Computer Industry Mourns DEC Founder Ken Olsen 172
alphadogg writes "Kenneth Olsen, the computer industry pioneer who co-founded and led minicomputer king Digital Equipment Corp. for 35 years, died at the age of 84 on Sunday in Indianapolis. As DEC's leader, Olsen oversaw the company's epic battles vs. IBM and its mainframes for the hearts and business of IT shops – a fight DEC eventually lost as the era of fast, cheap and networked PCs took hold in the 1980s and 1990s. During its heyday, DEC's PDPs, VAXes and DECnet network technology became staples in many organizations, and today's IT industry remains filled with companies whose founders once worked at DEC or with its gear. Digital was acquired in 1998 by Compaq. Dan Bricklin, co-creator of the VisiCalc spreadsheet and DEC alum, tweeted: 'Ken Olsen is in the elite club of tech founders w/Gates & Jobs, and set the stage for them. What he did we take for granted today.'"
Farewell (Score:3)
Re: (Score:2)
I have been looking around for a first person lunar lander but so far no such luck. I installed a simple 2D lunar lander game from the android marker but its not as good as that 1969 effort.
Re: (Score:3)
Re: (Score:1)
Re: (Score:2)
It wasn't long after those Fortran, Pascal, and COBOL classes that we were administering said DEC VAX computers. I think I can still find a record of one of my DECUS program submissions ("MCLS" #V00200) out there somewhere. That was back when we called "open source" by the name "public domain".
Similar (Score:2)
I played my first game of Dungeons & Dragons at the DEC office in Quad Cities, Iowa.
It was quite the experience for a little kid.
Windows NT = VMS (sort of) (Score:4, Interesting)
I do wonder what would have happened if DEC hadn't been taken over by the dead hand of Compaq. After all, IBM still sell plenty of big iron systems and there's a definite need these days for highly reliable and secure systems - of the type DEC made - for eCommerce applications.
Re: (Score:2)
I think that DEC had lost it long before they were taken over by Compaq. One thing was cosmetic: they insisted that they were called "Digital" instead of the name everybody knew and loved them by, "DEC". But more importantly, they put their effort into increasingly large VAXes instead of the low end machines they had made their fortune on. They killed off the PDP-11 line, and had to bring it back because of customer demand. They has made their fortune on relatively simple boxes that people could use and abu
Re: (Score:2)
Yup, it was odd that the summary said they lost the battle against IBM - they didn't lose it, the battle became irrelevant in the face of cheap commodity PCs. Mainframes are still very profitable for IBM, but they've gone from being the way that any company that needs computing power supplies it to being a very small niche market.
The Alpha did well though. There was a time when anyone who wanted a fast computer got an Alpha. Even with the overhead of emulating x86, it could often run Windows applicati
Re: (Score:2)
I have sort of a love/hate relation with VMS. It has some really nice ideas that are well engineered. VMS was just chock full of corporate stuff and was "enterprise" before that word came into vogue, whereas Unix had much more of a "let's get 'er done" approach. Ie, VMS was suit and tie, whereas Unix was tie-dye. It was often a struggle to get simple things done under VMS, whereas keeping-it-simple was the underlying theme in Unix.
I know that a VAX-780 in my university supported many more users simultan
Re: (Score:2)
But nobody who didn't really need it had a shelf-full of Vax book.
I still have a couple boxes of orange VMS manuals in binders, from when I did a VMS version upgrade for the robotics lab where I worked. Rather than toss the obsolete manuals, I brought them home... 24 years ago. And the last time I touched or logged in to a VMS system was probably 22 years ago.
Re: (Score:2)
Anyone who has ever worked with VMS (and RSX for that matter) know what programming freedom you had, and what the real programming is.
Yes. The freedom to spend all afternoon trying to count single and double quotes in DCL, making sure that they're properly mismatched in order to accomplish what you're trying to do.
And let's not forget QIO. "Hey, let's wrap an entire OS worth of functionality in one entry point, and just use a ginormous variable parameter list to sort it out!"
Re: (Score:2)
DEC like TI blew in a way blew it because they didn't want to kill their markets.
DEC had single chip versions of the PDP11 the F11 cpu. They could have combined them with one of their single users OSs and produced a computer that was more powerful and with a larger software base then CP/M or the PC had at launch.
DEC could have sold the CPUs and OS to other manufactures after the launched the first machine and became Intel and MIcrosoft combined.
Thing was they didn't want to not sell high margin PDP-11s to p
Re: (Score:2)
The market for the older minicomputer companies was still around, but in the workstation front. Digital just didn't do very well there with the VAXstations and DECstations; their later Alphas were good but a bit too little too late. But Sun managed to grow the small workstation market at the right time, so that the newcomer muscled out the veteran.
Digital's version of Unix, Ultrix was just a bit odd in some ways (though not as goofy as AIX) but SunOS felt more like a traditional Unix. What really got Sun
Re: (Score:2)
Ahh, to be the geek again;
And play around with old SYSGEN
And users who would worship you
Because I gave them TPU.
Goodnight, Mr. Olsen, sleep well :(
Except that VMS was rock solid, NT - not so much (Score:2)
Something obviously got lost in the translation to x86.
Re: (Score:2)
While you can criticize Windows all you want for the baroque Win32 API and it's assorted GUI and userland tools, there's no denying that the kernel itself is quite robust. In fourteen years (since NT 4.0), I've never seen a Windows machine crash that wasn't due to faulty hardware or bad third party drivers, a lord knows I've seen, used and setup a LOT of Windows boxen. For the amount of features and hardware it supports, I think it is an achievement that deserves more respect than it gets.
That's where I rec
Re: (Score:2)
Something obviously got lost in the translation to x86.
KESU. The four-tier address/command relationship responsible for VAX/VMS' armour-plated security. Intel couldn't support it, thus x86 user mode programs were writing to places they well nigh shouldn't. Welcome to the world of buffer overrun exploits.
Re:Windows NT = VMS (sort of) (Score:4, Interesting)
They didn't just "hire DEC engineers". David was one of the architects of VMS, and there were serious lawsuits about the theft of copyrighted material and trade secrets, which were settled out of court. (Why do you think NT ran on Alpha chips for a long time?)
DEC thought they were better off continuing to innovate instead of trying to bring down Microsoft's law-breaking new monopoly in court. But meanwhile Intel was stealing the Alpha CPU technologies for the Pentium chip: between the loss of both of those leading edge technologies, and their less effective but still cost-effective re-unication by the thieves in the "Wintel" architectures, DEC had little left to work with.
DEC made a lot of fabulous technologies which I found useful in the lab and in industry, but they weren't good at protecting their best assets from intellectual theft, and they refused to embrace the open source approach of various UNIX and later Linux technologies, and it left them without a large enough market to continue to build such wonderful hardware.
Re: (Score:2)
BTW, does anyone remember the OpenVMS/Alpha based laptop? It did exist - developed by a third party.
Wow! How long did it run on one charge of refrigerant?
Re: (Score:2)
The biggest legacy, to me, of DEC is Linux. Without a certain PDP-7 computer and some extremely brilliant engineers, Unix would never have existed, and Linus and his buddies would have nothing to "borrow" from.
Re: (Score:2)
The VMS parts are in the Kernel, not in WIN32. Probably the most recognizable parts would be the priority schemes and paging which are not exactly VMS but definitely inspired along that style. Most Windows programmers I think don't get down to the Kernel level and treat WIN32 as the lowest level. The original NT design technically allowed using something other than WIN32, like their POSIX layer (completely useless except for marking off a box on a checklist, since all the useful facilities you needed wer
Re: (Score:2)
fun fact (maybe just a coincidence after all, or an urban legend, but fun anyway ) : It was said that Windows NT naming comes from VMS.
(hint : VMS successor => WNT)
Re: (Score:2)
So
Rest In Peace, Mr Olsen (Score:2, Interesting)
I didn't had that much the priviledge of meeting your personnally, I only have a few memories of your very common car lost 'somwhere' on the parking lot at ZK-01 and you, just looking for it. Or spending a much-than-expected time you spent with the FS engineers setting up the systems for DECville in Cannes. Or even you fixing the washing machine of your neighbour's mother which neighbour was a DEC employee who almost had a stroke when she saw her CEO kneeled in soapy water in her basement. Obviously, you ma
Re: (Score:2)
zk - ha! spitbrook. been 15 or more years since I thought of that site.
my first 6 years in computing were with DEC. after college in boston (early 80's) - DEC was *the* place to apply for if you were comp-sci. everyone wanted to work at DEC and you'd be happy to get in any way you could (heh, I started at EduServices then found my way back into engineering).
I'm now out in calif (left DEC to relo out here in the 90's) and while I've worked for some giants (cisco, sgi, places like that) - nothing was anyt
Had a good innings (Score:3)
84 or as I prefer to say it, 124.
Re: (Score:2)
Re: (Score:2)
160010 400
160020 410
160030 420
160040 430
160050 440
160060 450
160070 460
160100 470
160110 500
and so on. Its the bus address and interrupt vectors for DZ11 MUX cards at my last job. Sorry, its just burned into my brain and I couldn't resist. On the PDP you could pretty much assume that a number would be octal, if it had anything to do with the hardware. If a card failed and brought the CPU down the register dump would almost always give you the bus address of the card.
Re: (Score:2)
Heh, try keeping up with a distro that do its release numbering in octal...
Re: (Score:2)
That's the reason for using zero as prefix... if he used 0124 it would be explicit.
Re: (Score:2)
That's the reason for using zero as prefix... if he used 0124 it would be explicit.
The hell it would! Some of us would like to use zero for padding decimal output, but we can't because when you then feed it into anything that uses libc atoi() or strto*(), the leading '0's make it get interpreted as octal. That, and *printf()'s inability to print a numerical value as a bit-string (i.e. "1111" (binary) in addition to "017" (octal), "f" (hex), and "15" (decimal)), are two of my biggest pet-peeves with C.
Re: (Score:2)
Octal was great for many of the PDPs though, because it used 3 bits for a lot of the instruction fields that lined up with octal nicely (ie, you could look at an instruction and pick out the registers that were being used).
Re: (Score:2)
DEC scared IBM in the 80's (Score:5, Interesting)
I remember reading an article in The Economist about DEC and their VAXes in the 80's. The point was the a VAX was cheap enough that a low level executive could approve the expenditure. An IBM mainframe purchase would require approval at the top executive level of the company. IBM responded by bringing out a mini-mainframe called the 9370 as a "VAX killer," but it was a flop. The minicomputer was killed by PCs. However, IBM still makes a lot of money with their mainframes, with folks who have tons of data, and need high availability: like banks and insurance companies.
For DEC they could have gone downscale to PCs, but the profit margins are too low: it's a commodity item. IBM doesn't build PCs anymore; they sold their PC business to Lenovo. Or they could have gone upscale, to compete with IBM mainframes. In the 90's, big Sun servers were causing IBM some grief. But we all see what happened to Sun.
I like to have choice. So the more vendors that are out there, the better. When I look at the passenger airplane industry, there are only two choices: Airbus or Boeing. I would welcome more competition, from say, Japan or Russia. Russia!?!?! Well, their Soyuz is the only way to get into space now, so they could probably be able to build good passenger airplanes.
Re: (Score:2)
Well into the 1990s the IBM minicomputer ran OS/2. It was okay as a department level server. DEC also had failures at the low end. There was the MicroVAX and a micro VMS to go along with it, but technology caught up too fast so cut down systems were soon not required.
Re: (Score:2)
> Well into the 1990s the IBM minicomputer ran OS/2.
System/1, System/34, System/36, and the AS/400 most certainly did not use OS/2 as their operating system(s).
sPh
Re: (Score:2)
No but they had kind of a high end PC running OS/2 as a server. Notes at my site used AIX as the main server with this smaller OS/2 boxes scattered around. Wide area networks then weren't what they are now.
Re: (Score:2)
Well into the 1990s the IBM minicomputer ran OS/2
Huh? When did IBM ever run OS/2 on anything other than PCs?
Re: (Score:2)
Well into the 1990s the IBM minicomputer ran OS/2
Huh? When did IBM ever run OS/2 on anything other than PCs?
Maybe I was too liberal with the term "minicomputer". I was referring to the high end PC hardware which OS/2 was deployed on to function as a local or department level server.
Re: (Score:2)
Actually IBM had plans for a micro-kernel called IBM Workplace OS or WPOS: http://en.wikipedia.org/wiki/IBM_Workplace_OS [wikipedia.org] . One of the ideas was to get it running on other non-PC architectures, so OS/2 could run on them. It was developed at IBM Boca Raton in Florida, where OS/2 came from. They actually shipped a product, OS/2 Warp 4, that ran on the micro-kernel.
I talked to an ex-IBM AIX developer years after the whole project was shit-canned. He told me he attempted to use WPOS with AIX on their POWER
Re: (Score:2)
LoB
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
but then getting a decent VT emulator running was very difficult
Ha! a former co-worker of mine (he could be reading) wrote a VT compatible emulator for windows. He went by the username "lec" so his emulator was called LECTerm.
Re:DEC scared IBM in the 80's (Score:4, Informative)
For DEC they could have gone downscale to PCs, but the profit margins are too low: it's a commodity item. IBM doesn't build PCs anymore; they sold their PC business to Lenovo.
You have forgotten the DEC Rainbow. But that's ok, everyone else has also forgoten the Rainbow.
Re: (Score:2)
You have forgotten the DEC Rainbow. But that's ok, everyone else has also forgoten the Rainbow.
Which is sad really. It was a dual-processor system - a Zilog Z80 and an Intel 8080 CPU. When it ran CP/M the Z80 did everything, but when it ran MS-DOS the 8080 was the primary CPU and the Z80 handled the IO.
The architecture was even better thought through and didn't break up the RAM like the IBM PC did (hence the 640K "limit"). I remember booting my Rainbow 100B and getting 720KB of usable RAM without trying very hard.
Sadly, the only real games that got ported to it were the Zork line of Infocom games,
Re: (Score:2)
My high school received a Rainbow the year after I graduated. I recall going in to look at it or help set it up. I believe it had a Pascal compiler, though I might be misremembering that.
Re: (Score:2)
Anyone remember "SCRAM"?
I remember it: it was the second video game I ever played, and the first one in color. The first game was LADDER on the DEC Rainbow. (I would have been like four at the time.)
Unfortunately there don't seem to be any DEC Rainbow emulators out there, meaning that all those old Rainbow games (...both of them...) are lost to time, I guess.
Re: (Score:2)
Oh, I remember the Rainbow.
Q: What's the difference between a DEC Rainbow and a bowling ball?
A: There's more software for the bowling ball.
Re: (Score:1)
Actually it was the PDP that got introduced under the nose of IBM. That's why it was called "Peripheral Data Processor", it could fly under the radar as office equipment. VAX was much later, after the DecSystem 10 and 20 (odd 36-bit machines).
The Dec Rainbow was the first DEC PC has a "real" PC unlike the previous Robin (a VT100 with the guts replaced with a PC board). That machine was dual processor and coud be taken apart and put back together with a dime. Very cool architecture. But DEC just didn't take
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
When I look at the passenger airplane industry, there are only two choices: Airbus or Boeing.
They're the ones you see more usually in the US and Europe, I suppose, but there are others, such as Embraer, Bombardier, Sukhoi, Tupolev, and Antonov.
Re: (Score:1)
Re: (Score:2)
For DEC they could have gone downscale to PCs, but the profit margins are too low: it's a commodity item. IBM doesn't build PCs anymore; they sold their PC business to Lenovo.
People, including Gordon Bell, like to mention Ken's "Nobody needs a computer in their home" quote to say he lacked the vision to get into the PC market. I don't agree - as much as it saddens me, the home computer was a fad that had died out by the 1990s. People have PCs, which is a business machine, in their home offices.
While DEC's PC efforts failed, this 1982 movie shows that they did try:
http://www.youtube.com/computerhistory#p/u/31/YKbnbvF_2Ew [youtube.com]
This approach, trying to leverage their PDP-11 and PDP-8 tec
The biggest little company in the world. (Score:5, Interesting)
Happy days,
We had the easynet, Dec's internal network, and we did Notes conferencing. I remember trying to explain to people about sitting stateside, dealing with my UK email and getting blank looks. Then we had a notes conference called 'the house' where each topic was someone's room. It felt pioneering back then in the 80's.
You could ask for help on the net, and get help. Then they grew too fast and brought in middle managers who blocked innovation.
We built some great things, global systems with cluster failover, self healing networks, global sync waves, bleeding edge leading edge database technology, all on VMS which was truly elegant.
That's when I really learnt how to build stuff.
Ken used to have a stuffed beaver in his office (now now) chewing a tree, the tree represented IBM.
I remember him acknowledging his biggest commercial mistake, which was when Bell Labs offered him Unix for free if he would only support it.
Goodbye Ken
Re: (Score:2)
VMS which was truly elegant
VMS is still very much alive and kicking. It is running on modern Itanium hardware, it is still fully supported by HP, and will continue to be supported for many years to come.
It has support for IPv6, and it's as elegant (and inscrutable) as ever.
You can even get personal use licenses for free if you have the hardware to run it.
I do miss DEC (Score:1)
Re: (Score:3)
Compaq really killed DEC after the takeover.
We'll see what's left over from Sun in a few years, after Larry Ellison is finished with it. After the Compaq takeover, I met some former DEC employees, who were then Compaq employees, in a hotel on a business trip. After a few drinks, they had some blunt advice: "If your company gets taken over by another company, quit as soon as you can, and get a job somewhere else. Takeovers always end in tears."
I am still scratching my head over how DEC ended that way: from being the heavyweight champion in the Uni
Thanks for the career Ken (Score:1)
Rest in peace Ken. Your DEC gear introduced me to computing. It changed my career path and it's enriched my life more than anything else I can think of. I wish I'd had the chance to say so in person. I still have a PDP-11 running RSX-11. I might run it for a little while this afternoon.
Jeff
Re: (Score:3)
Yes I have a DEC 3000 alpha server here. One of these days I want to find out if it really will make a good boat anchor. One thing I remember about the PDP 11/84s is that the electronics vastly outlasted the rubber padding inside the top cover of the CPU box. The rubber turned to dust and fell on to the backplane. This was okay until you re-seated a card, then the dust fell into the slots and you had to vacuum the whole thing out.
I have fond memories of lying prone under an 11/84 with a wire wrap tool in my
Rest in peace man (Score:1)
Re: (Score:2)
Yeah that had me puzzled for a bit. The IT department where I used to work would stick bar codes on to a VT240 because it had two parts and must be a computer, but they left a VT320 alone because that was just a "monitor". My hands ache at the thought of a VT100. Horrible keyboard.
Beginnings have Ends (Score:4, Interesting)
When I went off to college I thought I'd grow up to be a physicist. I loved the science, but half way through the first semester I was discovering that physics was not working for me. Struggling with calculus I was directed to a room filled with what I thought were tv screens. They were monitors hooked to a PDP 11/45 which had just been installed at the school, replacing the IBM mainframe.
In that moment, sitting in front of that terminal and working my first program, I fell in love with computers. I loved how I could imagine something, then create it. Working on the PDP introduced me to the mini world, programming, and my career. While never knowing the man, the mind that conceived and created the DEC PDP family will be one I certainly honor and respect.
RIP Ken (Score:2)
Re: (Score:2)
Loved that PDP-8. We ran mumps on it back in the 70's in a lab. Loved coding away on it.
Buried face down nine edge first?? (Score:3)
Im thinking that he would be one of the last folks this would apply to.
Of course there is a problem of the shrinking number of folks that will get the reference also.
Re: (Score:2)
Except, he founded DEC. The Hollerith card is an IBM invention, dating back to the late 20's.
Re: (Score:1)
Re: (Score:2)
The VAX (Score:2)
Back in my school days, we used to talk about the mythical VAX in hushed tones, due to its awesomeness (at least we thought so). I never really used one in anger, they were already on their way out by the time I left school and went to university, but the uni still had a VAX cluster (on which we were forced to write COBOL, which soured the experience somewhat).
I actaully have a small VAX now, something I take to vintage computing shows and use as a fileserver for a network of Sinclair Spectrums...
VAX 6000 Clusters (Score:1)
Inspiration (Score:4, Interesting)
I'm super inspired by the math. He was 31 when he founded his company. All we ever seem to hear about are the impossible situations of being born into wealth, stalking through the ivy league, founding a government funded start-up by age 18, (having the 'rents boot the bill for) article of incorporation at age 20 and being (due to the misled, ignorant millions) in charge of some pointless "dot-com" by 23-25. Here we have an innovator who saw an inroad at a certain date -- he could have been in his 40's or 50's when it happened but he got "lucky" -- and followed it, carried through with his idea using determination and resolve, saw his vision fulfilled and had the fun he predicted he would in elbowing aside giants like IBM. It could happen to anybody! The economy doesn't need to be in the shitter. Anybody can go back to college, re-socialize, swing and actually hit the ball, sometimes out of the park. That's something that will never, ever, ever be heard of again in a country that allows itself to lapse into one (1) complete generation of Gimme-Jobber clones. We're mere fractions away from being in that exact, dire situation, and right now is probably our last chance at a strong economy with our independence intact. We have to do like Ken Olson, stop trying to "look for a job", stop trying to compete-by-rote (dislodge the 24/7 vee-dee-yo holo-game controller implant) and relearn to socialize and do sound business with integrity and grit. Our country is turning into a bunch of antisocial, passive-aggressive fucktards with chips on their shoulders and not even the brains to know what the fuck they're such douchebags for in the first place, with tarnished, discount-antique-store, silver spoons up their asses. A bunch of whiney fucking nobodies looking up to Hollywoodization as the key to all knowledge, more film-reel upstairs than just plain real.
the founder of Route 128 (Score:1)
Back in the heyday of the "Massachusetts Miracle" ('80s), road signs bragged: "Route 128, America's Technology Highway". I think it started with Lincoln Labs, and MIT was the big feeder school (Harvard really came into play much later, when Zuckerman came along), but Olsen gave it the big push. Wave after wave of minicomputer and workstation startups sprung up around Route 128 and I-495 (beltways around Boston), often founded or staffed by ex-Digital engineers, starting with Data General. Many achieved
And remember, children: If it's not 36 bits... (Score:5, Funny)
PDP-10 into eternity!
----
@info ver
Bamboo Forest of the Lost, Eientei TOPS-20 Monitor 7.1(21733)
PANDA TOPS-20 Command processor 7.1(4453)-4
@systat
Tue 8-Feb-2011 07:43:09 Up 1958:51:50!
0+9 Jobs Load av 0.03 0.01 0.01
No operator in attendance
A real gentleman (Score:2)
I was fortunate enough to meet the man a few times during my short stay at DEC in the 80's. He was very gracious, intelligent and committed to the company. Ken was on the Ford Motor Company Board of Directors so he came to Detroit fairly often. I remember at one point he gave Henry Ford II a Rainbow PC and one of the guys I worked with had to go install it at the Duce's mansion. Henry gave Ken an Escort station-wagon which he drove for several years. Rest in Peace Ken.
R.I.P. Ken Olsen (Score:5, Interesting)
I worked at DEC for over a decade, in the 70's and early 80's and was there when he was forced out. The company was very much built around his charisma -- he was a big man, unassuming, but very charismatic -- and even in remote field offices, every new employee would soon know who the president of the company was, and hear a few stories about how he embarrassed one of the local sales reps by speaking too bluntly to a customer. Unless you were in sales, these were considered proof the President was a good guy, one of "us".
Once I had the good fortune to be able to visit the Mill, in Maynard, Mass, with a few others on training. On Friday when the class let out early, we wandered the complex (it was a campus of interconnected buildings), visiting the clock tower, and asking people where Ken Olsen's offices were.
Well, we found the executive offices, and tentatively asked one of the secretaries, which was Ken's. She pointed it out, and then, to our horror, picked up the phone and asked if he would come out and meet us. Son of a bitch, he did. He took the time to come out and shake our hands and speak to us lowly field employees, and he seemed as interested in meeting us as we were to meet the man himself.
When he left, it wasn't the same company. DEC had some serious marketing challenges at the time, granted, but I don't think many appreciate the technology it had. VMS in the 80s was a better operating system than any flavor of Unix, today. You could write programs with modules in C, Fortran, Cobol, Basic, or just about any other language, mix and match, and the architecture supported that. VMSclusters in the 80's were far easier to configure and run, and more functional than any Unix cluster I've seen today. The Alpha architecture had legs for twenty years, maybe more.
I was sorry to hear about your passing Ken, and I know heaven has a place for you.
ken (Score:2)
i spent 13 years writing about dec in decpro and rstspro magazines.
i often was critical. ken never failed to greet me and was ever gracious.
i think of him rolling out his g4 at bedford.
fare foward, ken
dave mallery
Thank You (Score:2)
I mourn the alpha chip (Score:2)
DEC had been working on the Alpha for a number years, from all reports a "world beater" in high performance chip. DEC was sold to Compaq which was bought by HP. HP had a close relationship with Intel, which was working their own high end chip the Itanium. By all reports the Alpha was much better than the Itanium. HP transferred the Alpha technology to Intel in exchange for some deals on chips and marketing. That's the last anyone heard of the Alpha.
Meanwhile the Itanium, dubbed the iTanic, sank. Here an art
Re: (Score:2)
Though, like many other DEC innovations, we continue to benefit from Alpha today. The next-generation Alpha EV8 was the first chip to implement SMT (hyperthreading) [wikipedia.org]. There was an internal legend as to how that came about: rumor was that the chip folks said to the compiler folks, "Hey, we just built an 8-issue processor," to which they responded, "Are you kidding? We can
Olsen's own staff turned on him, and killed DEC. (Score:4, Insightful)
Ken Olsen was an engineer's engineer, and he built a company that was based on innovative engineering and run by engineers.
DEC had 64-bit computing, virtual memory and virtual address extension, and dozens of other things we take for granted today literally twenty years before the competition! I worked routinely on inexpensive 64-bit machines in the 1980s, machines that simultaneously ran TCP/IP, SPX/IPX, LAT, and DECnet on the same wires, supporting 400 end users and huge databases with less processing power than you find on an nVidia card nowadays.
Sadly, the marketing and professional management people at DEC turned on Olsen, and engineered a financial crisis that allowed his ouster. Admittedly, those people were treated badly by Olsen, who viewed salesmen as a necessary evil and never really hid his opinion that business people were less valuable that the engineers and programmers. However, the salesforce rebellion was self-defeating, because from there the company entered a death spiral, as the bean counters' failure to maintain Olsen's unique corporate culture drove the top brains away to Microsoft (see wikipedia's entry on Cutler), Intel, Sun and Oracle.
After the disastrous Microsoft settlement, and the equally disastrous tech giveaway to Intel, DEC lost software and hardware primacy, and without Olsen at the helm the ship ran aground. A sad end to a mighty force for innovation; parted out to the highest bidder.
Goodbye, Ken. You were a good man, and it was an honor to have known you; I'll never forget you.
My First Computer Language, DEC-FORTRAN VI (Score:2)
I was there (Score:2)
I worked for DEC from 1977 till 1980. I left for warmer pastures (took a job in Florida) it would be a few more years before DEC would hit hard times. I think their biggest mistake was in not building a PC like machine. The Rainbow was just a "me too" machine, and not very well marketed. They could have boxed up their 11/70 on a chip (JAWS) into a killer machine that would have blown the AT away, and later on the VLSI VAX as well, but desktop versions of these super-minis never saw the light of day. Th
Worked for DEC, badge #48818 (Score:5, Insightful)
.
My Dad bought me a used PDP-8S, then a "straight 8" which were my first computers, a good bit before these newfangled microchips, and this is what I learned programming on, while also engineering my own peripherals. In fact, I wound up cancelling a charter subscription to Byte because they kept dissing the things, which at the time were a ton faster and better than any microchip. They were actually pretty nice machines, with a read-modify-write able to happen in one core cycle, and later when Silconix attempted a chip version, they were never able to get it as fast as the original, with that nifty Diode-Capacitor-Diode logic (which could create things like and gates where both the inputs didn't have to be there simultaneously as long as they were close enough).
I met Kenny, and he was a righteous dude, actually. The occasion was I was up in Mass taking a course on some new hardware, and talking in the company lunchroom to some Aussies at the table, who turned out to be buyers from some big retail outfit, and we were discussing the merits of this or that DEC product. At the time, the VAX was new, untested, a little flakey, and not as fast as a PDP-11/70 (particularly if the latter was maxed out) but cost more, so I steered them that way -- which would have (did) cost DEC some revenue, but they were nice guys, and it was the correct choice for them in their situation.
Kenny was standing behind me the whole time -- he'd come to the lunchroom to invite them to private talks in his office. Talk about my heart dropping into my gut -- this was my first really good job, and I'd just dissed the company's new flagship product to a very important customer, while the CEO was standing behind me.
Kenny grinned and shook my hand, and complimented me for being an honest guy, saying that was what DEC was all about, thanked me for helping promote that image! Soon after, *I* was promoted to Mid Atlantic support, one of the better jobs DEC had (free everything, expenses, flights on helicopters, full authority to make field-expedient decisions, all very nice).
That job was the basis of my career from then on. At that point I knew everyone big enough to be in computers at all (including the then-new ARPA and that crazy arpanet thing, node in Arlington) -- crap-tons of good contacts, and I never actually had to look for work ever again after that. From one beltway bandit to the next, to starting and running my own company with a nice customer list, that was what started it all.
We'll miss you Kenny, and my heartfelt condolences to the rest of the family. You weren't always right, but you were always good -- and that counts for more in my book.
favored platform for UNIX: "Linux of the 80s" (Score:2)
DEC in the family (Score:2)
Re: (Score:2)
A DECcie doesn't die (Score:2)
Any company that wrote it's training manuals with variables of $FOO and $BAR was my kind of company
- TTL
P.S. I still dream in TECO. Not that wussy VTEDIT full screen stuff... but writing programs in TECO and executing in MUNG. My therapist says I have closure issues;-)
Re: (Score:2)
Too bad that book [wikipedia.org] was about Data General, and not DEC.
The book opens with a turf war between two computer design groups within Data General Corporation, a minicomputer vendor in the 1970s. Most of the senior designers are assigned the "sexy" job of designing the next generation machine, which will be done in North Carolina. Their project (code-named "Fountainhead") is to give Data General a machine to compete with the new VAX computer from Digital Equipment Corporation (DEC), which is starting to take over the new 32-bit minicomputer market.
Re: (Score:2)
Re: (Score:2)
By a coincidence, before I heard about Ken Olsen's passing, I re-read "vmsbook.pdf" yesterday, which is a very cheery and optimistic history of VAXes and VMS up until 1997. Very sad that Compaq took them over the following year.
Here [slashdot.org] is a link to the PDF.
Re: (Score:2)
What the heck? My link got messed up. It should be to:
http://h71000.www7.hp.com/openvms/20th/vmsbook.pdf [hp.com]
Link [hp.com]
Oops, looks like the http:/// [http] got left off in my original post. Sorry.
Re: (Score:2)
I thought that was the Itanium.
Re:stolen (Score:4, Informative)
Steve Jobs is Eldon Tyrell. Steve Wozniak is JF Sebastian.
I really hope that's not a reference to Do Androids Dream of Electric Sheep. I think a lot of people would take offence at Woz being described as intellectually subnormal (a 'chickenhead' in the book's parlance).
Re: (Score:2)
I really hope that's not a reference to Do Androids Dream of Electric Sheep. I think a lot of people would take offence at Woz being described as intellectually subnormal (a 'chickenhead' in the book's parlance).
JF had to be good for Tyrell to play chess with him. I reckon that while Tyrell had vision, up there in his pyramid, playing with toys like Rachel, he needed JF to build things, hence the chess games as a way of keeping in touch.
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
.
Now if you want