asg@sage.cc.purdue.edu (The Grand Master) (06/11/91)
Hello.
I currently have a 286 running MINIX. Within the next 1.5 years however,
I plan on upgrading to either an Amiga (running Amiga SYS5r4) or a
486. The current 486 that I am looking at is a Gateway 2000 486-33.
I would get it with 40MB of memory, and two 200MB IDE drives. Now
my questions:
-Will the extensions to MINIX for a 386 work equally well on a 486?
-Is a 15-20MB Ramdisk feasable?
-Is anyone out there working on (or planning on) extensions to MINIX
for a 486, or will any 486 extensions be applicable to a 386 as well.
-Is anyone out there working on Job control, swapping&paging, and/or
sockets for MINIX?
-Is there a higher power UNIX(tm)-like OS out there with complete source
that would have Job Control, swapping/paging, sockets, etc?
-What are the problems associated with MINIX on IDE drives.
-I have heard that there is a 32M limit on MINIX partitions. If so,
4*32=128 which would leave 72MB/drive unoccupied. Any suggestions
on how to work around this?
-Would I be better off getting a more high-power (and more high-priced :-( )
UNIX(tm)-like OS (propbably without the source)?
Questions for my current configuration:
- I have a 287 math co-processor Is there anyone out there who has done
any work on getting MINIX to use the math coprocessor?
- Is there a way to split my HD into > 4 partitions? If so, what files would
I need to modify?
thanx for your time
Bruce
---------
### ##
Courtesy of Bruce Varney ### #
aka -> The Grand Master #
asg@sage.cc.purdue.edu ### ##### #
PUCC ### #
;-) # #
;'> # ##jds@cs.umd.edu (James da Silva) (06/11/91)
In article <13393@mentor.cc.purdue.edu> asg@sage.cc.purdue.edu (The Grand Master) writes: >-Will the extensions to MINIX for a 386 work equally well on a 486? Yes, I run Minix on my AMI Voyager 486-25. Works quite well. >-Is a 15-20MB Ramdisk feasable? Well, what do you mean by feasable? There isn't any fundamental reason that it can't be done, but a filesystem cache would be better: 20 meg is a lot to lose in a power outage. >-Is anyone out there working on (or planning on) extensions to MINIX > for a 486, or will any 486 extensions be applicable to a 386 as well. I don't know that Minix needs to explicitly deal with 486isms. There are a number of differences listed between the 486 FPU and the 387 that look at first glance that they might be significant to the operating system, so if someone does do FP exception support in the kernel, 486 specific code might be needed or desired. >-Is anyone out there working on Job control, swapping&paging, and/or > sockets for MINIX? Why turn Minix into BSD: BSD will be free "soon". >-Is there a higher power UNIX(tm)-like OS out there with complete source > that would have Job Control, swapping/paging, sockets, etc? If you can wait, BSD 4.4 should fit the bill for you. >-I have heard that there is a 32M limit on MINIX partitions. If so, > 4*32=128 which would leave 72MB/drive unoccupied. Any suggestions > on how to work around this? The limit on Minix 1.5 partitions is 64 Meg, not 32, so with 200M disks you are safe. Minix 1.6 should handle larger partitions; I don't recall what the limits are. Ideally, the whole disk should be one partition; there's no good reason to pigeon-hole the disk. >-Would I be better off getting a more high-power (and more high-priced :-( ) > UNIX(tm)-like OS (propbably without the source)? Well, what do you really want?? Minix isn't really designed for machines with 40 megabytes of memory; after all, you can't run X on it. :-) If you want to run X and big lisp jobs, or support 16 users, then you will want VM. Thus you probably don't want Minix. If you just want all that memory to make up for slow disks, perhaps you should run Sprite. >- Is there a way to split my HD into > 4 partitions? If so, what files would > I need to modify? You need to modify at_wini.c. There's nothing really sacred about the partition table, except compatibility with current tools and OSes. You could probably hack at_wini.c to use a different scheme, but then you'd need your own fdisk equivalent. Something simple would be to ignore the partition table and use fixed size partitions: /dev/hd1 is the first 64M of disk, hd2 the second 64M, etc. until you run out of disk space. The second hard disk could start at /dev/hd32 or whatever. I'm not recommending this as a general solution, but it's the kind of simple hack that can get you by on a rainy day when you are contemplating the unused chunks of your 340 Meg SCSI disk. :-) Jaime ........................................................................... : domain: jds@cs.umd.edu James da Silva : path: uunet!mimsy!jds Systems Design & Analysis Group
melling@cs.psu.edu (Michael D Mellinger) (06/11/91)
In article <35441@mimsy.umd.edu> jds@cs.umd.edu (James da Silva) writes: >-Is anyone out there working on Job control, swapping&paging, and/or > sockets for MINIX? Why turn Minix into BSD: BSD will be free "soon". No it won't. -Mike
awb@ed.ac.uk (Alan W Black) (06/11/91)
In article <13393@mentor.cc.purdue.edu> asg@sage.cc.purdue.edu (The Grand Master) writes: From: asg@sage.cc.purdue.edu (The Grand Master) > Hello. > I currently have a 286 running MINIX. Within the next 1.5 years however, > I plan on upgrading to either an Amiga (running Amiga SYS5r4) or a > 486. The current 486 that I am looking at is a Gateway 2000 486-33. > I would get it with 40MB of memory, and two 200MB IDE drives. Now > my questions: > > -Will the extensions to MINIX for a 386 work equally well on a 486? I have run Minix with Bruce's patches on a 486 with no problems (I've also heard of others doing this). > -Is a 15-20MB Ramdisk feasable? Yes (if you have enough memory). There is (possibly) a problem in using more 16M on minix but that should be fixed soon. However I see little point in having such a large ram disk. It is better to have a fast hard disk and a reasonably sized buffer (a few meg). > -Is anyone out there working on (or planning on) extensions to MINIX > for a 486, or will any 486 extensions be applicable to a 386 as well. I'm not sure what the real differences are. The 486 has a floating point processor as part of it. I've never heard of a compiler that takes advantage of it (gcc doesn't seem to have any 486 extensions). So I assume it is basically 386 with better cache, pipelining and floating point. Thus no real extensions to minix would have relevant. > -Is anyone out there working on Job control, swapping&paging, and/or > sockets for MINIX? Sockets are being worked on. Job control and swapping/pagin may be too but they'll be some time. As someone else said BSD will be available "soon" for almost free. (Someone disagreed but I'm not sure why -- the "soon" bit is the most arguable). CMU's MACH and GNU's system will come along, but I suspect they'll all be merged. These systems will be free (from the software licence point of view) but probably non-trivial to install. However if you want a unix with source at an affordable price now, Minix is the only option. > -Is there a higher power UNIX(tm)-like OS out there with complete source > that would have Job Control, swapping/paging, sockets, etc? Source really is the question. Only Minix has source sold to the masses. BSD and AT&T licences are possible (most universities have a source licence) but they are too expensive for the individual. There are other SysV based unix's for pcs but none of them will include source unless you pay lots of dollars. > -What are the problems associated with MINIX on IDE drives. I use a 210Meg IDE drive and have never noticed any problems, nor did anything special to minix to make it work. > -I have heard that there is a 32M limit on MINIX partitions. If so, > 4*32=128 which would leave 72MB/drive unoccupied. Any suggestions > on how to work around this? I have a 64Meg partition, larger than that it stops working. I vaguely remember discussion on the net about this so I suspect it can be made bigger. > -Would I be better off getting a more high-power (and more high-priced :-( ) > UNIX(tm)-like OS (propbably without the source)? > Well it depends. If you want X and want to program in it (rather than just want to use a window system). The bigger system's will usually allow X, tcp-ip etc which might be what you want but if you just want an interesting system to hack that runs a c compiler, debugger, emacs and many things from the net Minix is a good choice. Alan Alan W Black 80 South Bridge, Edinburgh, UK Dept of Artificial Intelligence tel: (+44) -31 650 2713 University of Edinburgh email: awb@ed.ac.uk
jds@cs.umd.edu (James da Silva) (06/11/91)
In article <yz6Hilos@cs.psu.edu> melling@cs.psu.edu (Michael D Mellinger) writes: >In article <35441@mimsy.umd.edu> jds@cs.umd.edu (James da Silva) writes: > Why turn Minix into BSD: BSD will be free "soon". > >No it won't. Why do you say that? Do you have some inside information? Please share it. If you are going to refute my statement, please provide some more detailed prose. First you have to define "soon"; to me, "soon" is any time within the next year. Perhaps that's the source of our disagreement. Jaime ........................................................................... : domain: jds@cs.umd.edu James da Silva : path: uunet!mimsy!jds Systems Design & Analysis Group
paul@ukpoit.co.uk (Paul Wood) (06/12/91)
In article <AWB.91Jun11091405@stoat.uk.ac.ed.aipna> awb@ed.ac.uk (Alan W Black) writes: >As someone else said BSD will be available >"soon" for almost free. (Someone disagreed but I'm not sure why -- the >"soon" bit is the most arguable). My guess is the "almost free" part is the problem. I understand that it is not going to be free to private individuals. In fact is going to be very expensive compared to Minix. Can someone clarify? -- Paul Wood [ e-mail: paul@ukpoit.co.uk or ...!ukc!ukpoit!paul ] [ address: iT, Barker Lane, Chesterfield, England S40 1DY ] [ phone: +44 246 214256, postline: 5403 4256, fax: +44 246 214353 ]
jds@cs.umd.edu (James da Silva) (06/13/91)
In article <1991Jun12.145309.29231@ukpoit.co.uk> paul@ukpoit.co.uk (Paul Wood) writes: >My guess is the "almost free" part is the problem. I understand that it >is not going to be free to private individuals. In fact is going to be >very expensive compared to Minix. Can someone clarify? Here's the copyright notice from a kernel file in the 4.3 Reno release: /* * Copyright (c) 1982, 1986 Regents of the University of California. * All rights reserved. * * Redistribution and use in source and binary forms are permitted provided * that: (1) source distributions retain this entire copyright notice and * comment, and (2) distributions including binaries display the following * acknowledgement: ``This product includes software developed by the * University of California, Berkeley and its contributors'' in the * documentation or other materials provided with the distribution and in * all advertising materials mentioning features or use of this software. * Neither the name of the University nor the names of its contributors may * be used to endorse or promote products derived from this software without * specific prior written permission. * THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED * WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. * * @(#)fs.h 7.10 (Berkeley) 6/28/90 */ Unless they back away from this copyright (why would they?), BSD _will_ be freely redistributable. Any charges would be distribution charges. You can expect to have to pay SOMETHING to get the full sources; either for a tape, or to your phone company to download the files. The AT&T-free portion of 4.3 reno are on UUNET right now; you can get to it via anonymous UUCP on a 900-number. It may be cheaper to buy the 4.4BSD tape, unless you are on the Internet, or can get the code via a local call or clone a tape from a friend. Another possibility is that not all of the next release will be covered by the above copyright. In particular I don't know anything about the copyright on the 386 specific stuff in the "386/BSD" project. Or maybe it won't be entirely AT&T-detoxified. I'm just speculating here; I haven't heard anything to indicate this. Again, I would love to hear from anyone with inside information. Jaime ........................................................................... : domain: jds@cs.umd.edu James da Silva : path: uunet!mimsy!jds Systems Design & Analysis Group
shawn@jpradley.jpr.com (Shawn Blair) (06/28/91)
In article <1991Jun12.145309.29231@ukpoit.co.uk> paul@ukpoit.co.uk (Paul Wood) writes: >In article <AWB.91Jun11091405@stoat.uk.ac.ed.aipna> awb@ed.ac.uk (Alan W Black) writes: >>As someone else said BSD will be available >>"soon" for almost free. (Someone disagreed but I'm not sure why -- the >>"soon" bit is the most arguable). >My guess is the "almost free" part is the problem. I understand that it >is not going to be free to private individuals. In fact is going to be >very expensive compared to Minix. Can someone clarify? > I recall seeing a small article announcing the new BSD version in this month's Unix World. If memory serves it mentioned a cost of $900 for the source with discounts for educational institutions. Availability of PC versions were due in September. Correct me if I'm wrong ;-). Regards, Shawn _______________________________________________________________________________ -- Shawn R. Blair ~ shawn@jpr.com -- -- ...murphy!jpradley!rpmc!{srblair!}shawn (hourly) -- -- uunet!areyes!admiral!rpmc!{srblair!}shawn (weekly) --