The reason policemen draw their guns when someone draws a knife is because that is the protocol of a fight: when you draw a weapon, you announce intention of lethality, which gives everyone else right to draw their weapons, which may be far more powerful than your own. Thus the phrase 'never bring a knife to a gun fight'.
I am not being optimistic about the capabilites of a human being. Take some serious martial arts classes. Secondly, such training will give you reaction time far superior to an untrained person, allowing you to easily disarm him/her of the knife. And even if you suck, the hijackers were vastly outnumbered, and only a few people with some training could have defeated them.
Oh gee, as if I didn't study medicine long enough. I am quite aware of how a stab wound can kill. I didn't need a movie to disavow me of that notion. Now you go look into deaths via beatings. And then remember that skilled martial artists tend to have discipline enough not to enter most fights. This is all besides the point anyway; I don't think we were trying to prove that hands or knives are harmless.
Your only correct point; yes, weapons in hand are scary. Primarily because of the above, the protocol.
I still say the hijackers were unarmed -- particularly in comparison to our expectations of weaponry. Their tools were equivalent to a shiv or a pointy stick. Once again, they conquered those planes solely by fear, and no amount of technology can change that.
Re:Time for some highly unpopular opinion...
on
Handling the Loads
·
· Score: 2
You'd have to be blind to see the American government change its mind mid-stride -- first by supporting a group (again, with weapons and money), then by turning face, cutting off support or even condemning the actions of the group they supported.
We have had a poor track record of withdrawing support for groups we provided for. The ideal situation entails choosing the right group, funding it with enough arms to win their local cause and no more, and tapering down arms to nothing in the case of withdrawing support after the cause is completed. If it ends too soon, you should try to get back some percentage of the weapons, such as the standing offer of 300 million US dollars for the stockpile of Stinger missiles still in Afghanistan. It doesn't help much that Afghanis are fetishists for guns, or that other Arab states are outbidding us for the missiles.
In the case of the Afghanis, there were really no good choices, except for Massoud, who didn't really want foreign support in his country - he is/was the pro-democrary guy in the mujahedin. We focused on Hetmakyar and bin Laden, who schmoozed better with the West. Hetmakyar, by the way, basically avoided fighting and stockpiled everything we sent to him; he knew that the war would end, then he would conquer the nation. So the war ended, and chaos continued as Hetmakyar, Dostum, and Massoud shook it up amongst themselves. Later, the Talibs, primarily religious students from the south and funded by merchants with Pakistani ties, grew sick of the corruption of the warlords and started raising hell. They eventually captured all Hetmakyar's territory, who then fled to Pakistan, then diplomacy & assassination delivered them Dostum's territory, who fled to Turkey. When the Taliban captured Hetmakyar's weapon's cache, they proclaimed that they "now have enough weapons for 25 years of war". I'm inclined to believe them.
I wrote this little rant to point something out: Afghanistan is probably the wierdest place we have ever intervened in. The forces in control have shifted numerous times the culture in place has changed radically, and its not easy to just say 'we supported this guy and now he is out to get us' when everything over there changes radically every 5 years. Finally, it just goes to show that if you give a man a gun, you never know who will eventually hold it, and who it will be used on...
Interested parties might want to read up on Afghanistan at dangerfinder, as well as at more conventional news sources.
Judging from your stance on this, I can confidently say that if you attacked me with a knife, I could kill you with my bare hands. If you have any training, or if you just outnumber the guy with the knife, you should consider it essentially equivalent to skilled hand-to-hand technique. A knife has similar range; within this range he can be disarmed; a crushing bare-hands blow to the throat or base of the sternum can kill as easily as a stab wound.
Fighting someone with a gun is totally different. The gunwielder has enormous range, each bullet is crippling, and he requires almost no training to wield it effectively in close combat. That would be a good time to run away.
When you realize that a knife and skilled empty hands are essentially equivalent, you grasp the audacity here -- that essentially unarmed men captured 4 planes and killed thousands of people. And that no amount of scanners and x-rays can stop that.
As an aside, a japanese translation: empty -> kara; hand -> te; empty hand -> karate.
The whole "terrorists of the future" techno-fear bunk completely misses the lessons given over the last few days. Let me repeat:
A small band of essentially unarmed men captured 4 airplanes by playing to passengers & pilots fears. They then drove these planes into tall buildings, killing several thousand. Their total cost was rudimentary flight training, plane tickets (did they buy in advance?), and room & board while planning. They brought no advanced weapons, hacked no computer systems. Once again, it has been shown that the unaided human mind is the most dangerous weapon in the known universe.
There was, save the existence of airplanes, no technology whatsoever in Tuesday's attacks. Just victims' fear and the terrorists' willingness to die. These are social problems, and all the techno-fear 'solutions' that have been bandered about over the last few days both here and in the mainstream media, are completely ineffective to affect these social problems.
How does changing our crypto laws fix that?? Take as an example bin Laden, which the investigation is leaning towards. Where is the ambiguity there? In 1996 he issued a fatwah declaring war on the United States. How could we assume that that was nothing; that something like this wouldn't eventually happen? There are so many ways to infiltrate these groups, there are existing ways to harass their activities both within the US and without. How does attacking the civil liberties of US citizens to use technology freely aid the capture of a group whose men can perform such audacities without the aid of technology??
In this case, the S/390 series. The zSeries mainframes (12 CPUs) are about half the height of a rack and a bit wider. And the power requirements are way lower.
Why do you need VM programmers? The port is already done, the logic for running Linux as a guest OS is there, and it's stable. Henceforth you should be coding on the Linux level, not the VM level.
I saw a fantastic film last year about youth in Lebanon (Beirut). I forget the name. The killer line was when one elderly Arab said something to the effect of 'the Arabs used to be a great and learned people. What happened to us.' This while standing in the ruins of Beirut (thank you, Israel and Syria).
Much as you keep bringing up the Sharia, I do not think that is a complete answer. The last thousand years has seen a thorough collapse of Arab culture. Laws and traditions formed at the height of Arab civilization no longer serve, nor are they even well implemented. Add all the oil wealth, and you have a familiar pattern of a sadly uneducated redneck culture that has just recently come into a great deal of prestige and power -- a sure receipe for screwing a place up. These disputes aren't the result of any coherent religious position, just the old, sad inevitabilities of ignorant hate in rapidly changing cultures.
OTOH, the Israelis could stop, you know, killing a dozen Arabs in retaliation for each Israeli death. That's certainly no way to make friends.
That's an interesting evaluation. I assume that is the standard Christian position?
I don't think I ever use terms like 'glory' in everyday context. What I found thus far is that, even ignoring 'glorifying' someone else, opening my code makes me strive to be better..."and in the perfection of the individual comes all good things." I guess I just like the phrasing Matthew uses.
First, programming is a craft. You may gain some theoretic backing in formal education, but you only improve by writing code. Lots of code. Lots of code that recieves criticism from intelligent peers. This is the first pattern I put on the table: that a craft skill improves only by repeated efforts that meet the demands of the outside world.
Second, take everything business management theorists have ever written about employee motivation and team cohesion. You might notice the theme "respect" repeated over and over again. There are many extremely good programmers in companies that don't release code, but there is still a wealth of top-notch coders who publish open-source code, read other people's code, and provide criticism to other coders. It is a big kick to hear an awesome coder praise your work, and this community bears a strong incentive to try to impress them.
Third, one gets to communicate one's successes and struggles to the outside world. Did you spend two weeks writing a really complex block of simulation code? If the project is closed, then people only see the little button to trigger the event. If the project is open everyone can see how cool you really are. Is this important? Compare to other professions -- would doctors or lawyers agree with closing all of their work, and never sharing research?
I am currently running 4 open-source projects, and contributing to 3 more. I work way too many hours, but the 3 previous points basically summarize my motivations and what I get out of it.
After college, most of us do not find ourselves in an environment that encourages education. Working with OSS helps build your own educational environment. And there's something else: somewhere in the Bible Christ encourages some multitude not to "hide your light under a bushel". (I am not Christian; perhaps someone else can provide the quote.) The language is archaic, but I think it has meaning in that from the moment I exposed my code to the outside world I improved as a coder. For whatever reasons, first among them being the consideration of others in previously private efforts, I have found that writing OSS code makes me more professional.
Oh, and please thank Boucher from some pleased Californians.
It's easy to say what might have happened. Particularly in AI, wherein everything thus far has been vapor.
More to the point, being 'uncannily perceptive' doesn't solve the core problem with Clippy, which was that no one likes forcible context switches away from their work. There is a great deal of needed research and implementation on how people interact with their computers, how they maintain continuity through an application, and how to present easy access to information. The idea that you can end-run around those problems by having an application interrupt you at odd times is hogwash, no matter how intelligent the application.
Re:-- Be fair about the processor lines --
on
HP Buys Compaq
·
· Score: 2
I'll call a spade whatever the hell I want to.
What you are missing here is that a Unix server vendor's processors' primary market is that vendor's operating system. Ultra Sparc for Solaris, Power4 for AIX, MIPS for IRIX, and Alpha for Digital Unix. Much as you might think the mass market desktop OSes are the only way to sustainability, each of these chips has/had done very well even before other OSes appeared on them (Linux, mostly). What Digital & Compaq/Digital did was cancel further Digital Unix development, ensuring a swift and painless death to the processors it depended on.
I couldn't give a shit about RedHat. How is RedHat supposed to be representative of all of us, let alone representative of my small niche of hardcore hardware geeks raised on big iron? You might not have noticed, but from almost the moment Linus wrote the Alpha port, we have been clamoring for desktop Alphas, small workgroup servers, anything. Never happened; Digital never made consumer motherboards. There was some hope Samsung would have done so, but for whatever legal & business reasons, they did not either. And then Compaq bought Digital and started chopping it up for parts.
Sufficient software forces were salivating at the prospect of working on the Alpha. What eventually crushed it was Digital's total incompetence at marketing and delivering to consumers, and Compaq's inability to sell to markets other than the WinTel field they were raised on, not Billy G deciding to take his ball and go home.
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 2
AMD started showing real numbers in SPECFP on the K6-2, which is when they had integrated the NexGen FP core. The Alpha people were set right to work on the K7, ie, the Athlon.
I agree the Alphas kick ass. Their most notable strength, IMHO, is their instruction dispatch rate is the highest in the industry. To clarify, modern CPUs can run 20 or so instructions per cycle, but most of them will be discared for timing issues. Most CPUs (like the P3) hover around 2; Alphas around 3.5 to 4 (IIRC). Incidentally, this inefficiency is the prime motivator behind VLIW and on-chip threading.
But I also think that it's better to have the engineers than the original tech on hand. The Alphas have always been fabbed on rather old processes (.25, I think, for the 833Mhz). Their power consumption is just insanely high, albeit less than the Itanium. Other chips are starting to catch up with it: Pentium4 posted SPECFP numbers that edged ahead of the latest Alpha. And of course there's that whole messy problem of instruction sets.
The consolidation of chip design is really wierd. This in a time of interpreted languages, where everyone is moving away from instruction set dependence, is even stranger. But the Unix vendors who are stepping into thrall to Intel and Microsoft would be wise to reconsider. All the vendors who have played in that direction: DEC, SGI, HP, have either suffered or been destroyed. Only Sun, who has fought WinTel tooth and nail, has thrived.
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 1
I think Steve has the notes on that. It's largely contextual.
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 2
If this is true, then why has Dell done so well?
Sorry, it was late and I should have clarified my language. No one has been able to run the high-margin 'workstation' market with WinTel boxes, at least not for long. By workstation I mean the sort of thing a professional runs AutoCAD on -- what you would replace an Ultra 10 or Onyx II with. The only 2 good attempts I've seen at this have been Intergraph and SGI's recent PCs (with that neat channeled memory architecture). Both sold for a short time, then fell back into the mire. There's just not enough differentiation.
As for Compaq/DEC, you're right that it's for the consulting arm. For its entire history, DEC was pure technical genius and absolute incompetence in sales/marketing. Acquiring that would be an easy win. Of course, alienating all your consulting customers isn't all that smart of a move either....
Did they dump their PA-RISC/HP-UX line, or just move many of their resources on to creating their next line of processors, which is Intel's IA-64.
They cancelled their planned next revision of PA-RISC and signed some intensive contracts with Intel binding them to make IA-64. So they didn't have much more incentive to make another PA-RISC once they tossed Belluzzio. I think the engineering was also intrigued at building the first mass-market VLIW processor -- shame it doesn't deliver. Maybe McKinley will.
As for your analysis, well, companies like having reorganizations. It makes it look like they are doing something.
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 2
1. But companies need to be really big
to sustain research.
2. I don't understand your fondness for
Tandem. Their h/w was incredibly expensive
and their OS was bizarre
Assertion 1 is totally unsupported, and IMHO, wrong. A big company only supports research if it wants to -- one of the first things Compaq said after buying DEC was "we will be stopping all research. We can buy technology we need." This position was met with great applomb by analysts. The notion that "big companies create research" is a myth -- they already get free patents on university research, and have much lower engineer/management ratios than small companies. In all of my professional experience, most computer research comes not from a Fortune 100 behemoth spreading funding across the land, but from a small group of geeks hacking late into the night at a start-up. The rare exception is companies like IBM, who target the market 20 years down the road, and have found that putting 1% into long-range research (10 year windows) ensures that when the future is invented, you'll be a part of it. Most corps are too heavily in the thrall of quarterly return reports to make that kind of an investment.
As for Tandems, well, I'm not fond of them like I am of IBM s/390s. But they did do their job well.
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 5, Interesting
You're thinking in the wrong realm. Intel isn't trying to compete with AMD with IA-64, they are trying to compete with Ultra SPARC-3, MIPS 10k, IBM's Power4, and HP's PA-RISC. Really big boxes that address 2.83 assloads of RAM and have several dozen processors (like up to a thousand in the case of SGI). Turns out there just no way a home user is going to buy one of these, and just no way a clone vendor is going to build one.
The real power with these systems is not the processor, it is the backplane: the buses, the memory, etc. That is where companies differentiate; that is what separates a million-dollar server from a desktop PC.
With this in mind, the processor is almost an afterthought. Why even develop the IA-64; why not use the P4? Well, you need to directly address more than 4 GB of RAM, which is the limit on 32 bits. Also you can operate on larger numbers in one operation, rather than several in a 32-bit chip. There's also a bit of black art involved in developing a chip to play well in a SMP or NUMA memory environment.
I expect AMD to be the Next Big Thing, and HPaq will declare bankruptcy within 2 years. Sledgehammer will run old 32-bit binaries fast, IA-64 will not. That alone will keep most people from buying IA-64. And with the alpha designers at AMD...all they need to do is license the alpha technology.
Pardon my saying this, but here you have walked from 'flights of fancy' into 'complete nonsense'.
Why would HP declare bankruptcy? And which kind, ie, Chapter? They have just reclaimed the title of 2nd largest computer company in the world. They have been on the Dow Jones for years. They have bukos of money. I don't think they have much going for them, but they aren't a dot.com, with no cash in the bank and set to blow away.
AMD already is the Next Big Thing. Haven't you been watching? They have been eating Intel's lunch in the desktop arena, then going home to Intel's house and raiding the fridge.
"most people" will never buy an IA-64. Read the above to see why. Switching all the desktops over to 64 bit is still 10 years out.
"license the alpha technology"? They already have!! The Athlon uses the same bus architecture as the Alphas. More to the point, they have the engineers; why do they need to license everything?
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 4, Interesting
As I heard it, the StrongARM team was based around Digital's New England foundry. Given the often idiosyncratic nature of the weather in the area, winter in particular, you would sometimes not see some engineers for several days. So it seemed that Intel's much less flexible culture might not look kindly on this kind of behavior. Speaking of culture, everyone in the chip biz seems to think of Intel as the place chip designers go to die -- overwork, mistreatment, malfeasance, etc, etc. I don't know, nor do I care that much. That's just the word on the street, and it seemed to be enough for them. OTOH, I'm glad to hear the Alpha team is doing so well.
I was building boards on StrongARM back in 1998, and when Intel bought them, it just sort of fell off the face of the earth for a while. I think it wasn't until 2000 that I started seeing StrongARM in anything higher than the 233MHz DEC had fabbed on.35 micron. I was really hopeful when Intel bought them; I thought we would see them move it to.22 or.18 as soon as possible. Imagine! 600+ MHz at <1 watt, in 1999! Didn't happen. With other assumptions and evidence in hand, I believe that Intel's short-term business was best protected by sitting on StrongARM until Intel's core chips had caught up. Of course, having the core team quit doesn't help them ramp up quickly either.
While you're here, could you tell me which ARM core they're building XScale with now? Do they have SMP enabled? (StrongARM (v4 core) had the SMP pin shorted).
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 1
C'mon, VAXman, surely you have a better rebuttal up your sleeve. Why not bring the full weight of backplane design on this young whippersnapper's head, and explain to him how server architectures differ from the desktop machines he is acquainted with?
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 2
Perhaps you are worth listening to; why don't you log in? I said I'm making a list of people worth listening to, not just people I agree with. I assign 'worth listening' by ability to coherently defend one's position with some element of neutral examination, combined with experience and domain knowledge. The standard set of logical fallacies are of course 'right out'.
As for your rebuttal, I contend that logic in businesses just does not move that fast. We in the computer industry enjoy flattering ourselves with homilies about how fast change is in the 'new economy', all the while using 20 year old tools. So I think the logic back in 1998 is still very much around, and most of yesterday's "markets and CEO's[sic] and investors and managers" are still today's; more importantly, the principles which govern the markets haven't changed that fast. I'd like to see you defend your assertions.
And if you're feeling thoughtful tonight, you might note that I haven't labeled your thoughts as "worth nothing".
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 1
then sub to my journal so I can have your name (do it after your mod points are gone or something). I am building a little list of slashdot users worth listening to.
Re:Hate to say, sounds like a dot-bomb strategy...
on
HP Buys Compaq
·
· Score: 5, Interesting
With this one, I'd have to say that Fiorina is a tool.
I'm glad someone brought up previous acquisitions. There's a bit of history worth examining here.
Compaq ate Digital, sold the StrongARM to Intel who buried it b/c it was an order of magnitude faster than Intel's low-power chips. Compaq/Digital then shed all their good engineers b/c their corporate culture sucked. Most of the Alpha guys went to AMD, which explains a great deal about the Athlon. (Incidentally, many of the StrongARM guys went to Cadence. Anyone know anything else?) They partnered the with Samsung, but for whatever reasons, Samsung has not been able or willing to sell Alphas here in the States. In op/sys, Compaq/Digital has tried several times to cancel the Digital Unix line; but hey, they renamed it to True64! Compaq/Digital told all their Unix customers that they were switching them over to NT; you can imagine how receptive their customers were about that. Thus, True64, marginal continued development, but most customers just left and went to Sun/IBM/Linux.
Final analysis? Fucking waste of money. The only people who benefited from this were the executives and the competition.
Round about the same time, Compaq bought Tandem. I used to run a Tandem in 96 -- nice boxes. The first thing Compaq did was gut the sales force. Compaq, a PC vendor, assumed that one needs one salesman to sell one machine (or some such). Turns out, you need a small army to sell a mainframe; lots and lots of handholding and a salespeep for each engineer. Tandem would often have several dozen salespeople working on a single client, for a multi-million dollar order. The inevitable response to gutting the salesforce? Yes, they lost all those orders.
Final analysis? Fucking waste of money. The only people who benefited from this were the executives and the competition.
Modern corporations are not innately designed to make money. They are innately designed to get bigger, driven by senior executives with Napoleon complexes. It does not help that standard management training teaches managers to seek larger fiefdoms rather than efficiency or productivity. This is not the usual Green-party ranting -- a survey of CEO salaries indicates an explosive growth over the last decade; even biz-school professors and analysts are worried.
Before I finish this, I should turn my cynicism on HP. In, I think, 1996 HP announced a new direction: dump their processors (PA-RISC) and their Unix (HP-UX), in exchange for Intel & NT. Of course, the customers fled to the other Unix vendors; they sold some nice NT boxes before realizing that no one can sustainably sell WinTel boxes on the margins that a big corp demands, since the clone makers can always build the same thing for less. HP fired the CEO who masterminded that FUBAR decision, and got back behind PA-RISC & HP-UX. Lasting fallout: fewer customers, multi-year development agreements with Intel (witness the Itanium & McKinley.). Is this the sort of company that can integrate a company like Compaq?
Technical acquisitions are perhaps the most complex of any company integration project. When I see an announcement like this, by two companies who have spent the last few years hurting while everyone else enjoyed the boom times, whose product lines overlap and present no clear engineering wins; I think 'golden parachute'. This is a way to manipulate the stock price. I see no clear way or reason for HP/Compaq to become anything more than an also-ran.
I'd like to think that a good journalist isn't limited to a small number of fields he/she is already trained in.. Their job is, after all, to digest and summarize large amounts of seemingly arcane or trivial content and render it comprehensible to their readership. If one has to be fully inside the technical community, the entering of which often takes several years of determined geekitude, in order to write a report, then something is horribly wrong.
Actually, I think I'll agree with my unstated question -- investigative journalism is dead, dead, dead. Geeks report on things out of their biases and previous training. Professional journalists plagarize from news releases and marketing copy. No one is actually going out and asking some engineer "what the fsck does this mean?"
I think the wisdom in the article can be distilled down to one of the closing lines:
All these
people wanted to be something other than reporters and for awhile, they got away with it. Because they wanted to be something they weren't while refusing to
recognize that greatness lies in doing their jobs. Journalism is a noble profession when done right. And people get killed doing it every year.
You might be right, you might have more raw computrons with 1.2 M $ of PCs than a z/390. But not everything lends itself well to distributed computation. So depending on what you run, you could see lower performance and hellish administration.
You're also forgetting the point of a mainframe - HA. PCs aren't designed for reliability. When you administer a cluster you expect a certain rate of hardware failures. A mainframe is expected to have 99.999+ percent uptime. The hardware is fully scoped, localized, and hot-swappable - right down to the processors. A company that's looking for that kind of uptime really has no other options. What all this Linux/390 stuff is about is selling Linux to groups who won't compromise on the uptime.
Re:It's the logical move
on
The New Zelda
·
· Score: 2
'-dono' is(was) only used to title ranking officials of the government or similar. It is also *totally* archaic. I have heard it used precisely zero times.
And yes, he is Miyamoto, so if you worship the ground he walks on, sure, use '-sama'. I hope you understand that worshipping the ground someone walks on doesn't look sane in any culture, OTOH. Be prepared to be scorned, and deservedly.
'-sensei' *could* be appropriate, but you're still on tricky ground. '-san' is best. Japanese are extremely polite, and it's not just a totally cosmetic effort with a bunch of self-debasing titles. Your politeness is evidenced by a wide swath of changes in your behavior and form of speech. Trying to find a suitable title to say "I'm not worthy", as you are doing, just makes you look like an idiot, and an insincere one at that.
Oh, and ignore The_Messenger. He appears to be on crack today.
Slashdot Mirror
I am not being optimistic about the capabilites of a human being. Take some serious martial arts classes. Secondly, such training will give you reaction time far superior to an untrained person, allowing you to easily disarm him/her of the knife. And even if you suck, the hijackers were vastly outnumbered, and only a few people with some training could have defeated them.
Oh gee, as if I didn't study medicine long enough. I am quite aware of how a stab wound can kill. I didn't need a movie to disavow me of that notion. Now you go look into deaths via beatings. And then remember that skilled martial artists tend to have discipline enough not to enter most fights. This is all besides the point anyway; I don't think we were trying to prove that hands or knives are harmless.
Your only correct point; yes, weapons in hand are scary. Primarily because of the above, the protocol.
I still say the hijackers were unarmed -- particularly in comparison to our expectations of weaponry. Their tools were equivalent to a shiv or a pointy stick. Once again, they conquered those planes solely by fear, and no amount of technology can change that.
In the case of the Afghanis, there were really no good choices, except for Massoud, who didn't really want foreign support in his country - he is/was the pro-democrary guy in the mujahedin. We focused on Hetmakyar and bin Laden, who schmoozed better with the West. Hetmakyar, by the way, basically avoided fighting and stockpiled everything we sent to him; he knew that the war would end, then he would conquer the nation. So the war ended, and chaos continued as Hetmakyar, Dostum, and Massoud shook it up amongst themselves. Later, the Talibs, primarily religious students from the south and funded by merchants with Pakistani ties, grew sick of the corruption of the warlords and started raising hell. They eventually captured all Hetmakyar's territory, who then fled to Pakistan, then diplomacy & assassination delivered them Dostum's territory, who fled to Turkey. When the Taliban captured Hetmakyar's weapon's cache, they proclaimed that they "now have enough weapons for 25 years of war". I'm inclined to believe them.
I wrote this little rant to point something out: Afghanistan is probably the wierdest place we have ever intervened in. The forces in control have shifted numerous times the culture in place has changed radically, and its not easy to just say 'we supported this guy and now he is out to get us' when everything over there changes radically every 5 years. Finally, it just goes to show that if you give a man a gun, you never know who will eventually hold it, and who it will be used on...
Interested parties might want to read up on Afghanistan at dangerfinder, as well as at more conventional news sources.
Fighting someone with a gun is totally different. The gunwielder has enormous range, each bullet is crippling, and he requires almost no training to wield it effectively in close combat. That would be a good time to run away.
When you realize that a knife and skilled empty hands are essentially equivalent, you grasp the audacity here -- that essentially unarmed men captured 4 planes and killed thousands of people. And that no amount of scanners and x-rays can stop that.
As an aside, a japanese translation: empty -> kara; hand -> te; empty hand -> karate.
A small band of essentially unarmed men captured 4 airplanes by playing to passengers & pilots fears. They then drove these planes into tall buildings, killing several thousand. Their total cost was rudimentary flight training, plane tickets (did they buy in advance?), and room & board while planning. They brought no advanced weapons, hacked no computer systems. Once again, it has been shown that the unaided human mind is the most dangerous weapon in the known universe.
There was, save the existence of airplanes, no technology whatsoever in Tuesday's attacks. Just victims' fear and the terrorists' willingness to die. These are social problems, and all the techno-fear 'solutions' that have been bandered about over the last few days both here and in the mainstream media, are completely ineffective to affect these social problems.
How does changing our crypto laws fix that?? Take as an example bin Laden, which the investigation is leaning towards. Where is the ambiguity there? In 1996 he issued a fatwah declaring war on the United States. How could we assume that that was nothing; that something like this wouldn't eventually happen? There are so many ways to infiltrate these groups, there are existing ways to harass their activities both within the US and without. How does attacking the civil liberties of US citizens to use technology freely aid the capture of a group whose men can perform such audacities without the aid of technology??
Why do you need VM programmers? The port is already done, the logic for running Linux as a guest OS is there, and it's stable. Henceforth you should be coding on the Linux level, not the VM level.
Much as you keep bringing up the Sharia, I do not think that is a complete answer. The last thousand years has seen a thorough collapse of Arab culture. Laws and traditions formed at the height of Arab civilization no longer serve, nor are they even well implemented. Add all the oil wealth, and you have a familiar pattern of a sadly uneducated redneck culture that has just recently come into a great deal of prestige and power -- a sure receipe for screwing a place up. These disputes aren't the result of any coherent religious position, just the old, sad inevitabilities of ignorant hate in rapidly changing cultures.
OTOH, the Israelis could stop, you know, killing a dozen Arabs in retaliation for each Israeli death. That's certainly no way to make friends.
I don't think I ever use terms like 'glory' in everyday context. What I found thus far is that, even ignoring 'glorifying' someone else, opening my code makes me strive to be better..."and in the perfection of the individual comes all good things." I guess I just like the phrasing Matthew uses.
Second, take everything business management theorists have ever written about employee motivation and team cohesion. You might notice the theme "respect" repeated over and over again. There are many extremely good programmers in companies that don't release code, but there is still a wealth of top-notch coders who publish open-source code, read other people's code, and provide criticism to other coders. It is a big kick to hear an awesome coder praise your work, and this community bears a strong incentive to try to impress them.
Third, one gets to communicate one's successes and struggles to the outside world. Did you spend two weeks writing a really complex block of simulation code? If the project is closed, then people only see the little button to trigger the event. If the project is open everyone can see how cool you really are. Is this important? Compare to other professions -- would doctors or lawyers agree with closing all of their work, and never sharing research?
I am currently running 4 open-source projects, and contributing to 3 more. I work way too many hours, but the 3 previous points basically summarize my motivations and what I get out of it.
After college, most of us do not find ourselves in an environment that encourages education. Working with OSS helps build your own educational environment. And there's something else: somewhere in the Bible Christ encourages some multitude not to "hide your light under a bushel". (I am not Christian; perhaps someone else can provide the quote.) The language is archaic, but I think it has meaning in that from the moment I exposed my code to the outside world I improved as a coder. For whatever reasons, first among them being the consideration of others in previously private efforts, I have found that writing OSS code makes me more professional.
Oh, and please thank Boucher from some pleased Californians.
Why in the living hell didn't Compaq discover that accounting discrepancy in due diligence?
More to the point, being 'uncannily perceptive' doesn't solve the core problem with Clippy, which was that no one likes forcible context switches away from their work. There is a great deal of needed research and implementation on how people interact with their computers, how they maintain continuity through an application, and how to present easy access to information. The idea that you can end-run around those problems by having an application interrupt you at odd times is hogwash, no matter how intelligent the application.
What you are missing here is that a Unix server vendor's processors' primary market is that vendor's operating system. Ultra Sparc for Solaris, Power4 for AIX, MIPS for IRIX, and Alpha for Digital Unix. Much as you might think the mass market desktop OSes are the only way to sustainability, each of these chips has/had done very well even before other OSes appeared on them (Linux, mostly). What Digital & Compaq/Digital did was cancel further Digital Unix development, ensuring a swift and painless death to the processors it depended on.
I couldn't give a shit about RedHat. How is RedHat supposed to be representative of all of us, let alone representative of my small niche of hardcore hardware geeks raised on big iron? You might not have noticed, but from almost the moment Linus wrote the Alpha port, we have been clamoring for desktop Alphas, small workgroup servers, anything. Never happened; Digital never made consumer motherboards. There was some hope Samsung would have done so, but for whatever legal & business reasons, they did not either. And then Compaq bought Digital and started chopping it up for parts.
Sufficient software forces were salivating at the prospect of working on the Alpha. What eventually crushed it was Digital's total incompetence at marketing and delivering to consumers, and Compaq's inability to sell to markets other than the WinTel field they were raised on, not Billy G deciding to take his ball and go home.
I agree the Alphas kick ass. Their most notable strength, IMHO, is their instruction dispatch rate is the highest in the industry. To clarify, modern CPUs can run 20 or so instructions per cycle, but most of them will be discared for timing issues. Most CPUs (like the P3) hover around 2; Alphas around 3.5 to 4 (IIRC). Incidentally, this inefficiency is the prime motivator behind VLIW and on-chip threading.
But I also think that it's better to have the engineers than the original tech on hand. The Alphas have always been fabbed on rather old processes (.25, I think, for the 833Mhz). Their power consumption is just insanely high, albeit less than the Itanium. Other chips are starting to catch up with it: Pentium4 posted SPECFP numbers that edged ahead of the latest Alpha. And of course there's that whole messy problem of instruction sets.
The consolidation of chip design is really wierd. This in a time of interpreted languages, where everyone is moving away from instruction set dependence, is even stranger. But the Unix vendors who are stepping into thrall to Intel and Microsoft would be wise to reconsider. All the vendors who have played in that direction: DEC, SGI, HP, have either suffered or been destroyed. Only Sun, who has fought WinTel tooth and nail, has thrived.
I think Steve has the notes on that. It's largely contextual.
As for Compaq/DEC, you're right that it's for the consulting arm. For its entire history, DEC was pure technical genius and absolute incompetence in sales/marketing. Acquiring that would be an easy win. Of course, alienating all your consulting customers isn't all that smart of a move either....
They cancelled their planned next revision of PA-RISC and signed some intensive contracts with Intel binding them to make IA-64. So they didn't have much more incentive to make another PA-RISC once they tossed Belluzzio. I think the engineering was also intrigued at building the first mass-market VLIW processor -- shame it doesn't deliver. Maybe McKinley will.As for your analysis, well, companies like having reorganizations. It makes it look like they are doing something.
As for Tandems, well, I'm not fond of them like I am of IBM s/390s. But they did do their job well.
The real power with these systems is not the processor, it is the backplane: the buses, the memory, etc. That is where companies differentiate; that is what separates a million-dollar server from a desktop PC.
With this in mind, the processor is almost an afterthought. Why even develop the IA-64; why not use the P4? Well, you need to directly address more than 4 GB of RAM, which is the limit on 32 bits. Also you can operate on larger numbers in one operation, rather than several in a 32-bit chip. There's also a bit of black art involved in developing a chip to play well in a SMP or NUMA memory environment.
Pardon my saying this, but here you have walked from 'flights of fancy' into 'complete nonsense'.I was building boards on StrongARM back in 1998, and when Intel bought them, it just sort of fell off the face of the earth for a while. I think it wasn't until 2000 that I started seeing StrongARM in anything higher than the 233MHz DEC had fabbed on .35 micron. I was really hopeful when Intel bought them; I thought we would see them move it to .22 or .18 as soon as possible. Imagine! 600+ MHz at <1 watt, in 1999! Didn't happen. With other assumptions and evidence in hand, I believe that Intel's short-term business was best protected by sitting on StrongARM until Intel's core chips had caught up. Of course, having the core team quit doesn't help them ramp up quickly either.
While you're here, could you tell me which ARM core they're building XScale with now? Do they have SMP enabled? (StrongARM (v4 core) had the SMP pin shorted).
C'mon, VAXman, surely you have a better rebuttal up your sleeve. Why not bring the full weight of backplane design on this young whippersnapper's head, and explain to him how server architectures differ from the desktop machines he is acquainted with?
As for your rebuttal, I contend that logic in businesses just does not move that fast. We in the computer industry enjoy flattering ourselves with homilies about how fast change is in the 'new economy', all the while using 20 year old tools. So I think the logic back in 1998 is still very much around, and most of yesterday's "markets and CEO's[sic] and investors and managers" are still today's; more importantly, the principles which govern the markets haven't changed that fast. I'd like to see you defend your assertions.
And if you're feeling thoughtful tonight, you might note that I haven't labeled your thoughts as "worth nothing".
then sub to my journal so I can have your name (do it after your mod points are gone or something). I am building a little list of slashdot users worth listening to.
I'm glad someone brought up previous acquisitions. There's a bit of history worth examining here.
Compaq ate Digital, sold the StrongARM to Intel who buried it b/c it was an order of magnitude faster than Intel's low-power chips. Compaq/Digital then shed all their good engineers b/c their corporate culture sucked. Most of the Alpha guys went to AMD, which explains a great deal about the Athlon. (Incidentally, many of the StrongARM guys went to Cadence. Anyone know anything else?) They partnered the with Samsung, but for whatever reasons, Samsung has not been able or willing to sell Alphas here in the States. In op/sys, Compaq/Digital has tried several times to cancel the Digital Unix line; but hey, they renamed it to True64! Compaq/Digital told all their Unix customers that they were switching them over to NT; you can imagine how receptive their customers were about that. Thus, True64, marginal continued development, but most customers just left and went to Sun/IBM/Linux.
Final analysis? Fucking waste of money. The only people who benefited from this were the executives and the competition.
Round about the same time, Compaq bought Tandem. I used to run a Tandem in 96 -- nice boxes. The first thing Compaq did was gut the sales force. Compaq, a PC vendor, assumed that one needs one salesman to sell one machine (or some such). Turns out, you need a small army to sell a mainframe; lots and lots of handholding and a salespeep for each engineer. Tandem would often have several dozen salespeople working on a single client, for a multi-million dollar order. The inevitable response to gutting the salesforce? Yes, they lost all those orders.
Final analysis? Fucking waste of money. The only people who benefited from this were the executives and the competition.
Modern corporations are not innately designed to make money. They are innately designed to get bigger, driven by senior executives with Napoleon complexes. It does not help that standard management training teaches managers to seek larger fiefdoms rather than efficiency or productivity. This is not the usual Green-party ranting -- a survey of CEO salaries indicates an explosive growth over the last decade; even biz-school professors and analysts are worried.
Before I finish this, I should turn my cynicism on HP. In, I think, 1996 HP announced a new direction: dump their processors (PA-RISC) and their Unix (HP-UX), in exchange for Intel & NT. Of course, the customers fled to the other Unix vendors; they sold some nice NT boxes before realizing that no one can sustainably sell WinTel boxes on the margins that a big corp demands, since the clone makers can always build the same thing for less. HP fired the CEO who masterminded that FUBAR decision, and got back behind PA-RISC & HP-UX. Lasting fallout: fewer customers, multi-year development agreements with Intel (witness the Itanium & McKinley.). Is this the sort of company that can integrate a company like Compaq?
Technical acquisitions are perhaps the most complex of any company integration project. When I see an announcement like this, by two companies who have spent the last few years hurting while everyone else enjoyed the boom times, whose product lines overlap and present no clear engineering wins; I think 'golden parachute'. This is a way to manipulate the stock price. I see no clear way or reason for HP/Compaq to become anything more than an also-ran.
All I want to see is just *one* benchmark from the MySQL folks that isn't blatant dishonesty or incompetence.
Actually, I think I'll agree with my unstated question -- investigative journalism is dead, dead, dead. Geeks report on things out of their biases and previous training. Professional journalists plagarize from news releases and marketing copy. No one is actually going out and asking some engineer "what the fsck does this mean?"
I think the wisdom in the article can be distilled down to one of the closing lines:
You're also forgetting the point of a mainframe - HA. PCs aren't designed for reliability. When you administer a cluster you expect a certain rate of hardware failures. A mainframe is expected to have 99.999+ percent uptime. The hardware is fully scoped, localized, and hot-swappable - right down to the processors. A company that's looking for that kind of uptime really has no other options. What all this Linux/390 stuff is about is selling Linux to groups who won't compromise on the uptime.
'-dono' is(was) only used to title ranking officials of the government or similar. It is also *totally* archaic. I have heard it used precisely zero times.
And yes, he is Miyamoto, so if you worship the ground he walks on, sure, use '-sama'. I hope you understand that worshipping the ground someone walks on doesn't look sane in any culture, OTOH. Be prepared to be scorned, and deservedly.
'-sensei' *could* be appropriate, but you're still on tricky ground. '-san' is best. Japanese are extremely polite, and it's not just a totally cosmetic effort with a bunch of self-debasing titles. Your politeness is evidenced by a wide swath of changes in your behavior and form of speech. Trying to find a suitable title to say "I'm not worthy", as you are doing, just makes you look like an idiot, and an insincere one at that.
Oh, and ignore The_Messenger. He appears to be on crack today.