax.commodore.com (Christian Ludwig - CATS) writes: > >> Networking is a biggie. I've heard, in one of the Amiga magazies, that >>Commodore is developing the Standard Amiga Networking Architecture (SANA). I >>think that they should, instead, call it Standard Amiga Networking Environment >>(SANE). It can be marketed as "The computer for the SANE mind." > > >SANE is a trademark of Apple Computer that stands for "Standard Apple >Numerics Environment" >[...] And "Apple" is a trademark of The Beatles. And Sun Microsystems was "asked" to cease using the trademarked "Yellow Pages" and is now using something that is a trademark of my company. Same for HP's use of "VUE" (which is also a trademark of my company). And you should see my company's lawyers jumping around and wringing their hands in glee! Gee, I've *NEVER* seen lawyers smile like that before, it's almost as diabolical as during the case "IRS vs. Margolis" ("we" won, by the way :-) Thad Floryan [ thad@btr.com (OR) {decwrl, mips, fernwood}!btr!thad ]
lord_zar@ucrmath.ucr.edu (wayne wallace) (06/12/91)
manes@vger.nsu.edu ((Mark D. Manes), Norfolk State University) writes: >Now, before June 31st comes, go "Power-up!". Ahem. "30 days hath september, april, JUNE, and november". :-) > | \/ | Mark D. Manes "The Most lopsided deal since ..." At this rate, he'll never have to powerup!...... Wayne
rknop@nntp-server.caltech.edu (Robert Andrew Knop) (06/12/91)
thad@public.BTR.COM (Thaddeus P. Floryan) writes: >And Sun Microsystems was "asked" to cease using the trademarked "Yellow Pages" >and is now using something that is a trademark of my company. Same for HP's >use of "VUE" (which is also a trademark of my company). Good grief, can anybody come up with a two, three, or four letter acronym that isn't trademarked by somebody, somewhere? I think it's time for copyright law reform; no more trademarking of abbrevaitions, just of the original titles and trademarks themselves. Tradmarking three letter abbreviations is just as bad as that guy who tried to copyright every 8x8 bit pattern. -Rob Knop rknop@tybalt.caltech.edu
peterk@cbmger.UUCP (Peter Kittel GERMANY) (06/13/91)
In article <1991Jun12.163101.6980@nntp-server.caltech.edu> rknop@nntp-server.caltech.edu (Robert Andrew Knop) writes: >thad@public.BTR.COM (Thaddeus P. Floryan) writes: > >>And Sun Microsystems was "asked" to cease using the trademarked "Yellow Pages" >>and is now using something that is a trademark of my company. Same for HP's >>use of "VUE" (which is also a trademark of my company). > >Good grief, can anybody come up with a two, three, or four letter acronym >that isn't trademarked by somebody, somewhere? I think it's time for copyright >law reform; no more trademarking of abbrevaitions, just of the original titles >and trademarks themselves. Tradmarking three letter abbreviations is just as >bad as that guy who tried to copyright every 8x8 bit pattern. Good idea. It was really bad when Philips back then forced Commodore to drop the name PET (at least here in Europw, I think). -- Best regards, Dr. Peter Kittel // E-Mail to \\ Only my personal opinions... Commodore Frankfurt, Germany \X/ {uunet|pyramid|rutgers}!cbmvax!cbmger!peterk
thad@public.BTR.COM (Thaddeus P. Floryan) (06/13/91)
In article <1991Jun12.163101.6980@nntp-server.caltech.edu> rknop@nntp-server.caltech.edu (Robert Andrew Knop) writes: >thad@public.BTR.COM (Thaddeus P. Floryan) writes: > >>And Sun Microsystems was "asked" to cease using the trademarked "Yellow Pages" >>and is now using something that is a trademark of my company. Same for HP's >>use of "VUE" (which is also a trademark of my company). > >Good grief, can anybody come up with a two, three, or four letter acronym >that isn't trademarked by somebody, somewhere? I think it's time for copyright >law reform; no more trademarking of abbrevaitions, just of the original titles >and trademarks themselves. Tradmarking three letter abbreviations is just as >bad as that guy who tried to copyright every 8x8 bit pattern. Well, if you haven't figured it out yet, the one that Sun is presently using is "NIS" which is my company's trademarked name (which is a variation of the other (longer) National Information Systems, Inc.) Personally, I think it's kinda neat that Sun is "advertising" (no pun! :-) my company so much in ALL their docs, esp. since my software will soon be on Sun platforms! But I don't involve myself in legal issues and cannot, in this forum, speak officially on behalf of my company in this regards though I can relate some interesting background material (which follows). Their (Sun's) prior use of "Yellow Pages" offended {"something" in the UK, perhaps British Post and Telegraph (whatever)} whose trademark THAT is. Regarding "VUE", that is a software program running on 25+ platforms, and HP's use of it (``HP-VUE'') is highly questionable, esp. if you see article in the last month's mag HP PROFESSIONAL. If you look at major software catalogs (I just happen to have a copy of the 1986 "AT&T COMPUTER SOFTWARE CATALOG" here on my shelf at home, on page 200 you'll find "VUE" and "NIS" (though we've since moved to San Jose CA, and "VUE" is up to version 8.* and is a LOT better on, say, a Silicon Graphics IRIS system or even on others running X and is now mostly in C)). And in response to some email re: the "IRS vs. Margolis" case and my use of ``"we" won'', it was delicious irony! The IRS was using "VUE" (of which they bought thousands of copies) to schedule personnel, court appearances, track document and people flow, etc etc and the Margolis Defense Team was using my 4GL (Accent R) to track, index, retrieve, report, etc. the tens of thousands of documents (and we, at that time several years ago, dedicated about 50% of a DEC-2060 to them for its processing power and file storage and other support software); Margolis won the case (and the IRS is *still* buying VUE (now also known as ``Accent VUE'' (tm)) and I sent a trainer to Washington DC just this past week to teach more classes to them). If you saw the front page of COMPUTERWORLD back in Dec. 1989, you'll have noted that Apple, too, uses my software for "Applicant Tracking" for which they daily read in the 600-700 *daily* incoming resumes using an optical scanner which feeds Accent R which then matches peoples' skills to job openings at Apple, etc. (or, now, probably, automatically sends out "Sorry, No Jobs Here" postcards to job-seekers (re: Apple's 10% layoffs :-)) Point being (in response to Robert Andrew Knop's "Tradmarking three letter abbreviations is just as bad as ..."): "NIS" and "VUE" (and "Accent R" and "Accent VUE") are NOT simply some casually-trademarked 3-char abbreviations and/or acronyms. And for those who like some fun, I've a number of non-trivial games written in Accent R (yeah, it's a complete programming language, too, for which I do the parsers/compilers/code-generators/etc. across platforms such as PDP-10, DEC-20, VAX, MC680*0 (and soon Sparc, MIPS, HP-PA, etc.) And for those who enjoy my pokes at Apple, the one that gave me the most pleasure was the time Apple personnel attended the "Advanced Technical Seminar" I give monthly (re: Accent R) and, as always, one of my office Amigas is front and center running VLT (for its VT100 and Tektronix emulations) while I'm demo'ing Fast Fourier Transforms (the Cooley-Tuki algorithm, which I wrote in Accent R) and explaining that audio CD-ROM sampling is deficient; the multi- tasking (on the Amiga) simply blew away the Apple people when they saw the 10+ programs running at the same time I was doing the FFT (from Accent R) displayed through VLT while running AudioTool (for its graphic-programmability of the Amiga's audio channels playing various waveforms in stereo demonstrating audio inter-modulation effects), along with PM, GfxMEM, etc. And the after-the- lecture finale was the showing of the pictures from Fred Fish Disk 196 which I "casually" threw up on the screen as an example of picture-data-base retrieval; they ALL gasped; it was a blast! :-) Thad Floryan [ thad@btr.com (OR) {decwrl, mips, fernwood}!btr!thad ]
pilgrim@daimi.aau.dk (Jakob G}rdsted) (06/14/91)
manes@vger.nsu.edu ((Mark D. Manes), Norfolk State University) writes: >Commodore built your dream machine. I believe action is stronger than >words. Commodore answered your letter with the A3000 and the A3000T. >Now, before June 31st comes, go "Power-up!". He'll have time allright. -- From the notorious Jakob Gaardsted, Computer Science Department Bed og arbejd ! University of Aarhus, Jylland (!) (Pray and work!) AMIGA! pilgrim@daimi.aau.dk | I'd rather play Moria.
kent@vf.jsc.nasa.gov (06/14/91)
In article <22335@cbmvax.commodore.com>, chrisl@cbmvax.commodore.com (Christian Ludwig - CATS) writes: > >> Networking is a biggie. I've heard, in one of the Amiga magazies, that >>Commodore is developing the Standard Amiga Networking Architecture (SANA). I >>think that they should, instead, call it Standard Amiga Networking Environment >>(SANE). It can be marketed as "The computer for the SANE mind." > > I hope not, Amiga needs to push the TCP/IP and novell cards they already produce for the Amiga. This is no time to come up with another new "standard". > > It is a toolkit that supports IEEE math on their platforms. -- Mike Kent - Lockheed Engineering and Sciences Company at NASA JSC 2400 NASA Rd One, Houston, TX 77058 (713) 483-3791 KENT@vf.jsc.nasa.gov
pilgrim@daimi.aau.dk (Jakob G}rdsted) (06/15/91)
peterk@cbmger.UUCP (Peter Kittel GERMANY) writes: >>Good grief, can anybody come up with a two, three, or four letter acronym >>that isn't trademarked by somebody, somewhere? I think it's time for copyright >>law reform; no more trademarking of abbrevaitions, just of the original titles >>and trademarks themselves. Tradmarking three letter abbreviations is just as >>bad as that guy who tried to copyright every 8x8 bit pattern. >Good idea. It was really bad when Philips back then forced Commodore to >drop the name PET (at least here in Europw, I think). I've always found those PET computers funny, as the "secret" intelligence service in Denmark is called PET... >-- >Best regards, Dr. Peter Kittel // E-Mail to \\ Only my personal opinions... >Commodore Frankfurt, Germany \X/ {uunet|pyramid|rutgers}!cbmvax!cbmger!peterk -- From the notorious Jakob Gaardsted, Computer Science Department Bed og arbejd ! University of Aarhus, Jylland (!) (Pray and work!) AMIGA! pilgrim@daimi.aau.dk | I'd rather play Moria.
limonce@pilot.njin.net (Tom Limoncelli +1 201 408 5389) (06/16/91)
In article <177458@netw23.uucp> val@netw23.uucp (Oberon Kenobi) writes: > I sent the following letter to Commodore (by U.S. mail) in February. > Commodore hasn't replied yet, but I thought that this letter would be of > general interest to the network community: I'm not going to include the entire text of the letter because it was long. Let me summarize it before I comment: > [ I haven't put a penny of profit into C-A's hands in more than 5 > years. > I'm asking for things that C-A already knows about or is re-argued > (with no new content) every 2 months on Usenet. ] Oh, oh never mind. Tom -- Tom Limoncelli tlimonce@drew.edu tlimonce@drew.bitnet 201-408-5389 Nothing is as easy as it looks. *I'm* not as easy as I look. :-)
kris@tpki.toppoint.de (Kristian Koehntopp) (06/16/91)
manes@vger.nsu.edu ((Mark D. Manes), Norfolk State University) writes: >Though I would like to see resource tracking, it would really hurt the >performance of the Amiga. If it were implemented, I would like it to >be able to be turned off via preferences. Go and see OS/9 on an Atari ST. OS/9 is an operation system with ressource tracking on a machine similar to the Amiga in performance. OS/9 is small and neat and is even fast enough for realtime applications. Resource tracking will in no way slow down system performance, if implemented properly. If you are going to allocate your memory in 16 byte chunks (as some versions of microemacs do), you will probably feel a performance impact, but any sane application will notice no difference, since it allocates memory and other system ressources in larger chunks or for longer times. Resource tracking will probably cost you some memory, but this is no longer an excuse for leaving out avital function of an operating system these days. >Talk about abandoning the A500/A1000/A2000! To do effective memory >protection would require the MMU of a 68020/68030. To be honest >however, the Amiga operating system is very stable (certainly 2.0 is!) >and this feature would really be the most effective on a developers >machine. Stability of an operating system does not help, when the applications are not stable and have no way of leaving gracefully, since there is no ressource tracking. And remember: One unstable application out of the many I run simultaneously is enough to force a reboot. When I am working with a machine, it is most likely that these applications work together on a certain set of files. If an Amiga application crashed, its file locks are not released and I cannot continue my work until I reboot. Memory protection will help in creating applications, that are reliable even on systems without this protection, since many bugs are revealed, that might have slipped through otherwise. But considering the internal operation of the Amigas OS, I guess, of all listed features, memory protection will be the hardest to implement. Even for well behaved applications it is common to forbid() and trace down some internal operating system structures and then to permit() again. A system with memory protection should offer an interface with function calls to retrieve all publicly available information about internal structures and hide the rest. >With the price of memory dropping like dirt what would be the real >advantage to this? The only things I can think of is handling large >24 bit graphics data and perhaps being able to run numerous multi-megabyte >applications. The advantage of virtual memory is: You have it immediately when you need it, even if its slow and need not run out and buy these little black silicon buggies. You can do your work (perhaps dtping a large book) in small, manageable portions on your small machine and then run one large, but slow, production run on this same machine without being forced to increase your amount of real memory. Kristian Kristian Koehntopp, Harmsstrasse 98, 2300 Kiel, +49 431 676689 "Painless installation." -- OS/2 1.3
mwm@pa.dec.com (Mike (My Watch Has Windows) Meyer) (06/17/91)
In article <4498@tpki.toppoint.de> kris@tpki.toppoint.de (Kristian Koehntopp) writes: manes@vger.nsu.edu ((Mark D. Manes), Norfolk State University) writes: >Though I would like to see resource tracking, it would really hurt the >performance of the Amiga. If it were implemented, I would like it to >be able to be turned off via preferences. Go and see OS/9 on an Atari ST. OS/9 is an operation system with ressource tracking on a machine similar to the Amiga in performance. OS/9 is small and neat and is even fast enough for realtime applications. Resource tracking isn't complete on OS/9. There are a few things that manage to get missed. Nor does OS/9 have the amount of process interconnnectedness and resource sharing that the Amiga has. >Talk about abandoning the A500/A1000/A2000! To do effective memory >protection would require the MMU of a 68020/68030. To be honest >however, the Amiga operating system is very stable (certainly 2.0 is!) >and this feature would really be the most effective on a developers >machine. Memory protection will help in creating applications, that are reliable even on systems without this protection, since many bugs are revealed, that might have slipped through otherwise. Actually, the developers tools on the Amiga already provide better facilities than protected environments. The protected environments I'm used to don't have things to scratch registers after system calls, don't have tools to verify memory allocations/frees and that you didn't write past the ends of an allocation in production code. Some of them don't even break when you read through a null pointer. But considering the internal operation of the Amigas OS, I guess, of all listed features, memory protection will be the hardest to implement. Even for well behaved applications it is common to forbid() and trace down some internal operating system structures and then to permit() again. A system with memory protection should offer an interface with function calls to retrieve all publicly available information about internal structures and hide the rest. The internal operation of the OS is a problem, but that's not the real one. Remember that all those lists are dynamically allocated, in small pieces. At least, they're small compared to the MMUs page size. You either wind up wasting large chunks of memory (both virtual and real), or rewriting the list management routines to know about the MMU page size. I don't know of a way to do this that won't fragment memory pretty badly. <mike -- It's been a hard day's night, Mike Meyer And I been working like a dog. mwm@pa.dec.com It's been a hard day's night, decwrl!mwm I should be sleeping like a log.
martin@cbmvax.commodore.com (Martin Hunt) (06/19/91)
In article <1991Jun14.081836.1@vf.jsc.nasa.gov> kent@vf.jsc.nasa.gov writes: >In article <22335@cbmvax.commodore.com>, chrisl@cbmvax.commodore.com (Christian Ludwig - CATS) writes: >> >>> Networking is a biggie. I've heard, in one of the Amiga magazies, that >>>Commodore is developing the Standard Amiga Networking Architecture (SANA). I >>>think that they should, instead, call it Standard Amiga Networking Environment >>>(SANE). It can be marketed as "The computer for the SANE mind." >> >> > >I hope not, Amiga needs to push the TCP/IP and novell cards they already >produce for the Amiga. This is no time to come up with another new >"standard". SANA2 defines a standard interface to networking device drivers. This allows multiple protocols to share networking cards. It also allows vendors to design new cards, write a SANA2 driver, then expect it to work with existing network software. It is similar (in concept) to the PC packet driver standard. TCP/IP and Novell are protocols that may run on several different media. They normally talk to some kind of ethernet or arcnet device driver, which can be proprietary or SANA2. Defining networking standards is exactly what Commodore needs to do. Networking is just too complex to expect any one company to write all the software and design all the hardware to solve everyone's connectivity needs. > >-- > >Mike Kent - Lockheed Engineering and Sciences Company at NASA JSC > 2400 NASA Rd One, Houston, TX 77058 (713) 483-3791 > KENT@vf.jsc.nasa.gov Martin Hunt martin@cbmvax.commodore.com Commodore-Amiga Networking, Siberia office
cmcmanis@stpeter.Eng.Sun.COM (Chuck McManis) (06/20/91)
In article <177458@netw23.uucp> val@netw23.uucp (Oberon Kenobi) writes: The following is a wish list that Oberon send to Commodore, I don't work for Commodore but I can guess what they would answer if they were in a mind to. (I've deleted the intro part) > ... Couldn't you make it so that V2.0 will run on the > Amiga 1000 by loading it into some RAM location? They did this, the developer versions work fine on a 1000. There are of course a couple of drawbacks to this. 1) The A1000 doesn't have anyway of protecting this memory from being written over (first 512K of a 2MB expansion board), so when your programs go astray they not only trash AmigaDOS they trash the "ROM" image as well. To reboot requires you to boot into 1.3, then load 2.0 into memory, and then reboot into 2.0. From power on that is four boot disks, 1.3 kickstart to 1.3 DOS, 2.0 kickstart to 2.0 Dos. Are you willing to pay that price everytime you hit Ctrl-Amiga-Amiga ? 2) If they stuck 2.0 in the internal 512K you wouldn't have any RAM for applications so you have to have a 2MB ram board attached to your A1000. Do You? Do you know what the percentage of A1000s is that has more than 512K of RAM? It is about 3,000 machines. If you do have an expansion RAM, is the company that made it still in business? 3) Support for A1000s costs developer resources that Commodore doesn't have. The people who took Commodore up on their upgrade offer would rather have fewer bugs than to have them spend time testing whether or not 2.0 breaks the "Spirit InBoard". > The current Amigas are wonderful, but there is insufficient > incentive for me to upgrade. 2.0 is an incentive. Don't upgrade and you can't run 2.0. Want to upgrade now? > - Resource tracking: When tasks terminate, they should return > all system resources back to the operating system even > under abnormal termination (i.e.: task deletion). This > most definately includes memory. You ask for two different things, but most people would recognize this request. Commodore has an OS that supports full resource tracking and the killing off of errant tasks. Its called UNIX System V. Believe people when they tell you that there are fundamental design decisions that went into the development of AmigaDOS that make what you ask for impossible to provide in a compatible fashion. > - Memory protection: One of the (occasional) problems with > programs on the Amiga is that if they are not perfectly > behaved, they crash the system. This isn't too bad most > of the time, but it happens EVERY time I try to > demonstrate it to someone who owns another computer > system. Sorry chap, no can do without hardware. This is another incentive to upgrade to a 68030 based system, and UNIX. You won't find your UNIX programs trashing each other. As for AmigaDOS it is a fact of life just like it is for Multifinder and Windows. > - Virtual memory: It would be nice, on Amigas with MMUs, to > have virtual memory possible. Special consideration must > be made for memory that was allocated with the MEMF_PUBLIC > bit. There should also be a way to allocate > non-paged/swapped (but not necessarily public) memory. Easy to say tough to do. People who have really understood the system looked at the problem and backed off. How about the UNIX, looking better every minute isn't it? Part of the attraction of AmigaDOS is that it is lightweight and practically real-time in its response. Adding virtual memory breaks that model for no good reason. > - Networking: Standard networking layers should be > distributed with the operating system. The network should > be accessed either as a library or a "device". This > networking "device" should be independant of the phyical > medium (i.e.: serial port, parallel port, ethernet, token > ring, etc.) that is used. And multiple protocols (i.e.: > TCP/IP, DECnet, AMIGAnet (?)) should be able to share the > same physical device. (A cheap ethernet interface would > be nice.) > Apple supplies networking with their Macintosh and even > includes the physical port with which to do it. > Networking support should be built into the file system > (as DECnet on VMS is), and the programs should not have to > worry about routing. The ability to cluster and/or remote > mount file systems should be included (i.e.: VAXcluster, > NFS, RFS) with the networking software. Again, look into that UNIX stuff. They've got everything you ask for. But you can't run Shadow of the Beast under UNIX. Everything costs in one way or another. > - Graphics: Support for varying sizes of graphics screens, > aspect ratios, and color depth should be included. > Different, and multiple simultaneous, display devices > should be supported. (The NeXT and Macintosh do this.) > This should include expanding the address space of the > custom chips to 8-megabytes. (As well as faster custom > chips.) Every seen OpenLook? How about the Xt intrinsics? Now there is a window system that trades off speed for generality. X runs on everything with a variety of colors and display architectures and guess what? Its included with UNIX. > Overall, however, I am pleased with the Amiga line of computers and > still recommend them to all my friends. Unfortunately, many of them > want to buy NeXT computers (when they can afford them). Guess what the NeXT computer runs? You guessed it, something like UNIX. :-) > Speaking of networking, would any hardware types out there be willing to > create a network adaptor for which a SANE driver could be written. It > should be a bus type network. To make it work with all Amigas and get a > reasonably fast throughput, the adaptor could be made to fit on the back > of the floppy drive chain. Any takers? You'll get better performance out of the parallel port (PARNET) then you will with the Floppy connector. And you can already get DNET sort of stuff from the software distillery. If you're going to a broadband type of net use the serial port with 422 drivers. > Most of the rest of the letter is self-explainatory. Some of the rest has > been done, but I want a Commodore standard that comes with the machine. Why should Commodore spend time making AmigaDOS a UNIX wannabe when they have a perfectly good UNIX? (and a machine that will boot either easily) Understand that what AmigaDOS is good at (real time interactive simulations) is what UNIX sucks at, but what UNIX is good at (software development and commercial application support) AmigaDOS sucks at. Attempting to make one into the other just ends up making it less good at what it does. Wait for the Amiga 4000 with the 68040 and mondoacclerated graphics (could be a long long wait) where you can boot into either system depending on what you are doing at the time. Commodore seems to be addressing all of your concerns, just possibly not in the fashion you would like. -- --Chuck McManis Sun Microsystems uucp: {anywhere}!sun!cmcmanis BIX: <none> Internet: cmcmanis@Eng.Sun.COM These opinions are my own and no one elses, but you knew that didn't you. "I tell you this parrot is bleeding deceased!"
nj@magnolia.Berkeley.EDU (Narciso Jaramillo) (06/21/91)
In article <15543@exodus.Eng.Sun.COM> cmcmanis@stpeter.Eng.Sun.COM (Chuck McManis) writes: 1) The A1000 doesn't have anyway of protecting this memory from being written over (first 512K of a 2MB expansion board), so when your programs go astray they not only trash AmigaDOS they trash the "ROM" image as well. To reboot requires you to boot into 1.3, then load 2.0 into memory, and then reboot into 2.0. From power on that is four boot disks, Actually, you could get away with three boot disks (just have a minimal 1.3 WB disk and put the kickstart image on that one); that would still require two disks on every kickstart-invalidating crash. (Since ZKICK preserves the image across ordinary reboots, you would only have to do the two-disk reboot if a program gurus really badly. I've only rarely been kicked back into 1.3.) If you have a hard disk, then it's really easy to set things up so you have a two-disk cold boot--a 1.3 Kickstart, then a minimal 2.0 boot disk that has both the kickit file and a small 2.0 startup that transfers you to the hard disk startup (this assumes a non-autobootable hard disk). 2) If they stuck 2.0 in the internal 512K you wouldn't have any RAM for applications so you have to have a 2MB ram board attached to your A1000. Do You? Do you know what the percentage of A1000s is that has more than 512K of RAM? It is about 3,000 machines. Er, that's not a percentage. Heh. Considering how few A1000s there are, this might not be such a small number. > The current Amigas are wonderful, but there is insufficient > incentive for me to upgrade. 2.0 is an incentive. Don't upgrade and you can't run 2.0. It's too early to make this statement. nj