Actually, for your typical "AOL Mom" (or "Joe Six-Pack"), Windows is even more of a nightmare than linux. If you want them to have a machine that's friendly to them, you should recommend that they buy a Mac.
Now, the usual reaction to this is "Macs are more expensive." Well, yes and no. There's been plenty written here and elsewhere explaining that, for comparable machines with comparable software, there is little if any price difference. The apparent difference comes from the fact that you can buy cheap, shoddy Windows boxes, but not cheap, shoddy Macs.
Recommending Windows to a non-expert is sorta like recommending a Yugo or Lada auto to a non-mechanic. Yeah, you could buy a car without locks on the doors, without any heat or air, and without a speedometer, spare tire, gas gauge or oil warning light. It would probably be cheaper.
Here in the US, it's illegal to sell such cars for street use. You can make one as a specialized vehicle, e.g. for racing. But you can't get a license to drive it on the street.
Maybe eventually the growing problems with Microsoft-supported viruses, worms, etc (and the somewhat unrelated problem of spam) will lead to similar laws preventing using unlicensed computers on the public Internet. It probably won't happen Real Soon Now, though.
Anyway, suggesting that novices use Windows is rather irresponsible. Novices should use machines that don't blame the users for the builtin problems. Such machines should come with normal security enabled by default. They shouldn't default to running code from strangers. They should warn users when they do something that will run new software from outside sources (especially email and web pages).
Of course, this means that novices shouldn't be buying the cheapest, shoddiest computer that's for sale at the neighborhood price-is-everything marketplace. And let's face it, that is pretty much a good description of mass-market Windows boxes.
... that the jobs I've had for the past 5 years or so have all been primarily developing software that runs on linux systems?
Funny thing is that these jobs have been paid for mostly by non-US companies who are trying to get out from under the thumb of either IBM or Microsoft (or both). And they're hiring Americans like me to help them do it.
A big selling point has been that N years from now I can guarantee that the software will still run and they'll still be able to read all their files. They've learned the hard way that this isn't always true with proprietary systems.
And I can easily explain to them how they can verify that there are no hidden tricks (trojans, backdoors, etc) in my code or in any of the lower-level software. Neither my code nor anything in "the system" can be sending their data off to some stranger's data warehouse. Granted, they'll have to keep around a staff of unix/linux geeks, who will both study the code and monitor the appropriate online fora. But they don't need to hire as many such geeks as they have on site now to keep their IBM/MS stuff running, so even that's a win.
Maybe eventually we'll see the day when all software has been written and no more is needed. But I suspect that day's still a long way off. And the world is growing more and more dependent on smaller and smaller computers to keep everything running.
So for the forseeable future, they'll still need lots of people who understand that, no matter what managers or marketing people say, 2+2 is always 4, not 5 or 3.95 or something desirable. (Except when it's 3.99999999998 of course, but any true geek will understand that, too.;-) You can't get software to work without a good understanding that computers don't respond to positive thinking or marketing, and such people will always be a tiny minority.
So I'll predict that people with the twisted (i.e., logical) minds required by programming will continue to have jobs until long after all of us are gone.
Of course, we may all have to move to India or China, as the patent system shuts down software development in the Western world.
A better question would be: What is preventing millions upon millions of people from switching to a free OS? Quality would be my guess.
Well, I've asked a lot of MS users about this, and so far every one of them has disagreed with you. Their answer is always of the form "The software I want/need is only available on DOS/Windows (depending on the year that I asked)."
So then I ask my followup question: "How many other kinds of computer systems did you look at?" And their answer is always the same: "None."
Invariably, MS users just know that software is only available on MS systems. They don't need to do any market research, because they already know the answer. There's no point in wasting time looking for something that doesn't exist.
There are, of course, lots of people who have done the obvious searches. They aren't MS users. They easily find alternatives, determine that the alternative is almost always of higher quality, and go with it.
But BillG and company (and IBM before them) have become rich betting that the great majority will never do even the slightest study of what's available. All it takes is a good-size marketing budget, and whatever you make will be the market leader, whether it's good or bad quality, because few people will ever look for alternatives.
Bill himself was in the enviable position to be able to use daddy's money to get into Harvard B-school, where he made the connections that allowed him to leverage an IBM marketing budget and do an end run against all those pipsqueaks who had demoed the viability of a "desktop" computer market. His marketing budget has remained greater than the total operating budget of all his competitors combined. So the great majority who just go with the "market leader" continue to buy from him, because they know there's no point at looking at anyone else's nonexistent software.
Anyway, try it yourself. Ask random MS users to name a single piece of software that runs on linux. I'll predict that, with very few exceptions, most of them will be unable to come up with anything at all. They have never looked, and they never will.
This shoots down the idea that they're buying based on any sort of "quality" determination. They're buying from the only software supplier that exists in their world.
(In the mainframe world, the same situation still exists, with "IBM" for "Microsoft" throughout.)
NOTHING should ever be backed up locally, because like you said, when fire hits, your backups are gone.
You hit that nail on the head. And even if you have all those tapes sitting in your basement, as I do, you'll find that you can't read them on the next machine you get. Even if you do order one of the tape drives that's available, they'll be physically incompatible with your old tapes. You could try moving the old tape drive from the dead machine to the new one, but you'll find that drivers are no longer available that work with the current OS.
What I do is follow Linus Torvalds' advice: Make your stuff sufficiently interesting and useful that others are willing to back them up on their disks. I have several "guest" accounts on machines in widely-separates places, with permission to back up all my stuff there. It's all visible via the web, though of course there are a number of "hidden" directories whose names you are highly unlikely to guess until long after the actuarial folks say I'll be gone.
That works a whole lot better than gambling that N years from now you'll have a working driver for a functional tape drive that can read a tape that hasn't deteriorated after N years.
Now, if I could afford $15,000 for a top-quality tape drive in the new machine that I buy every 5 years or so, I might feel differently. But mostly, if you suggest backup to tape, I'd just ask what planet you're from. It's obviously not one that I've lived on. Backup tapes have very rarely, in my experience, turned out to be actually readable on the machine where you need to read them.
Yeah, I remember the fuss over Hatch's web site using unlicensed "pirated" software. They apparently did get a license after this was brought to light.
But the current legislation is even more extreme, in a sense. Hatch and his people wouldn't be able to avoid the issue by belatedly applying for a license. Even if they had a license to use things like email and browsers, the license wouldn't stop the copyright violations that these tools make easy. You can't give someone a license to violate other people's copyrights.
We should be pointing out to them that email and browsers are primary examples of tools that enable copyright violation. If they really believe in what they're proposing, they should remove such software from their own computers. If they don't, they are openly telling the world that they consider themselves exempt from the rules that they want to impose on the rest of us.
I'm assuming their version of patent reform would be "Everyone else has to give up their bad patents, but we get to keep them."
Maybe, maybe not. Bezos' comment does sorta imply that he understands the idiocy of the situation but is trying to "fight it within the system" by forcing it to an absurd extreme. There is precedent for this approach.
OTOH, it's entirely possible that his understanding is that absurdities like this will have the effect of eliminating all those pesky little competitors, while leaving his bigger company intact. This would happen if the legislators can't be persuaded to improve the mess they've made.
There is some history of business leaders pushing for legislation that would outlaw their own business practices. There were interesting articles about a case several decades back, when the state of Oregon was considering the legislation that would eliminate most of the pollution from the waters of the Willamette River system. A number of the biggest polluters were publicly pushing for the legislation. Their reasoning was straightforward: "We can't stop our own pollution as long as our competitors are allowed to continue polluting; we'd be out of business. Our only choice is to push for laws that will force everyone to stop polluting." It worked, and the Willamette River is now a lot cleaner than it was back then. It helped a lot that the entire drainage basin was controlled by a single state government.
The corporate world does have a few leaders who are socially responsible. There's a good chance that Bezos is one of them. We'll have to watch him and see. I wonder how much real evidence we have now?
[A]re they going to ban FTP since you can distribute copyrighted material over FTP as well?
Probably. And they'll also ban SMB file sharing, since that clearly enables copyright violation.
Of course, this will make Windows boxes a lot more secure, so maybe it's a Good Thing.
And who'd have thought that Orrin Hatch would be the one that would finally force Microsoft to remove a major security hole from their software?
Actually, when you consider that unix-like systems are multi-user systems with a file system that encourages shared files and enables cooperative development, I'd wonder whether this bill would ban unix file systems.
Maybe what we should do is check out the computers in use by Hatch's political organization, and demand that they remove all software that enables copyright violation, starting with their email software. Maybe that would get the message across.
[M]anagement understands cost and benefit. They don't respond to features and feelings.
I would dispute this. I've seen any number of cases where management is dedicated to Microsoft and/or IBM. When I ask to see their numbers jjustifying this, they never have them. I've never seen anything approaching a financial study of the topic in any company. (Well, actually I did see one 15 years ago. That company converted over to Sun.;-)
Most managers just "know" that MS and/or IBM is the best. Since they have no numbers to back this, I'd conclude that they are indeed responding to features and feelings. Much of this is just going with what they are familiar with, which is pure "feelings".
I've found that a good way to get them really confused is to suggest that maybe MS or IBM is the best, but they don't actually know that because they haven't done the usual comparisons. If you're buying a car, you at least make an attempt to check on what's available; you don't just go with the biggest advertiser. Managers will routinely ask for price quotes from suppliers on things like paper. They'll compare telephone companies' services and prices (if they have more than one available). And so on for nearly everything their company pays for.
But when it comes to computers, most managers throw all this out and go with their feelings that are based primarily on who has the biggest marketing budget.
Interesting. I wonder if anyone has found a fix for the/. problem that I and a few others have reported, in which firefox simply stops responding to clicks on links within/. pages?
It does seem to be a rather bizarre bug, with no obvious correlation with the pages being viewed. But it's serious enough on my Mac (but not my linux box) that I simply use a different browser.
I just checked out msdn.microsoft.com from Camino. The only problem was that it forced a tiny font, and a quick CMD-+ fixed that. I have a collection of browsers on my linux and Mac boxen, because part of my job is testing web pages for usability.
We might note that Microsoft-related sites have a history of checking the client's ID string and sending garbled pages to non-IE browsers. There was a recent story here about the Opera folks getting a few million bucks from msn.com over this.
When someone complains about a site not rendering well on anything but IE, perhaps you should mention this practice. The traditional term for it is "sabotage", and sensible people wouldn't stand for it.
Checking a client's ID isn't necessarily a bad idea, of course. I do that a lot. Lately I've been testing my pages against a number of PDAs' browsers. Sometimes the only thing that works is to check the ID and generate a page that's formatted for a tiny screen. But it's also common to use this to discriminate against people who use non-MS software.
Then again, maybe he doesn't have a clue what he's talking about.
Maybe, maybe not.
Various historians of the topic have pointed out that until very recently, copyright only applied to publishers. There was no question about "consumers" sharing things they liked with friends; that was obvious ly legal, and anyone with the slightest sense of the Public Welfare would support it. A lot of authors have gone public with the comment that they like people sharing their writings with others. That's how you generate a reputation, dummy! Only an idiot (or a manager;-) would object to that.
In our household, we have a number of DVDs that we've purchased after either having a "pirate" copy that we made from a broadcast or having borrowed them from friends. If we don't particularly like something, we watch it once then erase the tape and forget about it. If we like it, we look for the commercial DVD and buy it.
Similarly with music. I have a number of CDs that I've bought after having heard them on the radio or on a CD borrowed from a friend. If I can't find the CD commercially (highly likely because I have a lot of "folk" material from around the world that's not available in the US), I make a "pirate" copy. If I then find it for sale, I buy a copy.
Last year we had the fun news that Eminem's latest CD was the top download. Then, after it was released, it became the top-selling CD in the world. (My wife has a copy in her car; she loves shocking her professional colleagues with it, not to mention whistling the tunes at work for the benefit of the few who recognize it.;-) So much for downloads hurting sales. Eminim himself has figured out that this is good publicity, and he remarked that he should pre-release his next album on the P2P sites.
If you want people to buy something, you have to make it available to them so they can decide whether they want to buy it. If I can't hear it or see it, I'm not about to spend my money on it. And if you threaten me when I try to determine whether I like it, I'll walk away and not bother with you.
I think Micheal Moore understands this. He wants his stuff seen by the public. He has figured out that the only way he'll ever succeed at his chosen profession is if the public is familiar with him.
We haven't seen Fahrenheit 9/11 (yet). My wife has already suggested that when it comes out on DVD, she'll probably order a copy just on general principles. It doesn't hurt that we've seen "Roger and Me" and "Bowling for Columbine", and really liked both of them. We have a DVD of the first, but not the second. This may change.
Indeed, we are a long way from making 40,000-km-long carbon nanotubes. One serious question is how practical it is to depend on something like this being built.
But it's only been a few years since the first carbon nanotubes were created. The first were only nanometers long. Then others reported making some that were micrometers long. And a couple of months ago, there was the first report of millimeter-long nanotubes.
This is rather rapid progress, around 9 orders of magnitude in about a decade. And the folks at Duke not only reached a 2-mm length, but they did it by continuously growing the tubes. And they can generate cross-connections between the tubes. Others are now experimenting with continuously-grown nanotubes. With funding to support the flock of researchers, it wouldn't be at all surprising to read about indefinitely-long nanotubes (or sheets of them with periodic cross-connections) within a year or two.
Then, instead of the first construction being done by sending up a huge spool of nanotubes and unwinding it in orbit, we'll read of them sending up a nanotube-manufacturing machine, which will extrude the tubes a few at a time and lower them to Earth.
A bigger problem, mentioned by Clark in Fountains of Paradise, is the cloud of space junk left over from thousands of earlier launches. The real expense will be the equipment to track every little particle passing through the Earth's neighborhood. To keep the elevator safe, we'll have to spot even tiny objects far enough in advance to send a wiggle down the rope just in time to move it aside when an object passes.
But, of course, this observing equipment will have huge scientific value itself, as it builds up a huge database of every little object in the solar system.
And the intermediate uses are developing. There are already sensors and drug-deliver devices being built that use nanotubes of various lengths. This is helping to get funding to the nanotube researchers.
Actually, if you go to dictionary.com, you'll see that they give 17 basic meanings of "free" in English. Several of those are groups of closely-related definitions, for a total of 27. Other English dictionaries will have a similar number of definitions.
The phrase "free beer" is meaning 7a, while "free speech" is 3c. I've long liked to respond to these phrases by mentioning "free disk space". I get some funny looks when I do that.
An old joke based on two different meanings (8b and 7a):
Person1: Are you free Friday evening? Person2: I'm not free, but for you I can be reasonably inexpensive.
Of course, there's also the venerable management principle that the more money (and people) you're in charge of, the more important you are. The capitalist ideologs like to claim that this is solely a problem in government. In fact, it's a generally-recognized management problem in all human organizations, and affects the corporate world as badly as any government.
It's interesting that so much linux/OSS/free-software news is coming from governments. It does seem easier to overcome this "the more I spend the better" attitude in governments than in private businesses. Probably because there's public oversight over at least some parts of some governments. Not nearly enough oversight, of course, but a lot more than what you find in the corporate world.
Not much irony there. Any DOD-validated commercial product will have to go through the same re-validation for every new version. If they don't, someone at the the DOD isn't doing their job (or is on the take).
They knew that OpenSSL had already been validated by several commercial vendors. So validating OpenSSL by itself should be a slam-dunk after they'd already done it N times. But suggesting that they just use OpenSSL for free rather than paying a commercial supplier for it is an "out of the box move" that "took guts"? As Dick Cheney might ask, WTF?
Furthermore, it would be a big surprise if other parts of the military didn't have copies of OpenSSL lying about on a few thousand machines already, so they wouldn't even have to go through the motion of downloading and verifying the public version. I'd bet that it's already mirrored on any number of.mil sites.
How can this idiocy be explained, other than by the theory that they shouldn't get something for free if they can spend money for the same thing and support a campaign contributor?
It does sorta go along with the old stories of the Navy using Windows NT to control their hardware...
In my experience, most Windows users can handle the job of turning off scripting. The GUI is sufficiently complex that someone without at least that many smarts isn't going to get much done anyway. Most Windows users really aren't stupid; they're at least smart enough to use Windows. And they've already done lots of configging, complaining the whole time about how nonsensical it is.
The only real problem is that the checkbox is hidden in different places in defferent releases of Windows. So you warn them of this, and give them a list of places they should look. They can do it; they just need to be told why they should do it.
10 years ago I was thinking how cool it was that I could have a windowing system of any kind on a cheap generic computer.
20 years ago, I was running unix on a PC/XT box, and installing the early X Window stuff.
Of course, at the time I was working down the hall from Jim Gettys, and he was constantly looking for victims^Wbeta testers. So I could have the fun of tying to figure out what this fancy new windowing stuff was all about. But MS Windows was still a few years off.
Meanwhile, on the other side of my desk was the cute Cadmus workstation, which was Mac hardware running Sys/V unix complete with Mac graphics and a terminal emulator so you could run shells. Nice system at the time. You could develop Mac software in a programmer-friendly environment, then copy the binaries over to a Mac and run them. Or just run them on the Cadmus and have a programmer-friendly GUI.
Even AT&T had a sort of windowing system, though it was based on ASCII graphics. It actually worked pretty well, considering.
As near as I can tell, Microsoft was the last company to move to a windowing system. Was there any company even later?
One of the general rules seems to be that any time a computer is added to anything, all precedent is forgotten, and everything we've learned from centuries (or millenia) of experience is forgotten and must be relearned.
The tipoff that this is happening is that someone utters the incantation "But that's different".
I have a wish-based package (network monitoring) that dates back to 1995. It has a "task bar" that at first glance seems similar to what people are talking about. It has a flock of buttons that trigger various actions. Some of the buttons are grouped within frames. Some of the buttons (and frames) are dynamic, appearing only when their actions are possible. Some disappear when not usable; others shrink to an "empty" button or frame as a placeholder. There are also a few entry widgets in the task bar with descriptive labels to their left; in such cases, an label and entry widget is within a frame to set them off visually. Of course, the border width of the frames is configurable, so setting it to 0 makes all the frame borders disappear.
So am I really in danger of being sued for patent infringement if Microsoft notices any of this stuff that I did about a decade ago?
Needless to say, I consider this all superficial cosmetics. It's rather trivial design work, to make the UI a bit more obvious to someone who doesn't use it every day. That's what the things described in other messages here also sounds like to me.
It's a sad day when superficial cosmetics can land you in a patent infringement lawsuit.
Oh, c'mon; I haven't written a call of gets() in years. That doesn't mean my code doesn't have security holes. If gets() were the only security issue, even Microsoft would have solved the problem a decade back.
Attributing all security problems to things like buffer overflows shows an incredibly shallow understanding of the issue. Most of the potential security holes can't be caught by a compiler (or linker) at all.
For example, most of the unix part of the industry has abandoned telnet and ftp in favor of ssh and scp. The problems with telnet and ftp have nothing to do with simple coding errors like buffer overflows. They are caused by sending information (such as passwords) across the Net in the clear. No compiler is going to correctly diagnose this sort of problem. And solving them isn't trivial; it requires an in-depth knowledge of the nuts and bolts of encryption.
We really need good information that goes a lot deeper than the current discussion. Unfortunately, most of the security advice for public consumption really is this shallow.
You've gotta include gets(), because without it, you can't claim POSIX compliance. Without that, you'll lose most of your contracts. Unless we can get gets() removed from the official standard, we have to provide it, no matter how bad an idea it turned out to be.
Of course, you can always use the GNU approach, of having tools (the linker in this case) give a loud warning when you use such dangerous tools. But you can't just delete them, not if you're selling in a market that requires compliance with standards.
Heh. Yeah, self-deprecating humor is always popular in nerd/geek circles.
But I've seen slashdot referenced in a lot of other news sites. This place it getting noticed by lots of non-nerds. In particular, media peoplea re learning that this is a good site to visit for leads to breaking stories. Along with news.google.com, of course.
Slashdot Mirror
Actually, for your typical "AOL Mom" (or "Joe Six-Pack"), Windows is even more of a nightmare than linux. If you want them to have a machine that's friendly to them, you should recommend that they buy a Mac.
Now, the usual reaction to this is "Macs are more expensive." Well, yes and no. There's been plenty written here and elsewhere explaining that, for comparable machines with comparable software, there is little if any price difference. The apparent difference comes from the fact that you can buy cheap, shoddy Windows boxes, but not cheap, shoddy Macs.
Recommending Windows to a non-expert is sorta like recommending a Yugo or Lada auto to a non-mechanic. Yeah, you could buy a car without locks on the doors, without any heat or air, and without a speedometer, spare tire, gas gauge or oil warning light. It would probably be cheaper.
Here in the US, it's illegal to sell such cars for street use. You can make one as a specialized vehicle, e.g. for racing. But you can't get a license to drive it on the street.
Maybe eventually the growing problems with Microsoft-supported viruses, worms, etc (and the somewhat unrelated problem of spam) will lead to similar laws preventing using unlicensed computers on the public Internet. It probably won't happen Real Soon Now, though.
Anyway, suggesting that novices use Windows is rather irresponsible. Novices should use machines that don't blame the users for the builtin problems. Such machines should come with normal security enabled by default. They shouldn't default to running code from strangers. They should warn users when they do something that will run new software from outside sources (especially email and web pages).
Of course, this means that novices shouldn't be buying the cheapest, shoddiest computer that's for sale at the neighborhood price-is-everything marketplace. And let's face it, that is pretty much a good description of mass-market Windows boxes.
... that the jobs I've had for the past 5 years or so have all been primarily developing software that runs on linux systems?
;-) You can't get software to work without a good understanding that computers don't respond to positive thinking or marketing, and such people will always be a tiny minority.
Funny thing is that these jobs have been paid for mostly by non-US companies who are trying to get out from under the thumb of either IBM or Microsoft (or both). And they're hiring Americans like me to help them do it.
A big selling point has been that N years from now I can guarantee that the software will still run and they'll still be able to read all their files. They've learned the hard way that this isn't always true with proprietary systems.
And I can easily explain to them how they can verify that there are no hidden tricks (trojans, backdoors, etc) in my code or in any of the lower-level software. Neither my code nor anything in "the system" can be sending their data off to some stranger's data warehouse. Granted, they'll have to keep around a staff of unix/linux geeks, who will both study the code and monitor the appropriate online fora. But they don't need to hire as many such geeks as they have on site now to keep their IBM/MS stuff running, so even that's a win.
Maybe eventually we'll see the day when all software has been written and no more is needed. But I suspect that day's still a long way off. And the world is growing more and more dependent on smaller and smaller computers to keep everything running.
So for the forseeable future, they'll still need lots of people who understand that, no matter what managers or marketing people say, 2+2 is always 4, not 5 or 3.95 or something desirable. (Except when it's 3.99999999998 of course, but any true geek will understand that, too.
So I'll predict that people with the twisted (i.e., logical) minds required by programming will continue to have jobs until long after all of us are gone.
Of course, we may all have to move to India or China, as the patent system shuts down software development in the Western world.
A better question would be: What is preventing millions upon millions of people from switching to a free OS? Quality would be my guess.
Well, I've asked a lot of MS users about this, and so far every one of them has disagreed with you. Their answer is always of the form "The software I want/need is only available on DOS/Windows (depending on the year that I asked)."
So then I ask my followup question: "How many other kinds of computer systems did you look at?" And their answer is always the same: "None."
Invariably, MS users just know that software is only available on MS systems. They don't need to do any market research, because they already know the answer. There's no point in wasting time looking for something that doesn't exist.
There are, of course, lots of people who have done the obvious searches. They aren't MS users. They easily find alternatives, determine that the alternative is almost always of higher quality, and go with it.
But BillG and company (and IBM before them) have become rich betting that the great majority will never do even the slightest study of what's available. All it takes is a good-size marketing budget, and whatever you make will be the market leader, whether it's good or bad quality, because few people will ever look for alternatives.
Bill himself was in the enviable position to be able to use daddy's money to get into Harvard B-school, where he made the connections that allowed him to leverage an IBM marketing budget and do an end run against all those pipsqueaks who had demoed the viability of a "desktop" computer market. His marketing budget has remained greater than the total operating budget of all his competitors combined. So the great majority who just go with the "market leader" continue to buy from him, because they know there's no point at looking at anyone else's nonexistent software.
Anyway, try it yourself. Ask random MS users to name a single piece of software that runs on linux. I'll predict that, with very few exceptions, most of them will be unable to come up with anything at all. They have never looked, and they never will.
This shoots down the idea that they're buying based on any sort of "quality" determination. They're buying from the only software supplier that exists in their world.
(In the mainframe world, the same situation still exists, with "IBM" for "Microsoft" throughout.)
Well, I have a policy that, when writing about mispellings, I always misspel the word "mispel".
;-)
(And it's fun when people attempt to correct me.
NOTHING should ever be backed up locally, because like you said, when fire hits, your backups are gone.
You hit that nail on the head. And even if you have all those tapes sitting in your basement, as I do, you'll find that you can't read them on the next machine you get. Even if you do order one of the tape drives that's available, they'll be physically incompatible with your old tapes. You could try moving the old tape drive from the dead machine to the new one, but you'll find that drivers are no longer available that work with the current OS.
What I do is follow Linus Torvalds' advice: Make your stuff sufficiently interesting and useful that others are willing to back them up on their disks. I have several "guest" accounts on machines in widely-separates places, with permission to back up all my stuff there. It's all visible via the web, though of course there are a number of "hidden" directories whose names you are highly unlikely to guess until long after the actuarial folks say I'll be gone.
That works a whole lot better than gambling that N years from now you'll have a working driver for a functional tape drive that can read a tape that hasn't deteriorated after N years.
Now, if I could afford $15,000 for a top-quality tape drive in the new machine that I buy every 5 years or so, I might feel differently. But mostly, if you suggest backup to tape, I'd just ask what planet you're from. It's obviously not one that I've lived on. Backup tapes have very rarely, in my experience, turned out to be actually readable on the machine where you need to read them.
Yeah, I remember the fuss over Hatch's web site using unlicensed "pirated" software. They apparently did get a license after this was brought to light.
But the current legislation is even more extreme, in a sense. Hatch and his people wouldn't be able to avoid the issue by belatedly applying for a license. Even if they had a license to use things like email and browsers, the license wouldn't stop the copyright violations that these tools make easy. You can't give someone a license to violate other people's copyrights.
We should be pointing out to them that email and browsers are primary examples of tools that enable copyright violation. If they really believe in what they're proposing, they should remove such software from their own computers. If they don't, they are openly telling the world that they consider themselves exempt from the rules that they want to impose on the rest of us.
I'm assuming their version of patent reform would be "Everyone else has to give up their bad patents, but we get to keep them."
Maybe, maybe not. Bezos' comment does sorta imply that he understands the idiocy of the situation but is trying to "fight it within the system" by forcing it to an absurd extreme. There is precedent for this approach.
OTOH, it's entirely possible that his understanding is that absurdities like this will have the effect of eliminating all those pesky little competitors, while leaving his bigger company intact. This would happen if the legislators can't be persuaded to improve the mess they've made.
There is some history of business leaders pushing for legislation that would outlaw their own business practices. There were interesting articles about a case several decades back, when the state of Oregon was considering the legislation that would eliminate most of the pollution from the waters of the Willamette River system. A number of the biggest polluters were publicly pushing for the legislation. Their reasoning was straightforward: "We can't stop our own pollution as long as our competitors are allowed to continue polluting; we'd be out of business. Our only choice is to push for laws that will force everyone to stop polluting." It worked, and the Willamette River is now a lot cleaner than it was back then. It helped a lot that the entire drainage basin was controlled by a single state government.
The corporate world does have a few leaders who are socially responsible. There's a good chance that Bezos is one of them. We'll have to watch him and see. I wonder how much real evidence we have now?
[A]re they going to ban FTP since you can distribute copyrighted material over FTP as well?
Probably. And they'll also ban SMB file sharing, since that clearly enables copyright violation.
Of course, this will make Windows boxes a lot more secure, so maybe it's a Good Thing.
And who'd have thought that Orrin Hatch would be the one that would finally force Microsoft to remove a major security hole from their software?
Actually, when you consider that unix-like systems are multi-user systems with a file system that encourages shared files and enables cooperative development, I'd wonder whether this bill would ban unix file systems.
Maybe what we should do is check out the computers in use by Hatch's political organization, and demand that they remove all software that enables copyright violation, starting with their email software. Maybe that would get the message across.
[M]anagement understands cost and benefit. They don't respond to features and feelings.
;-)
I would dispute this. I've seen any number of cases where management is dedicated to Microsoft and/or IBM. When I ask to see their numbers jjustifying this, they never have them. I've never seen anything approaching a financial study of the topic in any company. (Well, actually I did see one 15 years ago. That company converted over to Sun.
Most managers just "know" that MS and/or IBM is the best. Since they have no numbers to back this, I'd conclude that they are indeed responding to features and feelings. Much of this is just going with what they are familiar with, which is pure "feelings".
I've found that a good way to get them really confused is to suggest that maybe MS or IBM is the best, but they don't actually know that because they haven't done the usual comparisons. If you're buying a car, you at least make an attempt to check on what's available; you don't just go with the biggest advertiser. Managers will routinely ask for price quotes from suppliers on things like paper. They'll compare telephone companies' services and prices (if they have more than one available). And so on for nearly everything their company pays for.
But when it comes to computers, most managers throw all this out and go with their feelings that are based primarily on who has the biggest marketing budget.
Interesting. I wonder if anyone has found a fix for the /. problem that I and a few others have reported, in which firefox simply stops responding to clicks on links within /. pages?
It does seem to be a rather bizarre bug, with no obvious correlation with the pages being viewed. But it's serious enough on my Mac (but not my linux box) that I simply use a different browser.
Good thing there are a lot of browsers available.
I just checked out msdn.microsoft.com from Camino. The only problem was that it forced a tiny font, and a quick CMD-+ fixed that. I have a collection of browsers on my linux and Mac boxen, because part of my job is testing web pages for usability.
We might note that Microsoft-related sites have a history of checking the client's ID string and sending garbled pages to non-IE browsers. There was a recent story here about the Opera folks getting a few million bucks from msn.com over this.
When someone complains about a site not rendering well on anything but IE, perhaps you should mention this practice. The traditional term for it is "sabotage", and sensible people wouldn't stand for it.
Checking a client's ID isn't necessarily a bad idea, of course. I do that a lot. Lately I've been testing my pages against a number of PDAs' browsers. Sometimes the only thing that works is to check the ID and generate a page that's formatted for a tiny screen. But it's also common to use this to discriminate against people who use non-MS software.
Then again, maybe he doesn't have a clue what he's talking about.
;-) would object to that.
;-) So much for downloads hurting sales. Eminim himself has figured out that this is good publicity, and he remarked that he should pre-release his next album on the P2P sites.
Maybe, maybe not.
Various historians of the topic have pointed out that until very recently, copyright only applied to publishers. There was no question about "consumers" sharing things they liked with friends; that was obvious ly legal, and anyone with the slightest sense of the Public Welfare would support it. A lot of authors have gone public with the comment that they like people sharing their writings with others. That's how you generate a reputation, dummy! Only an idiot (or a manager
In our household, we have a number of DVDs that we've purchased after either having a "pirate" copy that we made from a broadcast or having borrowed them from friends. If we don't particularly like something, we watch it once then erase the tape and forget about it. If we like it, we look for the commercial DVD and buy it.
Similarly with music. I have a number of CDs that I've bought after having heard them on the radio or on a CD borrowed from a friend. If I can't find the CD commercially (highly likely because I have a lot of "folk" material from around the world that's not available in the US), I make a "pirate" copy. If I then find it for sale, I buy a copy.
Last year we had the fun news that Eminem's latest CD was the top download. Then, after it was released, it became the top-selling CD in the world. (My wife has a copy in her car; she loves shocking her professional colleagues with it, not to mention whistling the tunes at work for the benefit of the few who recognize it.
If you want people to buy something, you have to make it available to them so they can decide whether they want to buy it. If I can't hear it or see it, I'm not about to spend my money on it. And if you threaten me when I try to determine whether I like it, I'll walk away and not bother with you.
I think Micheal Moore understands this. He wants his stuff seen by the public. He has figured out that the only way he'll ever succeed at his chosen profession is if the public is familiar with him.
We haven't seen Fahrenheit 9/11 (yet). My wife has already suggested that when it comes out on DVD, she'll probably order a copy just on general principles. It doesn't hurt that we've seen "Roger and Me" and "Bowling for Columbine", and really liked both of them. We have a DVD of the first, but not the second. This may change.
Indeed, we are a long way from making 40,000-km-long carbon nanotubes. One serious question is how practical it is to depend on something like this being built.
But it's only been a few years since the first carbon nanotubes were created. The first were only nanometers long. Then others reported making some that were micrometers long. And a couple of months ago, there was the first report of millimeter-long nanotubes.
This is rather rapid progress, around 9 orders of magnitude in about a decade. And the folks at Duke not only reached a 2-mm length, but they did it by continuously growing the tubes. And they can generate cross-connections between the tubes. Others are now experimenting with continuously-grown nanotubes. With funding to support the flock of researchers, it wouldn't be at all surprising to read about indefinitely-long nanotubes (or sheets of them with periodic cross-connections) within a year or two.
Then, instead of the first construction being done by sending up a huge spool of nanotubes and unwinding it in orbit, we'll read of them sending up a nanotube-manufacturing machine, which will extrude the tubes a few at a time and lower them to Earth.
A bigger problem, mentioned by Clark in Fountains of Paradise, is the cloud of space junk left over from thousands of earlier launches. The real expense will be the equipment to track every little particle passing through the Earth's neighborhood. To keep the elevator safe, we'll have to spot even tiny objects far enough in advance to send a wiggle down the rope just in time to move it aside when an object passes.
But, of course, this observing equipment will have huge scientific value itself, as it builds up a huge database of every little object in the solar system.
And the intermediate uses are developing. There are already sensors and drug-deliver devices being built that use nanotubes of various lengths. This is helping to get funding to the nanotube researchers.
Actually, if you go to dictionary.com, you'll see that they give 17 basic meanings of "free" in English. Several of those are groups of closely-related definitions, for a total of 27. Other English dictionaries will have a similar number of definitions.
The phrase "free beer" is meaning 7a, while "free speech" is 3c. I've long liked to respond to these phrases by mentioning "free disk space". I get some funny looks when I do that.
An old joke based on two different meanings (8b and 7a):
Person1: Are you free Friday evening?
Person2: I'm not free, but for you I can be reasonably inexpensive.
Heh; I think you might have the "logic" right.
Of course, there's also the venerable management principle that the more money (and people) you're in charge of, the more important you are. The capitalist ideologs like to claim that this is solely a problem in government. In fact, it's a generally-recognized management problem in all human organizations, and affects the corporate world as badly as any government.
It's interesting that so much linux/OSS/free-software news is coming from governments. It does seem easier to overcome this "the more I spend the better" attitude in governments than in private businesses. Probably because there's public oversight over at least some parts of some governments. Not nearly enough oversight, of course, but a lot more than what you find in the corporate world.
Not much irony there. Any DOD-validated commercial product will have to go through the same re-validation for every new version. If they don't, someone at the the DOD isn't doing their job (or is on the take).
They knew that OpenSSL had already been validated by several commercial vendors. So validating OpenSSL by itself should be a slam-dunk after they'd already done it N times. But suggesting that they just use OpenSSL for free rather than paying a commercial supplier for it is an "out of the box move" that "took guts"? As Dick Cheney might ask, WTF?
.mil sites.
...
Furthermore, it would be a big surprise if other parts of the military didn't have copies of OpenSSL lying about on a few thousand machines already, so they wouldn't even have to go through the motion of downloading and verifying the public version. I'd bet that it's already mirrored on any number of
How can this idiocy be explained, other than by the theory that they shouldn't get something for free if they can spend money for the same thing and support a campaign contributor?
It does sorta go along with the old stories of the Navy using Windows NT to control their hardware
Right you are:
HTTP/1.1 302 Found
Date: Fri, 02 Jul 2004 23:08:06 GMT
Server: Apache/1.3.27 (Unix) (Red-Hat/Linux) PHP/4.1.2 mod_perl/1.29 mod_ssl/2.8.12 OpenSSL/0.9.6b
Location: http://www.johnkerry.com/
Connection: close
Content-Type: text/html; charset=iso-8859-1
They're even using mod_perl. Or at least, they have it installed.
In my experience, most Windows users can handle the job of turning off scripting. The GUI is sufficiently complex that someone without at least that many smarts isn't going to get much done anyway. Most Windows users really aren't stupid; they're at least smart enough to use Windows. And they've already done lots of configging, complaining the whole time about how nonsensical it is.
The only real problem is that the checkbox is hidden in different places in defferent releases of Windows. So you warn them of this, and give them a list of places they should look. They can do it; they just need to be told why they should do it.
10 years ago I was thinking how cool it was that I could have a windowing system of any kind on a cheap generic computer.
;-)
20 years ago, I was running unix on a PC/XT box, and installing the early X Window stuff.
Of course, at the time I was working down the hall from Jim Gettys, and he was constantly looking for victims^Wbeta testers. So I could have the fun of tying to figure out what this fancy new windowing stuff was all about. But MS Windows was still a few years off.
Meanwhile, on the other side of my desk was the cute Cadmus workstation, which was Mac hardware running Sys/V unix complete with Mac graphics and a terminal emulator so you could run shells. Nice system at the time. You could develop Mac software in a programmer-friendly environment, then copy the binaries over to a Mac and run them. Or just run them on the Cadmus and have a programmer-friendly GUI.
Even AT&T had a sort of windowing system, though it was based on ASCII graphics. It actually worked pretty well, considering.
As near as I can tell, Microsoft was the last company to move to a windowing system. Was there any company even later?
(Well, there are IBM mainframes, I suppose.
One of the general rules seems to be that any time a computer is added to anything, all precedent is forgotten, and everything we've learned from centuries (or millenia) of experience is forgotten and must be relearned.
The tipoff that this is happening is that someone utters the incantation "But that's different".
I have a wish-based package (network monitoring) that dates back to 1995. It has a "task bar" that at first glance seems similar to what people are talking about. It has a flock of buttons that trigger various actions. Some of the buttons are grouped within frames. Some of the buttons (and frames) are dynamic, appearing only when their actions are possible. Some disappear when not usable; others shrink to an "empty" button or frame as a placeholder. There are also a few entry widgets in the task bar with descriptive labels to their left; in such cases, an label and entry widget is within a frame to set them off visually. Of course, the border width of the frames is configurable, so setting it to 0 makes all the frame borders disappear.
So am I really in danger of being sued for patent infringement if Microsoft notices any of this stuff that I did about a decade ago?
Needless to say, I consider this all superficial cosmetics. It's rather trivial design work, to make the UI a bit more obvious to someone who doesn't use it every day. That's what the things described in other messages here also sounds like to me.
It's a sad day when superficial cosmetics can land you in a patent infringement lawsuit.
Oh, c'mon; I haven't written a call of gets() in years. That doesn't mean my code doesn't have security holes. If gets() were the only security issue, even Microsoft would have solved the problem a decade back.
Attributing all security problems to things like buffer overflows shows an incredibly shallow understanding of the issue. Most of the potential security holes can't be caught by a compiler (or linker) at all.
For example, most of the unix part of the industry has abandoned telnet and ftp in favor of ssh and scp. The problems with telnet and ftp have nothing to do with simple coding errors like buffer overflows. They are caused by sending information (such as passwords) across the Net in the clear. No compiler is going to correctly diagnose this sort of problem. And solving them isn't trivial; it requires an in-depth knowledge of the nuts and bolts of encryption.
We really need good information that goes a lot deeper than the current discussion. Unfortunately, most of the security advice for public consumption really is this shallow.
No wonder we have such problems.
You've gotta include gets(), because without it, you can't claim POSIX compliance. Without that, you'll lose most of your contracts. Unless we can get gets() removed from the official standard, we have to provide it, no matter how bad an idea it turned out to be.
Of course, you can always use the GNU approach, of having tools (the linker in this case) give a loud warning when you use such dangerous tools. But you can't just delete them, not if you're selling in a market that requires compliance with standards.
Heh. Yeah, self-deprecating humor is always popular in nerd/geek circles.
...
But I've seen slashdot referenced in a lot of other news sites. This place it getting noticed by lots of non-nerds. In particular, media peoplea re learning that this is a good site to visit for leads to breaking stories. Along with news.google.com, of course.
Now back to the humorous comments