Newsgroups: comp.databases.pick Path: bga.com!news.sprintlink.net!pipex!uunet!world!jws From: j...@world.std.com (jim w stephens) Subject: Re: Advanced Revelation Message-ID: Organization: The World Public Access UNIX, Brookline, MA References: <3c0bae$abg@ixnews2.ix.netcom.com> Date: Wed, 7 Dec 1994 01:06:03 GMT Lines: 51 In article , Henry Eggers wrote: >David Sigafoos (d...@ix.netcom.com) wrote: >: hegg...@netcom.com (Henry Eggers) writes: >: >David Sigafoos (d...@ix.netcom.com) wrote: >: >: hegg...@netcom.com (Henry Eggers) writes: >: >: >Garrett D. Hildebrand (g...@crl.com) wrote: >: >: >: apk (74774.1...@CompuServe.COM) wrote: >: >: >: : >Man, this is getting like the CI$ forum. I might have to check >: >: >: : >out of here too! >: >: >: : > > >This has been a problem since Microdata sold the first few machines, >decided that selling systems was fun, and went off to build a 'real' >system, called Express. Express didn't happen, but not for want of >spending most of the corporate development resources on it. ... .... many comments deleted > I know that the current alumni of MDCSC or whatever the thing should be called have a validly low opinion of Microdata. However, for all the problems that Don Fuller had and caused, he had an idea that was ahead of its time. Express was basically what the market finally defined as your basic IBM PC done with 1974 era mini hardware. It was very compact for a mini, and would have delivered much of what a basic business PC delivered. It was not for the lack of much argument on the part of some of us in engineering that the base provided by Reality was left out of the Express picture. However, they did and the rest as they say is history... I do note that the lack of support did pay a lot of Microdata programmers salaries for many years before the Pick market as we know it was born with multiple licensees to work for etc. Express was sold to Olivetti who did sell more than $1 billion in hardware derived from the basic Express design... >Regards hve. Jim Stephens j...@world.std.com (asbestos suit is on BTW) -- -------------------------------------------------------------------- Jim Stephens j...@world.std.com Irvine, Ca -------------------------------------------------------------------- Path: bga.com!news.sprintlink.net!pipex!uunet!nntp.crl.com!crl5.crl.com!not-for-mail From: g...@crl.com (Garrett D. Hildebrand) Newsgroups: comp.databases.pick Subject: Re: Advanced Revelation Date: 6 Dec 1994 23:04:45 -0800 Organization: Farcaster Portal Eight (No Tourists!) Lines: 99 Message-ID: <3c3mqd$bpj@crl5.crl.com> References: <3c0bae$abg@ixnews2.ix.netcom.com> Reply-To: g...@ACM.org NNTP-Posting-Host: crl5.crl.com X-Newsreader: TIN [version 1.2 PL2] jim w stephens (j...@world.std.com) wrote: : In article , : Henry Eggers wrote: [...] : many comments deleted : I know that the current alumni of MDCSC or whatever the thing : should be called The thing is constantly mutating: Microdata Microdata and CMC Microdata and CMC "on McDonnell Douglas" Microdata and CMC as one, "Microdata," "on McDonnell Douglas" McDonnell Douglas (Blue Sector, Gold Sector, et al...) McDonnell Douglas (fill in the blank, such as in: - Computer Systems Company (MDCSC) - Field Service Company (MDFSC) - International Systems, Inc. (MDISI) (Field Service was sold, became Novadyne) (Manufacturing was sold, leaving Marketing and Sales and Engineering) (Drafting and Hardware Engineering was gutted, leaving Henry and CO.) McDonnell Douglas Computer Systems Company (MDCSC) and MDISI (UK) "Jerry-Co" (No name could be had for any price!) McDonnell Information Systems Company (UK London Stock Exchange) (MDIS) (U.S. Software Engineering "ROS" group [Henry and CO.] is gutted) Today, it, whatever the heck it is, is known as MDIS. MDIS does PRO-IV... Good doggie. Twang your magic twanger (fill it in) : have a validly low opinion of Microdata. Which part? : for all the problems that Don Fuller had and caused, Don Fuller appears to have liberated Pick to do what he did! : he had an : idea that was ahead of its time. All Pick folks take note of this small history lesson... A long, long time ago, when Microdata and Dick Pick went separate ways, Microdata (Don Fuller) decided to try to build a microcomputer which would stand of its own merit and support all mainstream stuff of those days (read: COBOL, Fortran) and introduce new programming languages (read: MPL (Microdata Programming Language). It was called Express. A stack-architecture 32-bit machine with memory- mapped I/O. The maintenance panel on that thing would allow one to write programs with switches. It was cool. Mr. Stevens was a MAJOR player on this machine. For historical note, I worked on this machine in the Pilot/Manufacturing area with John Denney and several others. It was some machine. But wait! REALITY was EASIER to use. When Express was finally beginning to look good, Microdata had cash flow problems, and sold the rights to it to Olivetti. My bet is that MDC told them to, because MDC had just _bought_ Microdata. [time frame: circa 1980] : Express was basically what the : market finally defined as your basic IBM PC done with 1974 era : mini hardware. It was very compact for a mini, and would have : delivered much of what a basic business PC delivered. : It was not for the lack of much argument on the part of some : of us in engineering that the base provided by Reality was : left out of the Express picture. It was due to Don Fuller, N'est Pas? : However, they did and the : rest as they say is history... I do note that the lack of support : did pay a lot of Microdata programmers salaries for many years : before the Pick market as we know it was born with multiple : licensees to work for etc. : Express was sold to Olivetti who did sell more than $1 billion in : hardware derived from the basic Express design... : >Regards hve. : Jim Stephens : j...@world.std.com : (asbestos suit is on BTW) Garrett Hildebrand (similar asbestos suit is on, NFS) -- ---------------------------------------------------------------------------- Garrett Hildebrand | "'This man wishes to put a spell on us, for he is c.d.p. Monitor | immersing himself in water,' and they compel[led] g...@ACM.org | him to pay a fine." --Crichton, "Eaters of the Dead" Newsgroups: comp.databases.pick Path: bga.com!news.sprintlink.net!howland.reston.ans.net!gatech!udel!news.mathworks.com!news.kei.com!world!jws From: j...@world.std.com (jim w stephens) Subject: Re: Advanced Revelation Message-ID: Organization: The World Public Access UNIX, Brookline, MA References: Date: Thu, 8 Dec 1994 22:00:21 GMT Lines: 102 In article , Henry Eggers wrote: >jim w stephens (j...@world.std.com) wrote: >: In article , >: Henry Eggers wrote: >: >David Sigafoos (d...@ix.netcom.com) wrote: >: >: hegg...@netcom.com (Henry Eggers) writes: >: >: >David Sigafoos (d...@ix.netcom.com) wrote: >: >: >: hegg...@netcom.com (Henry Eggers) writes: >: >: >: >Garrett D. Hildebrand (g...@crl.com) wrote: >: >: >: >: apk (74774.1...@CompuServe.COM) wrote: >: >: >: >: : Dick Thiot said: >: >: >: >: : >Man, this is getting like the CI$ forum. I might have to check >: >: >: >: : >out of here too! >: >: >: >: : > >: > >With no intent to offend the people who worked on Express, with >the possible exception of those got paid for years for working >on Reality, while bitching interminably about how horrible, in >all conceivable and inconceivable ways Reality was, without any >suggestions as to what might be done about it, other than flushing >it, so far as I have been able to discover. Express could have run a Pick over Unix quite easily. EMOS was a simpler kernel than the Unix kernel, and could have run Reality processes. I wrote several processes that could have been models for such a task. > >Express, as I understand it, was a little bigger than a PC clone. > ... [deleted stuff] >Express was more like Dec meets IBM, with a 'standard' file system The file system had random access, sequential access and full keyed ISAM capabilities and outran all Unix systems at the time since our code was inside the kernel and theirs was plastered on top of the kernel. It did not have a "cheap" mode like Unix despite much argument that it was needed. You can't win all the arguments. The company that wrote the Cobol is now MicroFocus Cobol i believe... I don't defend the choices here, but they have proven rather good choices for business, even if they are abhorant to we worker bees in the trenches. > >So, Jim, with all due respect, they were trying to build a small, fast, >relatively inexpensive minicomputer (that's before UK marketing get's >their hands on the price book and multiplies by ...), which had all UK never entered into the deal. They were still owned by Pertec >the 'standard' features. On which, see today's previous comment in >this thread. But it wasn't a PC. Let me see... a small system (for 77 it was small) costing $10000 with business applications (full ansi cobol and fortran 77). wasn't that what IBM proposed in 1982 when they set out to do the PC? I said that it was the PC for 1977 technology, not what one would do with more resources 5 years later. Microprocessors were toys in those days and could not run an OS of any size reliably. I remember the early Mentor with its z8000's that would run for 2 or three min at a time. > >As I remember it, it was announced at a New York show (NCC?) (which would >have been April-June time frame) for delivery in '77 for $20k starting. >(This is based on the most fragmentary bits of gossip collected over >the last two decades.) And it didn't materialize. But I never heard It could be that the Reality group wasn't the only group that got the shaft at Microdata... I think I heard the current chorus of griping starting around 76 from a different set of bodies in the same building. >: Express was sold to Olivetti who did sell more than $1 billion in >: hardware derived from the basic Express design... > >They DID? Wow. I'd no idea. What do they call it? Why couldn't >John McDonnell do this? MD did not have the Dutch banks as clients. > >Thanks, Jim, and can you fill in some of the history, so I can >get over the view which I acquired while I was at Pick? > I will spare the pick people any more history but will be glad to offline... See you at plaza de cafe's next week sometime? >Regards, hve. -- -------------------------------------------------------------------- Jim Stephens j...@world.std.com Irvine, Ca -------------------------------------------------------------------- Newsgroups: comp.databases.pick Path: nntp.gmd.de!newsserver.jvnc.net!nntpserver.pppl.gov!princeton!rutgers!att-out!pacbell.com!ames!lll-winken.llnl.gov!enews.sgi.com!decwrl!netcomsv!netcom.com!heggers From: hegg...@netcom.com (Henry Eggers) Subject: Re: Advanced Revelation Message-ID: Organization: NETCOM On-line Communication Services (408 261-4700 guest) X-Newsreader: TIN [version 1.2 PL1] References: Date: Tue, 13 Dec 1994 18:14:50 GMT Lines: 140 jim w stephens (j...@world.std.com) wrote: : Henry Eggers wrote: : >jim w stephens (j...@world.std.com) wrote: : >: Henry Eggers wrote: : >: >David Sigafoos (d...@ix.netcom.com) wrote: : >: >: hegg...@netcom.com (Henry Eggers) writes: : >: >: >David Sigafoos (d...@ix.netcom.com) wrote: : >: >: >: hegg...@netcom.com (Henry Eggers) writes: : >: >: >: >Garrett D. Hildebrand (g...@crl.com) wrote: : >: >: >: >: apk (74774.1...@CompuServe.COM) wrote: : >: >: >: >: : Dick Thiot said: : >: >: >: >: : >Man, this is getting like the CI$ forum. I might have : >: >: >: >: : >to check out of here too! : >: >: >: >: : > : >: > : >on Reality, while bitching interminably about how horrible, in : >all conceivable and inconceivable ways Reality was, without any [Which, of course, it was.] : Express could have run a Pick over Unix quite easily. EMOS was : a simpler kernel than the Unix kernel, and could have run Reality : processes. I wrote several processes that could have been models : for such a task. Ok, so does my intuition suggest that EMOS (Erasable Metal Oxide ...) (Elfa Micro OS?) (Sorry, too much fun) is more like Multics? On the other hand, I know very little about the Unix kernel, because I suspect they've got all the wrong things on the inside, and vica versa, and it would just upset me. Of course EMOS was more sensible; it wasn't written by graduate students in something else in the night on 'requsitioned' machines, glued together on Arpanet. : > : >Express, as I understand it, was a little bigger than a PC clone. : > : The file system had random access, sequential access and full keyed : ISAM capabilities and outran all Unix systems at the time since our : code was inside the kernel and theirs was plastered on top of the : kernel. It did not have a "cheap" mode like Unix despite much : argument that it was needed. You can't win all the arguments. Unix doesn't have a file system. I'm not sure what to call it... On line disk storage allocation? Better than going and getting some sectors from the MISdirector's little black book? Did it need a 'cheap' mode like Unix? : I don't defend the choices here, but they have proven rather good : choices for business, even if they are abhorant to us worker bees : in the trenches. Boy, is _that_ an interesting question: Can there be software which both the market and the programmers like? : > : >So, Jim, with all due respect, they were trying to build a small, fast, : >relatively inexpensive minicomputer (that's before UK marketing get's : >their hands on the price book and multiplies by ...), which had all : UK never entered into the deal. They were still owned by Pertec Unnnnh... CMC UK was owned by Ray Parry, who'd bought it from CMC Santa Monica, in their first round of divestiture (to pay debt), in the early '70's(?); Pertec, the onshore company, is partially or wholly the result of the second round, a couple of, some, years later. They started shipping Reality, it is said in '75, and were bought by Don Fuller mid(?)-'77. My comment about UK marketing -- a market as lush as Dutch banks, it would appear (to jump ahead). : >the 'standard' features. On which, see today's previous comment in : >this thread. But it wasn't a PC. : Let me see... a small system (for 77 it was small) costing $10000 : with business applications (full ansi cobol and fortran 77). wasn't : that what IBM proposed in 1982 when they set out to do the PC? Really. Interesting. When a Reality cost $100,000, and 8k of memory, $8000 (list, 3k street). Lisa cost $10000, and that was years later. : I said that it was the PC for 1977 technology, not what one would : do with more resources 5 years later. Microprocessors were toys : in those days and could not run an OS of any size reliably. I remember : the early Mentor with its z8000's that would run for 2 or three min : at a time. That (in the fall of '81) was probably because I was showing Harold Ch....? various problems with serial printers in the hospitality suite. Harold was not amused. The other problem, of course, with early 16k chips, was Parity. They had parity checking. When a parity error would happen (less often than every 3 minutes, usually), the screen would display 'Parity Error' and stop. Adds got in a bit of twist about this (nothing like the 'dumb-ass record fixed by retix' flap), because the customers said that they never had this problem on Reality. Once Tim mentioned to them that Reality didn't check for parity errors, and that that's where the GFEs came from (mostly the 'FF's, which we were still finding late in the 7.0 machines, ariving from file save tapes) the Adds folks felt better. Then there was the problem that the programmers at Adds added half of each patch sent them, because 'We've already modified the other frame.' By the end of '82 the machine ground to halt, so Markell had to go fix it, which he did, in Hophog, while, I suspect, the Adds software management was off skiing.... :-) In any case the microprocessor worked just fine. It was everything else which was a bit otl. : > : >As I remember it, it was announced at a New York show (NCC?) (which would : >have been April-June time frame) for delivery in '77 for $20k starting. : >(This is based on the most fragmentary bits of gossip collected over : >the last two decades.) And it didn't materialize. But I never heard : It could be that the Reality group wasn't the only group that got the : shaft at Microdata... I think I heard the current chorus of griping : starting around 76 from a different set of bodies in the same building. So, Jim, when did they really start the Express program? When did each of the cancellations take place? Each of the resucitations? And I'm sure that everyone got the shaft. So, do you have a view on how they thought they were going to stay in business? : >: Express was sold to Olivetti who did sell more than $1 billion in : >: hardware derived from the basic Express design... : > : >They DID? Wow. I'd no idea. What do they call it? Why couldn't : >John McDonnell do this? : MD did not have the Dutch banks as clients. Would English police departments do? : > : I will spare the pick people any more history but will be glad : to offline... See you at plaza de cafe's next week sometime? Two part answer: The publishable stuff does just fine here; the other stuff we can have with some chewing fat over lunch. Now that it's Tuesday, when's good for you for lunch? Regards, hve. (4,195,835 * 256)/(3,145,727 / 256) Path: nntp.gmd.de!newsserver.jvnc.net!nntpserver.pppl.gov!princeton!rutgers!sgigate.sgi.com!olivea!uunet!news1.digex.net!ts.digex.net!not-for-mail From: j...@orange.digex.net (james w stephens) Newsgroups: comp.databases.pick Subject: Re: Advanced Revelation -> MD history Date: 14 Dec 1994 12:32:14 -0800 Organization: Express Access Online Communications, Greenbelt, MD USA Lines: 203 Message-ID: <3cnkof$h3s@orange.digex.net> References: NNTP-Posting-Host: orange.digex.net [Many levels of citation deleted] :Ok, so does my intuition suggest that EMOS (Erasable Metal Oxide ...) :(Elfa Micro OS?) (Sorry, too much fun) is more like Multics? On :the other hand, I know very little about the Unix kernel, because I :suspect they've got all the wrong things on the inside, and vica :versa, and it would just upset me. Express's kernel worked quite well managing a segmented machine architecture. The system did not have paging, but only segments and swapping. The process control was provided for in firmware which made it nice. :Of course EMOS was more sensible; it wasn't written by graduate :students in something else in the night on 'requisitioned' machines, :glued together on Arpanet. :: > :: >Express, as I understand it, was a little bigger than a PC clone. :: > :: The file system had random access, sequential access and full keyed :: ISAM capabilities and outran all Unix systems at the time since our :: code was inside the kernel and theirs was plastered on top of the :: kernel. It did not have a "cheap" mode like Unix despite much :: argument that it was needed. You can't win all the arguments. : :Unix doesn't have a file system. I'm not sure what to call it... :On line disk storage allocation? Better than going and getting some :sectors from the MISdirector's little black book? Did it need a :'cheap' mode like Unix? A system that is potentially going to host a system like Reality that already has its own file system needs to protect users from leaving the confines of allocated files on the host, but do little else. :: > : [some more deletions] :So, Jim, when did they really start the Express program? When did each :of the cancellations take place? Each of the resucitations? And I'm :sure that everyone got the shaft. So, do you have a view on how they :thought they were going to stay in business? Microdata had a business selling the equivalent of microprocessors to people. Their core business was selling 1600's to Basic4, their various mid level customers, and I think perhaps another third to a customer group like Microform who used them to do 411 work and Decision data that was one of ADP dealer services competitors before Pick. They had a business doing Pick that was brought in house by Max Malone, and was sold thru the Dealer network, who really called the shots, not thru MD and a salesman. All business inside Microdata had to be traced to a sales guy to be interesting, and the Reality stuff did not have someone making a commission. I may be wrong on this, but I think that there was no single guy getting a commission to generate a focus on the Reality business, hence the odd attitude to it. Express had several people with big stakes in it and they could drive the company to make it happen (or not as they later proved). Most of the fits and spurts of the Express project can be traced to the marketing and sales groups and various hot ideas (Gee wee need an Express, Don't we have a nice ready to go 32 bit micro in the basement?) Off you go with a machine that had a lot going for it, but needed some basic improvements before taking it to market, and going off half cocked. Express was going and in trouble when I started at MD in June 1976. The friday night massacre occurred in Feb 77. Olivetti came into the picture later that year. After the February meeting with all the express customers (there were 30 or so) five did not have the sense to cancel their orders, or that would have been it right there. Those five took possession of maybe 7 machines total. Olivetti took maybe 10 (Garrett?) . : : :Two part answer: The publishable stuff does just fine here; the :other stuff we can have with some chewing fat over lunch. Now that :it's Tuesday, when's good for you for lunch? : : [Many levels of citation deleted] : :: : Ok, campers, the question is, Why do the ad hoc things work, and the best :: : laid plans auger in? Or is this true? : :Express... : :: Because the software can leapfrog the dead or dying hardware, like :: a flea to some new host,.. : Not so with Express, : :A fascinating question due some consideration. : :: The boards were very sensitive to noise and timing, and the sucker ran :: so hot that if the components from a vendor weren't dead-center in the :: spec that a raft of problems would ensue. : :A close cousin of the 1600... A distant cousin of the 1600. The 1600 was a Ken Omohundro machine, and the 3200 was a Rick Reed machine. The 3200 hardware made many assumptions about the macro machine it was executing as to its instruction set and had 5 hardware stack registers that were manipulated by side affects to register movements having to do with moving operands to the hardware and memory. It was a cmos memory machine only, and had vastly different sets of side effects in the way the data moved. Much closer to Sequil or other 2900 based system that was wrapped around the upper machine, code but retained extensive microcode. The 1600 ran among other things: 1621/1610/810 firmware (blink and you would think it was a Varian 520/720) 32k Max memory, ordinary mini instruction set. Process control, automation + jim stephens education in how to microprogram... (MPS O/S) (this is only included to demonstrate my bias on things. The usual response to this machine is a yawn) Lisp machine (at University of Mo Rolla) IBM 1130 via Cliff Myers IBM 360 (No I/O) Basic 4, a variety of 1621 extended to have better features. Custom Processor for Mead Digit (included a nifty front panel card that was firmware supported to be indistingusable from the machine that was replaced.) Reality version 1 taken from 820 hardware version which had 16k max memory. Almost no variation possible in I/O (4 ports maybe 8?) Reality Version 2 (B1) This is 2.x reality firmware which ran quite a large variety of disks, 2 tape controllers, 1 8way 64k Memory Reality Version 3.x with expansion to include hardware registers, and more firmware space (and assorted other crap) ran faster (2x?) 128k memory Reality Version 4.x add 512k memory, and used modified Micro 1 single tick, single card cpu design rather than the original cpu's 2 tick design (anybody remember CPH1 and CPH2? which was the I/O clock) Anybody even heard of a Micro One? (I have 1 that works) R77 (Pick's version of the above) Supported 32 Ports R80 A cleaner version of the R77. Supported 64 ports Irvine Computer's R80 2 card version of R80. The 1600 was a very general and fast micromachine. : :: Interestingly, the second failure of Express was the diagnosticians, who :: were unable to replicate in diagnostics what abuse the OS could dish out. : :Never seen it not to be the case. Run Adventure (cf Tom West). : :: [Jim, JD, rememeber that blond babe?] : :Tell us more :-) : :: : entrails of Microdata, MD bought Microdata in mid-'79 -- probably with :: : the expectation the Express would replace all McDonnell's IBM :: : and DEC requirements, which were a lot.) Express was history (or nightmare) at that point. All interest was in Reality at that time, and Express was an operation to appease the original customers, and satisfy Olivetti : I heard that the approach was made to McD to open a line of operation credit via a private placement of stock, and McD's response was "oh, we couldn't loan that much, but we could buy you outright". I think they also had angles related to the large number of machines used by the McD health care systems division, and to a tiny degree the management services for phone company's (now owned by EDS), and since they still fancied diversification into the info processing field at McD wanted to build great and wonderful things at MD. :A late flash has it that they bought it to solve their internal word :processing problem.... and Mr Jim wanted something for John to do. :So, it is said, McAuto went with Wang instead, 'making' Wang... No, :if _that_ was the plan...? Anybody have any knowledge of people :working on an MD wordprocessing proposal in 1980? Textpro? No :offense meant, but I don't think they had wordmunch in mind. : : :Regards, hve. (4,195,835 * 256)/(3,145,727 / 256) : : Please respond by email if this is boring. Sorry for the length if it is. Jim Path: nntp.gmd.de!newsserver.jvnc.net!netnews.upenn.edu!news.amherst.edu!news.mtholyoke.edu!uhog.mit.edu!news.mathworks.com!hookup!swrinde!howland.reston.ans.net!vixen.cso.uiuc.edu!uwm.edu!lll-winken.llnl.gov!decwrl!nntp.crl.com!crl3.crl.com!not-for-mail From: g...@crl.com (G. D. Hildebrand) Newsgroups: comp.databases.pick Subject: Re: Advanced Revelation -> MD history Date: 16 Dec 1994 00:38:39 -0800 Organization: Farcaster Portal Eight (No Tourists!) Lines: 31 Message-ID: <3crjmf$8jj@crl3.crl.com> References: <3cnkof$h3s@orange.digex.net> Reply-To: g...@ACM.org NNTP-Posting-Host: crl3.crl.com X-Newsreader: TIN [version 1.2 PL2] james w stephens (j...@orange.digex.net) wrote: : [Many levels of citation deleted] [to get to the point I wanted to respond to, Jim's incredibly good history was clipped here...] : After the February meeting with all the express customers (there were 30 : or so) five did not have the sense to cancel their orders, or that would : have been it right there. Those five took possession of maybe 7 machines : total. Olivetti took maybe 10 (Garrett?) . I think maybe it was eleven, with a bunch of odd parts. But hey, that is a vague memory, so don't bet on the number. (JD?) [clip the rest...] : Please respond by email if this is boring. Sorry for the length if it is. : Jim Not so far! :-) Regards, Garrett -- ---------------------------------------------------------------------------- Garrett Hildebrand | "'This man wishes to put a spell on us, for he is | immersing himself in water,' and they compel[led] g...@ACM.org | him to pay a fine." --Crichton, "Eaters of the Dead" Path: bga.com!news.sprintlink.net!howland.reston.ans.net!pipex!uunet!nntp.crl.com!crl4.crl.com!not-for-mail From: g...@crl.com (Garrett D. Hildebrand) Newsgroups: comp.databases.pick Subject: Re: Advanced Revelation Date: 8 Dec 1994 23:20:05 -0800 Organization: Farcaster Portal Eight (No Tourists!) Lines: 105 Message-ID: <3c90f5$k9s@crl4.crl.com> References: <3c0bae$abg@ixnews2.ix.netcom.com> Reply-To: g...@ACM.org NNTP-Posting-Host: crl4.crl.com X-Newsreader: TIN [version 1.2 PL2] Henry Eggers (hegg...@netcom.com) wrote: : jim w stephens (j...@world.std.com) wrote: : : Henry Eggers wrote: : : >David Sigafoos (d...@ix.netcom.com) wrote: : : >: hegg...@netcom.com (Henry Eggers) writes: : : >: >David Sigafoos (d...@ix.netcom.com) wrote: : : >: >: hegg...@netcom.com (Henry Eggers) writes: : : >: >: >Garrett D. Hildebrand (g...@crl.com) wrote: : : >: >: >: apk (74774.1...@CompuServe.COM) wrote: : : >: >: >: : Dick Thiot said: {...} : Ok, campers, the question is, Why do the ad hoc things work, and the best : laid plans auger in? Or is this true? Well, REALITY/Pick has always been a virtual machine, and so that death of some hardware is not the end of the world. Because the software can leapfrog the dead or dying hardware, like a flea to some new host, all of the folks interested in the well-being of the software can continue to refine it. Not so with Express, which had - at the macro level - three incantations, as I recall. Express I, Express X, Express III. All were wet dreams by some fairly talented hardware types, and I very much liked the maintenance panel on the Express III, which allowed awesome control over the machine and very nice troubleshooting capability (not to mention the ability of writing all of memory in one tape block to tape with a few select switch combinations, as John Denney will attest to). Interlude: the Express X, incorrectly called the Express Ten, was actually 'X' because nobody was sure what to call it, but the name stuck. Jim, If I have gotten this wrong, please do correct me here in cdp so the oral history is correct. This is my best recollection. But, back to the main swerve: what killed Express was Express! I remember that the EO's came down from on high just so fast and furious that an additional rework person had to be hired. Let's see... there was Rose, and Maria, who would just as soon spit in your eye as do rework (but damn, she was good!), then they hired Don, who rode in from Mission Viejo on a bicycle and looked like a clone for one of the members of ZZ-Top. Anyway, the rework was done with green wire, and the etch side of the boards had so much wire on them they looked like a spider's nest. We used to have trouble getting the boards into the machine without snapping off green-wire. The boards were very sensitive to noise and timing, and the sucker ran so hot that if the components from a vendor weren't dead-center in the spec that a raft of problems would ensue. Interestingly, the second failure of Express was the diagnosticians, who were unable to replicate in diagnostics what abuse the OS could dish out. Legion are the stories about machines which passed all of the diagnostics then crapped out during normal use. The killer test was not a test at all, but a sort of ATP for Express, what we called an... EMOS run da, da da da, daaaaaaaaaaaa! It simply crashed any flakey hardware. Now, who do you suppose would wander down to the Pilot floor or the Manufacturing floor to go through the entrails of core dump (memory dump to printer)? None other than Jim Stevens. Ol' Jim would sit there pouring through this stuff, all in ASCII-HEX, and after some period of time he would say, "It's a hardware problem, probably the," [framitz] or whatever the problem was. He was right eight of ten times. Every so often he would find some kernel bug. ANYWAY, getting back to the earlier point, that of Express killing Express, the Olivetti folks, our biggest (and perhaps only) Express III customer, was in the process of helping us figure out just what the hardware problems were when Microdata got bought by the big MDC. I think that is the catalyst that caused Express to be sold to Olivetti. As Jim said, Olivetti made good with the system. They redesigned the hardware so that it worked. BTW, IMHO (YMMV) I think that was also the end of Express because the software went with the hardware, and the Italians eventually lost interest. [Jim, JD, rememeber that blond babe?] : : Express was sold to Olivetti who did sell more than $1 billion in : : hardware derived from the basic Express design... : They DID? Wow. I'd no idea. What do they call it? Why couldn't : John McDonnell do this? (For those less than interested in the : entrials of Microdata, MD bought Microdata in mid-'79 -- probably with : the expectation the Express would replace all McDonnell's IBM : and DEC requirements, which were a lot.) : Thanks, Jim, and can you fill in some of the history, so I can : get over the view which I acquired while I was at Pick? : Regards, hve. Regards, gdh. -- ---------------------------------------------------------------------------- Garrett Hildebrand | "'This man wishes to put a spell on us, for he is c.d.p. Monitor | immersing himself in water,' and they compel[led] g...@ACM.org | him to pay a fine." --Crichton, "Eaters of the Dead" Newsgroups: comp.databases.pick Path: nntp.gmd.de!newsserver.jvnc.net!nntpserver.pppl.gov!princeton!rutgers!att-out!pacbell.com!ames!hookup!newshost.marcam.com!news.mathworks.com!europa.eng.gtefsd.com!howland.reston.ans.net!news.moneng.mei.com!uwm.edu!lll-winken.llnl.gov!enews.sgi.com!decwrl!netcomsv!netcom.com!heggers From: hegg...@netcom.com (Henry Eggers) Subject: Re: Advanced Revelation Message-ID: Organization: NETCOM On-line Communication Services (408 261-4700 guest) X-Newsreader: TIN [version 1.2 PL1] References: <3c0bae$abg@ixnews2.ix.netcom.com> <3c90f5$k9s@crl4.crl.com> Date: Tue, 13 Dec 1994 18:35:51 GMT Lines: 51 Garrett D. Hildebrand (g...@crl.com) wrote: : Henry Eggers (hegg...@netcom.com) wrote: : : jim w stephens (j...@world.std.com) wrote: : : : Henry Eggers wrote: : : : >David Sigafoos (d...@ix.netcom.com) wrote: : : : >: hegg...@netcom.com (Henry Eggers) writes: : : : >: >David Sigafoos (d...@ix.netcom.com) wrote: : : : >: >: hegg...@netcom.com (Henry Eggers) writes: : : : >: >: >Garrett D. Hildebrand (g...@crl.com) wrote: : : : >: >: >: apk (74774.1...@CompuServe.COM) wrote: : : : >: >: >: : Dick Thiot said: : {...} : : Ok, campers, the question is, Why do the ad hoc things work, and the best : : laid plans auger in? Or is this true? Express... : Because the software can leapfrog the dead or dying hardware, like : a flea to some new host,.. : Not so with Express, A fascinating question due some consideration. : The boards were very sensitive to noise and timing, and the sucker ran : so hot that if the components from a vendor weren't dead-center in the : spec that a raft of problems would ensue. A close cousin of the 1600... : Interestingly, the second failure of Express was the diagnosticians, who : were unable to replicate in diagnostics what abuse the OS could dish out. Never seen it not to be the case. Run Adventure (cf Tom West). : [Jim, JD, rememeber that blond babe?] Tell us more :-) : : entrails of Microdata, MD bought Microdata in mid-'79 -- probably with : : the expectation the Express would replace all McDonnell's IBM : : and DEC requirements, which were a lot.) A late flash has it that they bought it to solve their internal word processing problem.... and Mr Jim wanted something for John to do. So, it is said, McAuto went with Wang instead, 'making' Wang... No, if _that_ was the plan...? Anybody have any knowledge of people working on an MD wordprocessing proposal in 1980? Textpro? No offense meant, but I don't think they had wordmunch in mind. Regards, hve. (4,195,835 * 256)/(3,145,727 / 256)