wolf@mel.cipl.uiowa.edu (09/09/90)
*** Ignorant User Alert *** Our department is buying new Macs (ci's) and new workstations (DEC5000's). We don't want to cut ourselves short by getting Macd HD's which would not be large enough for A/UX v(put most recent version number here). I read through several notes in this group. Someone had asked minimum disk size, but non one replied. I hear 80M is minimum, and 160 is a good size, correct? Also, I had heard that some software bundled with 3rd party drives will not partition correctly for A/UX, correct? If so, why is this and who has the 'good' software. I had not seen anything about this, and it will seem a stupid question but, can any of the files A/UX needs be placed on a unix device (like a DEC5000 disk)? Any help would be appreciated! M. Wolf -=-=-=-=-=-=-=-=-=-=- Dislcaimer? Heck, I don't even know her! -=-=-=-=-=-=-=-=-=-=-
rmtodd@servalan.uucp (Richard Todd) (09/10/90)
wolf@mel.cipl.uiowa.edu writes: >I hear 80M is minimum, and 160 is a good size, correct? Also, I had heard that Yep, 80M is pretty close to the minimum, and 160M is pretty good. >some software bundled with 3rd party drives will not partition correctly for >A/UX, correct? If so, why is this and who has the 'good' software. Why? Because despite Apple clearly documenting the format of the "new" format partition table in Inside Mac Vol. 5, in various technotes, and in practically every publication except the New York Times, there are still vendors so utterly braindead they haven't updated their software to handle the new-style partition table. Contact the LaCie company (they advertise in your major Macintosh magazines.) They sell a program called "SilverLining" which will format practically any drive, and which will handle the new format of partition table. >I had not seen anything about this, and it will seem a stupid question but, can >any of the files A/UX needs be placed on a unix device (like a DEC5000 disk)? You mean, like having the Mac mount an NFS partition off a disk on the DEC 5000? Haven't tried it, for obvious reasons (not having a DEC5000 to play with :-), but it oughta work. In which case you should be able to cut down that minimum disk space required by A/UX to just enough to give you enough of a minimal root partition to NFS mount everything else you need. -- Richard Todd rmtodd@uokmax.ecn.uoknor.edu rmtodd@chinet.chi.il.us rmtodd@servalan.uucp
liam@cs.qmw.ac.uk (William Roberts) (09/10/90)
In <1990Sep9.202622.21422@servalan.uucp> rmtodd@servalan.uucp (Richard Todd) writes: >wolf@mel.cipl.uiowa.edu writes: >>I had not seen anything about this, and it will seem a stupid question but, can >>any of the files A/UX needs be placed on a unix device (like a DEC5000 disk)? >You mean, like having the Mac mount an NFS partition off a disk on the DEC >5000? Haven't tried it, for obvious reasons (not having a DEC5000 to play >with :-), but it oughta work. In which case you should be able to cut down >that minimum disk space required by A/UX to just enough to give you enough >of a minimal root partition to NFS mount everything else you need. To give you some idea about a minimal root partition, consider the following: 1) A/UX 1.1.1 can be booted from three floppies (two if you have a superdrive) One 800k floppy holds a root partition without a kernel, and the kernel is booted from the Mac filesystem. This fits on a single 800K floppy, but if you have 1.4Meg floppies you can have Sash and a Mac system as well. This 800K root contains everything you need to be an NFS client. 2) A/UX 1.1.1 works OK with a 10 Meg root, 5 meg swap and 5 meg /tmp (we don't believe in diskless nodes). If you wanted to, you could run A/UX 1.1.1 on an SE/30 with a 20 Meg disk and have 7 meg of local disk not included in the above. NFS fileservers hold the rest of the stuff. 3) I am currently working on getting A?UX 2.0 to boot in 10 Meg, including the Mac system stuff. This is very tight indeed (you need 3 lots of 400K+ Mac SYstem files, amongst other reasons) and I am likely to offload the actual System Folder into that 5 meg /tmp mentioned above. If anyone is interested, we have a very rough-and-ready Hypercard stack which splits the A/UX 2.0 distribution into pieces, and can produce scripts for cpio-ing selected things from the CD-ROM. No instructions, no promises, but it might be something to look at if you are looking for things to throw away. I'd also appreciate it if Mr "millions of lines of code" doesn't start whingeing about "proprietary formats" again - I tend to use the tools available... -- William Roberts ARPA: liam@cs.qmw.ac.uk Queen Mary & Westfield College UUCP: liam@qmw-cs.UUCP Mile End Road AppleLink: UK0087 LONDON, E1 4NS, UK Tel: 071-975 5250 (Fax: 081-980 6533)
kaufman@Neon.Stanford.EDU (Marc T. Kaufman) (09/11/90)
In article <1990Sep9.202622.21422@servalan.uucp> rmtodd@servalan.uucp (Richard Todd) writes: >wolf@mel.cipl.uiowa.edu writes: ->some software bundled with 3rd party drives will not partition correctly for ->A/UX, correct? If so, why is this and who has the 'good' software. >Why? Because despite Apple clearly documenting the format of the "new" format >partition table in Inside Mac Vol. 5, in various technotes, and in practically >every publication except the New York Times, there are still vendors so >utterly braindead they haven't updated their software to handle the new-style >partition table. That's just not true. While the major fields of the Partition format were documented in IM-5, nowhere is there coherent documentation of the flag fields, the spare sector information, and the A/UX and Eschatology specific information. In fact, at A/UX 1.0 time, the MacOS folks didn't know what the A/UX folks wanted, and vice versa. Marc Kaufman (kaufman@Neon.stanford.edu)
thad@cup.portal.com (Thad P Floryan) (09/12/90)
liam@cs.qmw.ac.uk (William Roberts) in <2771@sequent.cs.qmw.ac.uk> writes:
[...]
I'd also appreciate it if Mr "millions of lines of code" doesn't start
whingeing about "proprietary formats" again - I tend to use the tools
available...
No problem. Some other companies during the past several weeks GAVE me some
SVR3.2 and SVR4 boxes upon which to port my software; if it also works under
A/UX, then fine, but I'm no longer immediately concerned about problems with
A/UX. Apple itself is using my product on its in-house VAXes and that's fine
with me.
Perhaps in the future you'll realize the benefit of ABI (Application Binary
Interface) and the other goodies of a more-modern UNIX; I don't see those
capabilities existing with an A/UX based on SVR2 and BSD4.2 circa 1983-1984.
What amazes me, though, is that in 21 of 27 emails I received (including some
from A/UX Development Team members), I was asked (paraphrased):
If you didn't care for or want the MacOS, why'd you get A/UX?
I simply wanted to run UNIX on some of the hardware I already have at hand.
Is that so difficult to comprehend? Or is there another vendor's UNIX also
available for the Mac II platforms?
-Thad
Thad Floryan [ thad@cup.portal.com (OR) ..!sun!portal!cup.portal.com!thad ]liam@cs.qmw.ac.uk (William Roberts) (09/13/90)
In <33817@cup.portal.com> thad@cup.portal.com (Thad P Floryan) writes: >Perhaps in the future you'll realize the benefit of ABI (Application Binary >Interface) and the other goodies of a more-modern UNIX; I don't see those >capabilities existing with an A/UX based on SVR2 and BSD4.2 circa 1983-1984. I do understand the benefit, and I also recognise that A/UX has missed the boat - unless the UNXI ABI for 680x0 includes the myriad different schemes for getting syscall arguments into the kernel (if you don't already know about this mess, consider pipe, read, semop and socketpair: yes - they are ALL different). >I simply wanted to run UNIX on some of the hardware I already have at hand. >Is that so difficult to comprehend? Or is there another vendor's UNIX also >available for the Mac II platforms? I understand that Mach is available for the Mac II. -- William Roberts ARPA: liam@cs.qmw.ac.uk Queen Mary & Westfield College UUCP: liam@qmw-cs.UUCP Mile End Road AppleLink: UK0087 LONDON, E1 4NS, UK Tel: 071-975 5250 (Fax: 081-980 6533)
thad@cup.portal.com (Thad P Floryan) (09/14/90)
liam@cs.qmw.ac.uk (William Roberts) in <2791@sequent.cs.qmw.ac.uk> writes:
I understand that Mach is available for the Mac II.
Thank you for that information! I'm also looking at the Nixdorf and Siemens
boxes per a lead I received today from my sales office in The Hague; I've had
it with single-vendor "solutions" which is why I'm also going along with the
HP-UX, CTIX, SVR4, and others' ... I'm just now "recovering" from having used
mostly DEC equipment the past 28 years.
Lest anyone receive the wrong idea: I still intend to contribute to this
newsgroup, and my month-ago offer to post PD and freely-redistributable stuff
(in source form) I've ported to A/UX is still valid. Having been absent for
two weeks, I must have missed any posting of archive sites to which material
can be uucp'd and/or ftp'd; were any sites agreed upon?
-Thad
Thad Floryan [ thad@cup.portal.com (OR) ..!sun!portal!cup.portal.com!thad ]coolidge@cs.uiuc.edu (John Coolidge) (09/14/90)
thad@cup.portal.com (Thad P Floryan) writes: >liam@cs.qmw.ac.uk (William Roberts) in <2791@sequent.cs.qmw.ac.uk> writes: >Lest anyone receive the wrong idea: I still intend to contribute to this >newsgroup, and my month-ago offer to post PD and freely-redistributable stuff >(in source form) I've ported to A/UX is still valid. Having been absent for >two weeks, I must have missed any posting of archive sites to which material >can be uucp'd and/or ftp'd; were any sites agreed upon? I've created a directory for contributions to the archive at wuarchive.wustl.edu: /archive/systems/aux/Incoming (systems/aux/Incoming via anonymous ftp). Feel free to place any A/UX related things there, and I'll move them into the proper place in the archive (also send me mail, so I know something's there! :-)). --John -------------------------------------------------------------------------- John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge Of course I don't speak for the U of I (or anyone else except myself) Copyright 1990 John L. Coolidge. Copying allowed if (and only if) attributed. You may redistribute this article if and only if your recipients may as well.
ml10+@andrew.cmu.edu (Michael A. Libes) (09/15/90)
>I understand that Mach is available for the Mac II.
Not quite yet. MacMach is under development here at Carnegie Mellon
University. It is unknown if or when it will available to the outside
world. The distribution rights to MacMach are owned by Apple, and they
have made no announcements concerning their plans.
If anyone has further questions, feel free to send email to
ml10+@andrew.cmu.edu. We don't need beta testers, so please don't ask.
------------------------------------------------------------------
Lunarmobiscuit (Luni) Internet Address:
MacMach Project ml10@andrew.cmu.edu
Carnegie Mellon University
------------------------------------------------------------------paul@taniwha.UUCP (Paul Campbell) (09/19/90)
In article <2791@sequent.cs.qmw.ac.uk> liam@cs.qmw.ac.uk (William Roberts) writes: >I do understand the benefit, and I also recognise that A/UX has missed the >boat - unless the UNXI ABI for 680x0 includes the myriad different schemes >for getting syscall arguments into the kernel (if you don't already know >about this mess, consider pipe, read, semop and socketpair: yes - they >are ALL different). Well yes and no (read on it doesn't really matter ...). 'Yes' because there are a couple of ABIs out there - the first one called the 68000BCS (Binary Compatability Standard) which was promoted by Motorola about 3 years ago codifies system call parameters, trap codes etc for Posix implementation it codified the system call practices in use at that time in the standard Unix porting base (from which A/UX is derived) [I wrote a large chunk of it so I remember it in terrible detail :-]. It was rather still-born because most of the 68K Unix vendors at that time were in the process of moving to RISC. There is a similar 88K BCS document. 'No' because the current evolving standards (new unixes etc) use shared libraries to handle this. The standards describe how applications find system calls in the libraries (ie 'read', shmop' etc as well as non-system call routines such as 'printf' etc) how the calls in the library actually call the kernel are up to a particular implementation and are irrelevant to the issue of ABIs. Note that one of the things that came out in the BCS development was that system call schemes are relatively independant of function - this means that a system can have more than one built-in system call scheme (for example it can have two system call traps - one that passes system parameters in registers, another that passes them on the stack (A/UX actually does have these) - or even traps with different numbers for system calls, or even different parameters etc etc). Paul -- Paul Campbell UUCP: ..!mtxinu!taniwha!paul AppleLink: CAMPBELL.P What most people don't realize is that those plastic cover slips that your 3 inch floppies come in are actually condoms for protecting your computer from harmfull computer viruses - practice safe computing ..... :-)