Amazon.com is far from profitable, though they move millions of dollars in book and other sales. No "profit" (as defined by paying back all their investments into infrastructure and making money past any running costs) in the foreseeable future.
Discuss why Linux didn't do as well as you hoped?
on
Web Server Comparisons
·
· Score: 4
Maybe you'll consider (Score -1: Troll), because I'm suggesting rational discourse over religious whining.
It appears from this thread of comments that the slashdot community is unhappy about all sorts of things that don't seem central to the issue. The comments becry ZDNet's advertising vs testing integrity and methodology, Caldera distro's shortcomings, Stronghold's performance vs other Apache releases, and why they didn't choose the EndOS-BeOS of solutions.
The issue, as I would see it, is "what can Linux do, to fare better in third-party comparisons?"
It appears from the article that there are several reasons that ZDNet listed why they felt Linux/Apache/Stronghold was limited. Let's start with those.
ZDNet chose to tune ALL the servers to have 68Mb of web source material and at least 68Mb of memory disk cache. Why did this give NT an unfair advantage? Why does Linux (or particularly the Caldera distro) solution not deal with RAM-rich servers as well as NT? I think the poster who complained of this meant the inverse: if it were a RAM-poor server, Linux would have the advantage in disk accesses.
ZDNet used multiprocessor servers. All religious handwaving aside, why did NT fare better by spinning threads than Apache could do by spinning processes? What is the big bottleneck in managing a process, that managing a thread doesn't have? They were using a brand-new MP kernel straight from Linus. Will the Linux kernel mature to deal with SMP situations and massive numbers of similar threads or processes better?
ZDNet suggested that in-process programming worked better for all the hairy e-commerce they decided to test. Though I think they should have configured some PHP or Mod_Perl into their mix, just as they had to bend to the SSL3-only constraint for another platform, they have a point. Writing modules is the way to go, to get inside the server and run fast. Besides PHP and Mod_Perl, where can Linux go to improve?
Most of these seem to suggest that ZDNet could have configured their Linux servers more like real-world Linux admins would, and would find better performance. This does talk poorly of Linux's ease-of-use, though they lay it on thick when they cry about config files. This is an education issue.
Slashdot is already slanted (pardon the pun) towards the Linux solution. Not every problem is a nail, and there are even different hammers for different nails. Let's be objective and constructive, instead of whining about every possible outside excuse. Improve the tool, and the tool will become the standard.
I fear that Open Source as a 'movement' has a few stigma to surmount before it can really be successful in *my* eyes.
I need to be able to depend on the authors who are out there to make good on their Open Source commitments and pledges. That is, if I hear that someone's planning to do something, that it actually get done. I can't make solid plans myself if I depend on someone else who doesn't follow through. This problem exists in Industrial/Commercial software too, but not to the same extent as the nascent Open Source community that is still figuring out how to be organized.
Example One: Whether for good or for bad, there's little doubt out there that Microsoft will ship Windows 2000, and if it's not in the promised February, it won't be far behind, and press releases are made occasionally when there've been slips or advances along that path. IS departments founded on Windows NT are making plans around the expected release of Windows 2000.
Example Two: the GNU pages list a lot of what has been written, but not a lot about what is planned. A few requests out there for hobbyists to slip in some changes, bug fixes, or pick up some project ideas, but not much in the way of "you can pick up GNU HotApp_1.4.605.3.tar.z in two weeks, and it will have HotFeature, HotGizmo and HotThingy completed. HotWishItem is planned for 1.5.000.0 in two months."
Example Three: I won't say where I've grabbed this quote, but you can find it yourself with one click from the page you're reading now. It was written by someone I think is safe to call an Open Source advocate. It was written before the company that was responsible for posting it announced a pending stock IPO. It was written about ten months ago, with no visible change since then.
You can download an ancient version of [the source code] below. Someday I'll post a new version, but honestly its a lower priority to me than it ought to be. I'm to busy ironing out kinks and adding features to take a couple days and create a distributable tar ball. It'll happen, but not tomorrow. I'm already working pretty much every waking second of the day.
It's almost as if we're the pre-1988 Olympics of software. No corporate branding, no professionals in the competition, nobody but amateurs who write code for the sport of it, for the achievement of pushing the envelope. I say almost because many if not most of the developers of Open Source code are doing it in their "copious free time," not as a job or profession. The phrase "copious free time" is always said with sarcasm, because there's never enough time to get to those lower-priority things like writing code that someone else might need. Packaging an after-dinner software project doesn't buy the dinner, while packaging software written at a day-job does.
I certainly don't want Open Source to be more Industrial or Commercial than it is, but until Open Source is treated as a professional pursuit, with published goals and objectives that can be met by the people who pick up projects, then we're stuck where we are... bad as it sounds, a "fringe" group instead of a powerful force that can succeed.
Who has any statistics or relevant experience with the Washington State laws that were passed to protect citizens from unwanted emails? I moved away from Washington State just before they enacted such legislation, and I've heard only a couple anecdotes of the aftermath.
A San Jose ISP (a2i)'s approach to "aiding" spammers figure out if recipient is a Washington citizen or not. Upshot: make it possible, make it consensual to the account holder, make it entirely too cumbersome for the spammer.
The logged activities of a law enforcement officer are legally admissible evidence of fact.
If YOU turn on logging in a channel, and sit there listening to slander/libel/threats, or receive DCC files that are stolen (unlicensed) property, the defendant can rightly point out how log files are highly suspect and can be forged.
If a law enforcement officer does precisely the same things, the log files will rarely come under the same scrutiny.
If you think that's not right, claim falsification of evidence next time you get a speeding ticket.:)
I got the QuickCam Pro USB, which is a 640x480 color camera. It's a nice little thing to plug into the USB port I've got on my laptop.
I've found that most webcam software (for Win32) cannot distinguish two identical devices on the USB "bus". That is, I can't plug two QuickCam Pro USBs to the machine, and have software choose which camera to watch.
I don't know enough about how USB works (does each device choose its own ID, or does the software tell a device to use a dynamic ID, or what).
Anyone else play around with this on Win32 or Linux? Same results? Same problems? Actual solutions?
In the same vein, perhaps as a solution, another device I'd like is a *controllable* USB turnswitch. Tell the switch (via USB) to turn on ports A and B, but turn off ports C and D. I know they'll have to make something like this if keyboards and mice all use USB; the five-servers-one-head cabinets all use Serial/PS2 switches now.
Will DVD-2 have a reasonable less-lossy compression scheme? I really hate putting up with NTSC, nevermind making the image even worse with highly lossy compression that adds obvious artifacts.
I had just gotten into laserdisc when DVD came out, and of course now it's impossible to find laserdiscs of good titles. (I know, whine whine.)
Just because my dad's 1934 antique Ford works doesn't mean he has no need for a more modern car. It works. When he can find leaded gasoline. When he chooses to cruise short distances without a shoulder seatbelt. Etc.
Any new version of Windows or any other OS will have some new features, and some new quirks that will have to be worked around or accepted; if you can't accept the quirks, you can't have the new features. Nothing surprising here. But the solution of "making do" with the status quo... that doesn't sound like the slashdotter nerd's approach to the world.
When WinNT4 was going into Release Candidate stages, there was the same sort of situation. The Wksta version would limit itself to 10 concurrent tcp sockets, while the Server version would allow any number of sockets.
There was a huge stink about such an arbitrary differentiation, and MSFT dropped that distinction before they shipped WinNT4.
Do recall that Windows 2000 is Windows NT 5.0, not Windows 98 +2. They assume you're in a deep pocket corporation with 30+ employees, a computer room for your T3, and an additional ~$1 or $2k per overpriced Compaq is just a drop in the MIS budget.
Being a tiny shop myself, I am not happy to hear that MSFT is announcing this plan again. Linux and AMD may make for a nice cheap server box, but I don't have the cash to afford a dedicated second box for that, and my business is in Win32 app development. If I can't serve a little HTML from the same box on which I write code, it'll be a problem.
A couple weeks after I cast my MSFT stockholders' votes via a secure website proxy, I read this article. I'm seeing a lot of nerds proclaim "but authenticating the votes as being genuine would be nigh impossible!"
And you call yourself nerds? Nerds are supposed to say, "Hey! It's software! We can make it do anything, we can solve any problem!"
What system do we have, today? Have you actually gone down to the local VFW or Baptist dance hall or Gymnasium? Watched a grandmother volunteer's wrinkled melanoma-covered hand as she followed the lines of barcodes and registered voters' names, until it fell on your own unprotected name and address? Wondered how a little checkmark near your name on this list was somehow more secure than what you could do in software?
We stand on the edge of deciding if we can export many-bit security methods to other countries. The politicians' argument: but we can't let them use our robust security methods! Shouldn't we be using our robust security methods?
For a voting method to succeed, it must match our system's ideals and exceed our current approach. It must exceed today's system's performance on security, voter's convenience, and counting accuracy.
Security: You have to validate the fact that each cast vote represents one registered voter; that each voter can, at their will, avoid coercion of choice from third parties; that any found fraudulent votes can be redacted within a reasonable period of time; and that each voter can keep their vote private if they desire.
(Partial solution: e-voter must specifically register for "license" to use e-vote methods; the use of said license implies the voter bears more responsibility over the e-vote security. Those who don't register for e-vote must use physical vote methods where government bears all responsibility over the p-vote security.)
Convenience: You have to make sure that those voters who find the gymnasiums and vfw rooms to be more convenient CAN use them. If an e-voter finds a website more convenient, that the website is open and available. It can show your current and past vote history. You can *change* your vote up to the vote deadline.
Accuracy: come on. Remember the old lady at the VFW using a checkbox? I'd trust a billion-transaction-per-minute server more than I'd trust Granny Mae.
No wonder that many slashdotters are assuming iD owns the whole franchise of Wolfenstein. People born after "Castle Wolfenstein" are legally able to vote and join the Army.
Anyone else remember the sound of a 1MHz 8bit processor trying to modulate an 8ohm speaker directly, to curse in German?
Kinda interesting that the entire top ten popular software were games, except for #7, which was an operating system. Talk about your "killer apps."
There are a lot of ways to post a photo that you have created but not stored, and still tell whether it was unmodified when you get a copy back from an untrusted reporter.
Off the top of my head,
Use a nonlossy compression method (GIF, etc), so you can embed authentication data IN the image.
Choose a 32-pixel area, and clear the low-order bit of the blue samples. Overlay a checksum for the rest of the image data into those 32 bits.
Use the same trick with the low-order bits of the green samples, and red samples (same or different pixels, your choice), using different hashing methods.
To get really funky, hide a pgp signature in the low-order color bits the same way.
Also, Photoshop has a digital signature filter which works on similar methods. I think it has lots of redundant information so that it won't break down with lossy compression (or even print-then-scan cycles). It was intended to FIND photos, not to DISCARD photos, that may be from a given source, such as porn CD-ROMs stockpiling illegal scans of Playboy (C) artwork.
The typical ubergeek rarely worries about the longevity of their computer hardware, and instead worries about whether or not the next model will have more horsepower or gamma rays.
However, I've heard that many of the random failures that creep in as a system gets older is not in fact due to bugs in vspazzod.386 or/etc/funky/.blah, but instead due to the relentless temperature cycling and subsequent warping of the many-layer fiberglass pcboards.
I can imagine that a bath of liquid nitrogen might help a processor operate at frequencies that annoy your neighbor's televisions, but can the fiberglass deal with it?
As an ex-Microsoftie, this quote hit me differently.
Then the big question dawned on me: What does it mean when Big Bill gives brand new P-III 450's running Linux to game-playing newbies who don't read reference books, manuals, How-To's or README's for a usability study? Can you say "viable desktop environment?"
So, Microsoft's been touted for years for hiring smart cookies. Even with the degradation of its standards and practices, and the complacency of being the largest corporation with an enviable bottom line, it's not easy to walk in and get a Microsoft job.
I still expect that the guy who called up wasn't an idiot. Sure, hadn't yet looked at the machine that Bill bought him, sure, hadn't used Linux before very much. Isn't that the perfect useability test case? And given that... how did Linux perform? The out-of-box experience seems to have failed.
I was on the team when Windows 95 was still called Windows 93, before it even grew the codename Chicago. At that time, the general manager of the desktop Windows Business Unit, Brad Silverberg, coined a mantra of the ideal in useability. He said that his [nontechnical] mother should be able to use Windows. Personally, I think we failed at reaching that ideal, but we made the right evolutionary step from Windows 3.1.
I've always contended that it would be stupid to outlaw flag-burning for the same reasons.
If you outlaw flag-burning, those who wish to lawfully disrespect government (uh, oxymoron?) would switch to bald-eagle-effigy burning. Then US-map burning. Then any number of other "national symbols."
Whites flaunting the word 'nigger' is a SYMBOL for racial divide. It's not the racial divide itself. The US flag is a SYMBOL of our freedoms that we love and want to uphold. It's not the freedoms themselves.
A domain name element can have, what, 32 characters from a set of at least 37 epsilon? (Geek:/[a-z0-9\-]{1,32}/i) That's 4.9e+55 second-level domain names. That doesn't count all the "f*ck.y*r.m*th*r.to" Tonganese and other offshore or forged DNS entries, or all the other creative "we.love.com/mitting-murder/" URLs you could devise.
Domain names. Eat all you want, we'll make more. Buy them for love. Buy them for hate. It doesn't really make a difference.
By default, the file is onle a "sample," called HOSTS.SAM. It has comments in it, to explain to the uninitiated. The location is different in Win95/98 and WinNT, but that's what Find Files is for, right?
Edit it with Notepad, and save it. Rename it to HOSTS. to remove the extension. (Note, some versions of Notepad won't let you Save As without an extension.)
It works as you should expect... it checks this before asking a RHCP or registered DNS server, and it's pretty good about rehashing itself after you modify it.
When I considered my first domain name, I kinda thought I'd use it forever. (I didn't.) When I considered my first public key, I kinda thought I'd use it forever. Reading about the GPG replacement for PGP, my first thoughts were
why in the heck would I relinquish or muddy my established identity contained within my existing uncompromised PGP public key and signature?
since stock PGP wouldn't handle GPG keys and methods, who could send me secure messages?
if I can't get around (1) and (2), why bother with GPG?
if GPG is OpenPGP, does that mean my rsa-patent-encumbered PGP5.5.5 or 6+ will be able to send to all those people who do make GPG keys?
Not directly experienced with this, but aren't most corporate email servers set up such that the clients do no local storage, and that clients' delete requests just "hide" the info? Otherwise, there's too much risk of other evidence-destruction liability, common with insider trading or espionage litigation for hi-tech companies.
Slashdot Mirror
Minor nitpick, and offtopic at that.
Amazon.com is far from profitable, though they move millions of dollars in book and other sales. No "profit" (as defined by paying back all their investments into infrastructure and making money past any running costs) in the foreseeable future.
Maybe you'll consider (Score -1: Troll), because I'm suggesting rational discourse over religious whining.
It appears from this thread of comments that the slashdot community is unhappy about all sorts of things that don't seem central to the issue. The comments becry ZDNet's advertising vs testing integrity and methodology, Caldera distro's shortcomings, Stronghold's performance vs other Apache releases, and why they didn't choose the EndOS-BeOS of solutions.
The issue, as I would see it, is "what can Linux do, to fare better in third-party comparisons?"
It appears from the article that there are several reasons that ZDNet listed why they felt Linux/Apache/Stronghold was limited. Let's start with those.
Why did this give NT an unfair advantage? Why does Linux (or particularly the Caldera distro) solution not deal with RAM-rich servers as well as NT? I think the poster who complained of this meant the inverse: if it were a RAM-poor server, Linux would have the advantage in disk accesses.
All religious handwaving aside, why did NT fare better by spinning threads than Apache could do by spinning processes? What is the big bottleneck in managing a process, that managing a thread doesn't have? They were using a brand-new MP kernel straight from Linus. Will the Linux kernel mature to deal with SMP situations and massive numbers of similar threads or processes better?
Though I think they should have configured some PHP or Mod_Perl into their mix, just as they had to bend to the SSL3-only constraint for another platform, they have a point. Writing modules is the way to go, to get inside the server and run fast. Besides PHP and Mod_Perl, where can Linux go to improve?
Most of these seem to suggest that ZDNet could have configured their Linux servers more like real-world Linux admins would, and would find better performance. This does talk poorly of Linux's ease-of-use, though they lay it on thick when they cry about config files. This is an education issue.
Slashdot is already slanted (pardon the pun) towards the Linux solution. Not every problem is a nail, and there are even different hammers for different nails. Let's be objective and constructive, instead of whining about every possible outside excuse. Improve the tool, and the tool will become the standard.
I fear that Open Source as a 'movement' has a few stigma to surmount before it can really be successful in *my* eyes.
I need to be able to depend on the authors who are out there to make good on their Open Source commitments and pledges. That is, if I hear that someone's planning to do something, that it actually get done. I can't make solid plans myself if I depend on someone else who doesn't follow through. This problem exists in Industrial/Commercial software too, but not to the same extent as the nascent Open Source community that is still figuring out how to be organized.
Example One: Whether for good or for bad, there's little doubt out there that Microsoft will ship Windows 2000, and if it's not in the promised February, it won't be far behind, and press releases are made occasionally when there've been slips or advances along that path. IS departments founded on Windows NT are making plans around the expected release of Windows 2000.
Example Two: the GNU pages list a lot of what has been written, but not a lot about what is planned. A few requests out there for hobbyists to slip in some changes, bug fixes, or pick up some project ideas, but not much in the way of "you can pick up GNU HotApp_1.4.605.3.tar.z in two weeks, and it will have HotFeature, HotGizmo and HotThingy completed. HotWishItem is planned for 1.5.000.0 in two months."
Example Three: I won't say where I've grabbed this quote, but you can find it yourself with one click from the page you're reading now. It was written by someone I think is safe to call an Open Source advocate. It was written before the company that was responsible for posting it announced a pending stock IPO. It was written about ten months ago, with no visible change since then.
It's almost as if we're the pre-1988 Olympics of software. No corporate branding, no professionals in the competition, nobody but amateurs who write code for the sport of it, for the achievement of pushing the envelope. I say almost because many if not most of the developers of Open Source code are doing it in their "copious free time," not as a job or profession. The phrase "copious free time" is always said with sarcasm, because there's never enough time to get to those lower-priority things like writing code that someone else might need. Packaging an after-dinner software project doesn't buy the dinner, while packaging software written at a day-job does.
I certainly don't want Open Source to be more Industrial or Commercial than it is, but until Open Source is treated as a professional pursuit, with published goals and objectives that can be met by the people who pick up projects, then we're stuck where we are... bad as it sounds, a "fringe" group instead of a powerful force that can succeed.
Who has any statistics or relevant experience with the Washington State laws that were passed to protect citizens from unwanted emails? I moved away from Washington State just before they enacted such legislation, and I've heard only a couple anecdotes of the aftermath.
A San Jose ISP (a2i)'s approach to "aiding" spammers figure out if recipient is a Washington citizen or not. Upshot: make it possible, make it consensual to the account holder, make it entirely too cumbersome for the spammer.
The same ISP gave a link to one service/sig on the issue: www.wa-state-resident.com
The logged activities of a law enforcement officer are legally admissible evidence of fact.
If YOU turn on logging in a channel, and sit there listening to slander/libel/threats, or receive DCC files that are stolen (unlicensed) property, the defendant can rightly point out how log files are highly suspect and can be forged.
If a law enforcement officer does precisely the same things, the log files will rarely come under the same scrutiny.
If you think that's not right, claim falsification of evidence next time you get a speeding ticket.
I got the QuickCam Pro USB, which is a 640x480 color camera. It's a nice little thing to plug into the USB port I've got on my laptop.
I've found that most webcam software (for Win32) cannot distinguish two identical devices on the USB "bus". That is, I can't plug two QuickCam Pro USBs to the machine, and have software choose which camera to watch.
I don't know enough about how USB works (does each device choose its own ID, or does the software tell a device to use a dynamic ID, or what).
Anyone else play around with this on Win32 or Linux? Same results? Same problems? Actual solutions?
In the same vein, perhaps as a solution, another device I'd like is a *controllable* USB turnswitch. Tell the switch (via USB) to turn on ports A and B, but turn off ports C and D. I know they'll have to make something like this if keyboards and mice all use USB; the five-servers-one-head cabinets all use Serial/PS2 switches now.
Will DVD-2 have a reasonable less-lossy compression scheme? I really hate putting up with NTSC, nevermind making the image even worse with highly lossy compression that adds obvious artifacts.
I had just gotten into laserdisc when DVD came out, and of course now it's impossible to find laserdiscs of good titles. (I know, whine whine.)
Just because my dad's 1934 antique Ford works doesn't mean he has no need for a more modern car. It works. When he can find leaded gasoline. When he chooses to cruise short distances without a shoulder seatbelt. Etc.
Any new version of Windows or any other OS will have some new features, and some new quirks that will have to be worked around or accepted; if you can't accept the quirks, you can't have the new features. Nothing surprising here. But the solution of "making do" with the status quo... that doesn't sound like the slashdotter nerd's approach to the world.
When WinNT4 was going into Release Candidate stages, there was the same sort of situation. The Wksta version would limit itself to 10 concurrent tcp sockets, while the Server version would allow any number of sockets.
There was a huge stink about such an arbitrary differentiation, and MSFT dropped that distinction before they shipped WinNT4.
Do recall that Windows 2000 is Windows NT 5.0, not Windows 98 +2. They assume you're in a deep pocket corporation with 30+ employees, a computer room for your T3, and an additional ~$1 or $2k per overpriced Compaq is just a drop in the MIS budget.
Being a tiny shop myself, I am not happy to hear that MSFT is announcing this plan again. Linux and AMD may make for a nice cheap server box, but I don't have the cash to afford a dedicated second box for that, and my business is in Win32 app development. If I can't serve a little HTML from the same box on which I write code, it'll be a problem.
A couple weeks after I cast my MSFT stockholders' votes via a secure website proxy, I read this article. I'm seeing a lot of nerds proclaim "but authenticating the votes as being genuine would be nigh impossible!"
And you call yourself nerds? Nerds are supposed to say, "Hey! It's software! We can make it do anything, we can solve any problem!"
What system do we have, today? Have you actually gone down to the local VFW or Baptist dance hall or Gymnasium? Watched a grandmother volunteer's wrinkled melanoma-covered hand as she followed the lines of barcodes and registered voters' names, until it fell on your own unprotected name and address? Wondered how a little checkmark near your name on this list was somehow more secure than what you could do in software?
We stand on the edge of deciding if we can export many-bit security methods to other countries. The politicians' argument: but we can't let them use our robust security methods! Shouldn't we be using our robust security methods?
For a voting method to succeed, it must match our system's ideals and exceed our current approach. It must exceed today's system's performance on security, voter's convenience, and counting accuracy.
Security: You have to validate the fact that each cast vote represents one registered voter; that each voter can, at their will, avoid coercion of choice from third parties; that any found fraudulent votes can be redacted within a reasonable period of time; and that each voter can keep their vote private if they desire.
(Partial solution: e-voter must specifically register for "license" to use e-vote methods; the use of said license implies the voter bears more responsibility over the e-vote security. Those who don't register for e-vote must use physical vote methods where government bears all responsibility over the p-vote security.)
Convenience: You have to make sure that those voters who find the gymnasiums and vfw rooms to be more convenient CAN use them. If an e-voter finds a website more convenient, that the website is open and available. It can show your current and past vote history. You can *change* your vote up to the vote deadline.
Accuracy: come on. Remember the old lady at the VFW using a checkbox? I'd trust a billion-transaction-per-minute server more than I'd trust Granny Mae.
No wonder that many slashdotters are assuming iD owns the whole franchise of Wolfenstein. People born after "Castle Wolfenstein" are legally able to vote and join the Army.
Anyone else remember the sound of a 1MHz 8bit processor trying to modulate an 8ohm speaker directly, to curse in German?
Kinda interesting that the entire top ten popular software were games, except for #7, which was an operating system. Talk about your "killer apps."
There are a lot of ways to post a photo that you have created but not stored, and still tell whether it was unmodified when you get a copy back from an untrusted reporter.
Off the top of my head,
Also, Photoshop has a digital signature filter which works on similar methods. I think it has lots of redundant information so that it won't break down with lossy compression (or even print-then-scan cycles). It was intended to FIND photos, not to DISCARD photos, that may be from a given source, such as porn CD-ROMs stockpiling illegal scans of Playboy (C) artwork.
Now all I need on this thing:
+ a flip-out dentist's mirror so I can peer behind a case that's against the wall
+ an LED to light the area because it's under the desk against the wall
+and a thingy that tells me if the socket I'm feeling is a DB15, DB9, DB25, or PS2, DIN, or RJ11 or RJ14
Something that magically flipped the cable I'm holding to orient said DB or DIN or RJ plug the right way would be a bonus.
The typical ubergeek rarely worries about the longevity of their computer hardware, and instead worries about whether or not the next model will have more horsepower or gamma rays.
However, I've heard that many of the random failures that creep in as a system gets older is not in fact due to bugs in vspazzod.386 or
I can imagine that a bath of liquid nitrogen might help a processor operate at frequencies that annoy your neighbor's televisions, but can the fiberglass deal with it?
As an ex-Microsoftie, this quote hit me differently.
Then the big question dawned on me: What does it mean when Big Bill gives brand new P-III 450's running Linux to game-playing newbies who don't read reference books, manuals, How-To's or README's for a usability study? Can you say "viable desktop environment?"
So, Microsoft's been touted for years for hiring smart cookies. Even with the degradation of its standards and practices, and the complacency of being the largest corporation with an enviable bottom line, it's not easy to walk in and get a Microsoft job.
I still expect that the guy who called up wasn't an idiot. Sure, hadn't yet looked at the machine that Bill bought him, sure, hadn't used Linux before very much. Isn't that the perfect useability test case ? And given that... how did Linux perform? The out-of-box experience seems to have failed.
I was on the team when Windows 95 was still called Windows 93, before it even grew the codename Chicago. At that time, the general manager of the desktop Windows Business Unit, Brad Silverberg, coined a mantra of the ideal in useability. He said that his [nontechnical] mother should be able to use Windows. Personally, I think we failed at reaching that ideal, but we made the right evolutionary step from Windows 3.1.
Now, how well can your mom use Linux?
I've always contended that it would be stupid to outlaw flag-burning for the same reasons.
/[a-z0-9\-]{1,32}/i) That's 4.9e+55 second-level domain names. That doesn't count all the "f*ck.y*r.m*th*r.to" Tonganese and other offshore or forged DNS entries, or all the other creative "we.love.com/mitting-murder/" URLs you could devise.
If you outlaw flag-burning, those who wish to lawfully disrespect government (uh, oxymoron?) would switch to bald-eagle-effigy burning. Then US-map burning. Then any number of other "national symbols."
Whites flaunting the word 'nigger' is a SYMBOL for racial divide. It's not the racial divide itself. The US flag is a SYMBOL of our freedoms that we love and want to uphold. It's not the freedoms themselves.
A domain name element can have, what, 32 characters from a set of at least 37 epsilon? (Geek:
Domain names. Eat all you want, we'll make more. Buy them for love. Buy them for hate. It doesn't really make a difference.
By default, the file is onle a "sample," called HOSTS.SAM. It has comments in it, to explain to the uninitiated. The location is different in Win95/98 and WinNT, but that's what Find Files is for, right?
Edit it with Notepad, and save it. Rename it to HOSTS. to remove the extension. (Note, some versions of Notepad won't let you Save As without an extension.)
It works as you should expect... it checks this before asking a RHCP or registered DNS server, and it's pretty good about rehashing itself after you modify it.
When I considered my first public key, I kinda thought I'd use it forever.
Reading about the GPG replacement for PGP, my first thoughts were
Not directly experienced with this, but aren't most corporate email servers set up such that the clients do no local storage, and that clients' delete requests just "hide" the info? Otherwise, there's too much risk of other evidence-destruction liability, common with insider trading or espionage litigation for hi-tech companies.