I'm really getting tired of running into immature brats playing online (M-rated) games. I'm not fond of censorship, but if it keeps some anti-social teen from going hostile on me in Diablo 2, I'm all for it.
Or just make them take a psych exam previous to buying the game. That way we'll know if they're serial killers trying to get their jollies.
I should probably comment on why I bought those cheap SyQuest and Imation drives, right after I said that all removable drives but MO pretty much suck. Basically, if they all suck, you might as well get the cheapest one.:) That way, I was able to take advantage of the technology, while not paying a premium for absolute crap hardware.
So, even though I will occasionally chastise people for buying low-end or cheaper mid-range stuff, I still reserve the right to buy them myself, should I decide that reliability (or functionality or whatever else the low-end or mid-range product is lacking) just isn't a factor in this purchase. I really hate advising people to use anything of debatable quality, because I know it will come back to haunt them. When it comes to me, I know that I can handle a couple floppies going bad from time to time. I don't have that same level of assurance when it comes to an anonymous stranger. I hope this explains why my post seemed to contradict itself.
Anyways, in response to your question, MO drives actually have taken off, but not in the desktop PC world. If you're worried about being compatible with your friends' PCs with ZIP drives, then you definitely would be better off buying a ZIP drive.
But why in the world would you want to buy multiple CD-RWs and create images of your hard drive? Wouldn't it be better to actually write to the hardware directly, at much, much higher speeds? Granted, a 10X CD-RW drive can write 640MB fairly quickly, but how long is that 10X drive going to take to write several gigabytes? Especially if it's over the EIDE ports, where you can only access one channel at a time, per port... which is a major problem if you have your hard drive and CD-RW on the same port (ie, master and slave respectively on the primary EIDE controller).
Argh. This is all getting much more complicated than I had intended, and I'm not sure that I'm making too much sense. The basic advantages are thus:
Advantages of MO over CD-RW
You can write to it like a hard drive (no creating images and burning to discs necessary)
Larger capacity (although, you can get smaller capacity MO drives, too)
More durable and reliable drives/media, since it's not a price-conscious, consumer-level product
It's like an external hard drive - there's no required software, drivers, or anything. True plug and play, as it were.
Disadvantages of MO
The media and drive are more expensive
You need a CDROM anyways
Some older MO drives are proprietary
MO media is more difficult to find locally
Basically, I'd say that MO is my favorite removable format. Tape isn't too bad, as it's damn cheap, but it suffers from other problems, most notably reliability and speed. You need to make sure that your backup tapes are good before you store them. Also, tapes are only good for backup, not much else. You can't really play MP3 files off of them...
I think your post is actually better than the article. You get the point across better than the journalist, who uses a sort of bad example for his hacker archetype. The Salon sysadmin seems a little clueless. I myself was impressed with the journalist's understanding of the tech world, though. They usually can't even figure out how to turn on a computer, much less write about computers.
It's true that floppy disks suffer from reliablity problems. Most cheap removable drives have had that plague them at least once in their life during the product development cycle.
I've heard so many horrible stories about Iomega Zip drives and/or disks going bad, I've stayed away from them. SyQuest (and their decendant Castlewood) offer ultra-cheap drives and media with about the same or slightly worse reliability. I decided to go with SyQuest when I bought a removable drive. At the time when Zip disks were 100MB, mine were 1GB.. and cheaper!
Eventually, I decided that playing around with parallel ports was pretty stupid, so I bought an Ultra2 Wide SCSI host adapter and some cool SCSI peripherals. Now, I've got four external U2W hard drives in an external rackmount chassis (bought it on ebay for a fraction of its actual worth), an external magneto-optical (MO) drive that reads and writes 2.6GB media at speeds near hard drives (I can actually install Windows 95 on a disc, without it being too slow to use), plus a very cheap 230MB magneto-optical (MO) 3.5" full-height drive. Unfortunately, as you might guess, the FH 230MB MO drive has given me some problems, as you don't really find too many FH 3.5" bays on PC cases. It doesn't really seem worth spending the money to make it external or fit a 5.25" bay, because the 2.6GB media is so damn cheap (~$30). The 230MB media is not much cheaper than that, even in quantity. My original plan was to replace my floppy drive with that MO drive, but it doesn't seem terribly likely that'll ever happen. So, I bought one of those Imation SuperDrives that stores 120MB or 1.44MB per floppy. It's very cool, except the media is way overpriced. Right now, I have only three superfloppies for it, and I don't plan on ever buying any more.
Remember, guys, there's an entire universe outside of the IDE world. Firewire and SCSI MO drives offer incredible advantages over IDE, floppy, and parallel port superfloppies, not to mention CD-Rs and CD-RWs. I hesitate to imagine making nightly backups to CD-RW or superfloppy. Eeechh.
The problem is that "real" enterprise servers don't ever go down, even when they change the hardware. Linux doesn't support many of the features that have been in more traditional UNIX servers for a decade already. Likewise, there are several companies making PCs that have hot swapable PCI cards, CPUs, etc. Does Linux support any of those? Nope. Does Linux support disconnection and reconnection of SCA hard drives? Does it even have a completed journaling file system?
Don't kid yourselves, guys.
Linux is awesome for hobbyists, good for workstations, and debatable for enterprise servers.
OpenBSD doesn't even support SMP, so don't feel all bad. Nobody has every feature. It's just a matter of priorities. Linus hasn't put enterprise features as his number one priority (yet?). Maybe in the future, Linux will compete better.
Get out into the "real world" and see what a real server can do before you start talking about Linux taking over.
(not a troll) (--- that is how you can tell it's not a troll) (would I lie to you?)
Seriously, I'm just trying to inject a little reality into the blind advocacy. I've run Linux for some 7 or 8 years now, so I obviously like it well enough.
That's silly. You're characterizing all religious people as ignorant morons. That doesn't really speak so well of you, unfortunately. Try to keep an open mind.
How do you think the Church explains children created with the help of fertility drugs? In vitro fertilization? Duh. Your post, although moderated up, shows only a lack of understanding and respect for others, rather than anything thoughtful.
If you don't trust the software, how can you trust the hardware? How do you know that 3COM, D-Link, Seagate, and Asus aren't all collecting information about you?
It'd be really easy for a network card to stick in one or two extra bytes every hour, until it had a sample of your processor's L2 cache (merely 64k on some processors). You wouldn't even notice it with a sniffer, unless you were paranoid.
Seriously, it'd be interesting to implement, just to see if anyone did notice it. And it'd be sort of fun to look over all those L2 cache dumps.
Anyways, lots of hardware manufacturers in the past ignored everyone but Microsoft customers; so much so, in fact, that it was either common knowledge or the equivalent (a strong, persistant rumor -- which is the best you can do when everything in the industry is covered by NDAs) that those companies were in bed with Microsoft big time. Or Intel. Why else would companies (say, 3COM) suddenly discontinue popular products (say, the Big Picture PCI video capture board) and replace them with brain-dead USB models that have half the features... and call it an upgrade? It doesn't make too much sense to me, except to believe that either Intel or Microsoft convinced them to 'retire' the PCI product.
So, of course it happens. And lots of corporations will happily ignore a significant minority (I'm sure that non-Win95/98/ME users make up over 10% of the desktop market; Win2k and NT must be somewhere around the size of the Mac or Linux desktop installations). Just look at the number of video card, sound cards, and USB peripherals that plain don't work outside of Win 98 or Win ME. Many more have seriously reduced functionality on Win NT4 or Win2k. Forget ever using them under a non-MS operating system.
Your best bet is to just hardware from manufacturers that are Linux friendly, ignore technology that does not have an international standards organization behind it, and wait until the second revision of any product before expecting stable (ie, in the stable kernel) Linux drivers.
Life sure is easy when you go with an all SCSI setup. Everything is a SCSI device. None of this weirdness with IDE, parallel port, or USB peripherals. Blech. They're all slower than SCSI, too. (Ultra DMA/66 and 100 are plain marketing bullshit. All you need is Ultra DMA/33 for 90% of the high performance EIDE hard drives.)
I suppose I just wanted to spout off.
Oh yeah, one more thing. There really aren't any I2O peripherals or motherboards out there, even though it's been years since Intel introduced it. There's the Asus P2B-D2, with integrated SVGA, SCSI, 10/100 NIC, and dual Pentium II support, but precious few other products. Mostly just triple channel Ultra 160 RAID adapters, which tend to cost as much as a used car.
Come on. This relentless Linux-can-do-and- support-everything-now! attitude gets a little annoying. I've been running Linux ever since I was old enough to drink (nigh on seven years), and I use Linux quite a lot more often than I drink. However, I remain just a little bit realistic about Linux's abilities and drawbacks.
First off, you can't just buy USB or 1394 peripherals and expect them to work under Linux. I don't care what kind of crazed advocate you are; you will never get a Logitech USB webcam to work under Linux. Believe me, I've tried. There are other, better-supported USB webcams out there, but I'd rather have mine supported. Oh well. Linux is a long way from supporting even half of the USB peripherals on the market. Look at the source code. We've got USB mice and keyboards (easy), a series of Kodak digital cameras (cool), some Palm type stuff (bleh), and a very few specific chipsets, like the ones used in the Creative Webcam series (good webcams, but not what I own).
I'm not trying to put down the USB support in Linux. I'm just saying be more realistic. Less hype and advocacy, more truth in advertising.
Windows 98 sucks in many ways, but USB support is not one of them.
p.s., I know that Logitech (may their mouse balls fall off) are the problem, not Linux programmers, but that doesn't change the fact that Logitech USB webcams aren't supported under Linux. I have the highest regard for the Linux USB programmers and the lowest regard for Logitech...
Give me a break. I'd like to have a PowerPC motherboard and CPU. Do you think I work for Microsoft? What the hell is wrong with you Apple people, anyways?
Actually, the C64 was a pretty awesome computer. I'm not the biggest fan of the x86 architecture, but at least you can upgrade it, buy stuff from more than one vendor, and assemble it yourself.
I don't want to buy an overpriced, pre-built system. I just want a motherboard and CPU.
Slashdot really sucks sometimes. I don't know why I bother posting. If you say anything controversial, it's obviously flamebait.
That's an interesting idea, but it does seem to be security through obscurity to me.:)
Another idea is to assemble a group of trusted hosts, with all traffic stopped by firewalls, keeping inbound traffic only from other trusted hosts.
In effect, you'd have your own network within the internet, about as secure as it could be. This way, unauthorized access would be near impossible. If you wanted to host important, sensitive, or controversial traffic, it wouldn't be open to inquisitive eyes.
Since all data is for internal purposes only, there's no reason not to use strong encryption. Even if someone did happen to intercept a packet, it would be useless to them.
Of course, if you want to host open information, the gopher idea is workable. Most people have never even heard of gopher, so they wouldn't know to look there if they were trying to cause trouble.
I read that the large cache Xeons will be released fairly soon. Probably at the end of the year. Until then, we'll have to make do with the small (256K) cache Xeons.
Sun engineers might be laughing at the Xeons, but the managers at Fortune 500 corporations are probably going "oooh" and "ahhhh" on cue from the Intel marketing.
I think what they do is use clusters of eight CPUs, so that if you want a 64 CPU system, it uses eight CPU daughtercards. The net result is a cluster of eight 8-way SMP daughtercards.
That's not exactly my number one choice for enterprise server. Still, it's quite innovative for an x86 vendor. IBM sure will go the distance for you if you really, really want to stick with the Wintel architecture.
But you're right about the domino effect. If we can get one Fortunate 500 company to see what's really going on, it's likely that several others will follow, and so on.
I don't care if they start using Linux or not. I just hope they stop using incompatible, proprietary software that claims to be compliant with the standard. It just means even yet more grief for those of us who have to interoperate with them.
So, are they going to stick to the Slot 2 format for the Xeons, then? I'd love to see a dual or quad FCPGA motherboard.
The Pentium Pros went from 256K all the way up to 1MB, which is pretty impressive, really. I remember all the articles about how Intel had poor yields on the Pros, especially the large cache chips. The 256k cache processors supposedly performed much worse than the 512k cache Pros, but I wonder how much of that was actually "benchmarketing" by Intel-friendly columnists.
I don't see why Intel, with all its vaunted engineers, had to give up so quickly on Socket 8. It's another example of Intel's defeatist attitude, I suppose.
Oh well. I've been thinking about grabbing a dual or quad processor PPro board, but so many of them are proprietary, it's an exercise in intense research to make an informed purchase.
Maybe I'll try to set up a cluster of two or three Athlons. I wonder if it would be worth it?
Haven't you heard? Slots are out again, sockets are in.:)
I have no idea why Intel went to Slot 1 and 2 in the first place. The L2 cache moved from on-die to off-die, then moved back on-die. What the hell? Is Intel just trying to make life difficult?
With the K6-III, AMD was able to implement on-die cache using the Socket 7 format. In fact, the K6-III used up to two megabytes of L3 cache, as well! It's too bad that the K6-III was such a spectacular failure. It could have been a very nice CPU.
Trivia: there's a socket-based Pentium II processor. Know what it is?
Damn. I meant to say, "it will NOT support 133 MHz FSB Coppermines". Repeat, the P2B-D will never support 133 MHz FSB Coppermines. It's a limitation of the BX chipset. Sorry.
It's incredibly difficult to do quad processing without a proprietary motherboard and/or case. The huge quad Xeon motherboards from Supermicro seem to be fairly standard; at least, they fit in the SC750A and SC760A cases, which are actually OEM Addtronics cases. (Check the Addtronics web site for details.)
There's an auction on eBay right now for a six-way Pentium Pro server, fully loaded, with all proprietary riser cards, proprietary case, and RedHat Linux 6.0. It looks interesting. I might bid on it, even though getting six VRMs will cost me a fortune! Not to mention paying for six 512K cache Pentium Pros...
Oh well. What good is cash if you just save it up?
p.s. AMD is never going to be anything more than a thorn in Intel's side. Intel is too big, too well-established, and too rich to suffer more than minor setbacks. But I like AMD a lot. I think AMD will become very profitable and carve themselves a nice niche in the x86 market: gamers and enthusiasts. This is a very good niche to have. Gamers and enthusiasts will spend big money to upgrade their processors every few months.
In my opinion, the best case scenario for AMD is that they take over 25% of the market: the low end, games and enthusiasts, and entry-level workstations (if AMD can ever manage to write an SMP chipset, like they promised a year ago).
I knew someone would challenge me on that statement, but I was too lazy to go back and reword it.
The original poster's claim was that you need dedicated hardware in order to get true random numbers. My claim is that you can sample any hardware, really. Sound cards, hard drive accesses, keyboard input, or whatever.
Is the BP-6 that good at dual processors, that you'd use it instead of a dual slot A board? I thought dual P-III mobos were only $20 more than the BP-6.
If anyone has links about the BP-6, please post. I'm considering buying one.
You can't use dual Slot A boards. They don't exist. The Athlon chipsets suck right now. Of course, that's mostly because VIA writes horrid chipsets.
If you want a real dual processor board, don't go with Abit. Abit makes cheap, four layer boards. You want something like Asus, Tyan, or Supermicro. They will cost more, but you get what you pay for, especially when it comes to motherboards. I learned my lesson a long time ago... if it's cheap, it's because they cut corners.
I'm using an Asus P2B-D, with is dual Slot 1. It supports up to 600 MHz Pentium III (Katmai) processors. Recent revisions of the P2B-D will also support 100 MHz FSB Coppermine processors (ie, the 500e, 550e, 600e, 650, 700, and 750). It will support the 133 MHz Coppermines (ie, 533eb, 600eb, 667, 733, 800, 866).
Get a Tyan or Supermicro i840 board if you want upgradability. i840 has some problems (like using a MTH, which slows down performance), but if you put two 733 MHz Coppermine processors in there, you'll smoke just about anything that AMD will be releasing in the next 12 months.
You're half right about most of the points you raised.
First, you can get true random numbers from software. Just sample the line-in from the SoundBlaster. It's vaguely possible that you'll get repeating values, due to the noise from the case fan(s), but I doubt that's really a concern for most people. This is the Wintel architecture, here.
Second, Intel did implement the hardware RNG. It's in the i810(e), i820, and i840 chipsets. It was never supposed to be on the CPU itself. There's no way to get hardware RNG on a BX chipset.
CPU IDs are mostly harmless... your computer has ten other hardware-generated, unique IDs in it already. What do you think the hard drive serial number is? Read the docs on your NIC. Etc, etc.
Slashdot Mirror
I'm really getting tired of running into immature brats playing online (M-rated) games. I'm not fond of censorship, but if it keeps some anti-social teen from going hostile on me in Diablo 2, I'm all for it.
Or just make them take a psych exam previous to buying the game. That way we'll know if they're serial killers trying to get their jollies.
How come this is still at a score of 2?
This sort of thing ends up in my local newspaper more than once a year, but not always with such a happy ending. And I live in the midwest...
So, even though I will occasionally chastise people for buying low-end or cheaper mid-range stuff, I still reserve the right to buy them myself, should I decide that reliability (or functionality or whatever else the low-end or mid-range product is lacking) just isn't a factor in this purchase. I really hate advising people to use anything of debatable quality, because I know it will come back to haunt them. When it comes to me, I know that I can handle a couple floppies going bad from time to time. I don't have that same level of assurance when it comes to an anonymous stranger. I hope this explains why my post seemed to contradict itself.
Anyways, in response to your question, MO drives actually have taken off, but not in the desktop PC world. If you're worried about being compatible with your friends' PCs with ZIP drives, then you definitely would be better off buying a ZIP drive.
But why in the world would you want to buy multiple CD-RWs and create images of your hard drive? Wouldn't it be better to actually write to the hardware directly, at much, much higher speeds? Granted, a 10X CD-RW drive can write 640MB fairly quickly, but how long is that 10X drive going to take to write several gigabytes? Especially if it's over the EIDE ports, where you can only access one channel at a time, per port... which is a major problem if you have your hard drive and CD-RW on the same port (ie, master and slave respectively on the primary EIDE controller).
Argh. This is all getting much more complicated than I had intended, and I'm not sure that I'm making too much sense. The basic advantages are thus:
Advantages of MO over CD-RW
Disadvantages of MO
Basically, I'd say that MO is my favorite removable format. Tape isn't too bad, as it's damn cheap, but it suffers from other problems, most notably reliability and speed. You need to make sure that your backup tapes are good before you store them. Also, tapes are only good for backup, not much else. You can't really play MP3 files off of them...
I think your post is actually better than the article. You get the point across better than the journalist, who uses a sort of bad example for his hacker archetype. The Salon sysadmin seems a little clueless. I myself was impressed with the journalist's understanding of the tech world, though. They usually can't even figure out how to turn on a computer, much less write about computers.
It's true that floppy disks suffer from reliablity problems. Most cheap removable drives have had that plague them at least once in their life during the product development cycle. I've heard so many horrible stories about Iomega Zip drives and/or disks going bad, I've stayed away from them. SyQuest (and their decendant Castlewood) offer ultra-cheap drives and media with about the same or slightly worse reliability. I decided to go with SyQuest when I bought a removable drive. At the time when Zip disks were 100MB, mine were 1GB.. and cheaper! Eventually, I decided that playing around with parallel ports was pretty stupid, so I bought an Ultra2 Wide SCSI host adapter and some cool SCSI peripherals. Now, I've got four external U2W hard drives in an external rackmount chassis (bought it on ebay for a fraction of its actual worth), an external magneto-optical (MO) drive that reads and writes 2.6GB media at speeds near hard drives (I can actually install Windows 95 on a disc, without it being too slow to use), plus a very cheap 230MB magneto-optical (MO) 3.5" full-height drive. Unfortunately, as you might guess, the FH 230MB MO drive has given me some problems, as you don't really find too many FH 3.5" bays on PC cases. It doesn't really seem worth spending the money to make it external or fit a 5.25" bay, because the 2.6GB media is so damn cheap (~$30). The 230MB media is not much cheaper than that, even in quantity. My original plan was to replace my floppy drive with that MO drive, but it doesn't seem terribly likely that'll ever happen. So, I bought one of those Imation SuperDrives that stores 120MB or 1.44MB per floppy. It's very cool, except the media is way overpriced. Right now, I have only three superfloppies for it, and I don't plan on ever buying any more. Remember, guys, there's an entire universe outside of the IDE world. Firewire and SCSI MO drives offer incredible advantages over IDE, floppy, and parallel port superfloppies, not to mention CD-Rs and CD-RWs. I hesitate to imagine making nightly backups to CD-RW or superfloppy. Eeechh.
The problem is that "real" enterprise servers don't ever go down, even when they change the hardware. Linux doesn't support many of the features that have been in more traditional UNIX servers for a decade already. Likewise, there are several companies making PCs that have hot swapable PCI cards, CPUs, etc. Does Linux support any of those? Nope. Does Linux support disconnection and reconnection of SCA hard drives? Does it even have a completed journaling file system?
Don't kid yourselves, guys.
Linux is awesome for hobbyists, good for workstations, and debatable for enterprise servers.
OpenBSD doesn't even support SMP, so don't feel all bad. Nobody has every feature. It's just a matter of priorities. Linus hasn't put enterprise features as his number one priority (yet?). Maybe in the future, Linux will compete better.
Get out into the "real world" and see what a real server can do before you start talking about Linux taking over.
(not a troll) (--- that is how you can tell it's not a troll) (would I lie to you?)
Seriously, I'm just trying to inject a little reality into the blind advocacy. I've run Linux for some 7 or 8 years now, so I obviously like it well enough.
That's silly. You're characterizing all religious people as ignorant morons. That doesn't really speak so well of you, unfortunately. Try to keep an open mind.
How do you think the Church explains children created with the help of fertility drugs? In vitro fertilization? Duh. Your post, although moderated up, shows only a lack of understanding and respect for others, rather than anything thoughtful.
If you don't trust the software, how can you trust the hardware? How do you know that 3COM, D-Link, Seagate, and Asus aren't all collecting information about you?
It'd be really easy for a network card to stick in one or two extra bytes every hour, until it had a sample of your processor's L2 cache (merely 64k on some processors). You wouldn't even notice it with a sniffer, unless you were paranoid.
Seriously, it'd be interesting to implement, just to see if anyone did notice it. And it'd be sort of fun to look over all those L2 cache dumps.
Anyways, lots of hardware manufacturers in the past ignored everyone but Microsoft customers; so much so, in fact, that it was either common knowledge or the equivalent (a strong, persistant rumor -- which is the best you can do when everything in the industry is covered by NDAs) that those companies were in bed with Microsoft big time. Or Intel. Why else would companies (say, 3COM) suddenly discontinue popular products (say, the Big Picture PCI video capture board) and replace them with brain-dead USB models that have half the features... and call it an upgrade? It doesn't make too much sense to me, except to believe that either Intel or Microsoft convinced them to 'retire' the PCI product.
So, of course it happens. And lots of corporations will happily ignore a significant minority (I'm sure that non-Win95/98/ME users make up over 10% of the desktop market; Win2k and NT must be somewhere around the size of the Mac or Linux desktop installations). Just look at the number of video card, sound cards, and USB peripherals that plain don't work outside of Win 98 or Win ME. Many more have seriously reduced functionality on Win NT4 or Win2k. Forget ever using them under a non-MS operating system.
Your best bet is to just hardware from manufacturers that are Linux friendly, ignore technology that does not have an international standards organization behind it, and wait until the second revision of any product before expecting stable (ie, in the stable kernel) Linux drivers.
Life sure is easy when you go with an all SCSI setup. Everything is a SCSI device. None of this weirdness with IDE, parallel port, or USB peripherals. Blech. They're all slower than SCSI, too. (Ultra DMA/66 and 100 are plain marketing bullshit. All you need is Ultra DMA/33 for 90% of the high performance EIDE hard drives.)
I suppose I just wanted to spout off.
Oh yeah, one more thing. There really aren't any I2O peripherals or motherboards out there, even though it's been years since Intel introduced it. There's the Asus P2B-D2, with integrated SVGA, SCSI, 10/100 NIC, and dual Pentium II support, but precious few other products. Mostly just triple channel Ultra 160 RAID adapters, which tend to cost as much as a used car.
Come on. This relentless Linux-can-do-and- support-everything-now! attitude gets a little annoying. I've been running Linux ever since I was old enough to drink (nigh on seven years), and I use Linux quite a lot more often than I drink. However, I remain just a little bit realistic about Linux's abilities and drawbacks.
First off, you can't just buy USB or 1394 peripherals and expect them to work under Linux. I don't care what kind of crazed advocate you are; you will never get a Logitech USB webcam to work under Linux. Believe me, I've tried. There are other, better-supported USB webcams out there, but I'd rather have mine supported. Oh well. Linux is a long way from supporting even half of the USB peripherals on the market. Look at the source code. We've got USB mice and keyboards (easy), a series of Kodak digital cameras (cool), some Palm type stuff (bleh), and a very few specific chipsets, like the ones used in the Creative Webcam series (good webcams, but not what I own).
I'm not trying to put down the USB support in Linux. I'm just saying be more realistic. Less hype and advocacy, more truth in advertising.
Windows 98 sucks in many ways, but USB support is not one of them.
p.s., I know that Logitech (may their mouse balls fall off) are the problem, not Linux programmers, but that doesn't change the fact that Logitech USB webcams aren't supported under Linux. I have the highest regard for the Linux USB programmers and the lowest regard for Logitech...
Wait a minute... doesn't Microsoft own a stake in Apple now, anyways?
I don't know why I'm still posting about this. I'm obviously dealing with morons.
Go ahead, mark this as flamebait, too. I don't give a shit.
Give me a break. I'd like to have a PowerPC motherboard and CPU. Do you think I work for Microsoft? What the hell is wrong with you Apple people, anyways?
If I don't want a Macintosh, it's flamebait?
I think I miss the logic in that.
Actually, the C64 was a pretty awesome computer. I'm not the biggest fan of the x86 architecture, but at least you can upgrade it, buy stuff from more than one vendor, and assemble it yourself.
I don't want to buy an overpriced, pre-built system. I just want a motherboard and CPU.
Slashdot really sucks sometimes. I don't know why I bother posting. If you say anything controversial, it's obviously flamebait.
That's an interesting idea, but it does seem to be security through obscurity to me. :)
Another idea is to assemble a group of trusted hosts, with all traffic stopped by firewalls, keeping inbound traffic only from other trusted hosts.
In effect, you'd have your own network within the internet, about as secure as it could be. This way, unauthorized access would be near impossible. If you wanted to host important, sensitive, or controversial traffic, it wouldn't be open to inquisitive eyes.
Since all data is for internal purposes only, there's no reason not to use strong encryption. Even if someone did happen to intercept a packet, it would be useless to them.
Of course, if you want to host open information, the gopher idea is workable. Most people have never even heard of gopher, so they wouldn't know to look there if they were trying to cause trouble.
I read that the large cache Xeons will be released fairly soon. Probably at the end of the year. Until then, we'll have to make do with the small (256K) cache Xeons.
Sun engineers might be laughing at the Xeons, but the managers at Fortune 500 corporations are probably going "oooh" and "ahhhh" on cue from the Intel marketing.
I think what they do is use clusters of eight CPUs, so that if you want a 64 CPU system, it uses eight CPU daughtercards. The net result is a cluster of eight 8-way SMP daughtercards.
That's not exactly my number one choice for enterprise server. Still, it's quite innovative for an x86 vendor. IBM sure will go the distance for you if you really, really want to stick with the Wintel architecture.
That wasn't communism. That was Stalinism.
But you're right about the domino effect. If we can get one Fortunate 500 company to see what's really going on, it's likely that several others will follow, and so on.
I don't care if they start using Linux or not. I just hope they stop using incompatible, proprietary software that claims to be compliant with the standard. It just means even yet more grief for those of us who have to interoperate with them.
So, are they going to stick to the Slot 2 format for the Xeons, then? I'd love to see a dual or quad FCPGA motherboard.
The Pentium Pros went from 256K all the way up to 1MB, which is pretty impressive, really. I remember all the articles about how Intel had poor yields on the Pros, especially the large cache chips. The 256k cache processors supposedly performed much worse than the 512k cache Pros, but I wonder how much of that was actually "benchmarketing" by Intel-friendly columnists.
I don't see why Intel, with all its vaunted engineers, had to give up so quickly on Socket 8. It's another example of Intel's defeatist attitude, I suppose.
Oh well. I've been thinking about grabbing a dual or quad processor PPro board, but so many of them are proprietary, it's an exercise in intense research to make an informed purchase.
Maybe I'll try to set up a cluster of two or three Athlons. I wonder if it would be worth it?
Haven't you heard? Slots are out again, sockets are in. :)
I have no idea why Intel went to Slot 1 and 2 in the first place. The L2 cache moved from on-die to off-die, then moved back on-die. What the hell? Is Intel just trying to make life difficult?
With the K6-III, AMD was able to implement on-die cache using the Socket 7 format. In fact, the K6-III used up to two megabytes of L3 cache, as well! It's too bad that the K6-III was such a spectacular failure. It could have been a very nice CPU.
Trivia: there's a socket-based Pentium II processor. Know what it is?
A chipset can't outperform another chipset. You're making nonsense claims. I don't really see any reason to reply to your other points.
Damn. I meant to say, "it will NOT support 133 MHz FSB Coppermines". Repeat, the P2B-D will never support 133 MHz FSB Coppermines. It's a limitation of the BX chipset. Sorry.
It's incredibly difficult to do quad processing without a proprietary motherboard and/or case. The huge quad Xeon motherboards from Supermicro seem to be fairly standard; at least, they fit in the SC750A and SC760A cases, which are actually OEM Addtronics cases. (Check the Addtronics web site for details.)
There's an auction on eBay right now for a six-way Pentium Pro server, fully loaded, with all proprietary riser cards, proprietary case, and RedHat Linux 6.0. It looks interesting. I might bid on it, even though getting six VRMs will cost me a fortune! Not to mention paying for six 512K cache Pentium Pros...
Oh well. What good is cash if you just save it up?
p.s. AMD is never going to be anything more than a thorn in Intel's side. Intel is too big, too well-established, and too rich to suffer more than minor setbacks. But I like AMD a lot. I think AMD will become very profitable and carve themselves a nice niche in the x86 market: gamers and enthusiasts. This is a very good niche to have. Gamers and enthusiasts will spend big money to upgrade their processors every few months.
In my opinion, the best case scenario for AMD is that they take over 25% of the market: the low end, games and enthusiasts, and entry-level workstations (if AMD can ever manage to write an SMP chipset, like they promised a year ago).
I knew someone would challenge me on that statement, but I was too lazy to go back and reword it.
The original poster's claim was that you need dedicated hardware in order to get true random numbers. My claim is that you can sample any hardware, really. Sound cards, hard drive accesses, keyboard input, or whatever.
It's easy.
You can't use dual Slot A boards. They don't exist. The Athlon chipsets suck right now. Of course, that's mostly because VIA writes horrid chipsets.
If you want a real dual processor board, don't go with Abit. Abit makes cheap, four layer boards. You want something like Asus, Tyan, or Supermicro. They will cost more, but you get what you pay for, especially when it comes to motherboards. I learned my lesson a long time ago... if it's cheap, it's because they cut corners.
I'm using an Asus P2B-D, with is dual Slot 1. It supports up to 600 MHz Pentium III (Katmai) processors. Recent revisions of the P2B-D will also support 100 MHz FSB Coppermine processors (ie, the 500e, 550e, 600e, 650, 700, and 750). It will support the 133 MHz Coppermines (ie, 533eb, 600eb, 667, 733, 800, 866).
Get a Tyan or Supermicro i840 board if you want upgradability. i840 has some problems (like using a MTH, which slows down performance), but if you put two 733 MHz Coppermine processors in there, you'll smoke just about anything that AMD will be releasing in the next 12 months.
You're half right about most of the points you raised.
First, you can get true random numbers from software. Just sample the line-in from the SoundBlaster. It's vaguely possible that you'll get repeating values, due to the noise from the case fan(s), but I doubt that's really a concern for most people. This is the Wintel architecture, here.
Second, Intel did implement the hardware RNG. It's in the i810(e), i820, and i840 chipsets. It was never supposed to be on the CPU itself. There's no way to get hardware RNG on a BX chipset.
CPU IDs are mostly harmless... your computer has ten other hardware-generated, unique IDs in it already. What do you think the hard drive serial number is? Read the docs on your NIC. Etc, etc.
Ignorant paranoia.
The best names:
Spitfire was such a cool name. AMD is run by a bunch of fools, but the engineers they picked up from Nexgen and Digital are great.