coleman@CS.UCLA.EDU (12/05/87)
I am currently in the market for a machine of my own and I have run across ads for UNIX/PCs and 3b1's for what I would consider to be tremendous prices. <$1300 for the PC and <$2500 for the 3b1 w/80M. I realize that these machines must be being dumped, but owning an orphan does not necessarily discourage me if they are otherwise good machines. Can anyone post or send information with regard to the following concerns? o Are they reliable? o How standard is the UNIX they run? o How fast are they compared to say, a 10MHZ AT o Is there any kind of memory limit? For instance, the 3b1 mentioned above comes w/2M. How much virtual memory can I get? o Are there technical references around which would allow me to write programs making use of the graphics system, etc.? o Do you think the machines (esp. 3b1) are worth the prices quoted above? If you had your choice, would you buy one (keeping in mind that that kind of money would buy a 12MHZ AT clone w/EGA and goodies. I really appreciate any help you can give me on this! Mike Coleman coleman@cs.ucla.edu
rcj@clyde.UUCP (12/06/87)
In article <9691@shemp.UCLA.EDU> coleman@CS.UCLA.EDU () writes: }o How standard is the UNIX they run? SVR3 (essentially); they follow the System V standards quite well from what I've seen. }o How fast are they compared to say, a 10MHZ AT We are using ours to offload register-level microsimulations from our poor overworked Vaxen. We ran tests on an 11/780 with only 3 users on it doing mail and editors and the 3B1 outperformed the 780!! }o Is there any kind of memory limit? For instance, the 3b1 mentioned } above comes w/2M. How much virtual memory can I get? 4Mb is as high as you can go, but it usually works out much cheaper to settle for one combo board (EIA) and just go for a total of 3.5 Mb. }o Are there technical references around which would allow me to write } programs making use of the graphics system, etc.? See an earlier article in this newsgroup on "3B1 HW". }o Do you think the machines (esp. 3b1) are worth the prices quoted } above? If you had your choice, would you buy one (keeping in mind } that that kind of money would buy a 12MHZ AT clone w/EGA and } goodies. We came, we saw, we bought (and we went outside the company for it; we could have bought anything; didn't have to be AT&T). But we weren't looking for graphics stuff; we want it as a workhorse. I was able to get it completely up and running, talking to our Vax over the LAN, running emacs, etc. etc. in one short afternoon. We've been very pleased so far and will probably purchase several more. Check misc.forsale for used ones; some good deals posted there. The MAD Programmer -- 201-386-6409 (Cornet 232) ^^^^ new extension alias: Curtis Jackson ...![ ihnp4 ulysses cbosgd allegra ]!moss!rcj ...![ ihnp4 cbosgd akgua watmath ]!clyde!rcj
kathy@bakerst.UUCP (Kathy Vincent) (12/10/87)
In article <18017@clyde.ATT.COM> rcj@moss.UUCP (Curtis Jackson) writes: >In article <9691@shemp.UCLA.EDU> coleman@CS.UCLA.EDU () writes: >}o How standard is the UNIX they run? > >SVR3 (essentially); they follow the System V standards quite well from >what I've seen. I believe that, if you check again, you'll find that the UNIX pc does not, in fact, run SVR3. It doesn't run real SVR2, either. It runs a sort of advanced form of SVR1. The latest version of the UNIX pc's UNIX is 3.51 - but that 3 doesn't relate to the 3 in SVR3. Kathy Vincent ------> {ihnp4|mtune|codas|ptsfa}!bakerst!kathy ------> { favourite AT&T gateway }!wruxe!unix
alex@umbc3.UMD.EDU (Alex S. Crain) (12/10/87)
In article <997@bakerst.UUCP> kathy@bakerst.UUCP (Kathy Vincent) writes: >In article <18017@clyde.ATT.COM> rcj@moss.UUCP (Curtis Jackson) writes: >>In article <9691@shemp.UCLA.EDU> coleman@CS.UCLA.EDU () writes: >>}o How standard is the UNIX they run? >> >>SVR3 (essentially); they follow the System V standards quite well from >>what I've seen. > >I believe that, if you check again, you'll find that the UNIX pc >does not, in fact, run SVR3. It doesn't run real SVR2, either. would you care to elaborate on this? Version 3.51 (as opposed to RELEASE 3) is supposed to pass SVVD for Sys5r2, and some of the optional network hacks in the kernal look to me to at least resemble Sys5r3 code, along with shared libs & loadable device-drivers, neither of which are Sys5r[012]. I know that there are some differences and some 3b1 specific goodies/features/bugs, but I don't know what they are exactly except that the optimizer doesn't work (well, sort of), and some commands are missing. Has anyone played with the Shared-libs much? Like getting around the duplicate symbol problems w/ curses? I would like to add code to the shlib. Has anyone done this? whats the format (or is their one? is it raw object code catted together?) I know about the symbol addresses in shlib.ifile, they stop ~128K past the 4meg mark, which leaves 375K of the 512K allocated. I would be interested in using this memory if I could have it. (Brant - have you considered putting most of emacs up there?) :alex. alex@umbc3.umd.edu
csf@mtunb.ATT.COM (C. Furchner) (12/10/87)
In article <997@bakerst.UUCP> kathy@bakerst.UUCP (Kathy Vincent) writes: >In article <18017@clyde.ATT.COM> rcj@moss.UUCP (Curtis Jackson) writes: >>In article <9691@shemp.UCLA.EDU> coleman@CS.UCLA.EDU () writes: >>}o How standard is the UNIX they run? >> >>SVR3 (essentially); they follow the System V standards quite well from >>what I've seen. > >I believe that, if you check again, you'll find that the UNIX pc >does not, in fact, run SVR3. It doesn't run real SVR2, either. >It runs a sort of advanced form of SVR1. The latest version of >the UNIX pc's UNIX is 3.51 - but that 3 doesn't relate to the 3 >in SVR3. Versions 3.5 and 3.51 complies with the UNIX System V Interface Definition for System V, Release 2 (SVR2) in UNIX commands, utilities, and libraries. They have some SVR3 features, such as support of flexnames in the C compiler. Note that "complies with the interface definition" means that it meets the inteface specification; it does not mean "identical." Versions 2.0 and 3.0 were essentially ports of a subset of SVR1. Version numbers in UNIX PC software have nothing to do with UNIX operating system version numbers, because the UNIX PC software includes many modules, like user agent, window manager, terminal emulator, phone manager, etc., that are not part of a standard UNIX operating system. -- C. Furchner ...ihnp4!mtune!mtunb!csf
kathy@bakerst.UUCP (Kathy Vincent) (12/15/87)
In article <633@umbc3.UMD.EDU> alex@umbc3.UMD.EDU (Alex S. Crain) writes: } In article <997@bakerst.UUCP> kathy@bakerst.UUCP (Kathy Vincent) writes: } >In article <18017@clyde.ATT.COM> rcj@moss.UUCP (Curtis Jackson) writes: } >>In article <9691@shemp.UCLA.EDU> coleman@CS.UCLA.EDU () writes: } >>}o How standard is the UNIX they run? } >> } >>SVR3 (essentially); they follow the System V standards quite well from } >>what I've seen. } > } >I believe that, if you check again, you'll find that the UNIX pc } >does not, in fact, run SVR3. It doesn't run real SVR2, either. } } would you care to elaborate on this? Version 3.51 (as opposed to } RELEASE 3) is supposed to pass SVVD for Sys5r2, and some of the optional | SVID See Carol Furchner's followup to my posting. I stand mostly corrected - about the SVR*2* part. Kathy Vincent ------> {ihnp4|mtune|codas|ptsfa}!bakerst!kathy ------> { favourite AT&T gateway }!wruxe!unix
dca@toylnd.UUCP (David C. Albrecht) (12/17/87)
> I believe that, if you check again, you'll find that the UNIX pc > does not, in fact, run SVR3. It doesn't run real SVR2, either. > It runs a sort of advanced form of SVR1. The latest version of > the UNIX pc's UNIX is 3.51 - but that 3 doesn't relate to the 3 > in SVR3. Enough speculation. According to my User Manual Vol II (and I quote). "Version 3.5 UNIX software passes SVVS for System V Release 2" The are a number of system maintenance commands in Rel 2 that the Unix PC doesn't have. As I don't have them I can't really speculate on what use they might be. More important is what user commands it supports and library routines it may be missing. Under user commands some of the commands don't support options available in Rel 2 I'm not going to try to exhaustively list them. The commands missing entirely, however, are: acctom, at, bs, calender, ct, ctrace, efl, f77, fsplit, gdev, ged, graph, graphics, greek, hpio, machid, mailx, news, nscstat, nsctorje, nusend, pg, ratfor, rjestat, sag, sar, send, sno, stat, stlogin, ststat, timex, toc, tplot, tput, trenter, troff, osend. System calls: No missing calls. Library calls: Basically, no fortran library. Missing from the c library: x25alnk, x25clnk, x25hlnk, x25ipvc Missing from the specialized libraries: ldgetname, plot This is a condensation of the rather exhaustive list given the front of the User's Manual vol II. Yes there are many differences between the Unix PC software and that in Rel 2. Do these differences cause problems with software meant for Rel 2? Not that I've run across. That's what's important isn't it. The rest is just picking nits. As for the missing user commands I haven't particularly missed them either. David Albrecht
stevens@hsi.UUCP (Richard Stevens) (12/18/87)
In article <192@toylnd.UUCP>, dca@toylnd.UUCP (David C. Albrecht) writes: > Missing from the specialized libraries: > ldgetname, plot > ^^^^ Actually someone at AT&T forgot to remove the plot(3) library from the 3b1 distribution. However, they didn't give out the man page, and most important, they don't have any plot(1) drivers. I'm working on a plot(1) driver for the 3b1's bit-mapped display, that I'll gladly post after the first of the year, if there is any interest. Does anyone know of a public domain version of the good-old Version 7 graph(1) program, as that along with a plot(1) for the 3b1 would be an easy way to generate fast plots. Richard Stevens Health Systems International, New Haven, CT { uunet | ihnp4 } ! hsi ! stevens
rich@oxtrap.UUCP (K. Richard Magill) (12/24/87)
In article <1146@mtunb.ATT.COM> csf@mtunb.UUCP (C. Furchner) writes: >In article <997@bakerst.UUCP> kathy@bakerst.UUCP (Kathy Vincent) writes: >>In article <18017@clyde.ATT.COM> rcj@moss.UUCP (Curtis Jackson) writes: >>>In article <9691@shemp.UCLA.EDU> coleman@CS.UCLA.EDU () writes: [ Lots about whether the 3b1 is really sV. The bottom line is that there is no System V unless you mean the one from at&t and you better give a date or we still won't know which one you mean. 3b1 Unix was written by Convergent Technologies. The C compiler is the most obvious clue. In other words saying system V is about as specific as saying UNIX. Agreed, they are a standard of sorts but there are *so* many of them. Like, SVID, SVVS, sV, sV.0, sV.1, sV1.0, etc, etc, and of course the 3b2 version of each of these is different from the 3b5 version is different from the....
ford@crash.cts.com (Michael Ditto) (12/27/87)
In article <2420@oxtrap.UUCP> rich@oxtrap.UUCP (K. Richard Magill) writes: > >3b1 Unix was written by Convergent Technologies. The C compiler is >the most obvious clue. Ok, this thread is getting pretty ridiculous... Now Unix was written by Convergent Technologies? The Unix PC comes with Genuine, Real-Live, AT&T Unix System V developed by AT&T/Bell Labs, and decended from code written by Kernighan, Ritchie, Thompson, and Bourne, themselves. It is a port of "System V Release 2" ported by people at Convergent Technologies, which is also where the hardware was designed. It also has some minor extentions from both System V Release 3 and Berkely Unix. For example, from SVR3 comes the "xt" multiplexing protocol and "layers" windowing system, and internal support for the "Streams" I/O system. From BSD comes the Berkeley network socket capability, although this is only included with the Ethernet software. In all normally visible aspects of the system, however, there are no significant deviations from System V Release 2. The C Compiler, assembler, and loader are the standard 68010 tools from the "SGS" (Software Generation System) from AT&T. They are essentially identical to the versions that would be found on any 68000 Unix System V (For example, the CounterPoint Computers 68020 systems, and the HP-350 68020 systems). There are some additions to the Unix PC's SVR2 made especially for the Unix PC: Loadable Device Drivers and the console windows. Loadable device drivers allow driver installation into a running kernel. The console windows are supported by a device driver in the kernel, and a few system programs have been modified to integrate well with the multiple windows. Then, on top of what is essentially a generic Unix box, you have the "User Agent" system, with the windows, menus, and automated system maintenence. This layer was implemented, in my opinion, in a very good way that is consistent with the design of Unix; most importantly it IS a "layer" of software that a user can choose to ignore. When the user agent is not used, what is left is a very complete port of System V Release 2 with a few extras thrown in. There are a very few things missing; these are described in the user's manuals as a command-by-command comparison between the Unix PC and the System V Interface Definition for SVR2. Someone on this group has already itemized the notable differences; the only one that has bothered me is the ommission of the "crash(1M)" command. In summary, I make very extensive use of my 3B1, I am used to System V release 2 and 3, and have not been disappointed in the functionality of the Unix of the Unix PC. I do wish it were a little more reliable (in terms of frequency of inexplicable crashes, etc.). -- Mike Ditto -=] Ford [=- P.O. Box 1721 ford%kenobi@crash.CTS.COM Bonita, CA 92002 ford@crash.CTS.COM
emike@richp1.UUCP (E. Mike Durbin) (12/29/87)
>The bottom line is that there is no System V unless you mean the one >from at&t and you better give a date or we still won't know which one >you mean. > ... >In other words saying system V is about as specific as saying UNIX. >Agreed, they are a standard of sorts but there are *so* many of them. >Like, SVID, SVVS, sV, sV.0, sV.1, sV1.0, etc, etc, and of course the >3b2 version of each of these is different from the 3b5 version is >different from the.... How true... Except for SVR3, which is as close to the same as you can get on any machine you find it on. (Or, at least it is now, 'till someone breaks it too :-). Lets see, I need 2 more lines to get this out. one more ok. -- ------------------------------------------------------------------------------ ...!cuuxb \ E. M. Durbin !richp1!emike Rich Inc. ...!ihnp4!laidbak!spl1 / Chicago
denny@mcmi.UUCP (Dennis Page) (12/29/87)
In article <2208@crash.cts.com> ford%kenobi@crash.CTS.COM (Michael Ditto) writes: >In all normally visible aspects of the system, however, there are no >significant deviations from System V Release 2. > >The C Compiler, assembler, and loader are the standard 68010 tools from the >"SGS" (Software Generation System) from AT&T. They are essentially identical >to the versions that would be found on any 68000 Unix System V (For example, >the CounterPoint Computers 68020 systems, and the HP-350 68020 systems). > >[...] When the user agent >is not used, what is left is a very complete port of System V Release 2 >with a few extras thrown in. Woa, woa, let's not get carried away here... The HP-350, while SVID, is based on Berkley. And it still comes a damn site closer to looking like a VR2 machine than the unix-pc does. No significant deviations? What about cron? Madre! -- Denny Page Martha, the Clones are loose again!
allbery@axcess.UUCP (Brandon S. Allbery) (12/31/87)
In article <2208@crash.cts.com>, ford@crash.cts.com (Michael Ditto) writes: +--------------- | In all normally visible aspects of the system, however, there are no | significant deviations from System V Release 2. > ... | "layer" of software that a user can choose to ignore. When the user agent | is not used, what is left is a very complete port of System V Release 2 | with a few extras thrown in. +--------------- I've had no problems with reliability. I *do* have a problem with the sorry excuse for UUCP and *no* way to change it (are you listening, AT&T? You could at least consider *selling* *binaries*, since everyone knows you have them!). And it'd be nice if they'd provided the new cron and left that functionality out of the Status Manager. WHY does /etc/smgr do cron's work?! It means that I can't even install the replacement cron from comp.sources.unix without wasting cycles on two different cron programs! -- ___ ________________, Brandon S. Allbery cbosgd \ ' \/ __ __, __, aXcess Company mandrill| __ | /__> <__ <__ 6615 Center St. #A1-105 !ncoast! / ` | \__. .__> .__> Mentor, OH 44060-4101 necntc | axcess!allbery \___/\________________. Moderator, comp.sources.misc hoptoad/
vanam@pttesac.UUCP (Marnix van Ammers) (01/03/88)
In article <125@axcess.UUCP> allbery@axcess.UUCP (Brandon S. Allbery) writes: >In article <2208@crash.cts.com>, ford@crash.cts.com (Michael Ditto) writes: >+--------------- > >I've had no problems with reliability. I *do* have a problem with the sorry >excuse for UUCP and *no* way to change it (are you listening, AT&T? You could >at least consider *selling* *binaries*, since everyone knows you have them!). Well, I'm finally going to say it: I *do* have problems with 3B1 reliability! When the fire sale came out I told all my friends and colleagues about it and about what a great machine the UNIX-PC is. I had had one on my desk for about 6 months and I had had no trouble with it whatsoever. Well, now that I finally bought one for personal use, I am very disappointed with it's reliability. I received my 3B1 around the beginning of August. AT&T is going to be coming out for the *FOURTH* time Monday (I hope -- I made my repair call Dec 31st). My first trouble was a horrible squawking sound, which I presumed to be the hard drive because I couldn't think of anything else that could make such a horrible sound. I figured that since the hard disk was the only thing moving, it must be the source of that horrible sound. So I had AT&T come out and change it. Well, the squawking sound continued, and after some articles on net news about squawking sounds, I figured out that the noise was coming from my speaker and had to do with incoming calls to my modem. Then my 3B1 started crashing with "PANIC: kernel addressing error" messages and my hard disk got trashed with hundreds of bad blocks. So they changed my mother board. The next thing was that my hard disk *did* start making a funny noise (kind of like a bearing going out). So AT&T came out and changed my hard disk again (this time for good reason). Then a few days ago, my 3B1 died. The fan is working, the hard disk is turning, but the screen remains dark and the system won't boot (not even off a floppy). I finally asked about getting the whole damn system replaced. The tech I talked to said I could try for that after they exchange my mother board one more time. I am *very* tired of having a flaky system. I'd like to know just what the repair call rate is for these units. Thank God I still have my Amiga. Marnix ----
rjg@sialis.mn.org (Robert J. Granvin) (01/04/88)
>>I've had no problems with reliability. I *do* have a problem with the sorry >>excuse for UUCP and *no* way to change it (are you listening, AT&T? You could >>at least consider *selling* *binaries*, since everyone knows you have them!). > >Well, I'm finally going to say it: I *do* have problems with 3B1 >reliability! > > [Maintenance horror story deleted :-)] Oh, but be very very very happy that you _are_ getting service. I'm currently in a battle with AT&T service for a wide variety of servicing problems. As it turns out, they seems to have "lost" or never recorded several Hotline calls. Now they tell me that my machine is out of warranty, and I have to pay for the repairs. (A tech wants to replace my mother board. $2,000. I refuse to pay that for something I reported in the first month of operation, and reported several times since. Unfortunately, the mother board may not be the final problem. Mine may be related to the Power Supply as well. And I won't even mention the disk whining as has been previously mentioned.. :-) My only guess is that it the Hotline techs did not record the problems because there was no 'physical resolution'. Every item that I have reported and physically received something because of it, has been recorded, but everything else is missing. Strange, now, isn't it? My personal suggestion is that if you try to resolve something through the Hotline, record the Ticket number, record the Tech's name, and all questions and resolutions. Even so, this may not help. The Hotline will not respond a warranty call to an out of warranty machine unless _they_ have recorded the problem in their database during the warranty period. So you should at least ask and make sure about that too. (Unless you are willing to pay for it, of course...) Maybe I'm just having bad luck. :-) In any case, I have to say that I'm pleased with the machine, and normally pleased with the Hotline, and thrilled with the service once I get it. However, this last little blast is going to go on for quite some while I can feel... Sigh. (P.S. For those AT&T folx that I've chatted with via email and other sources, you are no way included in this. Thank you for all your help and comments! I'm looking forward to more ... :-) If anyone out there has any comments or followup regarding this specifically, I'd much prefer an email response. If there is enough of an interest, I'd of course post a followup. I don't want to be accused of being the cause of starting comp.sys.att.hotline! :-) -- _____________________________________________________________________________ | Robert J. Granvin | INTERNET: rjg@sialis.mn.org | | | 2701 West 43rd Street | UUCP: ...ihnp4!meccts!sialis!rjg | "Whoops!!" | | Minneapolis, MN 55410 | ...uunet!rosevax!ems!sialis!rjg |________________|
wtm@neoucom.UUCP (Bill Mayhew) (01/04/88)
Hi, As Robert mentioned in his article, it is a good idea to ask for the "ticket number" each time you call the hotline with a bug. Also keep track of when you called the hotline. That way, you have something to go back to them with when they say that you never called while the machine was in warranty. Unfortunately, I almost learned too late about ticket numbers. My introduction was only recently when the machine crashed on a Saturday and I called to log the occurance. The person gave me the number and asked me to call back Monday and give the ticket number. Curiously, AT&T blamed my motherboard for the crashes I've been getting from wmgr blowing up. They tried to blame my modem, but I persisted and finally won. I did point out that it was the internal modem that liked to hang, not my external modem. The hotline person wanted to blame the external modem for toggling the CTS lead too often, but I pointed out that I had crashes when the modem hadn't been doing anything for over three hours. It is interesting that the normal problem resolution modality is to toss in a new motherboard. This is despite the fact that four or five other people on the Net running 3.5.1 are having problems similar to mine. Oh well, I guess the electronics are cheaper than expending human time -- obviously true for one person having a problem. When the hotline finally convinced itself that my motherboard was to blame, they said that a person would be here tomorrow to replace it. That is pretty decent response. It's just that working up to the response is the laborious part. So, all-in-all, AT&T service is pretty good, it is just convincing them that is the difficult part. If you'd ever had to suffer through Digital's (DEC's) so-called service, then you'd realize that AT&T's service really is nice. Of course getting service for IBM PCs is virtually impossible from IBM; they make the dealer channel do everything -- and I've seen stunningly incompetent dealer attempts at service. --Bill
allbery@axcess.UUCP (Brandon S. Allbery) (01/05/88)
In article <75@sialis.mn.org>, rjg@sialis.mn.org (Robert J. Granvin) writes: > final problem. Mine may be related to the Power Supply as well. And > I won't even mention the disk whining as has been previously > mentioned.. :-) I seem to remember that one brand of disk drive (Miniscribe?) has a rather badly designed static-draining tab: it tends to start squeaking. The maker's recommended fix is to CAREFULLY cut it off or bend it out of the way; the drive works fine without it. -- ___ ________________, Brandon S. Allbery cbosgd \ ' \/ __ __, __, aXcess Company mandrill| __ | /__> <__ <__ 6615 Center St. #A1-105 !ncoast! / ` | \__. .__> .__> Mentor, OH 44060-4101 necntc | axcess!allbery \___/\________________. Moderator, comp.sources.misc hoptoad/
erict@flatline.UUCP (eric townsend) (01/05/88)
In article <655@pttesac.UUCP>, vanam@pttesac.UUCP (Marnix van Ammers) writes: > Well, I'm finally going to say it: I *do* have problems with 3B1 > reliability! > I received my 3B1 around the beginning of August. AT&T is going to > be coming out for the *FOURTH* time Monday (I hope -- I made my repair > call Dec 31st). > [explains how AT&T Repair Clowns have show up 3 times already. Hey, > there *is* a commercial with an AT&T Clown in it. I can't resist.] The one time we've had a non-power supply problem they just replaced the whole machine. > I am *very* tired of having a flaky system. I'd like to know just > what the repair call rate is for these units. I would too, but I think it's higher than you think. We have 4 7300's and 4 3b1's at work, being used for development, and have only had to call the repairman *twice* in two years: 1 bad power supply and one hard drive that forgot what type it was -- they replaced the p/s on the first one, and the whole machine on the second one (they weren't sure if it was the hard drive or not.) I think that's a pretty good record for eight medium-high use machines... -- J. Eric Townsend ->uunet!nuchat!flatline!erict smail:511Parker#2,Hstn,Tx,77007 'Girls play with toys. Real women skate.' --Powell Peralta ad.| 'Hey, watch I disclaim all responsibility for others' ignorances. | me ollie this <whump>'
wtm@neoucom.UUCP (Bill Mayhew) (01/06/88)
I'd like to thank the people who are helping out at AT&T. Here is a little more information I've gotten. The PC=1FC0 (that number is from my memory, so you better not quote me) that was in the kernel addr fault panic register dump I got probably really was due to a motherboard oddity not directly related to the crash. I called the hotline on Monday and had the new motherboard Tuesday morning. Nice job!! I could have had the new board Monday, but I didn't want to take time off from work to go home and let the guy in to fix the 3b. By the way, I got a good chance to look over the hardware carefully while the machine was apart. The construction is pretty good. There are two jumpers on the motherboard- that's all. The power supply looks decent. I'd have to give at least a B+ on the quality of the electronics. The loser is the plastic case, which is made pretty chintzy. I'll give it a C-. By the way, doing a motherboard swap to replace the clock battery is reasonable, since the old motherboard will go back for bench service and get a new battery. Since board is good, it'll go back in the field the next time a battery fails somewhere else. At least, you'll be receiving a thoroughly tested board :-). Since the board has to come out to swap the battery, it might as well go in for a bench going-over. What would have been smarter would be to make the battery s plug-in type to save some work. There sure are a lot of screws holding the motherboard in! But-- it appears that hardware was not [the only] reson for the crashing. The last crash left uucico hung on ph1. In fact I logged in with the intent of seeing who was on my machine at 8:00 AM when nobody should have been on. It turned out that a uucp transfer had completed normally at about 2:50 AM and left the port hung since then. AT&T told me that they are beginning to suspect that there are some bugs in the serial/phone port driver(s) in the kernel. Apparently, the driver starts generating interrupts like mad if it receives a longbreak signal just at the end of a conversation. This apparently has something to do with the fact that the internal modem and serial port share some functions on the same chip. Since the serial port and modem interrupts are higher priority than the window stuff, the output locks up. Hitting a key does let windows get control for an instant and it will get a few characters out while it services the keyboard. Eventaully the line input buffer overflows or something like that, and the system goes out to lunch. The got-cha is that there is a good chance that the disconnect from another system that has called you is somewhat likely to generate a few trash characters that are perfect for inciting the lock-up. Looks like a temporary work-around would be to never let anybody call *in to* your 3b1. (If you are using 3.51) --Bill
david@ms.uky.edu (David Herron -- Resident E-mail Hack) (01/07/88)
In article <912@neoucom.UUCP> wtm@neoucom.UUCP (Bill Mayhew) writes: >By the way, doing a motherboard swap to replace the clock battery >is reasonable, since the old motherboard will go back for bench >service and get a new battery. Since board is good, it'll go back >in the field the next time a battery fails somewhere else. At >least, you'll be receiving a thoroughly tested board :-). Since >the board has to come out to swap the battery, it might as well go >in for a bench going-over. What would have been smarter would be >to make the battery s plug-in type to save some work. There sure >are a lot of screws holding the motherboard in! I have to STRONGLY disagree. What about after the machine is off of warranty and they want many hundreds of dollars to replace the motherboard? And don't tell me to get a service contract 'cause that's something like $50 per month ... gag! As for your crashes ... I've experienced a few crashes of that sort and have the details written down at home. I haven't called the hotline over 'em yet. All of my crashes have happened while something happening on the modem was in the process of finishing up. (i.e. a uucico or ATE) -- <---- David Herron -- The E-Mail guy <david@ms.uky.edu> <---- or: {rutgers,uunet,cbosgd}!ukma!david, david@UKMA.BITNET <---- <---- Winter health warning: Remember, don't eat the yellow snow!
allbery@axcess.UUCP (Brandon S. Allbery) (01/07/88)
In article <912@neoucom.UUCP>, wtm@neoucom.UUCP (Bill Mayhew) writes: +--------------- | then. AT&T told me that they are beginning to suspect that there | are some bugs in the serial/phone port driver(s) in the kernel. | Apparently, the driver starts generating interrupts like mad if it | receives a longbreak signal just at the end of a conversation. | | Looks like a temporary work-around would be to never let anybody | call *in to* your 3b1. (If you are using 3.51) +--------------- Hmmm. I have yet to allow dialins on my 3b1 (too little security), yet I have had one crash and some near misses. Maybe it's a combination of bugs? Certainly the printer on my 3b1 SEEMS to be the cause.... -- ___ ________________, Brandon S. Allbery cbosgd \ ' \/ __ __, __, aXcess Company mandrill| __ | /__> <__ <__ 6615 Center St. #A1-105 !ncoast! / ` | \__. .__> .__> Mentor, OH 44060-4101 necntc | axcess!allbery \___/\________________. Moderator, comp.sources.misc hoptoad/
wtm@neoucom.UUCP (Bill Mayhew) (01/08/88)
getting at the anti-static tab is a pain in the ... um, neck-- but it it is the source of drive squeal. The anti-static tab is a piece of copper with a carbon button that rubs on the bottom spindle of a hard drive. The idea is that the spinning platters can act like Vandergraf generators and discharge to ground through the head carriage. Not exactly a nice situation. The anti-static tab is supposed to prevent a charge build-up. The tab is usually soldered to the drive electronics PC board. As mentioned previously it can be removed. I've never heard of anything nasty happening, but you never know. The option I've used has been to stick a pice of electrical tape on the copper. It usually is enough to keep the tab from resonating. --Bill
rjg@sialis.mn.org (Robert J. Granvin) (01/09/88)
>I have to STRONGLY disagree. What about after the machine is off >of warranty and they want many hundreds of dollars to replace the >motherboard? $2,000 for the part. This price was just quoted to me by the technician who wanted to swap my motherboard to see if it would resolve my problem. We told him to go home for now. >And don't tell me to get a service contract 'cause that's something >like $50 per month ... gag! Yep. But seeing the prices, and how often I've already used the hotline, I know I'm gonna send in a check real soon now... sigh... one use of that $50/month can really really pay for itself in a hurry, even if you do have a perfectly happy and purring machine. >As for your crashes ... I've experienced a few crashes of that sort >and have the details written down at home. I haven't called the >hotline over 'em yet. All of my crashes have happened while something >happening on the modem was in the process of finishing up. (i.e. >a uucico or ATE) I've been reporting this or nearly identical problems to the hotline for months (as mentioned before). I am _still_ in a hassle with them over warranty service. This will continue, I know it, for some time. At least I am glad to see that _other_ people are finally reporting these problems. Maybe some sort of reasonable result will come of it someday. I shouldn't pick exclusively on the Hotline. When they come through, they come through like a charm. And I've discovered that they are apparently quite often _not_ supplied with the information they deserve. They reply "We've never heard of that problem before..." which is true, but many other internal areas have known about it for months or longer, but the information and solutions just don't always get to the front lines (like, where the users and purchasers become involved, and ultimately irritated). I know... I should support my claims... :-) -- ___________________________________________________________ o/` o/` o/` | Robert J. Granvin | INTERNET: rjg@sialis.mn.org | The hills are | 2701 West 43rd Street | UUCP: ...ihnp4!meccts!sialis!rjg | alive, and they | Minneapolis, MN 55410 | ...uunet!rosevax!ems!sialis!rjg | ate my mother...
rich@jolnet.UUCP (Rich Andrews) (01/09/88)
In article <916@neoucom.UUCP>, wtm@neoucom.UUCP (Bill Mayhew) writes: > getting at the anti-static tab is a pain in the ... um, neck-- but > it it is the source of drive squeal. The anti-static tab is a piece > of copper with a carbon button that rubs on the bottom spindle of > a hard drive. The idea is that the spinning platters can act like > Vandergraf generators and discharge to ground through the head > carriage. Not exactly a nice situation. The anti-static tab is > supposed to prevent a charge build-up. The tab is usually soldered > to the drive electronics PC board. > > --Bill Nasty things can happen ( and with my luck they usually do!) and when you remove the brush from the drive you are asking for trouble. Shortly after i got my 3b2/400, the drives started to squeal like a nest of crickets I took the drives out, took them apart, and lubricated the carbon brush with some moly grease. Since then no more squeal. I have heard of lots of bad things that have happen to drives that did not have the brush contacting the spindle motor. rich
rich@oxtrap.UUCP (K. Richard Magill) (01/09/88)
In article <2208@crash.cts.com> ford%kenobi@crash.CTS.COM (Michael Ditto) writes: >In article <2420@oxtrap.UUCP> rich@oxtrap.UUCP (K. Richard Magill) writes: >> >>3b1 Unix was written by Convergent Technologies. The C compiler is >>the most obvious clue. >The C Compiler, assembler, and loader are the standard 68010 tools from the >"SGS" (Software Generation System) from AT&T. They are essentially identical >to the versions that would be found on any 68000 Unix System V (For example, >the CounterPoint Computers 68020 systems, and the HP-350 68020 systems). Well, I've found 4 major bugs in the compiler on the 3b1/7300, aka ct safari, and the same bugs showed up on a burroughs xe550, aka ct megaframe. I do not see them on NCR tower's which look like 3b2 SysV. rich.
wtm@neoucom.UUCP (Bill Mayhew) (01/10/88)
I really like to have uucp from neoucom dial me, as neoucom has an fx line that doesn't cost them a lot to call me here. For me to call them is a toll call. That is why I wanted dial-in. Neoucom also won't forward mail if I poll them because my machine, "impulse" identifies itself as "impuls" to to the 6 character name limit hard coded into uucico (and several other places). Unfortunately, the name for my machine was picked before I was aware that more than six characters was verboten on the PC7300. Well, to make a long story short, after wrangling with the Hotline folks, with me and several other netters, the Hotline finally decided there was a problem in uucico. Somebodyat AT&T finally got around to fixing uucico, and they emailed me a new copy. The new uucico seems (knock on wood) to have fixed things-- or at least it hasn't broken anything additional. The old rendition had a file size of 71344, and the new version is 71372. Looks like only relatively minor brain surgery was necessary. Since the crashes were separated by up to a month, I guess only time is going to tell if this did the trick. Alas, the warranty will be gone by then. --Bill
rbl@nitrex.UUCP ( Dr. Robin Lake ) (02/02/88)
In article <2519@oxtrap.UUCP> rich@oxtrap.UUCP (K. Richard Magill) writes: >In article <2208@crash.cts.com> ford%kenobi@crash.CTS.COM (Michael Ditto) writes: >>In article <2420@oxtrap.UUCP> rich@oxtrap.UUCP (K. Richard Magill) writes: >>> >>>3b1 Unix was written by Convergent Technologies. The C compiler is >>>the most obvious clue. >>The C Compiler, assembler, and loader are the standard 68010 tools from the >>"SGS" (Software Generation System) from AT&T. They are essentially identical >>to the versions that would be found on any 68000 Unix System V (For example, >>the CounterPoint Computers 68020 systems, and the HP-350 68020 systems). > >Well, I've found 4 major bugs in the compiler on the 3b1/7300, aka ct >safari, and the same bugs showed up on a burroughs xe550, aka ct >megaframe. I do not see them on NCR tower's which look like 3b2 SysV. > > > >rich. Once upon a time, about 3 years ago, when we had a Motorola-CT 6300 on loan, we noted that the AT&T PC 7300 and Motorola 6300 were BINARY COMPATIBLE! Yep, stuff that wasn't even available (DBMS, word processing, etc.) on one could be directly executed on the other. A net note to this effect brought a reply from John Mashey (ex-Convergent) confirming that the biggest change was to have put some of the IC circuitry into PLAs to compact the board space. I have not tried the compatibility with the Motorola 6600 (CT MegaFrame) as the separate I/O processors seem baroque. Someday ..... (Really BP America R&D) -- Rob Lake {decvax,ihnp4!cbosgd}!mandrill!nitrex!rbl
rich@oxtrap.UUCP (K. Richard Magill) (02/07/88)
In article <658@nitrex.UUCP> rbl@nitrex.UUCP ( Dr. Robin Lake ) writes: >In article <2519@oxtrap.UUCP> rich@oxtrap.UUCP (K. Richard Magill) writes: >>In article <2208@crash.cts.com> ford%kenobi@crash.CTS.COM (Michael Ditto) writes: >>>In article <2420@oxtrap.UUCP> rich@oxtrap.UUCP (K. Richard Magill) writes: >>>> >>>>3b1 Unix was written by Convergent Technologies. The C compiler is >>>>the most obvious clue. [stuff deleted] >Once upon a time, about 3 years ago, when we had a Motorola-CT 6300 on loan, >we noted that the AT&T PC 7300 and Motorola 6300 were BINARY COMPATIBLE! [more deletions] >I have not tried the compatibility with the Motorola 6600 (CT MegaFrame) as >the separate I/O processors seem baroque. Someday ..... They work. In case you don't recognize me, I have been rich@cwruecmp, rich@hal, rich@rexago1, krm@cwruecmp, etc. and you once interviewed me. While I was rich@rexago1 I was running a csh from nitrex on 3b1's. I recently had the opportunity to run that same csh on a Burroughs XE550, basically a remarketted MegaFrame. rich. ps, the 3b1 now in my home, sendai, runs a legitimate copy of ksh.