diamond@hdchq.UUCP (Bill Diamond) (11/16/89)
We're using a Compaq 386/33 with 14 MB ram, and a 650 MB fixed drive with the DTP caching controller with 4.5 MB caching RAM on board, and the SunRiver fiber optic workstations. On the whole, we're pleased with ISC's support and assistance with problems, but on problem is terribly vexing for us. We know that ISC is working on it, but we're curious whether other organizations have experienced this problem, and whether they've found a work-around. The problem is when users are in VP/ix DOS emulation, whether using a particular DOS program or just at the DOS shell, on both the console and the SunRiver stations, the keyboard seems to "flip out". The caps lock locks in, and we cannot reset the keyboard. All keys pressed once this keyboard flip-flop occurs generate an equivalent scan code as if the <Shift> were pressed along with the single keystroke. ISC engineers have been forthright in telling us that this problem is number one on their list, but, until a permanent fix is available, I have quite a few anxious users dead in the water. Any ideas? Should there be sufficient interest, I will summarize for the net.
patrick@scrooge.ism.isc.com (Patrick Curran) (11/17/89)
In article <240@hdchq.UUCP> diamond@hdchq.UUCP (Bill Diamond) writes: > >The problem is when users are in VP/ix DOS emulation, whether using a >particular DOS program or just at the DOS shell, on both the console and >the SunRiver stations, the keyboard seems to "flip out". The caps lock >locks in, and we cannot reset the keyboard. All keys pressed once this >keyboard flip-flop occurs generate an equivalent scan code as if the ><Shift> were pressed along with the single keystroke. Try unplugging the keyboard and then plugging it back in again. This solves almost all of the "lockup" problems I experience. Patrick Curran (patrick@ism780c.isc.com) (uunet!ism780c!patrick) INTERACTIVE Systems Corp, Santa Monica, CA. (213) 453-8649
diamond@hdchq.UUCP (Bill Diamond) (11/18/89)
In article <36515@ism780c.isc.com>, patrick@scrooge.ism.isc.com (Patrick Curran) writes: > Try unplugging the keyboard and then plugging it back in again. > This solves almost all of the "lockup" problems I experience. > > Patrick Curran (patrick@ism780c.isc.com) > (uunet!ism780c!patrick) > INTERACTIVE Systems Corp, Santa Monica, CA. (213) 453-8649 Normally, I take these hints in the spirit given. In this case, I'm afraid I cannot. This is one of the most inutterably stupid ideas I've ever heard. Why don't you imagine the scenarios with which we, your clients are faced. "No problem Mr. CEO, just try unplugging your keyboard then plug it back in". Uh-huh. I like my job. I want to keep my job. My company demands top performance from me for each dollar. I demand top performance for the several thousands of dollars I have invested in your product. Either get your product right, or admit you can't do it and give us a refund. Just be honest with us; I can accept failure. I cannot and will not accept the continued levels of deception evidenced by the pervasive defects in the software.
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (11/18/89)
In article <290@minnie.UUCP> diamond@hdchq.UUCP (Bill Diamond) writes: | In article <36515@ism780c.isc.com>, patrick@scrooge.ism.isc.com (Patrick Curran) writes: | > Try unplugging the keyboard and then plugging it back in again. | > This solves almost all of the "lockup" problems I experience. | | Normally, I take these hints in the spirit given. In this case, I'm | afraid I cannot. | | This is one of the most inutterably stupid ideas I've ever heard. Why | don't you imagine the scenarios with which we, your clients are faced. "No | problem Mr. CEO, just try unplugging your keyboard then plug it back in". | Uh-huh. I like my job. I want to keep my job. My company demands top | performance from me for each dollar. I demand top performance for the | several thousands of dollars I have invested in your product. | | Either get your product right, or admit you can't do it and give us a | refund. Just be honest with us; I can accept failure. I cannot and will | not accept the continued levels of deception evidenced by the pervasive | defects in the software. When I read the posting I sent a note To Pat saying that he might get flamed for suggesting such action. I had in mind someone who was afraid of melting their keyboard, not this vitriolic attack on a useful suggestion. In my experience this lockup has been caused by problems in the hardware of the keyboard rather than the software, based on having seen a new keyboard cure the problem for one Xenix, two ix/386 and one MicroPort. I used a Northgate keyboard, but I'm sure other brands will do. One vendor supplied a new set of ROMs for the keyboard processor. Unplugging the keyboard resets the controller (or whatever) inside. Before flaming someone for giving you a helpful hint, be sure you understand the problem. I'm not claiming that there is no software problem, only that I haven't seen it, and I *have* seen lockup in several flavors of UNIX which responded to hardware fixes. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) "The world is filled with fools. They blindly follow their so-called 'reason' in the face of the church and common sense. Any fool can see that the world is flat!" - anon
dpi@loft386.UUCP (Doug Ingraham) (11/19/89)
In article <36515@ism780c.isc.com>, patrick@scrooge.ism.isc.com (Patrick Curran) writes: > In article <240@hdchq.UUCP> diamond@hdchq.UUCP (Bill Diamond) writes: > > > >The problem is when users are in VP/ix DOS emulation, whether using a > >particular DOS program or just at the DOS shell, on both the console and > >the SunRiver stations, the keyboard seems to "flip out". The caps lock > >locks in, and we cannot reset the keyboard. All keys pressed once this > >keyboard flip-flop occurs generate an equivalent scan code as if the > ><Shift> were pressed along with the single keystroke. > > Try unplugging the keyboard and then plugging it back in again. > This solves almost all of the "lockup" problems I experience. > > Patrick Curran (patrick@ism780c.isc.com) > (uunet!ism780c!patrick) > INTERACTIVE Systems Corp, Santa Monica, CA. (213) 453-8649 Don't do this! Your keyboard and or computer can be damaged by unplugging and re-plugging with the power on. They were not designed to allow this. I am surprised that anyone would recommend as crude a solution to any problem as this. I don't know how VP/ix treates the keyboard hardware, but I have written a correct interrupt handler for the keyboard hardware under DOS and it is not easy to get everything right. I doubt that many handlers for the keyboard make use of the error detection/correction sequences mostly because it is poorly documented and it seldom would make a difference. The keyboard software could be robust. I am guessing that contacting the correct people at SunRiver and Interactive would get this cleared up quickly. If others have this kind of problem, then its software. If you are alone, I would bet its hardware. -- Doug Ingraham (SysAdmin) Lofty Pursuits (Public Access for Rapid City SD USA) uunet!loft386!dpi
pcg@emerald.cs.aber.ac.uk (Piercarlo Grandi) (11/20/89)
In article <290@minnie.UUCP> diamond@hdchq.UUCP (Bill Diamond) writes:
This is one of the most inutterably stupid ideas I've ever heard. Why
don't you imagine the scenarios with which we, your clients are faced. "No
problem Mr. CEO, just try unplugging your keyboard then plug it back in".
Uh-huh. I like my job. I want to keep my job. My company demands top
performance from me for each dollar.
If I were your CEO I would fire you straight away. If your
company demands the best from you, why are you using any software
product that is explicitly described as not guaranteed fit for
any purpose, not merchantable, etc...? Why are you risking the
company's dollar on something that it must use at its own risk,
without knowing it?
I demand top performance for the
several thousands of dollars I have invested in your product.
If I were your CEO I would hold you accountable for investing
thousands of company dollars in products about which the
suppliers, from AT&T onwards, are not prepared to make any
representation at all, and not telling me; actually deluding me
that I should expect everything to work smoothly.
How amused will be your CEO to learn that the several thousands
of dollars invested give your company a right to replacement of a
set of floppies should they prove physically damaged!
Can you imagine your CEO face when you tell him that all those
thousands of dollars are invested in products that come from the
equivalent of a garage sale? And when you tell him that this is
the industry standard, and that you have misled him by not
telling him the vital business information that all the computer
software (and much hardware) you use comes like that?
If I were the CEO of your company I could actually live with the
idea that this software, about which suppliers make fantastic
claims about the high level of QA they make you pay for, is used
entirely at the company's risk (CEOs of oil companies do
authorize multimillion dollar expenditures for wildcat drilling,
but they get very annoyed indeed if material facts as to the
risks are not represented to them), but I would want to have a
practical idea of the risk, or at least I would want to know it
does exist.
If I were the CEO of your company I sure would not want to see
around me a guy like you that evidently has witheld from me this
vital piece of business information, as it is *me*, not you, that
has the executive's responsibility to assess risk and decide
whether it is worth taking. And I would be very cross that you
had not even told me that I should assess the risk.
Let your CEO read the "warranties" and "licenses" of the software
and hardware you have bought, tell him to take them *literally*
and try to find a good excuse for not having told him before.
--
Piercarlo "Peter" Grandi | ARPA: pcg%cs.aber.ac.uk@nsfnet-relay.ac.uk
Dept of CS, UCW Aberystwyth | UUCP: ...!mcvax!ukc!aber-cs!pcg
Penglais, Aberystwyth SY23 3BZ, UK | INET: pcg@cs.aber.ac.ukchasm@attctc.Dallas.TX.US (Charles Marslett) (11/26/89)
In article <525@loft386.UUCP>, dpi@loft386.UUCP (Doug Ingraham) writes: .. In article <36515@ism780c.isc.com>, patrick@scrooge.ism.isc.com (Patrick Curran) writes: .. > In article <240@hdchq.UUCP> diamond@hdchq.UUCP (Bill Diamond) writes: .. > > .. > >The problem is when users are in VP/ix DOS emulation, whether using a .. > >particular DOS program or just at the DOS shell, on both the console and .. > >the SunRiver stations, the keyboard seems to "flip out". The caps lock .. > >locks in, and we cannot reset the keyboard. All keys pressed once this .. > >keyboard flip-flop occurs generate an equivalent scan code as if the .. > ><Shift> were pressed along with the single keystroke. .. > .. > Try unplugging the keyboard and then plugging it back in again. .. > This solves almost all of the "lockup" problems I experience. .. > .. .. Don't do this! Your keyboard and or computer can be damaged .. by unplugging and re-plugging with the power on. They were not .. designed to allow this. I am surprised that anyone would recommend .. as crude a solution to any problem as this. I might add that if the keyboard micro is locked up (the usual situation), there is only one mechanism to free it (since there is no "RESET" line going from the computer to the keyboard, and even if there was, that doesn't always work). That is to turn of the power to the keyboard. You could also just turn off the power to the keyboard with the big red switch (and, by the way, turn off the power to the hard disk and the CPU). Neither the Unix operating system nor the keyboard were "designed" to be reset this way; however, some situations come up where we have to do it. [BTW, if unplugging the keyboard would cause serious problems, I wouldn't buy that keyboard -- it's bound to happen by accident often enough.] Unplugging the keyboard is quite comparable to turning the machine off (from the keyboard side), and from the processor's side it looks like a short, funny sequence of transitions on the input status lines (which could lock up a poorly written keyboard handler -- but not much else. .. . . . The keyboard software could be robust. I am .. guessing that contacting the correct people at SunRiver and Interactive .. would get this cleared up quickly. If others have this kind of .. problem, then its software. If you are alone, I would bet its hardware. Robustness does require the hardware to be equally (or more) robust. .. -- .. Doug Ingraham (SysAdmin) .. Lofty Pursuits (Public Access for Rapid City SD USA) .. uunet!loft386!dpi Charles Marslett chasm@attctc.dallas.tx.us