From: Rupert Goodwins Newsgroups: alt.folklore.computers Subject: The invention of the real-time interrupt Message-ID: X-Newsreader: Forte Free Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 7 Date: Fri, 10 Oct 2003 11:53:22 GMT NNTP-Posting-Host: 82.35.96.181 X-Complaints-To: abuse@blueyonder.co.uk X-Trace: news-text.cableinet.net 1065786802 82.35.96.181 (Fri, 10 Oct 2003 12:53:22 BST) NNTP-Posting-Date: Fri, 10 Oct 2003 12:53:22 BST Organization: blueyonder (post doesn't reflect views of blueyonder) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!news-hub.cableinet.net!blueyonder!internal-news-hub.cableinet.net!news-text.cableinet.net.POSTED!53ab2750!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153187 Having a discussion locally about who invented the real-time interrupt. Dijkstra's archive says it was invented in the late 50s, and we think it was probably one of the Cambridge, UK mob, but it would be nice to find out more. Rupert ###### From: paul125a@yahoo.com (Paul Ceruzzi) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 10 Oct 2003 10:36:08 -0700 Organization: http://groups.google.com Lines: 9 Message-ID: <28107548.0310100936.4ebd35ef@posting.google.com> References: NNTP-Posting-Host: 160.111.70.89 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Trace: posting.google.com 1065807368 28158 127.0.0.1 (10 Oct 2003 17:36:08 GMT) X-Complaints-To: groups-abuse@google.com NNTP-Posting-Date: Fri, 10 Oct 2003 17:36:08 +0000 (UTC) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!newsfeed.sunrise.ch!newsfeed.tiscali.ch!fu-berlin.de!postnews1.google.com!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153207 Rupert Goodwins > Having a discussion locally about who invented the real-time > interrupt. I believe it was The ERA (Later UNIVAC) 1103A computer, ca. 1955, that first had an interrupt. Paul Ceruzzi "A History of Modern Computing" (MIT Press, 2nd ed. 2003), p. 62. ###### From: aek@spies.com (Al Kossow) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Fri, 10 Oct 2003 10:46:11 -0700 Organization: Apple Computer, Inc. Lines: 4 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> NNTP-Posting-Host: haxrus.apple.com X-Trace: news.apple.com 1065807971 22078 17.205.21.66 (10 Oct 2003 17:46:11 GMT) X-Complaints-To: abuse@news.apple.com NNTP-Posting-Date: Fri, 10 Oct 2003 17:46:11 +0000 (UTC) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!HSNX.atgi.net!news.kjsl.com!news.spies.com!forum.apple.com!news.apple.com!haxrus.apple.com!user Xref: redlance.franklin.ch alt.folklore.computers:153208 > > Having a discussion locally about who invented the real-time > > interrupt. MIT Whirlwind (1951) ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 31 Date: Fri, 10 Oct 2003 20:38:07 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: pd7tw3no 1065818287 24.71.223.147 (Fri, 10 Oct 2003 14:38:07 MDT) NNTP-Posting-Date: Fri, 10 Oct 2003 14:38:07 MDT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!in.100proofnews.com!in.100proofnews.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!pd7cy1no!shaw.ca!pd7tw3no.POSTED!53ab2750!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153228 On Fri, 10 Oct 2003 11:53:22 GMT in alt.folklore.computers, Rupert Goodwins wrote: >Having a discussion locally about who invented the real-time >interrupt. Dijkstra's archive says it was invented in the late 50s, >and we think it was probably one of the Cambridge, UK mob, but it >would be nice to find out more. ISTR Manchester machines were first, but the following page indicates that EDSAC may have beaten them: http://www.cs.clemson.edu/~mark/interrupts.html The UNIVAC (Remington Rand/ERA) 1103A claim seems to be correct, but the timing seems off, as the feature was developed for NACA on an 1103, and 1102s were used for the wind tunnel tests, according to: http://www.cc.gatech.edu/gvu/people/randy.carpenter/folklore/v6n1.html The following page I turned up while checking is a terse history of OS features and developments presented as problems and solutions: http://www.cag.lcs.mit.edu/~rinard/osnotes/h1.html Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply ###### From: "John Varela" Message-ID: Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> User-Agent: ProNews/2 V1.53.cp050 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 8bit Lines: 25 Date: Sat, 11 Oct 2003 01:46:31 GMT NNTP-Posting-Host: 171.75.32.189 X-Complaints-To: abuse@earthlink.net X-Trace: newsread1.news.atl.earthlink.net 1065836791 171.75.32.189 (Fri, 10 Oct 2003 18:46:31 PDT) NNTP-Posting-Date: Fri, 10 Oct 2003 18:46:31 PDT Organization: EarthLink Inc. -- http://www.EarthLink.net Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!prodigy.com!logbridge.uoregon.edu!newshub.sdsu.edu!elnk-nf2-pas!elnk-pas-nf1!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread1.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153257 On Fri, 10 Oct 2003 17:46:11 UTC, aek@spies.com (Al Kossow) wrote: > > > Having a discussion locally about who invented the real-time > > > interrupt. > > MIT Whirlwind (1951) I don't know about Whirlwind, but I know for a fact that its successor, the AN/FSQ-7 SAGE computer, did not have interrupts. The SAGE computer worked on a timing cycle of three "subframes" making up a "frame". A frame was a little over 15 seconds long, which more or less matched the rotation time for the long range surveillance radar antennas. A set of "sequence parameters" told the monitor (not operating system!) what programs and data to roll in and out and what to execute when. The software--and system capacity--were structured to ensure that all processing could be done in less time than was allowed, with buffering out to the ends of the intervals. The feasibility demonstration for SAGE, the "Cape Cod System" (so called because it used a radar at Truro on Cape Cod) ran on Whirlwind. If Whirlwind had had interrupts, surely the AN/FSQ-7 would have had them too. Some sort of interrupt capability was added to XD-1, which was one of the prototypes of the AN/FSQ-7, around 1962. I recall reading a memorandum suggesting possible uses for the interrupt, though at this late date I don't recall what they were. What follows is an apparently authoritative quotation from http://interactive.sei.cmu.edu/news@sei/columns/watts_new/2001/3q01/watts-3q01.htm that implies that Whirlwind didn't have interrupts. "At Massachusetts Institute of Technology (MIT), where I wrote my first program for the Whirlwind Computer in 1953, we had only rudimentary programming support [Humphrey]. The staff at the MIT computing center had just installed a symbolic assembler that provided relative addressing, so we did not have to write for absolute memory locations. However, we did have to program the I/O and CRT display one character at a time. Whirlwind would run only one program at a time, and it didn't even have a job queue, so everything stopped between jobs. "Over the next 10 years, the design of both computing machines and operating systems evolved together. There were frequent tradeoffs between machine capabilities and software functions. By the time the IBM 360 system architecture was established in 1963, many functions that had been provided by software were incorporated into the hardware. These included memory, job, data, and device management, as well as I/O channels, device controllers, and hardware interrupt systems. Computer designers even used micro-programmed machine instructions to emulate other computer types." -- John Varela (Change "old" to "new" for email.) I apologize for munging the address but the spam is too much. ###### From: Joe Pfeiffer Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 11 Oct 2003 00:11:31 -0600 Organization: New Mexico State University Lines: 11 Message-ID: <1b1xtkpalo.fsf@cs.nmsu.edu> References: <28107548.0310100936.4ebd35ef@posting.google.com> NNTP-Posting-Host: viper.cs.nmsu.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: bubba.NMSU.Edu 1065852691 11167 128.123.64.113 (11 Oct 2003 06:11:31 GMT) X-Complaints-To: usenet@bubba.NMSU.Edu NNTP-Posting-Date: 11 Oct 2003 06:11:31 GMT User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!in.100proofnews.com!in.100proofnews.com!news-out.visi.com!petbe.visi.com!news.state.mn.us!lenny.tc.umn.edu!nunki.unm.edu!news.NMSU.Edu!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153262 I'm missing something on this whole thread. The question is about "real-time" interrupts, and the answers are all in terms of interrupts in general. Ummm, when I think about "real-time" interrupts I think in terms of a clock interrupt. I don't see any particular reason to think an interrupt necessarily means a "real-time" interrupt. Did I miss something early on? -- Joseph J. Pfeiffer, Jr., Ph.D. Phone -- (505) 646-1605 Department of Computer Science FAX -- (505) 646-1002 New Mexico State University http://www.cs.nmsu.edu/~pfeiffer Southwestern NM Regional Science and Engr Fair: http://www.nmsu.edu/~scifair ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 19 Date: Sat, 11 Oct 2003 08:10:21 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: pd7tw1no 1065859821 24.71.223.147 (Sat, 11 Oct 2003 02:10:21 MDT) NNTP-Posting-Date: Sat, 11 Oct 2003 02:10:21 MDT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!in.100proofnews.com!in.100proofnews.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!pd7cy1no!shaw.ca!pd7tw1no.POSTED!53ab2750!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153273 On 11 Oct 2003 00:11:31 -0600 in alt.folklore.computers, Joe Pfeiffer wrote: >I'm missing something on this whole thread. The question is about >"real-time" interrupts, and the answers are all in terms of interrupts >in general. Ummm, when I think about "real-time" interrupts I think >in terms of a clock interrupt. I don't see any particular reason to >think an interrupt necessarily means a "real-time" interrupt. Did I >miss something early on? ISTM a "real-time" would be characterized by an asynchronous external event of any sort, as opposed to an essentially synchronous "computer-time" program exception or trap of any sort. The word interrupt has been bandied about over the decades. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply ###### From: Rupert Goodwins Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> X-Newsreader: Forte Free Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 31 Date: Sat, 11 Oct 2003 12:46:44 GMT NNTP-Posting-Host: 82.35.96.181 X-Complaints-To: abuse@blueyonder.co.uk X-Trace: news-text.cableinet.net 1065876404 82.35.96.181 (Sat, 11 Oct 2003 13:46:44 BST) NNTP-Posting-Date: Sat, 11 Oct 2003 13:46:44 BST Organization: blueyonder (post doesn't reflect views of blueyonder) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!news2.euro.net!amsnews01.chello.com!news-hub.cableinet.net!blueyonder!internal-news-hub.cableinet.net!news-text.cableinet.net.POSTED!53ab2750!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153282 On Sat, 11 Oct 2003 08:10:21 GMT, Brian Inglis wrote: >On 11 Oct 2003 00:11:31 -0600 in alt.folklore.computers, Joe >Pfeiffer wrote: > >>I'm missing something on this whole thread. The question is about >>"real-time" interrupts, and the answers are all in terms of interrupts >>in general. Ummm, when I think about "real-time" interrupts I think >>in terms of a clock interrupt. I don't see any particular reason to >>think an interrupt necessarily means a "real-time" interrupt. Did I >>miss something early on? > >ISTM a "real-time" would be characterized by an asynchronous >external event of any sort, as opposed to an essentially >synchronous "computer-time" program exception or trap of any >sort. The word interrupt has been bandied about over the decades. > >Thanks. Take care, Brian Inglis Calgary, Alberta, Canada Yes, that's how I intended the question to read -- a non-deterministic physical signal that asynchronously invokes a routine. I've never fully understood why "software interrupts" are so-called, although they have some similarities, and surely they were never called that until hardware interrupts had been invented? Thanks, everyone, for the replies. Rupert ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Sat, 11 Oct 03 13:07:44 GMT Organization: UltraNet Communications, Inc. Lines: 22 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> X-Trace: UmFuZG9tSVZkfD1Hl2eq3hWE36Ujotjnjwmd3JgjREPdxmVbOsP7F5P5IHC1FWxs X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 11 Oct 2003 14:08:33 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-20 Xref: redlance.franklin.ch alt.folklore.computers:153288 In article <1b1xtkpalo.fsf@cs.nmsu.edu>, Joe Pfeiffer wrote: >I'm missing something on this whole thread. The question is about >"real-time" interrupts, and the answers are all in terms of interrupts >in general. Ummm, when I think about "real-time" interrupts I think >in terms of a clock interrupt. I don't see any particular reason to >think an interrupt necessarily means a "real-time" interrupt. Did I >miss something early on? Nope. You didn't miss anything. What is missing is a distinct definition of what everybody thinks is real-time. I guarantee you, you'll have as many different answers as posters. It sure seemed to me that nobody had the same definition that PDP-10 world had when you could have a real-time clock device installed in the -10. /BAH /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Sat, 11 Oct 03 13:12:12 GMT Organization: UltraNet Communications, Inc. Lines: 45 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> X-Trace: UmFuZG9tSVZ/NIKCS4eZXKt+qdt0gvs5MDWG+uMJpqPGZikLefc4qPRxN+XrVWvn X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 11 Oct 2003 14:13:00 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-20 Xref: redlance.franklin.ch alt.folklore.computers:153289 In article , Rupert Goodwins wrote: >On Sat, 11 Oct 2003 08:10:21 GMT, Brian Inglis > wrote: > >>On 11 Oct 2003 00:11:31 -0600 in alt.folklore.computers, Joe >>Pfeiffer wrote: >> >>>I'm missing something on this whole thread. The question is about >>>"real-time" interrupts, and the answers are all in terms of interrupts >>>in general. Ummm, when I think about "real-time" interrupts I think >>>in terms of a clock interrupt. I don't see any particular reason to >>>think an interrupt necessarily means a "real-time" interrupt. Did I >>>miss something early on? >> >>ISTM a "real-time" would be characterized by an asynchronous >>external event of any sort, as opposed to an essentially >>synchronous "computer-time" program exception or trap of any >>sort. The word interrupt has been bandied about over the decades. >Yes, that's how I intended the question to read -- a non-deterministic >physical signal that asynchronously invokes a routine. I've never >fully understood why "software interrupts" are so-called, although >they have some similarities, and surely they were never called that >until hardware interrupts had been invented? TOPS-10 had distinct definitions of all the above. I just don't know if that's what you're looking for. The routine RTTRP.MAC that was shipped with KA monitors was the device driver for a KA real-time clock. TOPS-10 implemented IPCF (interprocess communication facility) and considered that implementing software interrupts. I suppose one KL10 ringing the other KL10's doorbell might be considered a software interrupt. >Thanks, everyone, for the replies. You're welcome. So which answer did you pick? /BAH Subtract a hundred and four for e-mail. ###### Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Reply-To: Anne & Lynn Wheeler References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> From: Anne & Lynn Wheeler Message-ID: Organization: Wheeler&Wheeler User-Agent: Gnus/5.090024 (Oort Gnus v0.24) Emacs/21.3 (windows-nt) Cancel-Lock: sha1:zonvvR249ahmLvna9uAJwbAXeqg= MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Lines: 33 Date: Sat, 11 Oct 2003 14:31:02 GMT NNTP-Posting-Host: 209.245.14.4 X-Complaints-To: abuse@earthlink.net X-Trace: newsread3.news.pas.earthlink.net 1065882662 209.245.14.4 (Sat, 11 Oct 2003 07:31:02 PDT) NNTP-Posting-Date: Sat, 11 Oct 2003 07:31:02 PDT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!nf3.bellglobal.com!sjc70.webusenet.com!news.webusenet.com!elnk-nf2-pas!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!newsread3.news.pas.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153290 Rupert Goodwins writes: > Yes, that's how I intended the question to read -- a non-deterministic > physical signal that asynchronously invokes a routine. I've never > fully understood why "software interrupts" are so-called, although > they have some similarities, and surely they were never called that > until hardware interrupts had been invented? In 360, I/O interrupts and external interrupts not only interrupted the processor but also allowed for the processor state to be changed (typically from non-privilege or problem state to privilege or supervisor state, as part of the interrupt). The (software) supervisor call interrupts were mapped into the same infrastructure ... not that they were asyncronous ... but that they relied on the same architecture/infrastructure for state-change. The interrupts tended to result in entry into specific location in the supervisor ... and then the supervisor interrupt handler was responsible for saving the state at the time of the interrupt ... and decoding the meaning of the interrupt .... and invoking some process to handle that specific interrupt. Turns out from the supervisor side of interrupt handling ... the semantics were essentially identical whether it was an asyncronous I/O interrupt, asyncronous external interrupt or a supervisor call interrupt. In effect, the processing was identical ... even tho the cause possibly differed by whether it was software/syncronous or external/asyncronous. One can conjecture that since the processing appeared identical, there was a tendency to refer to the paradigms with similar terminology ... even tho the cause for the interrupts differ. -- Anne & Lynn Wheeler | http://www.garlic.com/~lynn/ Internet trivia 20th anv http://www.garlic.com/~lynn/rfcietff.htm ###### From: jimmydevice Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Sat, 11 Oct 2003 07:53:02 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.4) Gecko/20030624 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 50 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!news-xfer1.atl.newshosting.com!63.218.45.11.MISMATCH!newshosting.com!news-xfer2.atl.newshosting.com!38.144.126.75.MISMATCH!feed1.newsreader.com!newsreader.com!rip!sjc70.webusenet.com!news.webusenet.com!sn-xit-02!sn-xit-04!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153291 Rupert Goodwins wrote: > On Sat, 11 Oct 2003 08:10:21 GMT, Brian Inglis > wrote: > > >>On 11 Oct 2003 00:11:31 -0600 in alt.folklore.computers, Joe >>Pfeiffer wrote: >> >> >>>I'm missing something on this whole thread. The question is about >>>"real-time" interrupts, and the answers are all in terms of interrupts >>>in general. Ummm, when I think about "real-time" interrupts I think >>>in terms of a clock interrupt. I don't see any particular reason to >>>think an interrupt necessarily means a "real-time" interrupt. Did I >>>miss something early on? >> >>ISTM a "real-time" would be characterized by an asynchronous >>external event of any sort, as opposed to an essentially >>synchronous "computer-time" program exception or trap of any >>sort. The word interrupt has been bandied about over the decades. >> >>Thanks. Take care, Brian Inglis Calgary, Alberta, Canada > > > > Yes, that's how I intended the question to read -- a non-deterministic > physical signal that asynchronously invokes a routine. I've never > fully understood why "software interrupts" are so-called, although > they have some similarities, and surely they were never called that > until hardware interrupts had been invented? > > Thanks, everyone, for the replies. > > Rupert > I think that nost everybody called a software interrupt a "trap" up till the time microsoft and/or intel decided to pollute common understanding. Motorola may be guilty too. IMHO: Software "TRAP" ( indirect CALL+PSW or illegal instruction ) "Interrupt" ( something async yelling ) Jim Davis. As I typed that, I thought, where does an internal fault ( bus error/fault, div0 ) fall? Ack! ###### From: Joe Pfeiffer Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 11 Oct 2003 09:57:21 -0600 Organization: New Mexico State University Lines: 40 Message-ID: <1bad87ojha.fsf@cs.nmsu.edu> References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> NNTP-Posting-Host: viper.cs.nmsu.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: bubba.NMSU.Edu 1065887839 6327 128.123.64.113 (11 Oct 2003 15:57:19 GMT) X-Complaints-To: usenet@bubba.NMSU.Edu NNTP-Posting-Date: 11 Oct 2003 15:57:19 GMT User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!msc1.onvoy!onvoy.com!lenny.tc.umn.edu!nunki.unm.edu!news.NMSU.Edu!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153294 jmfbahciv@aol.com writes: > In article , > Rupert Goodwins wrote: > >On Sat, 11 Oct 2003 08:10:21 GMT, Brian Inglis > > wrote: > >> > >>ISTM a "real-time" would be characterized by an asynchronous > >>external event of any sort, as opposed to an essentially > >>synchronous "computer-time" program exception or trap of any > >>sort. The word interrupt has been bandied about over the decades. Where to me, a "real-time" interrupt would need to have something to do with time. > >Yes, that's how I intended the question to read -- a non-deterministic > >physical signal that asynchronously invokes a routine. I've never > >fully understood why "software interrupts" are so-called, although > >they have some similarities, and surely they were never called that > >until hardware interrupts had been invented? Well, OK, then other people are reading your question as intended. I've never had any problem with "software interrupt" as a term: after all, it's an instruction that causes the machine to do its normal interrupt handling. It does seem like a pretty fair bet that there were hardware interrupts at least as early as software interrupts, and I'd expect them to have been earlier. > TOPS-10 had distinct definitions of all the above. I just > don't know if that's what you're looking for. The routine RTTRP.MAC > that was shipped with KA monitors was the device driver for > a KA real-time clock. Exactly my thoughts. -- Joseph J. Pfeiffer, Jr., Ph.D. Phone -- (505) 646-1605 Department of Computer Science FAX -- (505) 646-1002 New Mexico State University http://www.cs.nmsu.edu/~pfeiffer Southwestern NM Regional Science and Engr Fair: http://www.nmsu.edu/~scifair ###### From: "lkrupp@pssw.NOSPAM.com.INVALID" <"lkrupp@pssw.NOSPAM.com.INVALID"> Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Sat, 11 Oct 2003 10:28:54 -0600 Organization: Posted via Supernews, http://www.supernews.com Message-ID: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.4) Gecko/20030624 Netscape/7.1 (ax) X-Accept-Language: en-us, he MIME-Version: 1.0 References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> In-Reply-To: Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 39 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!hermes.visi.com!news-out.visi.com!petbe.visi.com!64.62.191.80.MISMATCH!news-small.astraweb.com!news.astraweb.com!news-big.astraweb.com!207.217.77.102.MISMATCH!elnk-nf2-pas!newsfeed.earthlink.net!sjc70.webusenet.com!news.webusenet.com!sn-xit-02!sn-xit-06!sn-post-02!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153293 Rupert Goodwins wrote: > Yes, that's how I intended the question to read -- a non-deterministic > physical signal that asynchronously invokes a routine. I've never > fully understood why "software interrupts" are so-called, although > they have some similarities, and surely they were never called that > until hardware interrupts had been invented? FWIW, Burroughs Large Systems ALGOL featured software interrupt declarations (as Unisys MCP ALGOL does today) back in 1970 or so. These are asynchronous and triggered by external events and not exceptions triggered by something done by the current process. An example, cobbled together from bits out of the current Unisys ALGOL manual and my own hazy memory: ... TASK T; INTERRUPT I; BEGIN DISPLAY "OH WOW"; END EVENT E; TASK T; PROCEDURE P; BEGIN ... CAUSE(E); % Trigger interrupt in parent process ... END ... ATTACH I TO E; PROCESS P[T]; % Run P in a new thread ... These software interrupts were definitely modeled on hardware interrupts. Louis Krupp ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 50 Date: Sat, 11 Oct 2003 18:32:14 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: pd7tw1no 1065897134 24.71.223.147 (Sat, 11 Oct 2003 12:32:14 MDT) NNTP-Posting-Date: Sat, 11 Oct 2003 12:32:14 MDT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news-xfer1.atl.newshosting.com!63.218.45.11.MISMATCH!newshosting.com!news-xfer2.atl.newshosting.com!167.206.3.103.MISMATCH!news3.optonline.net!pd7cy1no!shaw.ca!pd7tw1no.POSTED!53ab2750!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153303 On Sat, 11 Oct 2003 07:53:02 -0700 in alt.folklore.computers, jimmydevice wrote: >Rupert Goodwins wrote: >> On Sat, 11 Oct 2003 08:10:21 GMT, Brian Inglis >> wrote: >> >> >>>On 11 Oct 2003 00:11:31 -0600 in alt.folklore.computers, Joe >>>Pfeiffer wrote: >>> >>>>I'm missing something on this whole thread. The question is about >>>>"real-time" interrupts, and the answers are all in terms of interrupts >>>>in general. Ummm, when I think about "real-time" interrupts I think >>>>in terms of a clock interrupt. I don't see any particular reason to >>>>think an interrupt necessarily means a "real-time" interrupt. Did I >>>>miss something early on? >>> >>>ISTM a "real-time" would be characterized by an asynchronous >>>external event of any sort, as opposed to an essentially >>>synchronous "computer-time" program exception or trap of any >>>sort. The word interrupt has been bandied about over the decades. ... >> Yes, that's how I intended the question to read -- a non-deterministic >> physical signal that asynchronously invokes a routine. I've never >> fully understood why "software interrupts" are so-called, although >> they have some similarities, and surely they were never called that >> until hardware interrupts had been invented? >> >> >I think that nost everybody called a software interrupt a "trap" up >till the time microsoft and/or intel decided to pollute common >understanding. Motorola may be guilty too. >IMHO: >Software "TRAP" ( indirect CALL+PSW or illegal instruction ) >"Interrupt" ( something async yelling ) >As I typed that, I thought, where does an internal fault >( bus error/fault, div0 ) fall? That's normally called an exception -- it's synchronous with the instruction stream and halts execution of an instruction in some way -- includes software faults and hardware failures. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <4cjgovs933i7hinfpv1igvvnmknlf5mn08@4ax.com> References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 30 Date: Sat, 11 Oct 2003 18:46:23 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: pd7tw1no 1065897983 24.71.223.147 (Sat, 11 Oct 2003 12:46:23 MDT) NNTP-Posting-Date: Sat, 11 Oct 2003 12:46:23 MDT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.telusplanet.net!newsfeed.telus.net!news3.optonline.net!pd7cy1no!shaw.ca!pd7tw1no.POSTED!53ab2750!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153306 On Sat, 11 Oct 03 13:07:44 GMT in alt.folklore.computers, jmfbahciv@aol.com wrote: >In article <1b1xtkpalo.fsf@cs.nmsu.edu>, > Joe Pfeiffer wrote: >>I'm missing something on this whole thread. The question is about >>"real-time" interrupts, and the answers are all in terms of interrupts >>in general. Ummm, when I think about "real-time" interrupts I think >>in terms of a clock interrupt. I don't see any particular reason to >>think an interrupt necessarily means a "real-time" interrupt. Did I >>miss something early on? > >Nope. You didn't miss anything. What is missing is a distinct >definition of what everybody thinks is real-time. I guarantee >you, you'll have as many different answers as posters. It >sure seemed to me that nobody had the same definition that >PDP-10 world had when you could have a real-time clock >device installed in the -10. Those are normally called clock or timer interrupts depending on the hardware involved. A clock interrupt would either be a facility like the IBM 3x0 clock comparator or the regular ticker in many systems providing system wall clock time. A timer interrupt would be a delay or timeout generated by a programmable interval timer. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply ###### Newsgroups: alt.folklore.computers From: William Hamblen Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> Organization: Utterly Disorganized User-Agent: slrn/0.9.8.0 (Linux) Lines: 16 Message-ID: Date: Sat, 11 Oct 2003 20:44:06 GMT NNTP-Posting-Host: 216.80.156.127 X-Complaints-To: abuse@earthlink.net X-Trace: newsread1.news.atl.earthlink.net 1065905046 216.80.156.127 (Sat, 11 Oct 2003 13:44:06 PDT) NNTP-Posting-Date: Sat, 11 Oct 2003 13:44:06 PDT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!sjc70.webusenet.com!news.webusenet.com!elnk-nf2-pas!elnk-pas-nf1!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread1.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153310 On 2003-10-11, jimmydevice wrote: > I think that nost everybody called a software interrupt a "trap" up > till the time microsoft and/or intel decided to pollute common > understanding. Motorola may be guilty too. Motorola definitely is guilty of putting an SWI (software interrupt) instruction on the 6800 microprocessor in the 1970s. It did more or less what a real interrupt would do: push the registers onto the stack and jump to a routine that usually would end with an RTI (return from interrupt) instruction. -- When the fog came in on little cat feet last night, it left these little muddy paw prints on the hood of my car. ###### From: "John Varela" Message-ID: Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> User-Agent: ProNews/2 V1.53.cp050 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 8bit Lines: 16 Date: Sat, 11 Oct 2003 20:51:36 GMT NNTP-Posting-Host: 171.75.32.81 X-Complaints-To: abuse@earthlink.net X-Trace: newsread2.news.atl.earthlink.net 1065905496 171.75.32.81 (Sat, 11 Oct 2003 16:51:36 EDT) NNTP-Posting-Date: Sat, 11 Oct 2003 16:51:36 EDT Organization: EarthLink Inc. -- http://www.EarthLink.net Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!feed.news.nacamar.de!newsfeed.icl.net!newsfeed.fjserv.net!newsfeed.icl.net!newsfeed.fjserv.net!news-out1.nntp.be!propagator2-sterling!news-in-sterling.nuthinbutnews.com!snoopy.risq.qc.ca!chi1.webusenet.com!sjc70.webusenet.com!news.webusenet.com!elnk-nf2-pas!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread2.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153311 On Sat, 11 Oct 2003 06:11:31 UTC, Joe Pfeiffer wrote: > Ummm, when I think about "real-time" interrupts I think > in terms of a clock interrupt. I don't see any particular reason to > think an interrupt necessarily means a "real-time" interrupt. FWIW, SAGE had a real-time clock but no interrupts. When the last program in the cycle, STM (System Timing and Miscellaneous), executed it did miscellaneous housekeeping, then kept fetching the clock until it was time to start the next cycle, whereupon it terminated itself and returned control to the monitor. -- John Varela (Change "old" to "new" for email.) I apologize for munging the address but the spam is too much. ###### From: Lon Stowell Reply-To: LonDot.Stowell@ComcastPeriod.Net User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.4) Gecko/20030624 Netscape/7.1 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> In-Reply-To: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 21 Message-ID: NNTP-Posting-Host: 12.240.77.188 X-Complaints-To: abuse@comcast.net X-Trace: sccrnsc01 1065906837 12.240.77.188 (Sat, 11 Oct 2003 21:13:57 GMT) NNTP-Posting-Date: Sat, 11 Oct 2003 21:13:57 GMT Organization: Comcast Online Date: Sat, 11 Oct 2003 21:13:57 GMT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news-out1.nntp.be!propagator2-sterling!news-in-sterling.nuthinbutnews.com!cyclone1.gnilink.net!wn14feed!wn13feed!wn11feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi_feed4!attbi.com!sccrnsc01.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153312 Approximately 10/11/03 13:51, John Varela uttered for posterity: > On Sat, 11 Oct 2003 06:11:31 UTC, Joe Pfeiffer wrote: > >> Ummm, when I think about "real-time" interrupts I think >> in terms of a clock interrupt. I don't see any particular reason to >> think an interrupt necessarily means a "real-time" interrupt. > > FWIW, SAGE had a real-time clock but no interrupts. When the last program in > the cycle, STM (System Timing and Miscellaneous), executed it did > miscellaneous housekeeping, then kept fetching the clock until it was time to > start the next cycle, whereupon it terminated itself and returned control to > the monitor. Was this on the original SAGE, or on the AVCO produced AN/FSS-7 that had one big univac computer for overall control plus a coupla floors full of miscellaneous stuff including "unit 10" that controlled the servos digitally [meaning it busted the spline gear a lot], the doppler processor/filter, etc. ? ###### From: "John Varela" Message-ID: Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> User-Agent: ProNews/2 V1.53.cp050 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 8bit Lines: 21 Date: Sat, 11 Oct 2003 23:16:55 GMT NNTP-Posting-Host: 171.75.32.82 X-Complaints-To: abuse@earthlink.net X-Trace: newsread1.news.atl.earthlink.net 1065914215 171.75.32.82 (Sat, 11 Oct 2003 16:16:55 PDT) NNTP-Posting-Date: Sat, 11 Oct 2003 16:16:55 PDT Organization: EarthLink Inc. -- http://www.EarthLink.net Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!newshub.sdsu.edu!elnk-nf2-pas!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread1.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153325 On Sat, 11 Oct 2003 21:13:57 UTC, Lon Stowell wrote: > Was this on the original SAGE, or on the AVCO produced > AN/FSS-7 that had one big univac computer for overall > control plus a coupla floors full of miscellaneous > stuff including "unit 10" that controlled the servos > digitally [meaning it busted the spline gear a lot], > the doppler processor/filter, etc. ? This was the AN/FSQ-7. A Google search reveals that the AN/FSS-7 was a radar. If you're unfamiliar with SAGE and the Q-7, Google will reveal quite a few web sites. A good one is http://history.acusd.edu/gen/20th/sage.html -- John Varela (Trade "OLD" lamps for "NEW" for email.) I apologize for munging the address but the spam is too much. ###### From: Lon Stowell Reply-To: LonDot.Stowell@ComcastPeriod.Net User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.4) Gecko/20030624 Netscape/7.1 (ax) X-Accept-Language: en-us, en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> In-Reply-To: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 61 Message-ID: NNTP-Posting-Host: 12.240.77.188 X-Complaints-To: abuse@comcast.net X-Trace: sccrnsc03 1065919414 12.240.77.188 (Sun, 12 Oct 2003 00:43:34 GMT) NNTP-Posting-Date: Sun, 12 Oct 2003 00:43:34 GMT Organization: Comcast Online Date: Sun, 12 Oct 2003 00:43:34 GMT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!in.100proofnews.com!in.100proofnews.com!cycny01.gnilink.net!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi_feed4!attbi.com!sccrnsc03.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153328 Approximately 10/11/03 16:16, John Varela uttered for posterity: > On Sat, 11 Oct 2003 21:13:57 UTC, Lon Stowell > wrote: > >> Was this on the original SAGE, or on the AVCO produced >> AN/FSS-7 that had one big univac computer for overall >> control plus a coupla floors full of miscellaneous >> stuff including "unit 10" that controlled the servos >> digitally [meaning it busted the spline gear a lot], >> the doppler processor/filter, etc. ? > > This was the AN/FSQ-7. A Google search reveals that the AN/FSS-7 was a radar. The Fuzzy-Seven was a rube goldberg combination. Built off a modified FPS-26, it normally functioned as a standard pulsed search/rotating radar for Sea Launch Ballistic Missiles, using 3 different beams. Famous for the "Twystron" a combination TWT and Klystron used to handle the somewhat unusual bandwidth of the power stages. Peak power somewhere in the 4-5 Megawatt but for very short pulses. All sorts of filtering and such done to both prevent jamming and to ignore non-missile and non-incoming events. If it got [mumble] a couple [/mumble] bangs off a target that also passed the filtering for direction and speed, it would pop into track mode [using only one of the 3 beams] in a really oddball looking scan pattern at the predicted location from the previous bangs. Once it got a trajectory on a ballistic missile, it would enter that into a virtual celestial sky map and fire off an impact warning to ADC, then go right back into search mode looking for other unknown targets. However, it also tied into the SAGE towers so by hitting a relay they could take it into a mode use for height finding of incoming bombers. This had a really bad habit of hitting the hydraulic driven azimuth gear so hard that Avco added a small spline gear as a mechanical fuse. Dunno if they ever managed to tame the movement inputs enough to prevent spline breakage or not, as you had to shut down the Twystrons to go up in the tower and replace the thing. One of the units on the Fuzzy-Seven tower was all tube logic, allegedly a piece of the old SAGE junk. The rest was mostly IC with a bit o discrete here and there. Still took up a couple floors pretty well. Suspect the older Q-7 was gone by the time the Fuzzy was in field testing, as this was around Lasnerian 63. > > If you're unfamiliar with SAGE and the Q-7, Google will reveal quite a few web > sites. A good one is > > http://history.acusd.edu/gen/20th/sage.html > -- My governor can kick your governor's ass ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Sun, 12 Oct 03 10:30:40 GMT Organization: UltraNet Communications, Inc. Lines: 82 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> <1bad87ojha.fsf@cs.nmsu.edu> X-Trace: UmFuZG9tSVaMI/F3xRgzglFTSSjFf741f1sNRwQdYst39MCkv3P64RfFoHwvTZvW X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 12 Oct 2003 11:31:41 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-234 Xref: redlance.franklin.ch alt.folklore.computers:153342 In article <1bad87ojha.fsf@cs.nmsu.edu>, Joe Pfeiffer wrote: >jmfbahciv@aol.com writes: > >> In article , >> Rupert Goodwins wrote: >> >On Sat, 11 Oct 2003 08:10:21 GMT, Brian Inglis >> > wrote: >> >> >> >>ISTM a "real-time" would be characterized by an asynchronous >> >>external event of any sort, as opposed to an essentially >> >>synchronous "computer-time" program exception or trap of any >> >>sort. The word interrupt has been bandied about over the decades. > >Where to me, a "real-time" interrupt would need to have something to >do with time. In real time; no delays can be tolerated. IIRC, the way we handled real time interrupts is that every thing else stopped (had lower PI level priority) until the code that dealt with the RT interrupt was finished and could dismiss it. I don't think this code was ever a few instrucions long _and_ had to always remain in core because it took too long to setup address tables and user data, etc. IIRC, most customers moved the real time stuff off to a PDP-11 or an -8 because a -10 was just too expensive to be a weighting boat anchor. They then used the -10 to process the data (it was really at doing that efficiently). > >> >Yes, that's how I intended the question to read -- a non-deterministic >> >physical signal that asynchronously invokes a routine. I've never >> >fully understood why "software interrupts" are so-called, although >> >they have some similarities, and surely they were never called that >> >until hardware interrupts had been invented? > >Well, OK, then other people are reading your question as intended. > >I've never had any problem with "software interrupt" as a term: after >all, it's an instruction that causes the machine to do its normal >interrupt handling. It does seem like a pretty fair bet that there >were hardware interrupts at least as early as software interrupts, and >I'd expect them to have been earlier. > >> TOPS-10 had distinct definitions of all the above. I just >> don't know if that's what you're looking for. The routine RTTRP.MAC >> that was shipped with KA monitors was the device driver for >> a KA real-time clock. > >Exactly my thoughts. Since I was the DEC expert in system accounting data capture, I had to give a presentation to a couple of prospective customers who worried a lot about cross-billing computing resources. These were people who were 100% IBM thinkers (people who unconsciously thought purely in terms of card processing). These two accounting types asked me if TOPS-10 could do downstream billing in real time. I got real puzzled (why would you want to use up all your computing resources to calculate charges for computer usage that won't exist because the cross-charger was using them all?). What they really meant was a transaction processing system that processed the usage data (note the word data) as it was stored. But they couldn't get real time out of their heads. They were so tied to batch-processing thinking that they thought of all computing services as 1. start service 2. end service 3. post charge of service. This is what had to be done when the data of using the one-deck-at-a-time computer was captured by the human operator in the log book. I used to keypunch this usage data when our center ran a 1620. /BAH /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Sun, 12 Oct 03 10:33:42 GMT Organization: UltraNet Communications, Inc. Lines: 42 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> <4cjgovs933i7hinfpv1igvvnmknlf5mn08@4ax.com> X-Trace: UmFuZG9tSVYoH8EfJt+rmkXLOstMFzl9AnSoYPnl5H+yOpYQ5bMe+4DTKpwzbS89 X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 12 Oct 2003 11:34:40 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!in.100proofnews.com!in.100proofnews.com!cycny01.gnilink.net!cyclone1.gnilink.net!peer01.cox.net!cox.net!bigfeed.bellsouth.net!bigfeed2.bellsouth.net!news.bellsouth.net!elnk-atl-nf1!newsfeed.earthlink.net!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-234 Xref: redlance.franklin.ch alt.folklore.computers:153343 In article <4cjgovs933i7hinfpv1igvvnmknlf5mn08@4ax.com>, Brian Inglis wrote: >On Sat, 11 Oct 03 13:07:44 GMT in alt.folklore.computers, >jmfbahciv@aol.com wrote: > >>In article <1b1xtkpalo.fsf@cs.nmsu.edu>, >> Joe Pfeiffer wrote: >>>I'm missing something on this whole thread. The question is about >>>"real-time" interrupts, and the answers are all in terms of interrupts >>>in general. Ummm, when I think about "real-time" interrupts I think >>>in terms of a clock interrupt. I don't see any particular reason to >>>think an interrupt necessarily means a "real-time" interrupt. Did I >>>miss something early on? >> >>Nope. You didn't miss anything. What is missing is a distinct >>definition of what everybody thinks is real-time. I guarantee >>you, you'll have as many different answers as posters. It >>sure seemed to me that nobody had the same definition that >>PDP-10 world had when you could have a real-time clock >>device installed in the -10. > >Those are normally called clock or timer interrupts depending on >the hardware involved. Not quite. There were two clocks if you had a real-time clock device. The real-time clock interrupt had a higher priority, IIRC, than the regular clock. I'm really talking about an extra device that was installed with a bona fide device driver to deal with it. > ...A clock interrupt would either be a >facility like the IBM 3x0 clock comparator or the regular ticker >in many systems providing system wall clock time. A timer >interrupt would be a delay or timeout generated by a programmable >interval timer. Not the real-time clock that a KA-10 could have. /BAH Subtract a hundred and four for e-mail. ###### From: "John Varela" Message-ID: Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> User-Agent: ProNews/2 V1.53.cp050 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 8bit Lines: 46 Date: Sun, 12 Oct 2003 17:28:43 GMT NNTP-Posting-Host: 171.75.32.92 X-Complaints-To: abuse@earthlink.net X-Trace: newsread1.news.atl.earthlink.net 1065979723 171.75.32.92 (Sun, 12 Oct 2003 10:28:43 PDT) NNTP-Posting-Date: Sun, 12 Oct 2003 10:28:43 PDT Organization: EarthLink Inc. -- http://www.EarthLink.net Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!cyclone.bc.net!newshub.sdsu.edu!elnk-nf2-pas!elnk-pas-nf1!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread1.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153354 On Sun, 12 Oct 2003 00:43:34 UTC, Lon Stowell wrote: > Approximately 10/11/03 16:16, John Varela uttered for posterity: > > > On Sat, 11 Oct 2003 21:13:57 UTC, Lon Stowell > > wrote: > > > >> Was this on the original SAGE, or on the AVCO produced > >> AN/FSS-7 that had one big univac computer for overall > >> control plus a coupla floors full of miscellaneous > >> stuff including "unit 10" that controlled the servos > >> digitally [meaning it busted the spline gear a lot], > >> the doppler processor/filter, etc. ? > > > > This was the AN/FSQ-7. A Google search reveals that the AN/FSS-7 was a radar. > > The Fuzzy-Seven was a rube goldberg combination. > One of the units on the Fuzzy-Seven tower > was all tube logic, allegedly a piece of the old SAGE junk. SAGE "junk"? > The rest was mostly IC with a bit o discrete here and there. > Still took up a couple floors pretty well. > > Suspect the older Q-7 was gone by the time the Fuzzy was in > field testing, as this was around Lasnerian 63. They already had ICs in production units in '63? Anyway, SAGE deployment wasn't completed until 1963. I suspect you're not familiar with what SAGE actually was; I suggest you take a look at one of the web sites dedicated to it. > > If you're unfamiliar with SAGE and the Q-7, Google will reveal quite a few web > > sites. A good one is > > > > http://history.acusd.edu/gen/20th/sage.html -- John Varela (Trade "OLD" lamps for "NEW" for email.) I apologize for munging the address but the spam is too much. ###### Sender: eric@ruckus.brouhaha.com From: Eric Smith Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> Organization: Eric Conspiracy Secret Labs X-Eric-Conspiracy: There is no conspiracy. Date: 13 Oct 2003 00:32:37 -0700 Message-ID: Lines: 5 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 64.62.206.2 X-Trace: 13 Oct 2003 00:42:34 -0700, 64.62.206.2 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!newsfeed00.sul.t-online.de!t-online.de!npeer.de.kpn-eurorings.net!newsfeed.kabelfoon.nl!195.129.110.21.MISMATCH!bnewsfeed00.bru.ops.eu.uu.net!bnewsinpeer01.bru.ops.eu.uu.net!emea.uu.net!ash.uu.net!newssorter-3001.bay.webtv.net!news.spies.com!64.62.206.2 Xref: redlance.franklin.ch alt.folklore.computers:153372 "John Varela" writes: > If Whirlwind had had interrupts, surely the AN/FSQ-7 would have had them too. Not if it was determined that SAGE didn't need interrupts, or possibly even that the use of interrupts was counter-productive for SAGE. ###### From: "John Varela" Message-ID: Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> User-Agent: ProNews/2 V1.53.cp050 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 8bit Lines: 19 Date: Mon, 13 Oct 2003 18:07:35 GMT NNTP-Posting-Host: 64.157.49.138 X-Complaints-To: abuse@earthlink.net X-Trace: newsread1.news.atl.earthlink.net 1066068455 64.157.49.138 (Mon, 13 Oct 2003 11:07:35 PDT) NNTP-Posting-Date: Mon, 13 Oct 2003 11:07:35 PDT Organization: EarthLink Inc. -- http://www.EarthLink.net Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!nf3.bellglobal.com!sjc70.webusenet.com!news.webusenet.com!elnk-nf2-pas!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread1.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153390 On Mon, 13 Oct 2003 07:32:37 UTC, Eric Smith wrote: > "John Varela" writes: > > If Whirlwind had had interrupts, surely the AN/FSQ-7 would have had them too. > > Not if it was determined that SAGE didn't need interrupts, or possibly even > that the use of interrupts was counter-productive for SAGE. Obviously SAGE didn't need interrupts because SAGE didn't have interrupts. No computer *needs* interrupts. Anything done with interrupts can be done instead by scanning devices, checking overflow bits, and whatever. Given the choice, which method would you use? -- John Varela (Trade "OLD" lamps for "NEW" for email.) I apologize for munging the address but the spam is too much. ###### Message-ID: <3F8B0C21.BF2B27EB@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 29 Date: Mon, 13 Oct 2003 20:51:56 GMT NNTP-Posting-Host: 12.90.173.170 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc05-news.ops.worldnet.att.net 1066078316 12.90.173.170 (Mon, 13 Oct 2003 20:51:56 GMT) NNTP-Posting-Date: Mon, 13 Oct 2003 20:51:56 GMT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!takemy.news.telefonica.de!telefonica.de!borium.box.nl!in.100proofnews.com!in.100proofnews.com!cycny01.gnilink.net!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!bgtnsc05-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153394 John Varela wrote: > wrote: > > "John Varela" writes: > > > > If Whirlwind had had interrupts, surely the AN/FSQ-7 would > > > have had them too. > > > > Not if it was determined that SAGE didn't need interrupts, or > > possibly even that the use of interrupts was counter-productive > > for SAGE. > > Obviously SAGE didn't need interrupts because SAGE didn't have > interrupts. > > No computer *needs* interrupts. Anything done with interrupts > can be done instead by scanning devices, checking overflow bits, > and whatever. Given the choice, which method would you use? >From another point of view, most computers are really running some sort of microprogram, which is interpreting the visible machine code. At regular intervals, which are tied to the execution of individual machine code instructions, it polls some sort of input for an interrups. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: Marco S Hyman Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 13 Oct 2003 14:14:57 -0700 Organization: S.N.A.F.U. -- http://www.snafu.org/ Lines: 13 Sender: marc@hana.snafu.org Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> NNTP-Posting-Host: dumbcat.snafu.org (208.201.244.209) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: news.uni-berlin.de 1066079824 23403171 208.201.244.209 (16 [97260]) X-Orig-Path: dumbcat.snafu.org!news User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.3 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news-fra1.dfn.de!fu-berlin.de!uni-berlin.de!dumbcat.snafu.ORG!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153395 "John Varela" writes: > No computer *needs* interrupts. Anything done with interrupts can be done > instead by scanning devices, checking overflow bits, and whatever. Given the > choice, which method would you use? Depends. If I want to process an event that doesn't occur often an interrupt is probably the way to go. If I'm trying to handle real time events I can often get better performance polling -- depends on how much time it takes to set up the interrupt handler. Some processors take forever to save state before calling the handler. // marc ###### From: mark@hubcap.clemson.edu (Mark Smotherman) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 13 Oct 2003 17:29:42 -0400 Organization: Clemson University Lines: 31 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> NNTP-Posting-Host: localhost X-Trace: hubcap.clemson.edu 1066080582 1806 127.0.0.1 (13 Oct 2003 21:29:42 GMT) X-Complaints-To: usenet@hubcap.clemson.edu NNTP-Posting-Date: 13 Oct 2003 21:29:42 GMT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.tele.dk!news.tele.dk!small.news.tele.dk!news-out.visi.com!petbe.visi.com!feed.news.qwest.net!emac1.ocs.lsu.edu!nntp.msstate.edu!finch!hubcap.clemson.edu!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153396 "John Varela" writes: > I don't know about Whirlwind, but I know for a fact that its successor, > the AN/FSQ-7 SAGE computer, did not have interrupts. The SAGE (started 1952, and operational in 1955) was the first computer to have DMA. On the SAGE, the I/O operations would start block transfers of data to/ from drum buffers that would proceed in parallel with further instruction execution. A controller generated the sequential memory addresses for the block being transfered and decremented a counter. The SAGE CPU had a conditional branch to test for completion of the transfer, and the I/O operations were also interlocked so that the CPU is stalled if a second I/O operation was attempted before the previous one ended. See J.F. Jacobs, The SAGE Air Defense System - A Personal History. Bedford, MA: MITRE Corp., 1986. R. Serrell, M.M. Astrahan, G.W. Patterson, and I.B. Pyne, "The Evolution of Computing Machines and Systems," Proc. IRE, vol. 50, no. 5, May 1962, pp. 1039-1058. (I've added this to my interrupt history web page cited in an earlier post. As other folks have corrections / additions, please post or email me directly. I'd like to capture all this history in one place.) -- Mark Smotherman, Computer Science Dept., Clemson University, Clemson, SC http://www.cs.clemson.edu/~mark/homepage.html ###### From: "John Varela" Message-ID: Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> User-Agent: ProNews/2 V1.53.cp050 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 8bit Lines: 53 Date: Mon, 13 Oct 2003 23:05:14 GMT NNTP-Posting-Host: 171.75.32.151 X-Complaints-To: abuse@earthlink.net X-Trace: newsread2.news.atl.earthlink.net 1066086314 171.75.32.151 (Mon, 13 Oct 2003 19:05:14 EDT) NNTP-Posting-Date: Mon, 13 Oct 2003 19:05:14 EDT Organization: EarthLink Inc. -- http://www.EarthLink.net Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!logbridge.uoregon.edu!newshub.sdsu.edu!elnk-nf2-pas!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread2.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153398 On Mon, 13 Oct 2003 21:29:42 UTC, mark@hubcap.clemson.edu (Mark Smotherman) wrote: > "John Varela" writes: > > I don't know about Whirlwind, but I know for a fact that its successor, > > the AN/FSQ-7 SAGE computer, did not have interrupts. > > > The SAGE (started 1952, and operational in 1955) was the first computer > to have DMA. Your dates are off by several years. The project was initiated in 1954 and deployment completed in 1963. I forget when and where the first site went operational. I believe it was McGuire around 1960. If anyone cares, I know people who were there, and could ask. http://history.acusd.edu/gen/20th/sage.html > On the SAGE, the I/O operations would start block transfers of data to/ > from drum buffers that would proceed in parallel with further instruction > execution. A controller generated the sequential memory addresses for > the block being transfered and decremented a counter. The SAGE CPU had > a conditional branch to test for completion of the transfer, and the I/O > operations were also interlocked so that the CPU is stalled if a second > I/O operation was attempted before the previous one ended. Yes. The interlock was turned off by the end-around-carry pulse from the counter when it stepped from +0 to -0 (the Q-7 used ones' complement arithmetic). There was also an Operate instruction to turn it off. I don't know what that was for. What does any of that have to do with interrupts? Terminology: SAGE wasn't a computer, it was a system. The central computer was the AN/FSQ-7. > See > > J.F. Jacobs, The SAGE Air Defense System - A Personal History. > Bedford, MA: MITRE Corp., 1986. I knew Jake, and worked for MITRE from 1960 to 1995. > R. Serrell, M.M. Astrahan, G.W. Patterson, and I.B. Pyne, > "The Evolution of Computing Machines and Systems," Proc. IRE, > vol. 50, no. 5, May 1962, pp. 1039-1058. > > (I've added this to my interrupt history web page cited in an earlier > post. As other folks have corrections / additions, please post or email > me directly. I'd like to capture all this history in one place.) -- John Varela (Trade "OLD" lamps for "NEW" for email.) I apologize for munging the address but the spam is too much. ###### Sender: eric@ruckus.brouhaha.com From: Eric Smith Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> Organization: Eric Conspiracy Secret Labs X-Eric-Conspiracy: There is no conspiracy. Date: 13 Oct 2003 17:12:43 -0700 Message-ID: Lines: 36 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 64.62.206.2 X-Trace: 13 Oct 2003 17:22:47 -0700, 64.62.206.2 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!news-stu1.dfn.de!news-fra1.dfn.de!npeer.de.kpn-eurorings.net!news.tele.dk!news.tele.dk!small.news.tele.dk!news-out.visi.com!petbe.visi.com!ash.uu.net!newssorter-3001.bay.webtv.net!news.spies.com!64.62.206.2 Xref: redlance.franklin.ch alt.folklore.computers:153402 "John Varela" writes: > If Whirlwind had had interrupts, surely the AN/FSQ-7 would have had them too. I wrote: > Not if it was determined that SAGE didn't need interrupts, or possibly even > that the use of interrupts was counter-productive for SAGE. "John Varela" writes: > Obviously SAGE didn't need interrupts because SAGE didn't have interrupts. Well, then it was pretty clever of the hardware designers to omit them. I'm not under any illusion that it saved the taxpayers money, but it may at least have increased the profits of the vendor with the contract (IBM?) by a few thousand dollars. There were probably a *lot* of features of Whirlwind that didn't make it into SAGE. > No computer *needs* interrupts. Anything done with interrupts can be done > instead by scanning devices, checking overflow bits, and whatever. That's true only in the sense that it's theoretically possible to reengineer any computer system using interrupts to produce an equivalent system not using interrupts. But the hardware requirements for the non-interrupt system may be drastically different than for the interrupt-driven system. Any particular computer may be able to do some things using interrupts that the same computer is NOT capable of doing without using interrupts, for instance, due to response time requirements being less than the minimum practical polling interval. > Given the choice, which method would you use? Depends on the problem to be solved, and the hardware that is available to solve it. For some problems, polling actually works better than interrupts. ###### From: mark@hubcap.clemson.edu (Mark Smotherman) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 15 Oct 2003 12:09:50 -0400 Organization: Clemson University Lines: 26 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> <1b1xtkpalo.fsf@cs.nmsu.edu> <4cjgovs933i7hinfpv1igvvnmknlf5mn08@4ax.com> NNTP-Posting-Host: localhost X-Trace: hubcap.clemson.edu 1066234190 7476 127.0.0.1 (15 Oct 2003 16:09:50 GMT) X-Complaints-To: usenet@hubcap.clemson.edu NNTP-Posting-Date: 15 Oct 2003 16:09:50 GMT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!in.100proofnews.com!in.100proofnews.com!news-out.visi.com!petbe.visi.com!feed.news.qwest.net!emac1.ocs.lsu.edu!nntp.msstate.edu!finch!hubcap.clemson.edu!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153450 Brian Inglis writes: >>... real-time clock ... >Those are normally called clock or timer interrupts depending on >the hardware involved. A clock interrupt would either be a >facility like the IBM 3x0 clock comparator or the regular ticker >in many systems providing system wall clock time. A timer >interrupt would be a delay or timeout generated by a programmable >interval timer. ... enter history mode again (i.e., just looked at Blaauw and Brooks textbook, section 7.6.1) ... progran-settable timers date back to at least IBM Stretch (1961), which had a 19-bit timer with 1 msec resolution (= 8 min) real-time clocks (used for accounting) also date back to at least Stretch (36 bits, 1 msec resolution = 2 years) Stretch also had a fixed interval watchdog timer (1 msec) to prevent inifinte loops in indirect addressing and in execute instructions -- Mark Smotherman, Computer Science Dept., Clemson University, Clemson, SC http://www.cs.clemson.edu/~mark/homepage.html ###### Message-ID: <3F8D89D1.472003E0@earthlink.net> From: jchausler X-Mailer: Mozilla 4.72 [en] (Win95; I) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 77 Date: Wed, 15 Oct 2003 17:57:28 GMT NNTP-Posting-Host: 63.178.115.232 X-Complaints-To: abuse@earthlink.net X-Trace: newsread2.news.atl.earthlink.net 1066240648 63.178.115.232 (Wed, 15 Oct 2003 13:57:28 EDT) NNTP-Posting-Date: Wed, 15 Oct 2003 13:57:28 EDT Organization: EarthLink Inc. -- http://www.EarthLink.net Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!news.tu-darmstadt.de!newsfeed.freenet.de!213.253.16.105.MISMATCH!mephistopheles.news.clara.net!news.clara.net!news2.euro.net!63.223.20.72.MISMATCH!sjc72.webusenet.com!sjc70.webusenet.com!news.webusenet.com!elnk-nf2-pas!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread2.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153453 Eric Smith wrote: > "John Varela" writes: > > Obviously SAGE didn't need interrupts because SAGE didn't have interrupts. > > Well, then it was pretty clever of the hardware designers to omit them. > I'm not under any illusion that it saved the taxpayers money, but it may > at least have increased the profits of the vendor with the contract (IBM?) > by a few thousand dollars. Not the reason. Remember this is the 50's. SAGE was based on vacuum tubes and as a system handling a real time process, running flat out at the technology level of the time doing a very specific process. Handling interrupts tends to be expensive in both hardware and time due to the need at some level to save machine state and then restore it at the end of the interrupt. It was not in any way a "general purpose computer". Interrupts (if they were thought of and that I don't know) would likely have had a negative impact on both cost and performance. Some early minicomputer real time systems I worked on (early 70's) even though the processor (DG Nova) had an interrupt were done without to keep the timing tight. Scanning was used instead. > There were probably a *lot* of features of Whirlwind that didn't make it > into SAGE. Although "Project Whirlwind" started life as a real time aircraft simulator as a goal, the goals changed long before much hardware had been built to be a "general purpose computer for real time interface". One of the reasons it was only 16 bits when most contemporary computers were more than 30 bits, sometimes much more, was that since the machine was expected to be interacting constantly with its surroundings, this "lack of precision" would be offset by "new information from the outside world" as the machine would constantly be recalculating what was going on. I have always argued that Whirlwind was the "first minicomputer" despite being big as a house as when the technology advanced far enough to make Nova's and PDP-8's possible, it finally became cost effective to employ them in such real time environments. Of course, in addition, Project Whirlwind was the incubator for a lot of people who eventually started the "Minicomputer Revolution". Of course, what became the Whirlwind computer was originally just to be a prototype for the final Whirlwind computer and 16 bits was considered enough of a test bed for evaluating whether the whole concept was workable. Remember these guys were treading where no one had gone before and spending a great deal of money in the process. They were lucky in that the Korean War came along when it did which made the military suddenly nervous about incoming bombers and in need of a command and control environment beyond what they had to handle it. The Whirlwind computer even though only a prototype and only just beginning operation was the only thing approaching a "real time" computer available at that time and thus the Project Whirlwind team was selected to coordinate the design of the SAGE. If this had not happened, it is likely that Project Whirlwind would have been shut down. When Whirlwind the computer was shut down in the late 50's, (for the first time, it had a strange second life in the 60s') it was still one of the fastest machines running at that time. Sadly, despite some effort to find out, I have only been able to get very general information on the actual design of Whirlwind. Also as it was built as a "breadboard" test environment rather than being built to be compact, it was easily changed over time and was. Whirlwind can claim a lot of "firsts" in the history of computers. Chris AN GETTO$;DUMP;RUN.ALGOL,TAPE $$ ###### Sender: eric@ruckus.brouhaha.com From: Eric Smith Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <3F8D89D1.472003E0@earthlink.net> Organization: Eric Conspiracy Secret Labs X-Eric-Conspiracy: There is no conspiracy. Date: 15 Oct 2003 20:08:32 -0700 Message-ID: Lines: 20 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 64.62.206.2 X-Trace: 15 Oct 2003 20:19:00 -0700, 64.62.206.2 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!HSNX.atgi.net!cyclone-sf.pbi.net!216.218.192.242!news.he.net!news.kjsl.com!news.spies.com!64.62.206.2 Xref: redlance.franklin.ch alt.folklore.computers:153470 "John Varela" writes: > Obviously SAGE didn't need interrupts because SAGE didn't have interrupts. I wrote: > Well, then it was pretty clever of the hardware designers to omit them. > I'm not under any illusion that it saved the taxpayers money, but it may > at least have increased the profits of the vendor with the contract (IBM?) > by a few thousand dollars. jchausler writes: > Not the reason. Remember this is the 50's. SAGE was > based on vacuum tubes and as a system handling a real > time process, running flat out at the technology level of the > time doing a very specific process. Handling interrupts tends > to be expensive in both hardware and time due to the need > at some level to save machine state and then restore it at the > end of the interrupt. How does that contradict my claim that omitting interrupts if they aren't needed saves (someone) money? ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 16 Oct 03 09:24:05 -0800 Organization: http://extra.newsguy.com Lines: 40 Message-ID: <1641.419T290T5643523@kltpzyxm.invalid> References: <3F8D89D1.472003E0@earthlink.net> NNTP-Posting-Host: p-706.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!newsfeed.berkeley.edu!ucberkeley!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews1 Xref: redlance.franklin.ch alt.folklore.computers:153495 In article eric-no-spam-for- me@brouhaha.com (Eric Smith) writes: >"John Varela" writes: > >> Obviously SAGE didn't need interrupts because SAGE didn't have >> interrupts. > >I wrote: > >> Well, then it was pretty clever of the hardware designers to omit >> them. I'm not under any illusion that it saved the taxpayers money, >> but it may at least have increased the profits of the vendor with the >> contract (IBM?) by a few thousand dollars. > >jchausler writes: > >> Not the reason. Remember this is the 50's. SAGE was >> based on vacuum tubes and as a system handling a real >> time process, running flat out at the technology level of the >> time doing a very specific process. Handling interrupts tends >> to be expensive in both hardware and time due to the need >> at some level to save machine state and then restore it at the >> end of the interrupt. > >How does that contradict my claim that omitting interrupts if they >aren't needed saves (someone) money? It doesn't. But I doubt that this cost was much of a consideration, especially since it would be a drop in the bucket compared to the cost of the entire project. Remember, this is the Department of Defense. If they really had to pinch pennies they'd probably cover the cost by cutting a couple of toilet seats from the budget. -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### Message-ID: <3F8EEA31.B6CB4D19@earthlink.net> From: jchausler X-Mailer: Mozilla 4.72 [en] (Win95; I) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <3F8D89D1.472003E0@earthlink.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 80 Date: Thu, 16 Oct 2003 19:00:56 GMT NNTP-Posting-Host: 63.178.89.148 X-Complaints-To: abuse@earthlink.net X-Trace: newsread2.news.atl.earthlink.net 1066330856 63.178.89.148 (Thu, 16 Oct 2003 15:00:56 EDT) NNTP-Posting-Date: Thu, 16 Oct 2003 15:00:56 EDT Organization: EarthLink Inc. -- http://www.EarthLink.net Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news-fra1.dfn.de!npeer.de.kpn-eurorings.net!news-out.nuthinbutnews.com!propagator2-sterling!In.nntp.be!newsfeed1.easynews.com!easynews.com!easynews!elnk-pas-nf1!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread2.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153498 Hi Eric, Eric Smith wrote: > "John Varela" writes: > > Obviously SAGE didn't need interrupts because SAGE didn't have interrupts. > > I wrote: > > Well, then it was pretty clever of the hardware designers to omit them. > > I'm not under any illusion that it saved the taxpayers money, but it may > > at least have increased the profits of the vendor with the contract (IBM?) > > by a few thousand dollars. > > jchausler writes: > > Not the reason. Remember this is the 50's. SAGE was > > based on vacuum tubes and as a system handling a real > > time process, running flat out at the technology level of the > > time doing a very specific process. Handling interrupts tends > > to be expensive in both hardware and time due to the need > > at some level to save machine state and then restore it at the > > end of the interrupt. > > How does that contradict my claim that omitting interrupts if they aren't > needed saves (someone) money? It doesn't... I was saying that the reason SAGE didn't have interrupts was not one of cost. (Any again, I'm not sure when "interrupts", and they may have been called something else at first, were first used or even thought of.) First off, it doesn't appear that any "cost" was much of a consideration for SAGE. What I am saying is that from what I have read on the subject, getting the SAGE system to work required many "tricks" (i.e. good engineering design) given the limits of the technology. It appears that money was spent freely. There were overlapping functions performed by the hardware but scanning was used to evaluate inputs. To back this up, I stated that in the early 70's when I started designing "real time" control systems using early minicomputers this was still a problem even though these machines (DG Nova) supported interrupts and thus, interrupts weren't used, direct scanning was. Machines today, of course, are much faster and as interrupts are a useful approach and might be used now for such an environment. "Real time", however, is a relative term and for sure some "real things" are still too fast for computers to follow directly. Then some external hardware (which can be made faster than software) is needed. For instance (one I know of) to capture the events of a nuclear explosion. The computer goes and reads the "hardware" long after (in the time frame of the explosion) the event is over in order to record what happened. As you stated, the designers were very clever but it wasn't because they omitted interrupts. They had to be very clever to get the job done at all with what they had to work with, technology wise, at the time. Then, of course, there's the argument that SAGE as a system was a failure as by the time it was installed the threat was no longer from incoming bombers which it could handle but incoming missiles which it couldn't. This, of course, forgets the deterrent value of it may have been the reason no bombers tried. Whatever you believe here, to get it to do what it did was no small feat. There was nothing else remotely like it of that scope at that time. Interrupts had nothing to do with it. Regards, Chris AN GETTO$;DUMP;RUN,ALGOL,TAPE $$ ###### Sender: eric@ruckus.brouhaha.com From: Eric Smith Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <3F8D89D1.472003E0@earthlink.net> <1641.419T290T5643523@kltpzyxm.invalid> Organization: Eric Conspiracy Secret Labs X-Eric-Conspiracy: There is no conspiracy. Date: 16 Oct 2003 16:33:28 -0700 Message-ID: Lines: 16 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 64.62.206.2 X-Trace: 16 Oct 2003 16:44:06 -0700, 64.62.206.2 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.tele.dk!news.tele.dk!small.news.tele.dk!news-stoc.telia.net!news-stoa.telia.net!telia.net!nntp.inet.fi!inet.fi!ash.uu.net!newssorter-3001.bay.webtv.net!news.spies.com!64.62.206.2 Xref: redlance.franklin.ch alt.folklore.computers:153522 "Charlie Gibbs" writes: > It doesn't. But I doubt that this cost was much of a consideration, > especially since it would be a drop in the bucket compared to the > cost of the entire project. Remember, this is the Department of > Defense. If they really had to pinch pennies they'd probably cover > the cost by cutting a couple of toilet seats from the budget. But why would they even bother designing interrupts in, if they weren't helpful? It would take more time, which translates to more engineering cost. I think both engineering and productions costs were likely to be significant factors in deciding what went into the system. After all, if cost wasn't a factor, they could have put megabytes of memory on the things. They could have run hundreds of CPUs in parallel. They could have done all sorts of interesting things, some of which would have actually made the system more useful. ###### Sender: eric@ruckus.brouhaha.com From: Eric Smith Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> Organization: Eric Conspiracy Secret Labs X-Eric-Conspiracy: There is no conspiracy. Date: 16 Oct 2003 16:36:39 -0700 Message-ID: Lines: 16 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 64.62.206.2 X-Trace: 16 Oct 2003 16:47:16 -0700, 64.62.206.2 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.telusplanet.net!newsfeed.telus.net!news3.optonline.net!ash.uu.net!newssorter-3001.bay.webtv.net!news.spies.com!64.62.206.2 Xref: redlance.franklin.ch alt.folklore.computers:153523 jchausler writes: > I was saying that the reason SAGE didn't have > interrupts was not one of cost. If cost wasn't the basis for deciding to omit features that aren't essential to the system, what basis would be used for such decisions? For example, it cost some non-zero amount of money to add ashtrays to the consoles, which clearly were not essential, but the cost was fairly low. Interrupts were also non-essential, but the cost would have been fairly high. So if not on the basis of cost, why would they choose to include ashtrays but not interrupts? ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 16 Oct 03 22:52:43 -0800 Organization: http://extra.newsguy.com Lines: 37 Message-ID: <1144.419T2195T13725488@kltpzyxm.invalid> References: <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> NNTP-Posting-Host: p-220.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.belwue.de!newsfeed.arcor-online.net!nntp-relay.ihug.net!ihug.co.nz!logbridge.uoregon.edu!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews4 Xref: redlance.franklin.ch alt.folklore.computers:153534 In article eric-no-spam-for- me@brouhaha.com (Eric Smith) writes: >jchausler writes: > >> I was saying that the reason SAGE didn't have >> interrupts was not one of cost. > >If cost wasn't the basis for deciding to omit features that aren't >essential to the system, what basis would be used for such decisions? > >For example, it cost some non-zero amount of money to add ashtrays to >the consoles, which clearly were not essential, but the cost was fairly >low. > >Interrupts were also non-essential, but the cost would have been fairly >high. > >So if not on the basis of cost, why would they choose to include >ashtrays but not interrupts? I think that in those days, tobacco was a lot more important than interrupts. :-) OS/MFT OS/MVT LS/MFT TGIF -- from Computer Lib by Ted Nelson -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: "Jim Mehl" Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Fri, 17 Oct 2003 02:06:05 -0700 Organization: The Diamond Lane Lines: 10 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> NNTP-Posting-Host: host-66-81-125-118.rev.o1.com X-Trace: news.tdl.com 1066381648 3195 66.81.125.118 (17 Oct 2003 09:07:28 GMT) X-Complaints-To: abuse@tdl.com NNTP-Posting-Date: Fri, 17 Oct 2003 09:07:28 +0000 (UTC) X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1158 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1165 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!nntp-relay.ihug.net!ihug.co.nz!news-out.newsfeeds.com!propagator2-maxim!news-in-maxim.spamkiller.net!news.he.net!news.tdl.com!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153538 > > Having a discussion locally about who invented the real-time > > interrupt. I always thought Morten Astrahan (and IBM) had a patent on the interrupt. I don't know for sure. It's probably best to let sleeping dogs alone. Jim Mehl ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Fri, 17 Oct 2003 17:14:31 +0100 Organization: Titanic Enterprises Unlimited Lines: 27 Message-ID: <1066407271.357531@saucer.planet.gong> References: <3F8D89D1.472003E0@earthlink.net> <1641.419T290T5643523@kltpzyxm.invalid> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1066407272 19677 80.177.7.220 (17 Oct 2003 16:14:32 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Fri, 17 Oct 2003 16:14:32 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1165 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1158 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@192.168.69.33 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.belwue.de!newsfeed01.sul.t-online.de!newsfeed00.sul.t-online.de!t-online.de!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153559 "Eric Smith" wrote in message news:qhad80iwqf.fsf@ruckus.brouhaha.com... > "Charlie Gibbs" writes: > > It doesn't. But I doubt that this cost was much of a consideration, > > especially since it would be a drop in the bucket compared to the > > cost of the entire project. Remember, this is the Department of > > Defense. If they really had to pinch pennies they'd probably cover > > the cost by cutting a couple of toilet seats from the budget. > > But why would they even bother designing interrupts in, if they > weren't helpful? It would take more time, which translates to more > engineering cost. I think both engineering and productions costs were > likely to be significant factors in deciding what went into the system. > > After all, if cost wasn't a factor, they could have put megabytes of > memory on the things. They could have run hundreds of CPUs in > parallel. They could have done all sorts of interesting things, > some of which would have actually made the system more useful. My guess is that they were sticking to what they knew would work rather than trying to add another layer of gold plating. KISS. :) Cheers, Rupert ###### From: mark@hubcap.clemson.edu (Mark Smotherman) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 17 Oct 2003 23:48:06 -0400 Organization: Clemson University Lines: 25 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> NNTP-Posting-Host: localhost X-Trace: hubcap.clemson.edu 1066448886 27777 127.0.0.1 (18 Oct 2003 03:48:06 GMT) X-Complaints-To: usenet@hubcap.clemson.edu NNTP-Posting-Date: 18 Oct 2003 03:48:06 GMT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.tele.dk!news.tele.dk!small.news.tele.dk!msc1.onvoy!ply1.onvoy!onvoy.com!feed.news.qwest.net!emac1.ocs.lsu.edu!nntp.msstate.edu!finch!hubcap.clemson.edu!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153578 "Jim Mehl" writes: >I always thought Morten Astrahan (and IBM) had a patent on the >interrupt. I don't know for sure. It's probably best to let sleeping >dogs alone. I imagine you mean US 3,319,230 Astrahan, et al. "Data Processing Machine Including Program Interrupt Feature" filed Sept. 1956 and granted May 1967 Despite the title, a quick sample (I didn't read all 105 pages) seems to indicate that it is describing the cycle-stealing feature of SAGE. If someone else has the time, maybe they could check to make sure. As to the original invention, NBS couldn't have filed a patent on the DYSEAC (and I doubt that Leiner or Alexander pursued it on their own, but don't know of a quick way to search for inventor names prior to 1975). Also NACA couldn't have filed a patent on the 1103 mods. I imagine the Brooks patent (US 3,048,332) is the earliest patent filed wrt interrupts as we know them today, but I haven't verified this. -- Mark Smotherman, Computer Science Dept., Clemson University, Clemson, SC http://www.cs.clemson.edu/~mark/homepage.html ###### From: aek@spies.com (Al Kossow) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Fri, 17 Oct 2003 22:29:34 -0700 Organization: Apple Computer, Inc. Lines: 11 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> NNTP-Posting-Host: haxrus.apple.com X-Trace: news.apple.com 1066454973 19188 17.205.21.66 (18 Oct 2003 05:29:33 GMT) X-Complaints-To: abuse@news.apple.com NNTP-Posting-Date: Sat, 18 Oct 2003 05:29:33 +0000 (UTC) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.tele.dk!news.tele.dk!small.news.tele.dk!lnewsoutpeer01.lnd.ops.eu.uu.net!lnewsinpeer01.lnd.ops.eu.uu.net!bnewsoutpeer01.bru.ops.eu.uu.net!bnewsinpeer01.bru.ops.eu.uu.net!emea.uu.net!ash.uu.net!newssorter-3001.bay.webtv.net!news.spies.com!forum.apple.com!news.apple.com!haxrus.apple.com!user Xref: redlance.franklin.ch alt.folklore.computers:153582 In article , aek@spies.com (Al Kossow) wrote: > > > Having a discussion locally about who invented the real-time > > > interrupt. > > MIT Whirlwind (1951) I was mistaken. I was able to locate my copy of the Dec 1951 Joint AIEE-IRE Computer Conference proceedings, and there is no mention of interrupts in Everett's paper starting on pg 70. ###### From: "Jim Mehl" Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Sat, 18 Oct 2003 00:24:54 -0700 Organization: The Diamond Lane Lines: 15 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> NNTP-Posting-Host: host-66-81-120-148.rev.o1.com X-Trace: news.tdl.com 1066461971 21494 66.81.120.148 (18 Oct 2003 07:26:11 GMT) X-Complaints-To: abuse@tdl.com NNTP-Posting-Date: Sat, 18 Oct 2003 07:26:11 +0000 (UTC) X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1158 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1165 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!in.100proofnews.com!tdsnet-transit!newspeer.tds.net!gail.ripco.com!localhost!news.tdl.com!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153615 > I imagine you mean > > US 3,319,230 Astrahan, et al. > "Data Processing Machine Including Program Interrupt Feature" > filed Sept. 1956 and granted May 1967 > > Despite the title, a quick sample (I didn't read all 105 pages) seems > to indicate that it is describing the cycle-stealing feature of SAGE. > If someone else has the time, maybe they could check to make sure. Yep, that's the one I was thinking of. I have no idea whether it's a legitimate patent. But I knew and worked with Morten. Jim Mehl ###### X-Trace-PostClient-IP: 24.83.56.193 From: genew@mail.ocis.net (Gene Wirchenko) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Reply-To: genew@mail.ocis.net Message-ID: <3f917ca3.1599307@shawnews> References: <28107548.0310100936.4ebd35ef@posting.google.com> <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> X-Newsreader: Forte Free Agent 1.21/32.243 Lines: 30 Date: Sat, 18 Oct 2003 18:49:31 GMT NNTP-Posting-Host: 24.69.255.206 X-Complaints-To: abuse@shaw.ca X-Trace: pd7tw1no 1066502971 24.69.255.206 (Sat, 18 Oct 2003 12:49:31 MDT) NNTP-Posting-Date: Sat, 18 Oct 2003 12:49:31 MDT Organization: Shaw Residential Internet Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!peer01.cox.net!peer02.cox.net!cox.net!pd7cy1no!shaw.ca!pd7tw1no.POSTED!53ab2750!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153604 Eric Smith wrote: >jchausler writes: >> I was saying that the reason SAGE didn't have >> interrupts was not one of cost. > >If cost wasn't the basis for deciding to omit features that aren't >essential to the system, what basis would be used for such decisions? > >For example, it cost some non-zero amount of money to add ashtrays to >the consoles, which clearly were not essential, but the cost was fairly >low. > >Interrupts were also non-essential, but the cost would have been fairly >high. > >So if not on the basis of cost, why would they choose to include >ashtrays but not interrupts? Perhaps, the unruly cost was time or complexity instead of merely money. Sincerely, Gene Wirchenko Computerese Irregular Verb Conjugation: I have preferences. You have biases. He/She has prejudices. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Sun, 19 Oct 03 09:31:49 GMT Organization: UltraNet Communications, Inc. Lines: 29 Message-ID: References: <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> <3f917ca3.1599307@shawnews> X-Trace: UmFuZG9tSVbMhfUk9TpmPTcoNihenS1tHcFay+MZSCJqAP/nocpEG1d7ylxIgTNX X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 19 Oct 2003 10:34:01 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.belwue.de!news-stu1.dfn.de!news-fra1.dfn.de!npeer.de.kpn-eurorings.net!news-out.nuthinbutnews.com!propagator2-sterling!news-in-sterling.nuthinbutnews.com!cyclone1.gnilink.net!peer01.cox.net!cox.net!bigfeed.bellsouth.net!bigfeed2.bellsouth.net!news.bellsouth.net!elnk-atl-nf1!newsfeed.earthlink.net!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-101 Xref: redlance.franklin.ch alt.folklore.computers:153620 In article <3f917ca3.1599307@shawnews>, genew@mail.ocis.net (Gene Wirchenko) wrote: >Eric Smith wrote: > >>jchausler writes: >>> I was saying that the reason SAGE didn't have >>> interrupts was not one of cost. >> >>If cost wasn't the basis for deciding to omit features that aren't >>essential to the system, what basis would be used for such decisions? >> >>For example, it cost some non-zero amount of money to add ashtrays to >>the consoles, which clearly were not essential, but the cost was fairly >>low. >> >>Interrupts were also non-essential, but the cost would have been fairly >>high. >> >>So if not on the basis of cost, why would they choose to include >>ashtrays but not interrupts? > > Perhaps, the unruly cost was time or complexity instead of merely >money. Especially the developers' time. /BAH Subtract a hundred and four for e-mail. ###### Message-ID: <3F94035B.A367EABE@yahoo.com> From: Peter Flass X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 17 Date: Mon, 20 Oct 2003 15:45:14 GMT NNTP-Posting-Host: 24.194.50.82 X-Complaints-To: abuse@rr.com X-Trace: twister.nyroc.rr.com 1066664714 24.194.50.82 (Mon, 20 Oct 2003 11:45:14 EDT) NNTP-Posting-Date: Mon, 20 Oct 2003 11:45:14 EDT Organization: Road Runner Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!c03.atl99!news.webusenet.com!diablo.voicenet.com!24.24.2.170.MISMATCH!news-rtr.nyroc.rr.com!news-out.nyroc.rr.com!twister.nyroc.rr.com.POSTED!53ab2750!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153678 Eric Smith wrote: > > jchausler writes: > > I was saying that the reason SAGE didn't have > > interrupts was not one of cost. > > If cost wasn't the basis for deciding to omit features that aren't > essential to the system, what basis would be used for such decisions? > > For example, it cost some non-zero amount of money to add ashtrays to > the consoles, which clearly were not essential, but the cost was fairly > low. > Every additional "thing" you add to a design increases complexity and unreliability. It's just one more thing to break. A good design includes all of those features deemed essential and nothing else. That's the lesson micro$loth has yet to learn. ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 20 Oct 03 10:53:54 -0800 Organization: http://extra.newsguy.com Lines: 29 Message-ID: <1332.423T2748T6535769@kltpzyxm.invalid> References: <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> <3F94035B.A367EABE@yahoo.com> NNTP-Posting-Host: p-824.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews1 Xref: redlance.franklin.ch alt.folklore.computers:153682 In article <3F94035B.A367EABE@yahoo.com> peter_flass@yahoo.com (Peter Flass) writes: >Every additional "thing" you add to a design increases complexity and >unreliability. It's just one more thing to break. A good design >includes all of those features deemed essential and nothing else. >That's the lesson micro$loth has yet to learn. They might know it very well for all I know. But that lesson has been overridden by the marketroids, who are aiming at lusers who get into pissing matches over feature lists. To such people, pretty screens are far more important than usability, which is something they can be fooled into believing they have anyway. (Watch someone fumbling around a screen full of doodads for a while - the only thing more painful is hearing them rave about how wonderful it is.) Yes, there are offices full of serious users who might wish that the system would get out of the way. Unfortunately, the purchasing decisions are made by PHBs who are dazzled by the flashy graphics and feature lists - real usability be damned. "So it goes." -- Kurt Vonnegut -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: "Don Chiasson" Newsgroups: alt.folklore.computers References: <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> <3F94035B.A367EABE@yahoo.com> <1332.423T2748T6535769@kltpzyxm.invalid> Subject: Re: The invention of the real-time interrupt Lines: 40 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1158 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1165 Message-ID: <2qWkb.99798$mf.57487@twister01.bloor.is.net.cable.rogers.com> Date: Mon, 20 Oct 2003 19:09:50 GMT NNTP-Posting-Host: 24.157.42.82 X-Complaints-To: abuse@rogers.com X-Trace: twister01.bloor.is.net.cable.rogers.com 1066676990 24.157.42.82 (Mon, 20 Oct 2003 15:09:50 EDT) NNTP-Posting-Date: Mon, 20 Oct 2003 15:09:50 EDT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!news-out1.nntp.be!propagator2-sterling!In.nntp.be!pd2nf1so.cg.shawcable.net!residential.shaw.ca!sjc70.webusenet.com!news.webusenet.com!cyclone01.bloor.is.net.cable.rogers.com!twister01.bloor.is.net.cable.rogers.com.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153763 "Charlie Gibbs" wrote in message news:1332.423T2748T6535769@kltpzyxm.invalid... > In article <3F94035B.A367EABE@yahoo.com> peter_flass@yahoo.com > (Peter Flass) writes: > > >Every additional "thing" you add to a design increases complexity and > >unreliability. It's just one more thing to break. A good design > >includes all of those features deemed essential and nothing else. > >That's the lesson micro$loth has yet to learn. > > They might know it very well for all I know. But that lesson has > been overridden by the marketroids, who are aiming at lusers who > get into pissing matches over feature lists. To such people, pretty > screens are far more important than usability, which is something > they can be fooled into believing they have anyway. (Watch someone > fumbling around a screen full of doodads for a while - the only thing > more painful is hearing them rave about how wonderful it is.) > > Yes, there are offices full of serious users who might wish that > the system would get out of the way. Unfortunately, the purchasing > decisions are made by PHBs who are dazzled by the flashy graphics > and feature lists - real usability be damned. A lot of the features seemed to be loudly emphasized in magazine reviews. The lack of a feature - even if practically no one would ever use it, let alone need to use it - was cause to give a low rating. Readers' choices for which software to purchase were far too often based on featuritis rather than performance, price and needed capabilities. IMHO simplistic reviews played a major role in which software failed to sell. Do we blame the reviewers or the software vendors who played the game or the people who took decisions based on marginal criteria? ++Don e-mail: it's not not, it's hot. ###### From: "Glen Herrmannsfeldt" Newsgroups: alt.folklore.computers References: <28107548.0310100936.4ebd35ef@posting.google.com> Subject: Re: The invention of the real-time interrupt Lines: 30 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4927.1200 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4927.1200 Message-ID: NNTP-Posting-Host: 12.207.204.17 X-Complaints-To: abuse@comcast.net X-Trace: sccrnsc03 1066690383 12.207.204.17 (Mon, 20 Oct 2003 22:53:03 GMT) NNTP-Posting-Date: Mon, 20 Oct 2003 22:53:03 GMT Organization: Comcast Online Date: Mon, 20 Oct 2003 22:53:03 GMT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.tele.dk!news.tele.dk!small.news.tele.dk!news-out.visi.com!petbe.visi.com!newsfeed2.dallas1.level3.net!news.level3.com!crtntx1-snh1.gtei.net!news.gtei.net!chcgil2-snh1.gtei.net!news.bbnplanet.com!wn11feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!sccrnsc03.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153698 "Mark Smotherman" wrote in message news:bmqd5m$r40$1@hubcap.clemson.edu... > "Jim Mehl" writes: > >I always thought Morten Astrahan (and IBM) had a patent on the > >interrupt. I don't know for sure. It's probably best to let sleeping > >dogs alone. > > I imagine you mean > > US 3,319,230 Astrahan, et al. > "Data Processing Machine Including Program Interrupt Feature" > filed Sept. 1956 and granted May 1967 > > Despite the title, a quick sample (I didn't read all 105 pages) seems > to indicate that it is describing the cycle-stealing feature of SAGE. > If someone else has the time, maybe they could check to make sure. In S/360, the "Program Interrupt" is the interrupt due to the execution of program instructions, such as overflow, divide by zero, and invalid operation, privileged operation, addressing, specification, segment translation, page translation, and related exceptions. The more usual interrupt is the I/O interrupt when I/O devices have finished a requested operation. (A series of channel commands are executed, and an interrupt indicated when the channel program ends.) -- glen ###### Message-ID: <3F9599B3.18364448@earthlink.net> From: jchausler X-Mailer: Mozilla 4.72 [en] (Win95; I) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 65 Date: Tue, 21 Oct 2003 20:43:29 GMT NNTP-Posting-Host: 63.178.89.111 X-Complaints-To: abuse@earthlink.net X-Trace: newsread1.news.atl.earthlink.net 1066769009 63.178.89.111 (Tue, 21 Oct 2003 13:43:29 PDT) jchausler@earthlink.net NNTP-Posting-Date: Tue, 21 Oct 2003 13:43:29 PDT Organization: EarthLink Inc. -- http://www.EarthLink.net Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newshub.sdsu.edu!elnk-nf2-pas!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!stamper.news.atl.earthlink.net!newsread1.news.atl.earthlink.net.POSTED!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153766 Hi Eric, Eric Smith wrote: > jchausler writes: > > I was saying that the reason SAGE didn't have > > interrupts was not one of cost. > > If cost wasn't the basis for deciding to omit features that aren't > essential to the system, what basis would be used for such decisions? You answered your own question, "interrupts were not essential to the system". In fact, they had to spend a lot of money just getting the essentials to work. And again, in that time frame, I'm not sure interrupts were a "common feature" so no one thought to add them "just because". They were trying to get a system which would work given the level of the technology of the time and although a simple interrupt system is not very complex, remember we're talking vacuum tubes here. > For example, it cost some non-zero amount of money to add ashtrays to > the consoles, which clearly were not essential, but the cost was fairly > low. I disagree. "Everybody" smoked back then. They're not essential to a car either but they all came with them (although the "lighter" was usually an an extra cost option). Where would you have them put the ashes, on the floor. And, of course, the cost was minimal, probably less than having the janitor cleaning the ashes off the floor (I've been in meetings long ago where there were no ashtrays, the floors were tile or linoleum and folks just started dropping them on the floor, not to mention crushing out their butts on the floor. It's amazing my lungs have survived all that second hand smoke.) > Interrupts were also non-essential, but the cost would have been fairly > high. Probably not "high" but not insignificant. Once again you have to think of SAGE as a system designed to do one specific task and one task only and with limited early technology. What I have read is that the designers had great trouble getting a design which would keep up with the "real world". Given a single task to do with no goal of generality, an interrupt solution would have made the system slower with no benefit. I mean, your toaster doesn't have interrupts (or at least not likely, they're putting computers in "everything" these days) cause it doesn't need them to make toast. Since that's all it will ever have to do, why put them in at any cost. And lastly, of course, we still haven't determined that the concept of interrupts was a a "well known thing" at that time. > So if not on the basis of cost, why would they choose to include > ashtrays but not interrupts? First off, it was a no brainer. That said, I wonder if they were in the specification. I've seen some specifications over the years with a lot of "non essential" or at least "not germane to the problem being addressed" things included just because "somebody important" wanted it that way. Chris AN GETTO$;DUMP;RUN,ALGOL,TAPE $$ ###### Sender: eric@ruckus.brouhaha.com From: Eric Smith Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt References: <28107548.0310100936.4ebd35ef@posting.google.com> <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> <3F9599B3.18364448@earthlink.net> Organization: Eric Conspiracy Secret Labs X-Eric-Conspiracy: There is no conspiracy. Date: 21 Oct 2003 16:15:43 -0700 Message-ID: Lines: 16 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 64.62.206.2 X-Trace: 21 Oct 2003 16:27:14 -0700, 64.62.206.2 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!in.100proofnews.com!in.100proofnews.com!news-xfer.cox.net!peer01.cox.net!cox.net!border3.nntp.aus1.giganews.com!border1.nntp.aus1.giganews.com!nntp.giganews.com!cyclone-sf.pbi.net!216.218.192.242!news.he.net!news.kjsl.com!news.spies.com!64.62.206.2 Xref: redlance.franklin.ch alt.folklore.computers:153769 jchausler wrote: > I was saying that the reason SAGE didn't have > interrupts was not one of cost. I wrote: > If cost wasn't the basis for deciding to omit features that aren't > essential to the system, what basis would be used for such decisions? jchausler writes: > You answered your own question, "interrupts were not essential > to the system". In fact, they had to spend a lot of money > just getting the essentials to work. So presumably they chose not to spend additional money to implement unnecessary features, which was my point. ###### From: Roland Hutchinson Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Wed, 22 Oct 2003 01:17:56 -0400 Lines: 28 Message-ID: References: <3F94035B.A367EABE@yahoo.com> <1332.423T2748T6535769@kltpzyxm.invalid> <2qWkb.99798$mf.57487@twister01.bloor.is.net.cable.rogers.com> NNTP-Posting-Host: pool-151-198-151-12.nwrk.east.verizon.net (151.198.151.12) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7Bit X-Trace: news.uni-berlin.de 1066799877 31038295 151.198.151.12 (16 [99522]) User-Agent: KNode/0.6.1 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.tele.dk!news.tele.dk!small.news.tele.dk!fu-berlin.de!uni-berlin.de!pool-151-198-151-12.nwrk.east.verizon.NET!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153781 Don Chiasson wrote: > A lot of the features seemed to be loudly emphasized in magazine > reviews. The lack of a feature - even if practically no one would ever > use it, let alone need to use it - was cause to give a low rating. > Readers' choices for which software to purchase were far too > often based on featuritis rather than performance, price and > needed capabilities. IMHO simplistic reviews played a major > role in which software failed to sell. Do we blame the reviewers > or the software vendors who played the game or the people > who took decisions based on marginal criteria? You mean the reviewers who had an inherent conflict of interest because they were being paid (in part) out of the revenues from advertisements for the software products they were reviewing, and whose reviews were largely guided by press materials distributed by the software publishers? How could one possibly blame such people? They OBVIOUSLY had to be utterly morally incorruptible in order to even consider undertaking such work! It's capitalism in action, folks. People get what they are willing to pay for. -- Roland Hutchinson Will play viola da gamba for food. NB mail to my.spamtrap [at] verizon.net is heavily filtered to remove spam. If your message looks like spam I may not see it. ###### From: mark@hubcap.clemson.edu (Mark Smotherman) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: 22 Oct 2003 10:04:52 -0400 Organization: Clemson University Lines: 14 Message-ID: NNTP-Posting-Host: localhost X-Trace: hubcap.clemson.edu 1066831492 23633 127.0.0.1 (22 Oct 2003 14:04:52 GMT) X-Complaints-To: usenet@hubcap.clemson.edu NNTP-Posting-Date: 22 Oct 2003 14:04:52 GMT Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!msc1.onvoy!ply1.onvoy!onvoy.com!feed.news.qwest.net!emac1.ocs.lsu.edu!nntp.msstate.edu!finch!hubcap.clemson.edu!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153797 Does anyone know the details of how interrupts were handled on the Electrologica X-1? (Knuth credits Dijkstra with independent development of an interrupt system in 1957/58.) Also, corrections welcome on my current writeup on the history of interrupts at http://www.cs.clemson.edu/~mark/interrupts.html -- Mark Smotherman, Computer Science Dept., Clemson University, Clemson, SC http://www.cs.clemson.edu/~mark/homepage.html ###### From: bhk@dsl.co.uk (Brian {Hamilton Kelly}) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Thu, 23 Oct 2003 14:44:17 +0000 (UTC) Organization: Dragonhill Systems Ltd Lines: 17 Message-ID: <1066863197snz@dsl.co.uk> References: <3F8EEA31.B6CB4D19@earthlink.net> X-Trace: mail2news.demon.co.uk 1066920257 20919 10.0.0.1 (23 Oct 2003 14:44:17 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Thu, 23 Oct 2003 14:44:17 +0000 (UTC) X-Received: from dsl.demon.co.uk ([158.152.92.150]) by news.demon.co.uk with smtp (Exim 4.12) id 1ACghC-0005R7-00 for mail2news@news.demon.co.uk; Thu, 23 Oct 2003 14:44:15 +0000 X-Path: dsl.co.uk!bhk X-To: mail2news@news.demon.co.uk X-Newsgroups: alt.folklore.computers X-Newsreader: Demon Internet Simple News v1.31 X-Lines: 16 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!newsfeed00.sul.t-online.de!t-online.de!kibo.news.demon.net!mutlu.news.demon.net!demon!mail2news.demon.co.uk!bhk Xref: redlance.franklin.ch alt.folklore.computers:153845 In article <3F8EEA31.B6CB4D19@earthlink.net> jchausler@earthlink.net "jchausler" writes: > I was saying that the reason SAGE didn't have > interrupts was not one of cost. (Any again, I'm > not sure when "interrupts", and they may have > been called something else at first, were first used When the Ferranti Atlas came out, it had an interrupt capability. The signals themselves on the tracings were annotated LAM, for "Look At Me". -- Brian {Hamilton Kelly} bhk@dsl.co.uk "We can no longer stand apart from Europe if we would. Yet we are untrained to mix with our neighbours, or even talk to them". George Macaulay Trevelyan, 1919 ###### From: Philip Nasadowski Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Thu, 23 Oct 2003 21:52:00 -0400 Organization: Biker/Metalhead from HELL!!! Lines: 17 Message-ID: References: <28107548.0310100936.4ebd35ef@posting.google.com> <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> <3F9599B3.18364448@earthlink.net> X-Trace: UmFuZG9tSVYXhP2gC8+3r7PDO+HGkJ8Sl/3f7NvQnzLDowMv1vdGB/J7tzaYjNuu X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 24 Oct 2003 01:52:00 GMT User-Agent: MT-NewsWatcher/3.1 (PPC) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!feed2.news.rcn.net!feed3.news.rcn.net!feed1.news.rcn.net!rcn!nasadowsk Xref: redlance.franklin.ch alt.folklore.computers:153878 In article <3F9599B3.18364448@earthlink.net>, jchausler wrote: > I mean, > your toaster doesn't have interrupts (or at least not likely, they're > putting computers in "everything" these days) cause it doesn't > need them to make toast. But it'd make implementing the timer for the toaster a lot easier, at least if the microcontroller running things was a PIC... Gah analog input, PWM for firing a triac, a few lines for i/o to the time and relay and popupper and 'on' light. You could do this on a 16F870 easily. Anyway, having an interrupt everytime Timer2 rolled over running you real time clock would be easier than just doing hard loops. -- To email me, chage 'usermale' to 'usermail'. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Fri, 24 Oct 03 10:09:28 GMT Organization: UltraNet Communications, Inc. Lines: 24 Message-ID: References: <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> <3F9599B3.18364448@earthlink.net> X-Trace: UmFuZG9tSVakpKEAeZ+SyvwvJgFCiX2OImLMejNThuWsR9WBXzX+pywCrT99fPFT X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 24 Oct 2003 11:12:32 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!nntp.abs.net!feed3.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-96 Xref: redlance.franklin.ch alt.folklore.computers:153891 In article , Philip Nasadowski wrote: >In article <3F9599B3.18364448@earthlink.net>, > jchausler wrote: > >> I mean, >> your toaster doesn't have interrupts (or at least not likely, they're >> putting computers in "everything" these days) cause it doesn't >> need them to make toast. > >But it'd make implementing the timer for the toaster a lot easier, at >least if the microcontroller running things was a PIC... > >Gah analog input, PWM for firing a triac, a few lines for i/o to the >time and relay and popupper and 'on' light. You could do this on a >16F870 easily. Anyway, having an interrupt everytime Timer2 rolled over >running you real time clock would be easier than just doing hard loops. You forgot the Musak. /BAH Subtract a hundred and four for e-mail. ###### From: iddw@hotmail.com (Dave Hansen) Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Fri, 24 Oct 2003 13:36:11 GMT Lines: 36 Message-ID: <3f99289d.651119409@News.CIS.DFN.DE> References: <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> <3F9599B3.18364448@earthlink.net> NNTP-Posting-Host: mail.nartron.com (216.65.187.224) X-Trace: news.uni-berlin.de 1067002357 33263017 216.65.187.224 (16 [97677]) X-Newsreader: Forte Free Agent 1.21/32.243 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news-fra1.dfn.de!fu-berlin.de!uni-berlin.de!mail.nartron.COM!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153895 On Fri, 24 Oct 03 10:09:28 GMT, jmfbahciv@aol.com wrote: >In article >, > Philip Nasadowski wrote: [...re: computer-controlled toaster...] >>But it'd make implementing the timer for the toaster a lot easier, at >>least if the microcontroller running things was a PIC... >> >>Gah analog input, PWM for firing a triac, a few lines for i/o to the >>time and relay and popupper and 'on' light. You could do this on a >>16F870 easily. Anyway, having an interrupt everytime Timer2 rolled over Way overkill. A 12F675 would be plenty (as long as you can live with only 5 or 6 I/O lines). Of course, you'd have to generate the PWM yourself (no CCP unit), which would require an interrupt (at least at higher frequencies). >>running you real time clock would be easier than just doing hard loops. If you did have a PIC with a CCP, you could still avoid interrupts by simply polling a free-running timer. Avoiding inpterrupts can greatly simplify the code. > >You forgot the Musak. I haven't heard about any PICs running a (software) MP3 decoder, but there are reports of units running TCP/IP. Even a book out there with source code... Regards, -=Dave -- Change is inevitable, progress is not. ###### From: Roland Hutchinson Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Fri, 24 Oct 2003 11:48:33 -0400 Lines: 31 Message-ID: References: <3F9599B3.18364448@earthlink.net> NNTP-Posting-Host: pool-151-198-151-12.nwrk.east.verizon.net (151.198.151.12) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7Bit X-Trace: news.uni-berlin.de 1067010514 32908341 151.198.151.12 (16 [99522]) User-Agent: KNode/0.6.1 Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!pool-151-198-151-12.nwrk.east.verizon.NET!not-for-mail Xref: redlance.franklin.ch alt.folklore.computers:153899 jmfbahciv@aol.com wrote: > In article > , > Philip Nasadowski wrote: >>In article <3F9599B3.18364448@earthlink.net>, >> jchausler wrote: >> >>> I mean, >>> your toaster doesn't have interrupts (or at least not likely, >>> they're putting computers in "everything" these days) cause it >>> doesn't need them to make toast. >> >>But it'd make implementing the timer for the toaster a lot easier, at >>least if the microcontroller running things was a PIC... >> >>Gah analog input, PWM for firing a triac, a few lines for i/o to the >>time and relay and popupper and 'on' light. You could do this on a >>16F870 easily. Anyway, having an interrupt everytime Timer2 rolled >>over running you real time clock would be easier than just doing hard >>loops. > > You forgot the Musak. And the AM radio. -- Roland Hutchinson Will play viola da gamba for food. NB mail to my.spamtrap [at] verizon.net is heavily filtered to remove spam. If your message looks like spam I may not see it. ###### From: Philip Nasadowski Newsgroups: alt.folklore.computers Subject: Re: The invention of the real-time interrupt Date: Fri, 24 Oct 2003 23:36:22 -0400 Organization: Biker/Metalhead from HELL!!! Lines: 51 Message-ID: References: <3F8D89D1.472003E0@earthlink.net> <3F8EEA31.B6CB4D19@earthlink.net> <3F9599B3.18364448@earthlink.net> <3f99289d.651119409@News.CIS.DFN.DE> X-Trace: UmFuZG9tSVaZc3MFljZHCpltBbW4izUrFfaxpky3RC5pb7j0a13cMjuLHRk+JeSG X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 25 Oct 2003 03:36:23 GMT User-Agent: MT-NewsWatcher/3.1 (PPC) Path: redlance.franklin.ch!pfaff2.ethz.ch!news-zh.switch.ch!irazu.switch.ch!switch.ch!news.maxwell.syr.edu!nntp.abs.net!feed3.news.rcn.net!feed1.news.rcn.net!rcn!nasadowsk Xref: redlance.franklin.ch alt.folklore.computers:153918 In article <3f99289d.651119409@News.CIS.DFN.DE>, iddw@hotmail.com (Dave Hansen) wrote: > Way overkill. A 12F675 would be plenty (as long as you can live with > only 5 or 6 I/O lines). Of course, you'd have to generate the PWM > yourself (no CCP unit), which would require an interrupt (at least at > higher frequencies). Ok ok, I'm not a big fan of the 12 series :) But I bet one would do the trick fine here. > If you did have a PIC with a CCP, you could still avoid interrupts by > simply polling a free-running timer. That's a little messy. > Avoiding inpterrupts can greatly simplify the code. Can. But not always. My concept of a RTC on a PIC would be a 2 byte (prefferably in bank 0) counter that's incremented with every processor int, and a timer set to int at a known rate. Thus, the ISR wouldn't need to do anything with W, though STATUS would be lost. (at least I think it'd be lost). Thus, if we assume that the timer's the only thing that'll interuppt the PIC, it's maybe 4 or 5 lines or so of code. Then all you're doing in your waiting loop is waiting for a match of your RTC Vs the your time value. Actually, this is probbably too simplistic for a toaster, I origionally thought of this for a N20 timer for a friend's Harley. If you could get a timer that interrupts every .25 sec or so (yes, I know, but then, how fast does this thing need to run anyway), it'll work though. Hey, this is theoretical. I'm spolied by how cheap the 16F870 is :/ I suppose if you wanted bare minimum, a 12 could do it, and you could do a lot of magic in SW, or just dump PWM and temerature regulation and just have a timer. But then, a 555 does the trick, and is argueably simpler :) I wonder what the optimal temperature for toast is, anyway? > I haven't heard about any PICs running a (software) MP3 decoder, but > there are reports of units running TCP/IP. With a PWM channel, audio playback on a PIC is easy, though MP3 might be out of the realm of a PIC's abilities. Then again... > Even a book out there with source code... Must be a fun read :) -- To email me, chage 'usermale' to 'usermail'.