ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar) (10/10/90)
Recently, at a visit to CBM Australia, Macquarie was given a demo of the A3000 UX. In further discussions were were talking about the utilisation of disk space on the 100M machine, and how it just was not practical to put in an AmigaOS partition as the Unix system needed around 60M with a 10M swap partition etc. It just did not leave enough room for user files. So here is a small suggestion that I submit to the net for consideration. Why not, through the use of a loadable handler, let AmigaOS share the UNIX partition. The user's perception of the UNIX FS is similar enough to the AmigaOS FS (file names etc), and I can see few problems writing an I-node handler to run under AmigaOS. You would lose file comments and possibly a couple of other rarely used features, but that presents no terrible difficulty. Remember that there will be no sharing problems, as both OS's cannot at present run simultaneously (alas!) The beauty of this system is that you now only need to allocate a 2M (or even 1M) partition to store the AmigaOS boot files, then you assign everything to the newly mounted UNIX filing system. I can see no fundamental technical difficulty with this idea, and I personally would find it very useful. Comments, anyone? -- Ian Farquhar Phone : 61 2 805-9403 Office of Computing Services Fax : 61 2 805-7433 Macquarie University NSW 2109 Also : 61 2 805-7205 Australia EMail : ifarqhar@suna.mqcc.mq.oz.au
griffith@eecs.cs.pdx.edu (Michael Griffith) (10/11/90)
ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar) writes: >So here is a small suggestion that I submit to the net for >consideration. Why not, through the use of a loadable handler, let AmigaOS >share the UNIX partition. The user's perception of the UNIX FS is similar >enough to the AmigaOS FS (file names etc), and I can see few problems >writing an I-node handler to run under AmigaOS. I can think of no reason why this cannot be done, fundamentally (although, it would involve a bit of work by someone...), yet I doubt that there is much practical reason as far as the 100 Mb hard drive, because with 70 megs already used up... Well, Unix source code and executables tend to be fairly large. Remember, they haven't stuck any applications on it at this point. So, although such a utility would be nice I don't think it is neccesary for that particular reason. As for running AmigaOS and Unix at the same time, I can't personally think of a reason why you would want to. The difficulties would be tremendous to say the least, and if I was running Unix I doubt I would ever use AmigaOS. (Note: this isn't a flame on AmigaOS, but let's face it, Unix has had more time to develop and has become more refined, although it's also too d*mn big) You would either have to make a virtual Amiga running under Unix (can you say slow?) or completely remap memory back and forth while swapping between to operating systems (slow as well, plus remember that the AmigaOS doesn't support virtual memory yet which would probably be a definite neccesity). | Michael Griffith | If I had an opinion it certainly | | griffith@eecs.ee.pdx.edu | wouldn't be the same one as | | ...!tektronix!psueea!eecs!griffith | Portland State University anyways. |
ken@cbmvax.commodore.com (Ken Farinsky - CATS) (10/11/90)
In article <606@macuni.mqcc.mq.oz> ifarqhar@sunc.mqcc.mq.oz.au writes: > >...Why not, through the use of a loadable handler, let AmigaOS >share the UNIX partition... > >I can see no fundamental technical difficulty with this idea, and I >personally would find it very useful. Comments, anyone? How are you going to handle protection issues (user/group/other) and file ownership (user name space)? -- -- Ken Farinsky - CATS - (215) 431-9421 - Commodore Business Machines uucp: ken@cbmvax.commodore.com or ...{uunet,rutgers}!cbmvax!ken bix: kfarinsky
peter@sugar.hackercorp.com (Peter da Silva) (10/12/90)
In article <298@pdxgate.UUCP> griffith@eecs.cs.pdx.edu (Michael Griffith) writes: > As for running AmigaOS and Unix at the same time, I can't personally think > of a reason why you would want to. There are a lot more software packages available for AmigaOS. > You would either have to make a virtual Amiga running under Unix (can > you say slow?) Hey, Apple can do this on the Mac... and the Mac system software is pretty screwed up. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
david@twg.com (David S. Herron) (10/13/90)
In article <6782@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: >In article <298@pdxgate.UUCP> griffith@eecs.cs.pdx.edu (Michael Griffith) writes: >> You would either have to make a virtual Amiga running under Unix (can >> you say slow?) > >Hey, Apple can do this on the Mac... and the Mac system software is pretty >screwed up. But Peter ... MacIntosh never made any pretention of being a Real Time system like AmigaDOS does. AmigaDOS programs must be able to control, exactly & precisely, the screen. But if you have to do a context switch at every system call you can pretty much kiss real-time goodbye. -- <- David Herron, an MMDF & WIN/MHS guy, <david@twg.com> <- Formerly: David Herron -- NonResident E-Mail Hack <david@ms.uky.edu> <- <- Remember: On System V it's "tar xovf", not "tar xvf"!
griffith@eecs.cs.pdx.edu (Michael Griffith) (10/13/90)
peter@sugar.hackercorp.com (Peter da Silva) writes: >In article <298@pdxgate.UUCP> griffith@eecs.cs.pdx.edu (Michael Griffith) writes: >> As for running AmigaOS and Unix at the same time, I can't personally think >> of a reason why you would want to. >There are a lot more software packages available for AmigaOS. I didn't mean that you wouldn't want to run AmigaOS at all, just not at the same time. There's no reason why you couldn't choose which to boot under. >> You would either have to make a virtual Amiga running under Unix (can >> you say slow?) >Hey, Apple can do this on the Mac... and the Mac system software is pretty >screwed up. Really? Hmmm. I would think it would be hard to run A/UX (or AU/X or whatever) and the Mac OS at the same time due to the nature of Macintosh ROM calls and whatnot (according to my 68030 reference guide it's pretty awful). Oh well. I still don't see the advantage. Now sharing files I can see the desire for. >-- >Peter da Silva. `-_-' ><peter@sugar.hackercorp.com>. | Michael Griffith | If I had an opinion it certainly | | griffith@eecs.ee.pdx.edu | wouldn't be the same one as | | ...!tektronix!psueea!eecs!griffith | Portland State University anyways. |
peter@sugar.hackercorp.com (Peter da Silva) (10/14/90)
In article <8081@gollum.twg.com> david@twg.com (David S. Herron) writes: > AmigaDOS programs must be able to control, exactly & precisely, the screen. > But if you have to do a context switch at every system call you can > pretty much kiss real-time goodbye. This is true. Standard UNIX and real-time are pretty much incompatible. But there's a hell of a lot of Amiga software that doesn't require real-time, and more that would be usable on a 3000 despite the lack of realtime. Do you really *care* if Blood Money occasionally pauses? Not OK for MIDI work, but fine for the amount of realtime most people do. Non-realtime display update for the normal UI is acceptable if it's fast enough. Look at the NeXT or Intergraph UIs. Plus, an aggressive AmigaOS compatibility box might be able to give you realtime response behind UNIX' back. Remember, you'd only need to be able to support *one* AmigaOS compatibility box... it's multitasking!... so you don't have to worry about having the AmigaOS drivers preempting the UNIX kernel. You could almost think of this as running UNIX under AmigaOS. (though I'd want to turn this off unless I was willing to let AmigaOS hang UNIX... which I might well be willing to do). -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
peter@sugar.hackercorp.com (Peter da Silva) (10/14/90)
In article <324@pdxgate.UUCP> griffith@eecs.cs.pdx.edu (Michael Griffith) writes: > I didn't mean that you wouldn't want to run AmigaOS at all, just not at the > same time. There's no reason why you couldn't choose which to boot under. The thing is, UNIX is multiuser. If I'm running UNIX I don't *want* to reboot it. Like, ever. > Really? Hmmm. I would think it would be hard to run A/UX (or AU/X or > whatever) and the Mac OS at the same time due to the nature of Macintosh > ROM calls and whatnot (according to my 68030 reference guide it's pretty > awful). "32-bit clean" Mac applications can run under A/UX. > Oh well. I still don't see the advantage. Now sharing files I can > see the desire for. If I wanted to reboot the computer to run a program I'd have bought an IBM-PC. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
peter@sugar.hackercorp.com (Peter da Silva) (10/14/90)
In article <15069@cbmvax.commodore.com> ken@cbmvax.commodore.com (Ken Farinsky - CATS) writes: > How are you going to handle protection issues (user/group/other) > and file ownership (user name space)? You have a program called "login", that you have to run to get non-world access to the UNIX file system. It goes into /etc/passwd and does all the password verification type stuff, then lets you "be" the user you selected. Sure, it's insecure as all hell. But once you've booted up under AmigaOS you've blown security away anyway. And it's a model that works well in other PC-UNIX file sharing arrangements. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
navas@cory.Berkeley.EDU (David C. Navas) (10/14/90)
In article ken@cbmvax.commodore.com (Ken Farinsky - CATS) writes: >In article <606@macuni.mqcc.mq.oz> ifarqhar@sunc.mqcc.mq.oz.au writes: >> >>...Why not, through the use of a loadable handler, let AmigaOS >>share the UNIX partition... >How are you going to handle protection issues (user/group/other) >and file ownership (user name space)? Why with that multi-user resource-based AmigaOS 4.0 you're writing ;) But seriously, there are a number of technical hurdles.... What to do with Unix executables is the next question.... What about /dev stuff? How would you go about handling that 19meg data-file on your 18meg A3000 anyway? :) Let's get Unix OUT (anyone know who to bug?), then worry about improving it. >-- >-- >Ken Farinsky - CATS - (215) 431-9421 - Commodore Business Machines >uucp: ken@cbmvax.commodore.com or ...{uunet,rutgers}!cbmvax!ken >bix: kfarinsky David Navas navas@sim.berkeley.edu "Excuse my ignorance, but I've been run over by my train of thought." -me
DXB132@psuvm.psu.edu (10/14/90)
I don't understand why people keep referring to the Amiga OS as "real time". What's real time about it? I don't see timing specifications in the Rom Kernal Manual. In theory, any task or interrupt may be held off for an infinitely long time (I guess you could call that a timing specification...:-)) -- Dan Babcock
valentin@cbmvax.commodore.com (Valentin Pepelea) (10/15/90)
In article <90287.120957DXB132@psuvm.psu.edu> DXB132@psuvm.psu.edu writes: > > I don't understand why people keep referring to the Amiga OS as "real time". > What's real time about it? I don't see timing specifications in the Rom > Kernal Manual. In theory, any task or interrupt may be held off for an > infinitely long time (I guess you could call that a timing specification... > :-)) Write a program that disables all interrupts, counts up to 10000, re-enables all interrupts, and waits for 0.1 secs. All in a loop. Now start typing on the keyboard. Neat feedback eh? Or better yet, start performing disk I/O. Most impressive, eh? So the true answer is no, an interrupt may not be held off for a long time, unless you are a masochist. There may be implementation errors which prevent interrupts from being processed when they should be, and tasks from being switched at the proper time, but that does not remove the real time characteristics of the operating system. If you look at the operating system description from the Exec manual, you will notice features found uniquely in real time operating systems. Like task exceptions, which are normally used in case a critical error occurrs to perform some recovery actions. An operating system may be referred to as real time if it guarantees that an event will be processed within a specified amount of time, in case of 'hard' constraints. In case of soft constraints, it merely implies that the usefullness of the system degrades with the amount of time that passes by. In this case, the fact that events are deterministic (they are guaranteed to be executed - eventually) allows us to call this a real time operating system. Strictly by the book. Valentin -- The Goddess of democracy? "The tyrants Name: Valentin Pepelea may destroy a statue, but they cannot Phone: (215) 431-9327 kill a god." UseNet: cbmvax!valentin@uunet.uu.net - Ancient Chinese Proverb Claimer: I not Commodore spokesman be
kent@swrinde.nde.swri.edu (Kent D. Polk) (10/15/90)
In article <90287.120957DXB132@psuvm.psu.edu> DXB132@psuvm.psu.edu writes: >I don't understand why people keep referring to the Amiga OS as "real time". >What's real time about it? I don't see timing specifications in the Rom Kernal >Manual. In theory, any task or interrupt may be held off for an infinitely >long time (I guess you could call that a timing specification...:-)) > >-- Dan Babcock It's not 'Real-Time' in the traditional sense of having a guaranteed response time. It does work nicely in many 'real-time' applications since the typical response time is usually a couple of orders of magnitude faster under load than 'Real-Time' systems which use a comparable processor platform. What does this mean? Don't use it where a guaranteed response time is required, but where you really need to respond to lots of events in a very timely fashion. Kent Polk: Southwest Research Institute Internet : kent@swrinde.nde.swri.edu UUCP : $ {cs.utexas.edu, gatech!petro, sun!texsun}!swrinde!kent
kent@swrinde.nde.swri.edu (Kent D. Polk) (10/15/90)
In article <6792@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: >In article <15069@cbmvax.commodore.com> ken@cbmvax.commodore.com (Ken Farinsky - CATS) writes: >> How are you going to handle protection issues (user/group/other) >> and file ownership (user name space)? > >You have a program called "login", that you have to run to get non-world [...] >you've blown security away anyway. And it's a model that works well in other >PC-UNIX file sharing arrangements. I agree. How would this be different than mounting an NFS partition under AmigaNFS? Actually, I think it would be great to have a login program just like this available under AmigaNFS to be used on an outgoing basis. Have it use the same passwd resources that are used when ftp'ing into the Amiga. I have been wanting to provide for a 'multi-user' environment which uses a passwd arrangement on one of the machines here since different people use it - just to set up the different environments everyone uses. Currently we just execute a script to set these things up. This does nothing for permissions, etc. but at least you get your home directory, default editor, etc. & your own set of background programs at your disposal. Would SANA be able to resolve some of this stuff? I.e. have a passwd program which sets variables internal to SANA, and then have SANA handle the file ownership issues when dealing with a 'locally' mounted Unix filesystem and an NFS mounted filesystem. Kent Polk: Southwest Research Institute Internet : kent@swrinde.nde.swri.edu UUCP : $ {cs.utexas.edu, gatech!petro, sun!texsun}!swrinde!kent
lron@easy.hiam (Dwight Hubbard) (10/16/90)
>In article <324@pdxgate.UUCP> griffith@eecs.cs.pdx.edu (Michael Griffith) writes: >peter@sugar.hackercorp.com (Peter da Silva) writes: > [junk deleted] >I didn't mean that you wouldn't want to run AmigaOS at all, just not at the >same time. There's no reason why you couldn't choose which to boot under. > Why, I wouldn't want to reboot the machine everytime I want to run PageStream. >>> You would either have to make a virtual Amiga running under Unix (can >>> you say slow?) > Yes, but it's running under Unix can you say Protected Mode and Virtual Storage? People are asking for both of these capabilites and this way we wouldn't have to wait till the 21st century. Besides, even if it is 5 Times slower running an AmigaDos Application under Amiga Unix on a 25Mhz 3000 it will still run faster than it runs on an Amiga 2000. >awful). Oh well. I still don't see the advantage. Now sharing files I can >see the desire for. Well, the Advantage I see is that it will give users some software for the O/S until the developers can catch up, and if it gives AmigaDos applications the Ability to run in an Enviroment with REAL Virtual Memory it would be a very powerful selling point as well. I do however see a real problem with security with this type of compatiblity option. Maybe make it a SU command?? -- -Dwight Hubbard, |-Kaneohe, HI -USENET: uunet.uu.net!easy!lron |-Genie: D.Hubbard1 lron@easy.hiam |-GT-Power: 029/004
perley@galaxy (Donald P Perley) (10/16/90)
In article <298@pdxgate.UUCP>, griffith@eecs (Michael Griffith) writes: >much practical reason as far as the 100 Mb hard drive, because with 70 megs >already used up... Well, Unix source code and executables tend to be fairly >large. Remember, they haven't stuck any applications on it at this point. Did anyone say we would get source code for Unix? -don perley perley@trub.crd.ge.com
ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar) (10/17/90)
In article <298@pdxgate.UUCP> griffith@eecs.cs.pdx.edu (Michael Griffith) writes: >As for running AmigaOS and Unix at the same time, I can't personally think >of a reason why you would want to. The difficulties would be tremendous to >say the least, and if I was running Unix I doubt I would ever use AmigaOS. >(Note: this isn't a flame on AmigaOS, but let's face it, Unix has had more >time to develop and has become more refined, although it's also too d*mn >big) You would either have to make a virtual Amiga running under Unix (can >you say slow?) or completely remap memory back and forth while swapping >between to operating systems (slow as well, plus remember that the AmigaOS >doesn't support virtual memory yet which would probably be a definite >neccesity). Actually, I said that AmigaOS and UNIX would *not* be running at the same time, but never mind. My point was the development of an AmigaOS handler to read the inactive UNIX partition on the disk. Anyway, let's deal with the question of running AmigaOS and UNIX simultaneously. As I see it, you would have to partition memory. AmigaOS would have to be given it's own section (including CHIP ram that is *never* swapped out for obvious performance reasons), and the exec.library would have to be rewritten so as to only grab the memory it is meant to get. Virtual memory would not be necessary on the Amiga side (though it would be nice). I can see several other problems though, for instance screen handling. It would be an interesting project to consider, and I have no doubt that several people at CBM have been considering it for some time. An amusing though (though not a terribly serious one), when we get device independent graphics maybe an AmigaOS graphics driver could write to an X-windows window as a screen, with Intuition windows inside it. I cannot see why swapping the OS's should introduce very severe penalties. This is similar to the AVM that runs on the Facom. We have both MSP and UTS (Unix 5.2 !!!) running on our 380S here. Each partition shares the hardware and resources to an extent, and the OS's are given a maximum amount of CPU time it can get under a fully loaded system (if the machine is not fully loaded, they grab what they can get). As for the comments about who would want to run AmigaOS when UNIX is available, the answer is a great many people. You are only looking at the OS (and I would say the AmigaOS has several advantages over Amiga UNIX/X, GUI speed for one). You are neglecting the applications that cam be run on each, and it is going to be an awful long time before things like Deluxe Paint and AmigaVision will appear on UNIX! -- Ian Farquhar Phone : 61 2 805-9403 Office of Computing Services Fax : 61 2 805-7433 Macquarie University NSW 2109 Also : 61 2 805-7205 Australia EMail : ifarqhar@suna.mqcc.mq.oz.au
griffith@eecs.cs.pdx.edu (Michael Griffith) (10/17/90)
perley@galaxy (Donald P Perley) writes: >In article <298@pdxgate.UUCP>, griffith@eecs (Michael Griffith) writes: >>much practical reason as far as the 100 Mb hard drive, because with 70 megs >>already used up... Well, Unix source code and executables tend to be fairly >>large. Remember, they haven't stuck any applications on it at this point. >Did anyone say we would get source code for Unix? You miss the point. The original author stated that 70 megs _were_ already used. If I stick any source code on in addition to that (NetHack, X windows, whatever...) I can probably fill the remaining 30 megs in a very short amount of time as is. Therefor, it is not likely to use a 100 meg HD for both Amiga OS and Unix (besides, you could just use partitions and have file transfer programs between the two). On the other hand, I have come to the realization that an Amiga OS emulator would be nice to prevent uneccesary (sp?) reboots. However, how would this work with a game, for example? It seems to me that either the hardware would have to be (a) shared, (b) owned exclusively, or (c) emulated on some end. It seems that each of these possible solutions would risk machine crashes or very bad speed reductions. Sissy Mac software doesn't have to worry about such things, after all. :-) I haven't found out about the new Mac machines yet, so no flames please. (I reserve the right to dump on the new Macs after I find out completely what is involved. Of course, it won't change my opinions anyways...) | Michael Griffith | If I had an opinion it certainly | | griffith@eecs.ee.pdx.edu | wouldn't be the same one as | | ...!tektronix!psueea!eecs!griffith | Portland State University anyways. |
palmermg@infonode.ingr.com (Michael G. Palmer) (10/17/90)
In article <lron.1046785@easy.hiam> lron@easy.hiam (Dwight Hubbard) writes: >>In article <324@pdxgate.UUCP> griffith@eecs.cs.pdx.edu (Michael Griffith) writes: >>peter@sugar.hackercorp.com (Peter da Silva) writes: >> >[junk deleted] >>I didn't mean that you wouldn't want to run AmigaOS at all, just not at the >>same time. There's no reason why you couldn't choose which to boot under. >> >Why, I wouldn't want to reboot the machine everytime I want to run PageStream. > >>>> You would either have to make a virtual Amiga running under Unix (can >>>> you say slow?) >> >Yes, but it's running under Unix can you say Protected Mode and Virtual >Storage? People are asking for both of these capabilites and this way we CAn you say 'several virtual pcs running on one '386 machine'? >Well, the Advantage I see is that it will give users some software for the >O/S until the developers can catch up, and if it gives AmigaDos applications >the Ability to run in an Enviroment with REAL Virtual Memory it would be >a very powerful selling point as well. UNIX has VM and protected modes, but AmigaDOS doesn't. So unix may be protected from rampant amiga applications, but the amiga apps still aren't protected from each other -- unless you run multiple copies of AmigaDOS - one for each application you want to protect from each other. I guess that you would also need multiple copies of all utilities that you also liked to have running in the Amiga background...... can you say slower still, and a memory/swap space hog! >-Dwight Hubbard, |-Kaneohe, HI >-USENET: uunet.uu.net!easy!lron |-Genie: D.Hubbard1 > lron@easy.hiam |-GT-Power: 029/004 Michael Palmer mpalmer.zaphod.b11@ingr.com uunet!ingr!b11!zaphod!mpalmer
david@starsoft.UUCP (Dave Lowrey) (10/18/90)
>In article <12792@crdgw1.crd.ge.com> perley@galaxy (Donald P Perley) writes: >In article <298@pdxgate.UUCP>, griffith@eecs (Michael Griffith) writes: > > >>much practical reason as far as the 100 Mb hard drive, because with 70 megs >>already used up... Well, Unix source code and executables tend to be fairly >>large. Remember, they haven't stuck any applications on it at this point. > >Did anyone say we would get source code for Unix? Do you have $100,000.00 for an AT&T System 5.4 source license? -- ---------------------------------------------------------------------------- These words be mine. The company doesn't care, because I am the company! :-) Dave Lowrey | david@starsoft or {uhnix1,moray}!starsoft!david Starbound Software Group | Houston, TX | "Dare to be stupid!" -- Weird Al Yankovic
lron@easy.UUCP (Dwight Hubbard) (10/18/90)
[junk deleted] >UNIX has VM and protected modes, but AmigaDOS doesn't. So unix may >be protected from rampant amiga applications, but the amiga apps still >aren't protected from each other -- unless you run multiple copies of >AmigaDOS - one for each application you want to protect from each other. >I guess that you would also need multiple copies of all utilities that you >also liked to have running in the Amiga background...... can you say slower >still, and a memory/swap space hog! > That's only true if your running AmigaDos under Unix, not AmigaDos applications. Why would you want to run AmigaDos under Unix? You only need to provide the AmigaDos applications with their normal AmigaDos function calls and libraries. The AmigaDos application is going to have no way of knowing if it is using the normal AmigaDos libraries under AmigaDos or if it is calling a series of routines that take the AmigaDos function parameters and call the necessary Unix routines to do the same thing. This would slow things down most notable when it comes to passing messages since protected mode is in all likelyhood going to require passing the entire message and not just a pointer to it. This is the same kind of thing IBM is supposed to do with the new release of OS/2 (ability to run Windows apps under OS/2 by mapping the windows functions to the OS/2 routines) The one thing about this is that the AmigaDos apps are only going to require one copy of the code to do the mapping for all the AmigaDos Apps and it will let the AmigaDos applications use the Unix capabilites as well. -- -Dwight Hubbard, |-Kaneohe, HI -USENET: uunet.uu.net!easy!lron |-GT-Power: 029/004
peter@sugar.hackercorp.com (Peter da Silva) (10/18/90)
In article <1990Oct17.141554.6314@infonode.ingr.com> palmermg@infonode.ingr.com (Michael G. Palmer) writes: > UNIX has VM and protected modes, but AmigaDOS doesn't. So unix may > be protected from rampant amiga applications, but the amiga apps still > aren't protected from each other -- unless you run multiple copies of > AmigaDOS - one for each application you want to protect from each other. So? At worst it's no worse than we have now as far as protection for Amiga programs goes. It'd be very useful under development: can you imagine running ADB on the core image left by AmigaOS after a guru? Or, "Oh, this virtual Amiga has gurued: write to location 0. Damn, a null pointer...". -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
peter@sugar.hackercorp.com (Peter da Silva) (10/19/90)
In article <lron.10469034@easy.UUCP> lron@easy.UUCP (Dwight Hubbard) writes: > That's only true if your running AmigaDos under Unix, not AmigaDos > applications. Why would you want to run AmigaDos under Unix? To provide a compatibility box for Amiga applications. > This would slow things down most notable when it comes to passing messages > since protected mode is in all likelyhood going to require passing the entire > message and not just a pointer to it. Since (a) most anything you do other than actually drawing images involves passing messages, and (b) you don't know the length of a message for copying purposes, this will realisticly mean you'll have to pass messages within a single address space using Amiga scheduling. This means you'll need to run most of AmigaOS in a UNIX process, restricting UNIX context switches to *real* I/O. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
lron@easy.UUCP (Dwight Hubbard) (10/19/90)
[stuff deleted] >Since (a) most anything you do other than actually drawing images involves >passing messages, and (b) you don't know the length of a message for copying >purposes, this will realisticly mean you'll have to pass messages within a >single address space using Amiga scheduling. > I realize that any program that sets any of the IDCMP flags to report the mouse position (MOUSEMOVE for example) will really run slow. As for not being able to find out the message length, isn't there a field - mn_Length in the message structure that is supposed to define the lenght of the message structure and it's data? -- -Dwight Hubbard, |-Kaneohe, HI -USENET: uunet.uu.net!easy!lron |-GT-Power: 029/004
joseph@valnet.UUCP (Joseph P. Hillenburg) (10/20/90)
Wouldn't it be easier to run UNIX under AmigaDOS instead of the other way around? -Joseph Hillenburg UUCP: ...iuvax!valnet!joseph ARPA: valnet!joseph@iuvax.cs.indiana.edu INET: joseph@valnet.UUCP
peter@sugar.hackercorp.com (Peter da Silva) (10/21/90)
In article <lron.10469225@easy.UUCP> lron@easy.UUCP (Dwight Hubbard) writes: > As for not being able to find out the message length, isn't there a field - > mn_Length in the message structure that is supposed to define the lenght of > the message structure and it's data? Yes, but it's not used consistently. More importantly, what about messages that contain pointers? -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
peter@sugar.hackercorp.com (Peter da Silva) (10/22/90)
In article <oN0BR5w163w@valnet> joseph@valnet.UUCP (Joseph P. Hillenburg) writes: > Wouldn't it be easier to run UNIX under AmigaDOS instead of the other way > around? Yes. Next question? Oh, why aren't they doing that? Well, that's a damn good question. I wonder how hard it would be to convert AmigaMINIX to a hosted implementation... -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar) (10/22/90)
In article <15069@cbmvax.commodore.com> ken@cbmvax.commodore.com (Ken Farinsky - CATS) writes: >In article <606@macuni.mqcc.mq.oz> ifarqhar@sunc.mqcc.mq.oz.au writes: >>...Why not, through the use of a loadable handler, let AmigaOS >>share the UNIX partition... ... >How are you going to handle protection issues (user/group/other) >and file ownership (user name space)? Well, two points to make: Firstly, as lots of other people have said, this is not much different from NFS. A handler which could access the shadow passwd and group files within the UNIX partition could determine access rights (presuming that the user was able to provide valid username and passwords), and UNIX file security is fairly basic. All you need is the login UID and GID, which are quite easy to determine. Secondly, few people have pointed out the basic security flaw in Amiga UNIX: anybody with a fairly simple program can access *anything* in the UNIX partition as long as this program can read sectors from the disk under AmigaOS. The scsi.device (or whatever is being provided) should be able to do this nicely, so without much work a file transfer program could be written that would lay the UNIX filing system wide open. Maybe I've missed something here or made an invalid assumption, but this seems a worry to me! -- Ian Farquhar Phone : 61 2 805-9404 Office of Computing Services Fax : 61 2 805-7433 Macquarie University NSW 2109 Also : 61 2 805-7205 Australia EMail : ifarqhar@suna.ocs.mq.oz.au
a665@mindlink.UUCP (Anthon Pang) (10/22/90)
gilgalad@caen.engin.umich.edu writes: > You guys are forgetting something aren't you? While, it is easy to > go and reboot an Amiga with the three fingered salute, you just don't > got and reset a UNIX box at a whim. /etc/shutdown In a computer lab (which shall go unnamed :) students regularly subjected the Apollo's to such abuse--diskless or server...didn't matter--I've seen students hit the "reset" switch (SW1), turn it off & on (really fast), and "/etc/xdmc ex"...[shrug]...I've also seen a row of nodes (hooked by token ring) go "pop"...[sigh]...Of course, this wouldn't happen if the systems weren't swiss cheese. Here's hoping that 3000UX users _never_ have to nerve pinch their Amiga...
abrown@hpcvca.CV.HP.COM (Allen Brown) (10/23/90)
>> Wouldn't it be easier to run UNIX under AmigaDOS instead of the other way >> around? > Yes. Next question? > Oh, why aren't they doing that? > -- > Peter da Silva. `-_-' My guess: memory protection. If Unix was a process under AmigaDOS wouldn't any AmigaDOS process be able to write into any Unix memory? If so, that would be totally unacceptable. -- Allen Brown abrown@cv.hp.com or abrown%hpcvca@hplabs.hp.com or hplabs!hpcvca!abrown or "Hey you!" Not representing my employer. "Oy yam wot oy yam, an at's all wot oy yam." --- Popeye
gilgalad@caen.engin.umich.edu (Ralph Seguin) (10/23/90)
In article <1410053@hpcvca.CV.HP.COM> abrown@hpcvca.CV.HP.COM (Allen Brown) writes: >>> Wouldn't it be easier to run UNIX under AmigaDOS instead of the other way >>> around? > >> Yes. Next question? > >> Oh, why aren't they doing that? >> -- >> Peter da Silva. `-_-' > >My guess: memory protection. If Unix was a process under AmigaDOS You guys are forgetting something aren't you? While, it is easy to go and reboot an Amiga with the three fingered salute, you just don't got and reset a UNIX box at a whim. /etc/shutdown >-- > Allen Brown abrown@cv.hp.com or abrown%hpcvca@hplabs.hp.com > or hplabs!hpcvca!abrown or "Hey you!" > Not representing my employer. > "Oy yam wot oy yam, an at's all wot oy yam." --- Popeye gilgalad@dip.eecs.umich.edu gilgalad@caen.engin.umich.edu Ralph Seguin | "You mean THE Zaphod Beeblebrox?" 536 South Forest | Apartment 915 | "No. Haven't you heard, I come in six packs!" Ann Arbor, MI 48104 | (313) 662-4805
peter@sugar.hackercorp.com (Peter da Silva) (10/23/90)
In article <1990Oct23.060629.6537@engin.umich.edu> gilgalad@caen.engin.umich.edu (Ralph Seguin) writes: > You guys are forgetting something aren't you? While, it is easy to > go and reboot an Amiga with the three fingered salute, you just don't > got and reset a UNIX box at a whim. /etc/shutdown You BSD heretic. /etc/init 0! More to the point, I'm not talking about a full UNIX in a box under AmigaOS. I'm talking about a UNIX emulation in a hosted environment. UNIX tasks would be AmigaOS tasks with extra information attached. The UNIX file system would be mapped onto the Amiga file system (using the comment field for that extra UNIX file-system info). A unix.library (or posix.library). -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
fnf@riscokid.UUCP (Fred Fish) (10/24/90)
In article <643@macuni.mqcc.mq.oz> ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar) writes: >Secondly, few people have pointed out the basic security flaw in Amiga >UNIX: anybody with a fairly simple program can access *anything* in the >UNIX partition as long as this program can read sectors from the disk >under AmigaOS. This of course has nothing to do with either security flaws in Unix or the Amiga. The statement applies equally well to any machine and any operating system. If you have physical access to the machine and sufficient technical expertise to get to the data, there is NO security. Some setups may just reduce the amount of expertise needed. -Fred
louie@sayshell.umd.edu (Louis A. Mamakos) (10/24/90)
In article <6875@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: >> got and reset a UNIX box at a whim. /etc/shutdown >You BSD heretic. /etc/init 0! C'mon, what we all used in V7 was sync; sync; sync <click!-power off>
jap@convex.cl.msu.edu (Joe Porkka) (10/25/90)
Now to reduce the expertise needed to nil, somebody has got to write a UNIX filesystem for AmigaDOS. Then, "mount unix:" and away we go! This would make xfers between the two OS's easier too. It would also make for a great way to compare filesystem speeds of UNIX vs AmigaDOS-FFS. Any volunteers?
dale@boing.UUCP (Dale Luck) (10/25/90)
In article <643@macuni.mqcc.mq.oz> ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar) writes: > >Secondly, few people have pointed out the basic security flaw in Amiga >UNIX: anybody with a fairly simple program can access *anything* in the >UNIX partition as long as this program can read sectors from the disk >under AmigaOS. The scsi.device (or whatever is being provided) should >be able to do this nicely, so without much work a file transfer program >could be written that would lay the UNIX filing system wide open. >Maybe I've missed something here or made an invalid assumption, but this >seems a worry to me! Don't know why this worries you and security on suns does not worry you? It has the exact same security flaw. Anyone that can write programs to access the /dev/sd?? can do this too so what is the difference? There is no difference between the two machines in this regard. As long as you can prevent the user from rebooting their machine from Amiga-Unix to Amiga-Dos you will not have the security problem. Conversely as long as you can keep a user from hitting L1-A on his sun workstation and rebooting -s you also will not have this security problem. > -- Dale Luck GfxBase/Boing, Inc. {uunet!cbmvax|pyramid}!amiga!boing!dale
dale@boing.UUCP (Dale Luck) (10/25/90)
In article <6875@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: > >More to the point, I'm not talking about a full UNIX in a box under AmigaOS. >I'm talking about a UNIX emulation in a hosted environment. UNIX tasks would >be AmigaOS tasks with extra information attached. The UNIX file system would >be mapped onto the Amiga file system (using the comment field for that extra >UNIX file-system info). A unix.library (or posix.library). This is essentially how the X Window environment is able to run under AmigaDOS. AmigaDOS is sufficiently similar to Unix to get most generic X programs to compile and run under AmigaDOS with absolutely no change to the source code. All you need is a good ANSII C compiler and C support libaries. We add additional routines to the standard libraries as blink complains about unresolved function calls. Most of the difficulties are found in spawning new tasks. (Using fork instead of system calls) or assumptions in file hierarchy. T: versus /tmp/ Sophisticated use of signal() also sometimes is a problem as well as the pipe() or popen() call. I think a lot of the user interface problems regarding unix vs AmigaDOS would be fixed by a properly supported unix like shell from Commodore-Amiga. If unix wildcard expansion were done by the shell, processes started in background via '&', all processes were run as separate tasks instead of gloming onto the current cli process, redirect should work anywhere on the command line, redirect and append '>>', a real '|' for pipes, and a command separator for multiple commands on aline. Maybe this is just a wish list, maybe it already exists in the PD. ;-) >-- >Peter da Silva. `-_-' ><peter@sugar.hackercorp.com>. -- Dale Luck GfxBase/Boing, Inc. {uunet!cbmvax|pyramid}!amiga!boing!dale
Chuck.Phillips@FtCollins.NCR.COM (Chuck.Phillips) (10/28/90)
>>>>> On 25 Oct 90 14:49:13 GMT, dale@boing.UUCP (Dale Luck) said: Dale> In article <643@macuni.mqcc.mq.oz> ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar) writes: > >Secondly, few people have pointed out the basic security flaw in Amiga >UNIX: anybody with a fairly simple program can access *anything* in the >UNIX partition as long as this program can read sectors from the disk >under AmigaOS. ... Dale> Don't know why this worries you and security on suns does not worry Dale> you? It has the exact same security flaw. With few exceptions, computers from PCs to multi-million dollar mainframes can be compromised by a knowlegable user with enough access to boot the system from their own media. Only in the last few years have systems requiring a password to boot become commonly available and then usually as an added cost option. The problem is you have to be able to initially load the operating system somehow. If you are able to do this, you can load something else instead. Dale> Anyone that can write programs to access the /dev/sd?? can do this Dale> too so what is the difference? I can write programs to access /dev/sd??, but unless I'm _already_ root, they won't buy me much. Most UNIX systems provide limited access to disk devices by default. I'd bet yours are only writable by root and have limited read access unless you've deliberately changed permissions while logged in as root. But in any case, the boot hole remains. Dale> There is no difference between the two machines in this regard. There _is_ a difference, but unless you paid the extra $$ for secure SunOS or are using a password protected boot prom, the difference isn't _huge_. The problem with AmigaOS and most single user OS's, is that all users (since there is no concept of multiple users) have the equivalent of UNIX's root permissions, making compromise that much easier. Dale> Conversely as long as you can keep a user from hitting L1-A on his Dale> sun workstation and rebooting -s you also will not have this security Dale> problem. L1-A rebooting can easily be turned off, though it's often not done in practice for a good reason: What happens if the boot disk gets corrupted and you _need_ to be able to do a special reboot to recover? The best protection is to keep hostile users away from your machine. Most of the alternatives will cost you some money. Perhaps boot EPROMS should be offered as an option for the security concious, perhaps with an encrypted serial number. ;^) "It's like deja vu all over again." - Yogi Berra #include <std/disclaimer.h> -- Chuck Phillips MS440 NCR Microelectronics chuck.phillips%ftcollins.ncr.com 2001 Danfield Ct. Ft. Collins, CO. 80525 ...uunet!ncrlnk!ncr-mpd!bach!chuckp
ferry@chorus.fr (Ferry de Jong) (10/28/90)
From article <6875@sugar.hackercorp.com>, by peter@sugar.hackercorp.com (Peter da Silva): > I'm talking about a UNIX emulation in a hosted environment. UNIX tasks would > be AmigaOS tasks with extra information attached. The UNIX file system would > be mapped onto the Amiga file system (using the comment field for that extra ^^^^^^^^^^^^^ And what about security, bring the system up under AmigaDOS and change the comment. Isn't the security needed (It's _just_ an Amiga?) or am I overlooking something? > UNIX file-system info). A unix.library (or posix.library). > -- > Peter da Silva. `-_-' > <peter@sugar.hackercorp.com>. ferry
peter@sugar.hackercorp.com (Peter da Silva) (10/29/90)
In article <6653@chorus.fr> ferry@chorus.fr (Ferry de Jong) writes: > From article <6875@sugar.hackercorp.com>, by peter@sugar.hackercorp.com (Peter da Silva): > > I'm talking about a UNIX emulation in a hosted environment. UNIX tasks would > > be AmigaOS tasks with extra information attached. The UNIX file system would > > be mapped onto the Amiga file system (using the comment field for that extra > ^^^^^^^^^^^^^ > And what about security, bring the system up under AmigaDOS and change the > comment. What *about* security? If you have direct access to the hardware there is no such thing. It's no different than booting your UNIX System V/386 box under DOS and running Norton Utilities... (not to mention that the system isn't brought up "under UNIX" or "under AmigaOS"... what I'm talking about is running UNIX pretty much in a console.device window). -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
zerkle@iris.ucdavis.edu (Dan Zerkle) (10/29/90)
In article <643@macuni.mqcc.mq.oz> ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar) writes: > >Secondly, few people have pointed out the basic security flaw in Amiga >UNIX: anybody with a fairly simple program can access *anything* in the >UNIX partition as long as this program can read sectors from the disk >under AmigaOS. Why is everybody having such a fit about this? Assuming you want any sort of security, you need a multi-user system -- i.e. Unix not AmigaDos. Fine. Put a Unix partition on the disk and not an AmigaDos partition. If you are really worried, you can take out the internal floppy, too. Seems simple enough to me. Dan Zerkle zerkle@iris.ucdavis.edu (916) 754-0240 Amiga... Because life is too short for boring computers.
msawyer@hokulea.hig.hawaii.edu (Michael Sawyer (REU)) (10/30/90)
In article <6653@chorus.fr> ferry@chorus.fr (Ferry de Jong) writes: >From article <6875@sugar.hackercorp.com>, by peter@sugar.hackercorp.com (Peter da Silva): [...] >> be mapped onto the Amiga file system (using the comment field for that extra > ^^^^^^^^^^^^^ >And what about security, bring the system up under AmigaDOS and change the comment. >Isn't the security needed (It's _just_ an Amiga?) or am I overlooking something? > Well, you can always bring Unix up in single user mode and change the files that way on just about ANY Unix box. That is really no more secure than using the comment field. Many systems have a novram inside allowing the operator to set the system up to ask for a password on single user boots. Maybe C= could provide us with a ROM which won't come up in AmigaDos and we could modify the Unix boot track to ask for password on single-user... Alas, a number of times I have had to halt the system and take it to single user because I didn't know the root password (Once I forgot it, blush; once they forgot to tell me) and have never seen a system with the secure console on. --- return mail to: msawyer@io.soest.hawaii.edu Michael Sawyer, Univ of Hawaii Physical Oceanography (They don't even know I am using rn, so I sure don't speak for UH!)
joseph@valnet.UUCP (Joseph P. Hillenburg) (10/30/90)
ferry@chorus.fr (Ferry de Jong) writes: > From article <6875@sugar.hackercorp.com>, by peter@sugar.hackercorp.com (Pete > > I'm talking about a UNIX emulation in a hosted environment. UNIX tasks woul > > be AmigaOS tasks with extra information attached. The UNIX file system woul > > be mapped onto the Amiga file system (using the comment field for that extr > ^^^^^^^^^^^^^ > And what about security, bring the system up under AmigaDOS and change the co > Isn't the security needed (It's _just_ an Amiga?) or am I overlooking somethi > > > UNIX file-system info). A unix.library (or posix.library). > > -- > > Peter da Silva. `-_-' > > <peter@sugar.hackercorp.com>. > > ferry Well, if their going to be using these machines for UNIX, it's likely that they will be running UNIX full time, and they could eliminate the AmigaDOS partition totally. -Joseph Hillenburg UUCP: ...iuvax!valnet!joseph ARPA: valnet!joseph@iuvax.cs.indiana.edu INET: joseph@valnet.UUCP
dave@unislc.uucp (Dave Martin) (10/30/90)
From article <6852@sugar.hackercorp.com>, by peter@sugar.hackercorp.com (Peter da Silva): > In article <lron.10469225@easy.UUCP> lron@easy.UUCP (Dwight Hubbard) writes: >> As for not being able to find out the message length, isn't there a field - >> mn_Length in the message structure that is supposed to define the lenght of >> the message structure and it's data? > > Yes, but it's not used consistently. More importantly, what about messages > that contain pointers? > -- > Peter da Silva. `-_-' > <peter@sugar.hackercorp.com>. Well, if this memory both in the message, and being pointed to by the message was declared MF_PUBLIC as it should have been, then there wouldn't be a problem. MF_PUBLIC memory could all be allocated out of a common pool, that all amiga tasks see. This is also a good reason to go ahead and implement VM under AmigaDOS; MF_PUBLIC has always been specified as the way to get shared memory in the RKMs. P.S. Will someone (anyone) please send me mail if they see this message. I am not convinced that our news reader knows how to post properly. Thanks in Advance (TiA) -- VAX Headroom Speaking for myself only... blah blah blahblah blah... Internet: DMARTIN@CC.WEBER.EDU dave@saltlcy-unisys.army.mil uucp: dave@unislc or use the From: line. Now was that civilized? No, clearly not. Fun, but in no sense civilized.
dave@unislc.uucp (Dave Martin) (10/31/90)
From article <643@macuni.mqcc.mq.oz>, by ifarqhar@sunc.mqcc.mq.oz.au (Ian Farquhar): > UNIX: anybody with a fairly simple program can access *anything* in the > UNIX partition as long as this program can read sectors from the disk > under AmigaOS. The scsi.device (or whatever is being provided) should > be able to do this nicely, so without much work a file transfer program > could be written that would lay the UNIX filing system wide open. > Maybe I've missed something here or made an invalid assumption, but this > seems a worry to me! Yes, but this requires access to the console (meaning the hardware and "main" terminal) directly; many systems allow you to get away with murder if you are at the console and know what you are doing. Of course, being a workstation environment people are going to have this access so lock up the floppies and make it not possible for just anyone to boot into amigados. This presumes that you are more interested in security on the unix side than in allowing users to use AmigaDOS, and you think one of them might try accessing the unix partition. Most mini and mainframe systems tend to keep the console locked up, or rely on the users not having the knowledge to "take over". -- VAX Headroom Speaking for myself only... blah blah blahblah blah... Internet: DMARTIN@CC.WEBER.EDU dave@saltlcy-unisys.army.mil uucp: dave@unislc.uucp or use the Path: line. Now was that civilized? No, clearly not. Fun, but in no sense civilized.
peter@sugar.hackercorp.com (Peter da Silva) (10/31/90)
In article <ceLTR2w163w@valnet> joseph@valnet.UUCP (Joseph P. Hillenburg) writes: > Well, if their going to be using these machines for UNIX, it's likely > that they will be running UNIX full time, and they could eliminate the > AmigaDOS partition totally. AAAAARRRRRRGGGGGGHHHHHH!!!!!!!! There, that's better. Is there an echo in here? If you're not going to run AmigaOS, why get an Amiga? (a good question... I wonder whether Amiga UNIX isn't the Stealth Bomber of the Commodore world: a product without a mission) Anyway, *assuming* they're going to run AmigaOS *and* UNIX, wouldn't it be nice to use the UNIX file system from AmigaOS? -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
s900657@otto.bf.rmit.oz.au (Felicity Jones [Ice]) (11/01/90)
test please ignore -- Felicity Jones Department of BIS The opinion/s expressed [Ice] RMIT/VUT above are my own, so s900657@otto.bf.rmit.oz.au Melbourne, Australia don't blame them!
david@twg.com (David S. Herron) (11/06/90)
In article <6944@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: >In article <ceLTR2w163w@valnet> joseph@valnet.UUCP (Joseph P. Hillenburg) writes: >> Well, if their going to be using these machines for UNIX, it's likely >> that they will be running UNIX full time, and they could eliminate the >> AmigaDOS partition totally. > >AAAAARRRRRRGGGGGGHHHHHH!!!!!!!! > >There, that's better. Is there an echo in here? > >If you're not going to run AmigaOS, why get an Amiga? Y'know.. we've got all these '386 clones scattered around the company on which the *ONLY* OS is System V .. I'm sure a die hard PC-DOS fanatic would have the same reaction. Yet, in general, people don't bat an eyelash to the idea of running Unix on a PC. (I personally find it less than satisfying since the hardware/software fit just isn't "quite right".. sigh) As for Unix on the A3000 I'll (probably) be buying Real Soon Now.. I'm kinda up in the air about it. I wanna have a Unix machine at home to work with & I can see that an Amiga would be a really really really good machine to run Unix on (given the fast bus & accelerated graphics). Better than just about anything else that I could afford. Yes, AmigaDOS multi-tasks and I should be happy with that. But having Unix on my home system is vewwy vewwy important -- and I've grown tired of the administrative hassle of having two systems at home to take care of. Soooo... -- <- David Herron, an MMDF & WIN/MHS guy, <david@twg.com> <- Formerly: David Herron -- NonResident E-Mail Hack <david@ms.uky.edu> <- <- Use the force Wes!
peter@sugar.hackercorp.com (Peter da Silva) (11/06/90)
In article <8222@gollum.twg.com> david@twg.com (David S. Herron) writes: > In article <6944@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: > >If you're not going to run AmigaOS, why get an Amiga? > Y'know.. we've got all these '386 clones scattered around the company > on which the *ONLY* OS is System V .. Which is *why* it doesn't make sense to run UNIX on am Amiga... because those 386 clones are out there, do all the same things, and are *CHEAPER*. I expect Amiga with UNIX to cost about as much as a NeXT, and the NeXT has a much better UNIX (no, not because it's BSD. BSD sucks. It's because it's Mach). -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
rjtatz@hpuxa.ircc.ohio-state.edu (Robert J. Tatz) (11/07/90)
Assuming that the A3000 gets a working bridgeboard or that the
A2X00 gets UNIX, would it be feasible to run MS-DOS applications
on the bridgeboard under UNIX? What about SOFT-PC running as
a UNIX application?
--
Bob rjtatz@hpuxa.ircc.ohio-state.edu
-> AMIGA (n.) - The BETAMAX of personal computers.david@twg.com (David S. Herron) (11/15/90)
In article <6998@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: >In article <8222@gollum.twg.com> david@twg.com (David S. Herron) writes: >> In article <6944@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: >> >If you're not going to run AmigaOS, why get an Amiga? >> Y'know.. we've got all these '386 clones scattered around the company >> on which the *ONLY* OS is System V .. >Which is *why* it doesn't make sense to run UNIX on am Amiga... because those >386 clones are out there, do all the same things, and are *CHEAPER*. I expect >Amiga with UNIX to cost about as much as a NeXT, and the NeXT has a much better >UNIX (no, not because it's BSD. BSD sucks. It's because it's Mach). I expect that Unix running on Amiga would give me a much faster/nicer environment than a 386 clone or a NeXT (tho' comparing to NeXT is much closer in those terms). The NeXT dies in my book because I'm interested in working with X, not NeXTStep. Yes I can get X for it, but that's not a Standard Feature of the system and, most importantly, the Interface Builder doesn't generate X code. So to compare an Amiga with Unix to a '386 -- The A3000 has a much faster bus (20 MByte/sec vs. ~1-3 MByte/sec; unless I get a Tyan machine but that runs me $10,000). The Amiga also has SCSI-II coming in at somewhere around 1MB/sec - 5 MB/sec (depending on drive); what the speed of ESDI interfaces? I was depressed a bit when cbm.com guys said recently that the X server on Unix doesn't use the custom chips. Another consideration is hardware reliability. These 386 clones around here simply haven't impressed me with their Solidness and Reliability... -- <- David Herron, an MMDF & WIN/MHS guy, <david@twg.com> <- Formerly: David Herron -- NonResident E-Mail Hack <david@ms.uky.edu> <- <- Use the force Wes!
peter@sugar.hackercorp.com (Peter da Silva) (11/16/90)
In article <8284@gollum.twg.com> david@twg.com (David S. Herron) writes: > I expect that Unix running on Amiga would give me a much faster/nicer > environment than a 386 clone or a NeXT (tho' comparing to NeXT is much > closer in those terms). Hmmm... I suspect it'll be faster than a run-of-the-mill 386, but not a comparably priced machine (with likely a 486 and EISA or MCA bus). And I can't see how it could be afster than the NeXT, given the NeXT has a faster processor and a better base operating system (Mach vs. System V). And remember that extra performance costs extra money, in either case. If you're looking for an entry-level UNIX workstation a clone is the clear price leader. As for the environment... have you *used* a NeXT? It's to other workstations what the Amiga is to other PCs. It's really got a nice integrated environment with graphics and sound, and it's cheaper than equivalent non-NeXT machines. For the clones, I can't see how the software can be significantly different. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
daveh@cbmvax.commodore.com (Dave Haynie) (11/17/90)
In article <8284@gollum.twg.com> david@twg.com (David S. Herron) writes: >So to compare an Amiga with Unix to a '386 -- The A3000 has a much faster >bus (20 MByte/sec vs. ~1-3 MByte/sec; unless I get a Tyan machine but that >runs me $10,000). Some of the '386s these days are MCA or EISA based, either of which is in the same ballpark as the A3000 expansion bus. You do, in general, pay for it in those machines too. >The Amiga also has SCSI-II coming in at somewhere around 1MB/sec - 5 MB/sec >(depending on drive); what the speed of ESDI interfaces? That's actually SCSI-I running at those speeds. The SCSI-I limit is around 5MB/sec. Most drives don't go that fast, and for those that do, they don't go fast for long. The extra transfer speed of SCSI can really win for very high end drives, and for multiple drive systems. EDSI goes at about 1.5 MB/s, which is about the speed of the standard asynchronous SCSI transfer. What's really going to make a difference under UNIX is the multitasking performance of your hard disk, which most PClones don't address. Since the Amiga OS itself is multitasking, we tend to think in those terms from the ground up. The question you ask for determining multitasking hard disk performance is "how much CPU time is available when the hard disk controller is running full speed". For the A3000 and a 1.5 MB/s hard disk going full speed, you get about 96% of the CPU time for tasks that aren't disk bound. >Another consideration is hardware reliability. These 386 clones around >here simply haven't impressed me with their Solidness and Reliability... I'm sure the quality of the more established Clone vendors is up there with that of the Amiga. I'm equally sure that a good number of the cheaper ones aren't things I would feel safe investing in. Its apparently common practice at the bottom end to violate timing parameters and other specs in the design to eek out a bit more performance performance, at least in a cool room, for the same price. As a computer designer, I get real nervous when I see timing specs violated -- that's just plain begging for trouble. ><- David Herron, an MMDF & WIN/MHS guy, <david@twg.com> -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Standing on the shoulders of giants leaves me cold -REM
joseph@valnet.UUCP (Joseph P. Hillenburg) (11/18/90)
Have you seen the SPARCstation II? It blows away the NeXT, and one of the
only competitors would be the DECstation or maybe the A3500 w/68040 @ 50
mhz. (There are other workstations that would do, but with the SPARC II,
NeXT is ouw of the picture.
Joseph Hillenburg
Secretary, Bloomington Amiga Users Group
joseph@valnet.UUCP ...!iuvax!valnet!joseph
"Only Apple could slow down a 68030 chip." -Computer Shoppermelling@cs.psu.edu (Michael D Mellinger) (11/18/90)
In article <Xcqss8w163w@valnet> joseph@valnet.UUCP (Joseph P. Hillenburg) writes:
Have you seen the SPARCstation II? It blows away the NeXT, and one of the
only competitors would be the DECstation or maybe the A3500 w/68040 @ 50
mhz. (There are other workstations that would do, but with the SPARC II,
NeXT is ouw of the picture.
And a Cray blows away a SPARCstation II!!!
The SPARCstation II costs three times as much as a NeXT Station. The
SPARCstation II has a SPECmark of 21. I haven't seen any SPECmarks for
the new NeXT, but it's probably around 14. You also seem to have
forgotten about the IBM RS/6000 series. They have a very good
price/performance ratio. And then there is the MIPS Magnum...
-Mike
BTW: Motorola doesn't even have the bugs out of the 25MHz 040. Don't
you think the A3500 w/ a 50MHz 040 is wishful thinking at this time?peter@sugar.hackercorp.com (Peter da Silva) (11/18/90)
In article <Xcqss8w163w@valnet> joseph@valnet.UUCP (Joseph P. Hillenburg) writes: > Have you seen the SPARCstation II? It blows away the NeXT, On speed, maybe. But by that criterion a 386SX clone blows away the Amiga 2000, so why are we all here? Yes, I believe SunOS and X windows looks as badly next to a NeXT as MS-DOS and Microsoft Windows looks next to an Amiga. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
daveh@cbmvax.commodore.com (Dave Haynie) (11/20/90)
In article <7057@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: >In article <8284@gollum.twg.com> david@twg.com (David S. Herron) writes: >> I expect that Unix running on Amiga would give me a much faster/nicer >> environment than a 386 clone or a NeXT (tho' comparing to NeXT is much >> closer in those terms). >And I can't see how it could be afster than the NeXT, given the NeXT has a >faster processor and a better base operating system (Mach vs. System V). Uh, if you're thinking of NeXTs you can actually buy today, they have the same speed processor -- a 25MHz 68030. I like the Mach kernel, and I'd say that even if I didn't get my degree at CMU. But the overall NeXT OS environment, which includes NeXTStep as well, is awfully slow. Though I've never used it from a real hard disk, or as a plain UNIX machine -- it should approach the speed of the A3000 at most UNIXish things. AmigaOS is faster at quite a few things than UNIX, but also of course unprotected. I think one of the main advantages of Amiga UNIX is that it's Standard UNIX. You have a zillion '386 machines out there, and if they're not running one of the ugly MS operating systems, they're running System V, probably 3.x but soon moving to 4.0. If you're thinking of Workstations and GUIs, according to Personal Workstation there are nearly twice as many shipping OpenLook applications as those for any other UNIX GUI, thanks to the large installed base of Sun machines. >As for the environment... have you *used* a NeXT? It's to other workstations >what the Amiga is to other PCs. It's alot like the A1000 -- nice and pretty, but not, out of the box, comparable to other Workstations. For instance, its the only machine in the Workstation category in recent memory without external cache (except the original Sun 4s, which had 100% 0-wait state memory, which in a uniprocessor system certainly eliminates the need for any cache). Fortunately, like the A1000, it has a very sound architecture, so I suspect the '040 machines, if they're out in time, will be pretty cool. The software environment is really cool. And, like most modern machines (including both the Mac and the Amiga), the first NeXT struggled to run this OS. But it looks pretty, and it comes with lots of development tools, some of the best on any machine. At least for the moment. >Peter da Silva. `-_-' ><peter@sugar.hackercorp.com>. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Standing on the shoulders of giants leaves me cold -REM
peter@sugar.hackercorp.com (Peter da Silva) (11/21/90)
In article <15973@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: > Uh, if you're thinking of NeXTs you can actually buy today, they have the > same speed processor -- a 25MHz 68030. With all due respect, the A3000UXes I can buy today don't even have that. I've played with the NeXT on an 040, but haven't seen hide nor hair of Amiga UNIX. Seriously, you can't make that sort of insinuation about the NeXT until you start shipping. > I like the Mach kernel, and I'd say that even if I > didn't get my degree at CMU. But the overall NeXT OS environment, which > includes NeXTStep as well, is awfully slow. Not from what I've seen. It's got the usual program startup delays that I've seen on every UNIX GUI, but otherwise is quit as snappy as X or SunView or whatever. > Though I've never used it from a real hard disk... Ah-ha! Well, that would tend to make it seem pretty slow. The new ones don't even come with an "optical floppy" and it makes a serious difference to the speed. > I think one of the main advantages of Amiga UNIX is that it's Standard UNIX. > You have a zillion '386 machines out there, and if they're not running one > of the ugly MS operating systems, they're running System V, probably 3.x but > soon moving to 4.0. And you think that applications developers are going to keep an A3000UX around to compile a copy of their code on for the relatively few 68000 ABI machines out there? If they have a 68000 based machine it'll probably be a Sun: and SunOS is BSD derived: not System V. In the workstation world BSD is the standard. (I personally think BSD sucks, but facts are facts) The GUI wars are, and remain, a red herring. I'm not going to get into that here, though. You guys are gung-ho about it, and I feel happy for you, but would you please finish 2.0 (or 2.1 or whatever you call it) because I really think it's going to do a lot more for you. Speaking of 2.0, I really really dislike the new DisplayBeep (it now flashes Color 0 into Color 1 instead of inverting the palette), because it makes all my text windows turn black. My wife finds it particularly distracting. Please change it back before you burn ROMs. Please. On the plus side 36.207/36.69 is definitely more robust than what you were shipping back in August. -- Peter da Silva. `-_-' <peter@sugar.hackercorp.com>.
mndaily@ux.acs.umn.edu (Linda Seebach) (11/22/90)
Disclaimer: Linda Seebach doesn't own an Amiga. Peter Seebach would. Thanks Mom! Why would someone have an amiga and not run AmigaDos? Well... I like unix. I was *raised* on unix, pretty much. I miss a lot of it... However, I like the amiga. Admittedly, if (more likely when) I get UNIX, I will probably maintain AmigaDos as part of my system. Among other things, there are a decent number of system calls which require the amigados system to run properly. Because of that, and the large number of normal people out there who will have amigas without unix, I intend to keep using Ados and writing programs that will work on it... On the other hand, I want to make as many of them as possible UNIX portable.. Speaking of ports... Is there any way to port Lattice/Aztec C *libraries*? I got curses.lib for lattice, but I have aztec. Grr... And, of course, one more obnoxious question: WHEN THE HELL IS UNIX COMING OUT? (And is unix, in fact, gay? ;-) Also, while I'm bitching about UNIX.... Does anyone have specs for this? Like, POSIX standards, what kind of extra software it will come with, compatability, etc... I know that there are UNIX machines out there that take forever to get to work with anything. Is the Amiga unix going to be (one would hope) compatible enough that I can connect my serial port to Joe Unix Box, and expect them to communicate after I set up a few files? Or will I have to write protocols and all that? --SeebS-- Prophet in Exile
tell@oscar.cs.unc.edu (Stephen Tell) (11/27/90)
In article <7100@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: > [After quoting Dave] >In article <15973@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: >And you think that applications developers are going to keep an A3000UX around >to compile a copy of their code on for the relatively few 68000 ABI machines >out there? If they have a 68000 based machine it'll probably be a Sun: and >SunOS is BSD derived: not System V. In the workstation world BSD is the >standard. (I personally think BSD sucks, but facts are facts) A minor nit, but it seems to be a common misconception. SunOS came from BSD, true, but AT&T and Sun are merging their unixes. SunOS and SVR4 are both really SVR3 + BSD. We have all sorts of machines around here, and I *know* that the SunOS 4.1 I'm using has most of what I rememember from SVR2 and is quite different from the latest 4.3BSD and also different from Ultrix 4.0, I'm not sure if it is correct to say that SunOS 4.1 *is* SVR4, but if it isn't its close and SunOS 4.x, x>1 probably will be. Now if they both only had the neat stuff out of Bell Labs research Unix, like /dev/fd and /proc... Again: SVR4 is not you're father's System V.... >Peter da Silva. `-_-' ><peter@sugar.hackercorp.com>. -------------------------------------------------------------------- Steve Tell e-mail: tell@wsmail.cs.unc.edu usmail: #5L Estes Park apts CS Grad Student, UNC Chapel Hill. Carrboro NC 27510
daveh@cbmvax.commodore.com (Dave Haynie) (11/28/90)
In article <7100@sugar.hackercorp.com> peter@sugar.hackercorp.com (Peter da Silva) writes: >In article <15973@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: >> Uh, if you're thinking of NeXTs you can actually buy today, they have the >> same speed processor -- a 25MHz 68030. >With all due respect, the A3000UXes I can buy today don't even have that. "A3000UX" is simply a marketing created A3000 UNIX bundle. It is the same A3000 you've been able to buy since June, only it comes with 8 Meg of RAM and an optional 200MB hard disk (100MB standard). And you get UNIX with it. At the very moment, no, you can't go to the store and get UNIX, you'd have to live with the faster AmigaOS for awhile until UNIX ships. >Not from what I've seen. It's got the usual program startup delays that I've >seen on every UNIX GUI, but otherwise is quit as snappy as X or SunView or >whatever. BYTE this month claims that Amiga UNIX seems faster to them than the NeXT OS on a cube or A/UX on a Mac. So it's not only Amiga heads who see this... >And you think that applications developers are going to keep an A3000UX around >to compile a copy of their code on for the relatively few 68000 ABI machines >out there? There are few 680x0 based now. If everyone with a compatible 680x0 machine bought a compatible ABI UNIX, 680x0 ABI UNIX would be the dominant UNIX. I don't expect the installed base of 680x0 ABI UNIX system to remain low, and I don't expect everyone who could to run out an buy UNIX as soon as it's available; they certainly didn't for Mac UNIX. Reality will be somewhere in the middle. >If they have a 68000 based machine it'll probably be a Sun: and SunOS is BSD >derived: not System V. Old SunOS is BSD derived, sure. But Sun's no longer supporting software upgrades on the 680x0 machines, either, I hear. >In the workstation world BSD is the standard. Maybe the Sun world. I've been using Apollos for 7 years, and until recently you couldn't even get UNIX for them. It's still not the dominant OS on that platform. And now, while there's kind of UNIX available within the Domain/OS system, you have BSD and System V stuff, function calls, etc. C= originally went with AT&T System V because it was the business standard for UNIX. Now it's very rapidly becoming The Standard, even if it did have to suck in some of the BSD features to get there (not all of which, certainly, are Bad Things). >You guys are gung-ho about it, and I feel happy for you, but would you please >finish 2.0 (or 2.1 or whatever you call it) because I really think it's going >to do a lot more for you. Well, I got the hardware going for them.... >Peter da Silva. `-_-' -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy ONLY 230 MILES TO GO
ssd@sugar.hackercorp.com (Scott Denham) (12/02/90)
In article <16094@cbmvax.commodore.com>, daveh@cbmvax.commodore.com (Dave Haynie) writes: > > "A3000UX" is simply a marketing created A3000 UNIX bundle. It is the same > A3000 you've been able to buy since June, only it comes with 8 Meg of RAM > and an optional 200MB hard disk (100MB standard). And you get UNIX with it. > At the very moment, no, you can't go to the store and get UNIX, you'd have > to live with the faster AmigaOS for awhile until UNIX ships. > Ah! You C= folks need to watch out for your dealer network, they may be your worst enemy! I heard from a local dealer the statement that an A3000UX had something "extra" on the motherboard, and that a standard A3000 or an A2500/30 would NOT be able to run Unix. It sounded like a ridiculous enough idea that I didn't really believe it, but you never know; C= marketing HAS been known to make a few "strange" decisions. I'd hate to see a return to the "ready, fire, aim" days! I take it from your article that this rumor is in fact untrue, and that sooner or later current A3000 owners (sadly, I am not one!) (and A2500/30 owners?) will be able to run Amiga UNIX as well?
daveh@cbmvax.commodore.com (Dave Haynie) (12/06/90)
In article <7185@sugar.hackercorp.com> ssd@sugar.hackercorp.com (Scott Denham) writes: >In article <16094@cbmvax.commodore.com>, daveh@cbmvax.commodore.com (Dave Haynie) writes: >> "A3000UX" is simply a marketing created A3000 UNIX bundle. It is the same >> A3000 you've been able to buy since June, only it comes with 8 Meg of RAM.. Well, actually 9 Meg if you include Chip RAM... > I take it from your article that this rumor is in fact untrue, and that >sooner or later current A3000 owners (sadly, I am not one!) (and A2500/30 >owners?) will be able to run Amiga UNIX as well? Technically speaking, certainly. Far as I know, the UNIX folks don't have any special "A3000UX" machine in their offices, and until last summer, they all had A2500/30s there (the A3000 does offer the advantage of easily having more than 4Meg of 32 bit RAM, important for memory hungry UNIX). Whether C= will sell a stand-alone UNIX package, or only offer the A3000UX bundle, is of course purely a marketing decision. -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Wheeeeeeeeeeeeeeee...........
etxtomp@eos.ericsson.se (Tommy Petersson) (12/07/90)
In article <16325@cbmvax.commodore.com> daveh@cbmvax.commodore.com (Dave Haynie) writes: -In article <7185@sugar.hackercorp.com> ssd@sugar.hackercorp.com (Scott Denham) writes: --In article <16094@cbmvax.commodore.com>, daveh@cbmvax.commodore.com (Dave Haynie) writes: - ->> "A3000UX" is simply a marketing created A3000 UNIX bundle. It is the same ->> A3000 you've been able to buy since June, only it comes with 8 Meg of RAM.. - -Well, actually 9 Meg if you include Chip RAM... - -> I take it from your article that this rumor is in fact untrue, and that ->sooner or later current A3000 owners (sadly, I am not one!) (and A2500/30 ->owners?) will be able to run Amiga UNIX as well? - -Technically speaking, certainly. Far as I know, the UNIX folks don't have -any special "A3000UX" machine in their offices, and until last summer, they -all had A2500/30s there (the A3000 does offer the advantage of easily having -more than 4Meg of 32 bit RAM, important for memory hungry UNIX). Whether C= -will sell a stand-alone UNIX package, or only offer the A3000UX bundle, is -of course purely a marketing decision. - --- -Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" - {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy - Wheeeeeeeeeeeeeeee........... ...and a question of honesty, since A2620's and A2630's have been sold under the assumption that they would be able to run Commodore's Unix when it RSN will be available, for a long time now. Is it technically possible to run the current implementation of Unix on an A2620 equipped Amiga 2000? Tommy Petersson
milamber@caen.engin.umich.edu (Daryl Scott Cantrell) (12/07/90)
In article <1990Dec6.191222.12243@ericsson.se> etxtomp@eos.ericsson.se writes: [...] >Is it technically possible to run the current implementation of Unix on >an A2620 equipped Amiga 2000? Should be, certainly. The 2620 was the "original" intended Unix platform, and I'd wager that's one of the big reasons they put the 68851 on it. The 2630 is basically the same board with asynchronous timing, and of course the MMU incorporated into the 030.. >Tommy Petersson -- +---------------------------------------+----------------------------+ | // Daryl S. Cantrell | These opinions are | | |\\\ milamber@caen.engin.umich.edu | shared by all of // | | |// Evolution's over. We won. | Humanity. \X/ | +---------------------------------------+----------------------------+
ford@amix.commodore.com (Mike "Ford" Ditto) (12/08/90)
In article <1990Dec6.191222.12243@ericsson.se> etxtomp@eos.ericsson.se writes: >Is it technically possible to run the current implementation of Unix on >an A2620 equipped Amiga 2000? Yes, I do it all the time. Our "standard" Beta release includes an A2620/A2630 boot floppy, an A3000 boot floppy, an install floppy, and a tape. I beleive that for the initial A3000UX machines, they have left out the A2620 floppy (why not?). By the time we release, we hope to have merged the boot procedures so that the A2000/A3000 boot floppies will be replaced by one universal boot floppy. That way, regardless of what marketing decides, you A2620 owners can get Unix because it'll be the same as for the A3000. :-) Right now, an A2620 or A2630 system uses its special boot ROM to boot Unix, while an A3000 boots Unix using the improved boot procedure in the 2.0 AmigaDOS ROM. When we merge, this may have the ironic effect of requiring you to select "AmigaDOS", not "Unix" in order to boot Unix on an A2620. -=] Ford [=- "The number of Unix installations (In Real Life: Mike Ditto) has grown to 10, with more expected." ditto@amix.commodore.com - The Unix Programmer's Manual, uunet!cbmvax!ditto 2nd Edition, June, 1972. ford@kenobi.commodore.com
jita@polaris.utu.fi (Marko Katajisto) (12/10/90)
In article <598@amix.commodore.com> ford@amix.commodore.com (Mike "Ford" Ditto) writes: >Yes, I do it all the time. Our "standard" Beta release includes an >A2620/A2630 boot floppy, an A3000 boot floppy, an install floppy, and >a tape. I beleive that for the initial A3000UX machines, they have >left out the A2620 floppy (why not?). By the time we release, we hope >to have merged the boot procedures so that the A2000/A3000 boot >floppies will be replaced by one universal boot floppy. That way, >regardless of what marketing decides, you A2620 owners can get Unix >because it'll be the same as for the A3000. :-) Nice, very nice indeed! Maybe I'll forget that DECStation 2100 and buy UNIX for my current A2000/A2620/A2320 (A2410 maybe?) configuration instead... :) Still, a couple of questions: In what format will that tape come, standard UNIX tape format I guess, but what's the size of the tape? 60 or 150 megs? And should I sell my current SCSI-adapter (Microbotics HardFrame) and get an A2091 or can I get drivers for other host adapters too (any bundled with current beta-relase?)? PS. Any ideas how to break this annoying 4 meg barrier on this board, maybe Dave Haynie could hack out a replacement of these 1 mb chips to 4 mb chips? X Windows with Open Look in 4 MB machine is swapper's nightmare I suppose? :) > -=] Ford [=- > >"The number of Unix installations (In Real Life: Mike Ditto) >has grown to 10, with more expected." ditto@amix.commodore.com >- The Unix Programmer's Manual, uunet!cbmvax!ditto > 2nd Edition, June, 1972. ford@kenobi.commodore.com ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Hehhee.... :) -- Jita - Marko Katajisto University of Turku Department of Computer Science jita@utu.fi jita@nic.funet.fi jita@firien.bitnet Tel: +358 21 325910
ford@amix.commodore.com (Mike "Ford" Ditto) (12/10/90)
In article <JITA.90Dec9215738@polaris.utu.fi> jita@polaris.utu.fi (Marko Katajisto) writes: >Still, a couple of questions: In what format will that tape come, standard >UNIX tape format I guess, but what's the size of the tape? 60 or 150 megs? QIC-150 format (150 Meg). >And should I sell my current SCSI-adapter (Microbotics HardFrame) and get >an A2091 or can I get drivers for other host adapters too (any bundled with >current beta-relase?)? It will be possible to add a device driver to the Unix kernel for a SCSI adapter of your choice, but only after you have Unix installed and running on your system. Catch 22. We are thinking about this problem, but for now, an A2090 or A2091 is required (A2091 recommended). >X Windows with Open Look in 4 MB machine is swapper's nightmare I suppose? :) Yes. But it does work. :-) X, being the memory hog that it is, is the only application I'm aware of for which it actually helps to add some 16 bit memory cards to an A2000. -=] Ford [=- "A just machine to make big decisions (In Real Life: Mike Ditto) programmed by fellows with compassion ditto@amix.commodore.com and vision." - Donald Fagen, "IGY" uunet!cbmvax!ditto ford@kenobi.commodore.com
scot@amigash.UUCP (Scot L. Harris) (12/11/90)
>In article <598@amix.commodore.com> ford@amix.commodore.com (Mike "Ford" Ditto) writes: >In article <1990Dec6.191222.12243@ericsson.se> etxtomp@eos.ericsson.se writes: >>Is it technically possible to run the current implementation of Unix on >>an A2620 equipped Amiga 2000? > >Yes, I do it all the time. Our "standard" Beta release includes an So how about those of us with GVP 68030 boards? Will a special boot floppy be required/provided? > > -=] Ford [=- > >"The number of Unix installations (In Real Life: Mike Ditto) >has grown to 10, with more expected." ditto@amix.commodore.com >- The Unix Programmer's Manual, uunet!cbmvax!ditto > 2nd Edition, June, 1972. ford@kenobi.commodore.com -- _ /// /_\ Scot L. Harris ...!tarpit!bilver!amigash!scot \XX/ / \ M I G A Orlando, FL (407)273-1759 [Falcon Mission Disk II. Must be great, haven't gotten any work done for days]
daveh@cbmvax.commodore.com (Dave Haynie) (12/11/90)
In article <JITA.90Dec9215738@polaris.utu.fi> jita@polaris.utu.fi (Marko Katajisto) writes: >In article <598@amix.commodore.com> ford@amix.commodore.com (Mike "Ford" Ditto) writes: >PS. Any ideas how to break this annoying 4 meg barrier on this board, maybe >Dave Haynie could hack out a replacement of these 1 mb chips to 4 mb chips? For the A2620, its possible to replace the memory decoding PAL with one that allows you to move the 4 Megs of 32 bit memory from being autoconfigured to sitting fixed at $01000000. There's no simple way to get more addressing on the A2620 without some serious hacking to it, since the extra address lines needed don't go anywhere, etc. So you can have more 16bit RAM, but no more 32 bit RAM. Maybe if I had the 4MB DRAM specs back then I would have thrown something in. The A2630 doesn't support this relocation, though it does in theory have this daughterboard support, should someone ever feel inclined to build such a device. The current board would need about as much hacking for 4MB DRAM support as the A2620. >Jita - Marko Katajisto University of Turku Department of Computer Science -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy Wheeeeeeeeeeeeeeee...........