lance@embassy.nsi.com (Lance N. Antrim) (10/15/90)
I am considering installng the 'ufm' upgrade to move my system from 2.3.2 to 2.3.3. However, my system seems to be working fine (except for daily cores in usr/spool/uucp). Having already installed the SLS's for my tape drive and mouse I am not sure whether the benefits outweigh the potential for problems from the upgrade. I would appreciate the experiences of others who have nstalled ufg (which is xnx155b(vols 1 & 2) from SCO's SLS files). thanks, lance (lance@nsi.com)
ronald@robobar.co.uk (Ronald S H Khoo) (10/16/90)
In article <107@embassy.nsi.com> lance@embassy.nsi.com (Lance N. Antrim) writes: > I am not sure whether the benefits outweigh the > potential for problems from the upgrade. [UFM (xnx155b)] The UFM upgrade procedure is pretty painless, the only thing which is slightly inconvenient is that you have to merge your local changes to /etc/termcap into the new one provided. On balance, I'd say, yeah, install it. You get improved COFF binary compatibility (COFF cxenix() calls are now supported) and the new cpio(C) binary has much better multi-volume backup support. The Archive tape driver is also *much* better (if you have an Archive tape drive -- I do :-) MHO: Cost: low. Benefit: medium to high. Conclusion: I'd go for it unless you have a "production" as opposed to a "development" or "hacking" system. -- ronald@robobar.co.uk +44 81 991 1142 (O) +44 71 229 7741 (H)
alex@grian.cps.altadena.ca.us (Alex Pournelle) (10/23/90)
ronald@robobar.co.uk (Ronald S H Khoo) writes: >In article <107@embassy.nsi.com> lance@embassy.nsi.com writes: >> I am not sure whether the benefits outweigh the >> potential for problems from the upgrade. [UFM (xnx155b)] >The UFM upgrade procedure is pretty painless, the only thing >which is slightly inconvenient is that you have to merge your local >changes to /etc/termcap into the new one provided. >On balance, I'd say, yeah, install it. >MHO: Cost: low. Benefit: medium to high. Conclusion: I'd go for it unless >you have a "production" as opposed to a "development" or "hacking" system. I have just installed the XNX155SLS on my computer, and have much to say about it. Pull up a chair, this stuff's important. System in question is an 386SX with 3.5 Megs, a 40 Mbyte MFM drive and an EGA card. Plus 1.2 and 1.44 Mbyte floppy drives. 1) There are some BIOS incompatibilities with this release. I had an old (Aug 89) AMI BIOS on my system and as soon as I hit multi-user mode I would get a "Panic: parity error" on the system. Worked (mostly) fine in single-user, though. When I switched to an 1988 Award BIOS, the system came up multi-user without a problem. Also :-) the mysterious coredumps in Urogue went away. Greg Fores, a really knowledgeable tech. at SCO, tells me that there is a new release of the SLS which probably works better on the older BIOSes. This is stuck in QC and will be released (ahem) Real Soon Now. Meanwhile, one can comment out the command to load the new RAM-checker (oem.o) and use the old (2.3.2) checker. 2) the 3.5" high-density disk support from 2.2.3 to 2.3.2 and later is different. Under the "older" system one would use /dev/fd096ds18 to read 3.5" HD disks. Under 2.3.2 and later, one uses /dev/fd0135ds18. To read older 3.5" HDs, mknod entries for fd096ds18 and rfd096ds18, lInK them to /dev/fd0 and /dev/rfd0 and THEN tar or de-archive. 3) The SCO manual is wrong; use "custom -r /dev/install1" to run custom from the second floppy device, not "custom -r /dev/fd096ds15" or whatever. Make sure you have properly installed the second floppy device. 4) Floppy read-write speed isn't improved any. Sorry. Nor are there any plans to change the floppy drivers (&*^*&^*(&^!!) -- Alex Pournelle, freelance thinker Also: Workman & Associates, Data recovery for PCs, Macs, others ...elroy!grian!alex; BIX: alex; voice: (818) 791-7979 fax: (818) 794-2297 bbs: 791-1013; 8N1 24/12/3
max@mthvax.cs.miami.edu (Max Southall) (10/23/90)
In sympathy with alex@grian.cps.altadena.ca.us: We had a similar problem with BIOS incompatibilities here. What I did was take the sources for a 286 bilingual BIOS we had developed, modified the initialization code and added some stuff for the 386 support chipset, and after a few fights with the documentation, the correct incantation was achieved and it's been off and running. The nice thing is that now it reboots Ok if there's a power failure (Florida is the lightning strike and outage capital of America) as the CMOS checks are more robust to recovery after that sort of thing. So why did AMI die? Still don't know, but homegrown works Real Well Now. -- Max Southall - max@mamia.UUCP "Rationality's failure is its dependance on human reasoning."