From gls at Think.COM Mon Aug 22 19:30:32 1988 From: gls at Think.COM (gls at Think.COM) Date: Aug 22 88 13:30:32 EDT Subject: No subject In-Reply-To: Alan Bawden's message of Sun, 14 Aug 88 22:15:58 EDT <427454.880814.ALAN@AI.AI.MIT.EDU> Message-ID: <8808221730.AA03248@joplin.think.com> Date: Sun, 14 Aug 88 22:15:58 EDT From: Alan Bawden AI is running an ITS patched so that .HANG with a non-zero AC is an illegal UUO. System hackers should be on the lookout for any programs that appear to be broken by this. The idea is to learn if anybody uses the AC field in .HANG for any private purposes before I give it a meaning. I propose to make .HANG with a non-zero AC behave like the disjunction of .HANG and .SLEEP, so that you can wait for a location to change -or- for a timeout to occur. So movei 10,5*30. skipl lock .hang 10, ; (sic) jumpe 10,timout ; C(AC) = 0 if and only if timer expired. ; If not, C(AC) is negative of time to wait ; for (just like .SLEEP). will wait for 5 seconds for C(LOCK) to be negative, and will then time out. Gee, why didn't we do this ten years ago??? --Guy From ALAN at AI.AI.MIT.EDU Mon Aug 15 04:15:58 1988 From: ALAN at AI.AI.MIT.EDU (Alan Bawden) Date: Aug 14 88 22:15:58 EDT Subject: No subject Message-ID: <427454.880814.ALAN@AI.AI.MIT.EDU> AI is running an ITS patched so that .HANG with a non-zero AC is an illegal UUO. System hackers should be on the lookout for any programs that appear to be broken by this. The idea is to learn if anybody uses the AC field in .HANG for any private purposes before I give it a meaning. I propose to make .HANG with a non-zero AC behave like the disjunction of .HANG and .SLEEP, so that you can wait for a location to change -or- for a timeout to occur. So movei 10,5*30. skipl lock .hang 10, ; (sic) jumpe 10,timout ; C(AC) = 0 if and only if timer expired. ; If not, C(AC) is negative of time to wait ; for (just like .SLEEP). will wait for 5 seconds for C(LOCK) to be negative, and will then time out. From SRA at XX.LCS.MIT.EDU Fri Aug 12 00:00:00 1988 From: SRA at XX.LCS.MIT.EDU (Rob Austein) Date: Fri 12 Aug 88, 00:00 Subject: When the KL is shut down, and taken apart. In-Reply-To: <7J14CB@KICKI.STACKEN.KTH.SE> Message-ID: <12421931449.38.SRA@XX.LCS.MIT.EDU> Date: Fri, 12 Aug 88 12:41:06 +0200 From: ROLL%SESTAK.BITNET at MITVMA.MIT.EDU (Peter Lothberg) When MX finally is going to be turned off and brought out from the 9:th floor to Sweden, where it hopfully will become alive again. I thought that it might be a god idea to have a 'burial cermony', or a small party. Or..?? Yeah, let's throw a wake. Even though we plan for the machine to come back to life in Sweden, it'll be the end of an era here, and god knows the machine deserves a wake after all we've put it through. (I want to have pictures of most of the people that has been involved in the ITS development, so we can have a 'history book' along with the machine, a this might be the right time to shoot them.) Which are you shooting, the pictures or the people? ------- From ROLL%SESTAK.BITNET at MITVMA.MIT.EDU Fri Aug 12 12:41:06 1988 From: ROLL%SESTAK.BITNET at MITVMA.MIT.EDU (Peter Lothberg) Date: Aug 12 88 12:41:06 +0200 Subject: When the KL is shut down, and taken apart. Message-ID: <7J14CB@KICKI.STACKEN.KTH.SE> When MX finally is going to be turned off and brought out from the 9:th floor to Sweden, where it hopfully will become alive again. I thought that it might be a god idea to have a 'burial cermony', or a small party. Or..?? (I want to have pictures of most of the people that has been involved in the ITS development, so we can have a 'history book' along with the machine, a this might be the right time to shoot them.) Peter From CENT at AI.AI.MIT.EDU Wed Aug 10 08:03:22 1988 From: CENT at AI.AI.MIT.EDU (Pandora B. Berman) Date: Aug 10 88 02:03:22 EDT Subject: KL hardware bug of the week club Message-ID: <425456.880810.CENT@AI.AI.MIT.EDU> two weeks ago, the KL HARDWARE BUG OF THE WEEK CLUB brought you the repeated crash of the console 11. we followed up with several days of incessant non-fatal parity errors which invariably led to fatal ones or ucode hangs, and topped it off with an admixture of crashing 11 again. after a day or two to let you catch your breath, we now present: DEATH OF THE CAPSTAN MOTOR the symptoms you will encounter include: tape not loading because the take-up reel won't spin to wind it and take-up reel ceasing to spin, for no apparent reason, in the middle of a copy operation, leading to tape piling up in the vacuum columns, the drive taking itself offline, and the copy stopping dead in its tracks because DUMP can no longer talk to the drive This is serious; the MX tape drive is becoming unusable. I can deal with the machine crashing; i can't fix that motor. This problem just started occurring this evening and it seems to be getting rapidly worse; it's now sufficiently bad that i can't get through copying a single old tape successfully. This must be fixed (competently) for me to do any more tape copying. Call in DEC if you must, but get it done. From CENT at AI.AI.MIT.EDU Mon Aug 8 11:10:44 1988 From: CENT at AI.AI.MIT.EDU (Pandora B. Berman) Date: Mon, 8 Aug 88 05:10:44 EDT Subject: ai console broken Message-ID: <424212.880808.CENT@AI.AI.MIT.EDU> ai's console makes obnoxious noises and lights its paper-out indicator whenever it tries to type more than 18 chars on a line. From CENT at AI.AI.MIT.EDU Sat Aug 6 11:26:22 1988 From: CENT at AI.AI.MIT.EDU (Pandora B. Berman) Date: Sat, 6 Aug 88 05:26:22 EDT Subject: ominous new turn Message-ID: <423574.880806.CENT@AI.AI.MIT.EDU> MX has been getting parity errors right and left whenever i copy tapes. then the system console went west again (the 11 crashed), so i had to patch the sys job and move my dump job to the vt52. when i decided to punt the dump, i tried to boot the system to restore function to the console. this went OK until i tried to load MTPITS. then everything stopped. the state is characterized more by what it is not: MX does not respond to the disk boot button; the disk is not unsafe; the fault light is not on; the breakers are not tripped; the par err lights are not on. normally my next trick would be to power cycle the machine, but in its current fragile state, i don't want to do that without a real hacker around to help pick up the pieces. someone please check it out.