Slashdot Mirror
Slashback: Memory, Constancy, Triumph
Why not put 'em on Freenet while you're at it ... Imran Ghory writes: "Google has put out an appeal to get NetNews CDs (produced by Sterling Software and CD Publishing Corporation) which archived usenet between 1992 to 1995. Looks like Google is reviving Deja's idea of a total usenet archive."
This sounds like a worthy objective, worth rooting around for -- maybe they'll even give you a credit somewhere.
They know that of which they speak. Hot on the heels of the inexorable GCC project's 3.0.1 release, zealot (and a number of other people) wrote with the news that "Intel will release its latest compilers (the ones that optimize for P4 and can do some auto-vectorization of code) for Linux this Thursday. I'd love to see some performance numbers for compiled code on a P4 if anyone gets their hands on this ... maybe the autovectorization could help some gimp plugins speed up."
You cannot stop the chess updates Álvaro Begué writes: "Junior is the new World Micro Computer Chess Champion, Shredder won in the single processor category (five years in a row) and Goliath won the blitz tournament. Congratulations to all of them. Check out the official website."
Maybe the durned things will stick around forever. In addition to the IBM research on making ultra-slim CRT monitors, an Anonymous Coward points to another article on the future of CRTs: "This is a new technology that can integrate into existing production lines and can halve the depth of a CRT type tube. A TV normally 22 inches deep would be only 11 inches."
Crikey! Usenet archives going back even FURTHER? Great, now people will be able to trace me back to my great Usenet roots....
"My name is Dave Rhodes. In September 1988 my car was reposessed and the bill collectors were hounding me like you wouldn't believe"....
Seriously. My current CRT takes up way too much room on my desk. I'm looking forward to having larger screens using less deskspace.
LCDs are nice, but you miss out on the flood of radiation pouring out the front.
The topic is Slashback: Memory, Constancy, Triumph. Yet there isn't any mention about memory. Maybe they forgot. :)
Would you be surprised if Intels compiler produced faster code than GCC? I believe Linus has stated that GCC is a bit "bloated". I wonder if you can compile the Linux kernel with it (minus assembly of course). That might be interesting, particularly for P4. Linux could get an instant speed boost. And such a radical switch in compiler might expose flaws in the code. Definately a worthwhile excercise if nothing else. And even though the average user isn't going to buy it to compile their kernel, the distro's might for their precomiled kernels (err, wonder how that would work
they do, and it works well. IIRC, they submitted benchmark results to SPEC where a Pentium chip (not sure what one) smoked several others in many benchmarks. SPEC rejected those benchmarks because Intel used a special proprietery compiler with the tests and not a normal compiler a developer would use.
Hence, Intel has compilers of their own that work very well, but why they aren't made public like this Linux one is, I wish I knew, as it could undermine MS-VC in terms of compiled code performance.
The One Rule Of Chess You'll Ever Need: Don't play someone who carries a kit in their bookbag.
Basically it has some builtin optimizations that try to make your code run as fast as possible on the P4. In fact to get decent performance on a P4 you pretty much HAVE to use a P4 optimized compiler, or assemble it yourself (assuming you know the ins and outs of the P4).
Intel probably has a document somewhere that will at least enumerate exactly how you should write your code for the P4 (at the machine level), the Intel compiler just follows that standard.
I read the internet for the articles.
I said some foolish things on USENET, but fortunately it seems to be just before the Google archives. I'm really hoping that none of my postings will be discovered...
I guess I can be accountable for my youth.
I've never used them, but Intel does provide high-performance math libraries. So, their compilers probably have real technical optimizations as well (not just marketing fluff).
How come we haven't seen quickies in a long time?
In this CNET article about the release of Intel's Linux compilers, they quoted the purchase price as $399 for a download, $499 for a CD. Somebody should tell them that blank CDs are a lot cheaper than they used to be...
(I know, I know. The boxed version probably also comes with some printed documentation, supposedly justifying the higher price. It still seemed funny to me..)
The new corperate america, u would think Intel would be giving this away seeing how AMD is kicking their butts and without this optimization the P$ is a slug.
Now they want to charge to make their dog chip work right?
and no one else sees this?
* Carthago Delenda Est *
Compare this cost to what it would cost you to pay an engineer to optimize his code.
The optimizations that an engineer would make would have a much more dramatic effect than tickling some opcodes.
Understood. Actually, I just realized the Makefiles would present an insurmountable problem in themselves.
Patent 6,274,978 patent looks likely. They seem to be saying they pass the electrons along a fiber.
Image 7b is the most useful; which isn't saying much.
Candescent Technologies has been working on this technology since 1991 and it looks like its about ready to go prime time with it. It has the same brightness, contrast, refresh time, and viewing angle that normal CRTs have but uses less power than LCDs in the same size package. Can't wait to hang one of these on the wall.
Science is the Real TRUTH!
Funny how everyone wants what they don't have:
"I hate this stupid CRT. I wish i had an LCD monitor. Cheapskate boss."
"I can't wait 'till i get this laptop back to the office so i can plug it into a CRT instead of having to squint at a stupid LCD."
--
Mod up a post Rob doesn't like and you'll never mod again
I know what you mean. Years of computer work have altered by biological structure to the point that if I'm away from a CRT for more than an hour or two I begin to feel weak.
Links:
IBM/Delphion
US Patent Office
I'm not exactly sure what compiler bloat is supposed to mean since what matters is the assemby the compiler generates and not how many lines of code the compiler was written in. Secondly it is very likely that a compiler written by Intel engineers for an Intel chipset will perform better than a general purpose compiler written by volunteers on Intel chipsets. Finally there are many that would argue that the Intel compiler has been of higher quality than gcc for quite sometime especially with regards to C++.
:(
PS: The fact that a post as empty as yours is at +4 is a sure sign that all the good posters have either left Slashdot or no longer actively partcipate. Sad.
NOOOOOOO! As a young, stupid college freshman in 1992, I discovered usenet and made a fool out of myself several times. I have been resting peacefully at night for the last decade, thinking that my past was safely hidden from the present, believing that nobody would be able to hold me responsible for the misdeeds of my youth. I guess I'm going to have to change my name now.
Since Intel is mainly a hardware producer, you'd think they would give away their compiler and even open up the source for it - and thereby boost their chip sales. Apparently, a program sold today is more worth to intel than 10 P4's in a week. Of course, they will have to offer support and maintain the compiler as well - which will cost money in the future. They're making a lot of strange decisions over in Santa Clara these days.
I have a friend who runs his own encryption company now and he's lamented to me that he wishes that he could excise some his posts made in earlier years from Bugtraq and USENET archives because he now receive several emails a day from script kiddies asking him to teach them how to steal AOL passwords and hack into hotmail.
SPEC rejected those benchmarks because Intel used a special proprietery compiler with the tests and not a normal compiler a developer would use.
Do you have evidence of this? I see plenty of SPEC CPU benchmarks using Intel compilers.
Yes, they cost more, but what are you really paying for?
I'd also be curious about recycle potential. There is much less material in an LCD, how about polution from disposal? How much of that can be reused and recycled? How about compared to a CRT?
Bob-
The Ludwig von Mises Institute. The reasoning individuals economics
How is there any "may" about this? IBM would have to be nuts to not license this technology to a mass-producer or two, they'll rake in the dough from licensing fees!
Every freelance graphic designer who has up until now had to surrender a big chunk of their living space to a hulking 19" or 21" CRT (because of finances or because of LCD color issues) will be flinging wads of money at the makers of slim CRT monitors. Not to mention the regular joes who just want a 17" or 18" LCD, but can't justify spending ~$1000 on a display.
Hell, I'd pony up for two of the things, just to replace what I have now and get my desk to stop bowing in the middle from the weight of my old-school 17" and 14".
~Philly
Really? Since using a 13.3" TFT LCD (on a Toshiba 2805) I've hated switching back to CRTs. In fact, I use my laptop at work rather than a company-provided desktop w/ CRT. (And until Ricochet died I would often not plug in to the company network, but that's another story.)
-- @rjamestaylor on Ello
... why don't they try searching for it?-)
(I'd love to see JMS's preproduction Netnews postings about Babylon 5, myself.)
Stupid job ads, weird spam, occasional insight at
This is a new technology that can integrate into existing production lines and can halve the depth of a CRT type tube. A TV normally 22 inches deep would be only 11 inches
This is nothing new, but it's an incremental improvement. I'd like some technical info before I can decide whether or not this is just a marketing stunt or other dubious improvement.
When TV sets first came out in the 1940s, their CRTs more resembled oscilloscopes. They were long, and with small screens. Their deflection angles were about 25 degrees.
As the early 1950s dawned, TV sets started to feature electromagnetic deflection. New, horizontal and vertical ouput tubes were suddenly able to support the current requirements of deflecting the beam 45 degrees towards a new big-screen 17" display.
The 1960s saw the beginning of the embrace of color television. As there are three electron beams in color TV sets, the neck was bigger than in monochrome sets. More deflection current was required to drive a 17" color set than a 17" black and white. High-tech new beam power amplifier tubes were developed to deal with the loads - compactron tubes like the 6LU8 and 21GY5 replaced the venerable 6BQ6. The spillover was that the mass-produced new high-power deflection tubes could also be used to make tighter deflection angles on black and white sets; the 19DUP4 was a Philco B&W picture tube released in 1965. It had a whopping 110 degree deflection angle, making for a TV set that had a 19" display but was only a foot deep.
Solid state TV sets using high-power MOSFET transistors have been able to handle the bigger current to drive new tight-deflection 110 degree color tubes. So far, it's been incremental.
But there remains a problem. A TV set's deflection yoke has to be driven with a sawtooth wave. There's a slow ramp up in voltage, then it quickly snaps down to off. Then another slow ramp and another quick snap. This corresponds to the beam sweeping sideways across the screen and then resetting to the left hand side very quickly.
Because the output amplifiers are neither fully on nor fully off, they're running in linear mode. All the energy not actually used to drive the yoke during the ramp is simply wasted as heat. But that energy isn't free... won't these things be meant to deal with Energy Star and other certifications? Tighter deflection means more deflection current means more wasted power in the amplifiers... and if the EPA buckles by defining a new guideline for thin monitors like these will purport to be, they'll be in competition with LCD monitors.
LCD will win.
The CRT will always be with us, but its time in the mainstream is coming to an end. This sounds too much like a marketing ploy, and goes too far against physics to be anything else.
Fire and Meat. Yummy.
The standard compiler is great for fucking around but you really want the real one for production systems. At my office the Windows weenies have a MSDN Universal subsrciption so they have all the cool toys anyway. If you aren't familiar with the wacky world of Windows, the MSDN Universal subscription is about $3000 per year and includes monthly (!!) shipments of the latest patched Microsoft OSs (all of them... Win2k Pro and Server, Me, et cetera), Visual Studio Enterprise (which includes VC++, InterDev and a whole bunch of other shit), plus beta releases of upcoming products. If you're a MS shop it's pretty sweet.
It may come with other toys; I'm not really sure, I'm not in the Windows group (I'm in the "web" group, we run AIX) and just use their VC++ install media on my NT workstation.
Come on, NT Server licenses cost $600-800 a piece. You think they're going to practically give away their fast compiler?
--
I like to watch.
The compiler crashed and burned. Their techical support site (which you get to by clicking on a creepy NDA) didn't contain much information. The links that did look interesting were broken
Eventually I found a document contained a list of known bugs. One of them was " was not included in the distribution. This will be fixed in the next update." Fantastic!
Has anyone out there successfully installed this compiler? My employers are very interested in using it (we want fast code for our intel machines), and I am very interested in trying it out.
The middle mind speaks!
The description reads like one of those analog devices that takes way too many alignment adjustments. But some of that can be automated, and components are stable enough now that many of the values can be fixed at the factory.
My guess is that the new scheme has some of the same elements of this one, and involves multiple correction coils to fix the beam distortions introduced when you deflect an electron beam through huge angles.
Point is, I'm betting that a compiler written for a specific chip and specific language (i.e. Intel's compiler) will perform better (i.e. produce better code) than a "compiler collection" wuth multiple pluggable front- and back-ends, all other things being equal. (Not that all other things necessarily are equal in this case (Go GNU!).)
This is an illogical statement. Apache and IIS support using multiple language to develop apps while my homemade webserver only supports C++. Does this mean my webserver is of higher quality than Apache or IIS? gcc is written a modular manner and the different language compilers are written by different people so talking about compiler bloat (whatever that means) is moot.
The important point is that Intel engineers with access to all sorts of internal Intel resources wrote a compiler that optimizes specifically for Intel chipsets while the gcc folk wrote a compiler that optimizes for x86 as well as other chipsets. The fact that the Intel guys spent 100% of their efforts on Intel chipsets while the gcc guys didn't is more likely to be the reason that Intel's compiler will outperform gcc and not because of any nebulous concept as compiler bloat.
That's what I saw here, on every human's face while they waited for their computers to figure out what to do next.
You'd think they'd at least have had a foosball or ping pong table or something. If I ever get into something like that, I'll remember to bring a copy of War and Peace.
I got my Linux laptop at System76.
Sometimes, yes. In some cases though, the compiler has more tricks up its sleeve. The issue is that a mature compiler has been programmed using the combined optimisation strategies from many _really_ good engineers; unless you've learnt every trick that all those engineers know, the compiler may be able to out-perform you, given the same piece of code to implement.
Of course, the compiler can only work with the C that it's given by the coder. There's things you can do like making structures an even power-of-two size which will speed the code up; this is a trade-off against memory usage which only the coder can make. But after that, it's up to the compiler to make it as efficient as possible. For instance, on some processors a compiler may implement a integer multiply by 9 as "shift-by-3, add original value" which is often faster than a single multiply instruction, and most engineers wouldn't write their C this way.
Grab.
I have a backup of the first few years. I have posted it here for posterity, but removed the header:
Test
This is a Test
TEST
Test!
anymore i have left out??
Once all corps finally merge into once, and they confiscate the slashdot servers, what mechanisms could be used to track each of us down?
Slashdot might be keeping our IP addresses. But maybe they are smart enough not to keep this?
Those who would give up liberty in exchange for security and DRM should switch to Microsoft Palladium!
I wouldn't mind having a Series III to sit in my office now, but they're probably beyond the dumpster phase...off to eBay!
This next song is very sad. Please clap along. -- Robin Zander
You can still use the ricochet devices peer to peer, over short range (up to a mile?) in Hayes emulation mode, or Starmode.
then as the field in the yoke decays, the Damper (say a 6AX4) starts conducting to finish the sweep.
Uhhh... The damper works simply to recapture energy that would otherwise cause barkhausen (sp?) oscillations. Pull out the damper, you'll still have full sweep (and maybe flyback, output and yoke damage). There'll be a hell of a ring at the right hand side of the screen, though.
What happens is that as the flux collapses (mostly in the flyback), the output circuit, which is designed to be resonant at 15,750Hz, rings like a bell at a higher harmonic.
During the tube era, if you look at a TV schematic, most of the dampers were even run off a separate winding on the flyback. The separate winding damped the oscillation and provided a nice bonus: the damper rectified the induced voltage, and it was fed to many output stages in the set as something called "B+ Boost".
I've got a fairly comprehensive collection of old TV sets. Of course, almost all the paper and early electrolytic capacitors are shot when you get an antique TV set. In one of them, a 1953 General Electric, I measured the B+ boost as 550V. It was used to drive the vertical and audio output stages. Symptomatically, it's like a modern TV set: if anything is wrong with the horizontal circuit, the set plays dead. Tube filaments with their cheery glow, but nothing else.
Fire and Meat. Yummy.
Yet they are in the black. Meanwhile, ambitious efforts like Infoseek, Lycos, Yahoo, and NBCi are floundering or defunct. Perhaps there is a lesson in that. I certainly hope so.