They've got your user ID, so what? OK, they see what site you are watching, but their point is what you want to _buy_, right? Or more precisely: They want to know what you are likely to be clicking at. If you don't click _anything_, you are likely to keep them in the dark.
I used to do something I thought was really smart at that time: I just set up an IP chain to block adserver*.doubleclick.net (via some IP ranges). Then, I noticed User Friendly used DoubleClick. Suddenly, I was responsible for Illiad losing money. I turned the filter back on.
I think you have three real choices: 1. Don't accept the cookies. Irritating every time, unless you have some auto-option. Go Lynx for this!:-) 2. Accept the cookie, but don't click the banner. They won't be getting any useful info from you. 3. Accept the cookie, and _let them_ build a profile on you. Sure, you will get ads that are more interesting to you -- so what? If you don't want or need the product, don't click... Of course, spam is more irritating -- guard that e-mail address, or get a spam filter (mail me if you're interested in betatesting my own). I thought commercials (ads in this case) actually were supposed to be a Good Thing(TM) for the _user_ as well as the company? If you get ads for products you want, is there anything wrong with that?
Have to agree. At least the old page was a bit relaxing. It didn't really stir you up like this one does. Well, I guess I'll get used to it (and HotWired _still_ has worse colors), just like I got used to Freshmeat.
I wonder who they consult for such things? External designers? Was this choice as hard to do as what way PacMan would be facing (left)? And most important of all, is this some kind of marketing stuff? (Santa is red because of Coca-Cola, you know...)
Perhaps we can finally get rid of always buying new IBMs at school; not surprised that they lose to Dell. The PCs are OK, but they are _always_ having their own weirdnesses. You have to make holes inside just to get a CD-ROM cable working! Their BIOS is awful -- stick to Award, people. And the machines tend to have all kinds of weird compatibility problems. Their support seems OK, though -- some stupid guy set the `power switch' on a 486 to 110V (we use 220V), and the machine went OOMPH and up in smoke. They fixed it for free... A 486...
A sad thing for all retro gamers, is that you'll have to get a legal copy of a game if you want to play it, even 10 years later. Some very nice people (mostly Amiga folks) have freed their ultra-cool games (now that the Amiga is dead), but you can't force anybody to do it. So I'm afraid you can't just spread the levels around -- their copyright is still valid. However, I myself wouldn't slam you for that -- as long as there's no easy way to get the game, what can you do?
At least the main share of the Debian profits goes to SPI (Software in the Public Interest), which is a charity set up to maintain Debian etc. I'm not sure if they donate to the FSF or not, but I wouldn't be surprised if they do.
And yes, Debian is 100% free, at least the stuff in the `main' section. They even have special definitions of what `free' implies -- you can't have anything exclusive to Debian, for instance.
The problem with `mainstream hardware' (mice) in this case, was that they killed off the poor guy's hand. He had a bad hand, and had to use a big trackball-like thing instead.:-)
And no, _I_ don't like RPM too much. I've installed it twice on my Slackware system (to get some RPM-only stuff), and both times, it messed up my machine completely. The big problem, in fact, is that it _isn't_ centralized -- and there seems to be no real standard for names, versioning etc. At least not in my experience..deb files, on the other hand:-)
I think you've addressed most of my points, so I'll just concentrate on the first one (incompabilities).
Sure, installing a new library may break things, but so may installing a new DLL. If a program really needs an old library, there are ways to get around that (LD_LIBRARY_PATH, for instance), even if the libraries have to same version number. Second, in Linux, it's very uncommon to see programs installing their own libraries. This is quite common on NT. If you browse through the other comments on this story, you'll see that people report strange things like `If I install SQL Server, my Visual InterDev menus goes crazy'. In Linux, you won't have to care so much about individual apps. I hear that W2K is going to `solve' such problems, by forcing each program to have all of their (non-system) DLLs in their own folder. Result: You will need a bigger HD.
You are definitely right in that NT is better than 9x. But the last NT machine I saw still needed reboots 4-5 times a day. Not because of a crash, but because the mouse driver faulted all the time. In NT, the user had to restart. In Linux, I do `killall gpm', then `gpm'. Or perhaps even `killall gpm ; rmmod psaux ; insmod psaux ; gpm' (except that I don't have a PS/2 mouse, this was just an example;-) ). Modularity at work:-)
Although you have some valid points, you seem to have overlooked some things. First, managing to keep a system secure _and_ stable at the same time (ie. deciding what `fixes' to install) is a pain for even an experienced user. In Windows, you have dozens of things conflicting with each others (something you notice yourself), and make things terribly fragile. I have yet to see a single such conflict in Linux, after having used lots of different distributions on lots of different machines.
Second, using `mainstream hardware' isn't always feasible. Some PCs simply don't come with mainstream hardware, and you'll just have to live with it. Sure, you should have chosen another (and more expensive) brand, but that's too late. Linux, on the other hand, seems (in my experience) to handle most hardware equally well, but there are of course some exceptions. BTW, a crashing module does _not_ crash the kernel. I've had the FAT filesystem module crash under an experimental kernel -- the filesystems were just no longer accessible, I could simply reload the module and remount the drives, and everything was back up. And there was no bluescreen halting my machine -- all my services etc. went on as usual without requiring me to do a thing.
Third, what's the point of a bugfix if you shouldn't install it?
While it's important to have stable software, you should also keep in mind that the time spent recovering from a crashing program is much lower than that of a crashed machine. You also mention that there will be more `core' applications that can crash -- I don't really see how, as programs are generally much less tangled into each other in Linux. (Take GIMP, for instance -- most plugin communication is done via a simple pipe.)
I think the Great Internet Mersenne Prime Search (GIMPS) should be included in this list. We're (I'm a part of it) have a couple thousand users going, and we're one of the few projects who actually have results to show (the three largest primes known to mankind, in just a few years). It has some advantages over some of the others (like very low bandwidth requirements), but it has some disadvantages as well (no fancy GUI, just a working one), but this is not intended as a comparison between those five.
This shows only that projects like GIMPS (which IMHO beats distributed.net), distributed.net and SETI@Home has done what part their original intention is: Lead the way for other distributed efforts.
Without having looked at the article, though, it looks like this will be much less accessible for the masses. GIMPSers (check http://www.mersenne.org if you're interested) have months or even years to complete a single assignments -- this one sounds like it will need the data in time. In other words, only 24/7 (or almost-24/7)-online computers will be able to participate effectively. I wonder how it will cope with having a variable amount of information available as well. OK, off to read the article:-)
They claim Windows NT has had C2 security since 3.51. In that case, somebody should fix the C2 standard. Easy example: take the MTF (Microsoft Tape Format) specs -- I believe it's somewhere on Seagate's web site. Throughout the document, there are multiple provisions for "C2-level security". And their password protection? Oh, that's a byte to or from. That's right: FLIP ONE BYTE, and ALL security on the tape goes away... Great! Go NT security!
>Of course new users are still left to install all 21.
Why?
Just install off the 'Net, and if your distribution can do it (at least dpkg/dselect can do it), then the older packages will never be downloaded at all -- security right away.
If you install from CD, of course, you will have to do a network update. That's approx 10-15 keypresses, for setting up the servers and hitting Update.
I think that Microsoft should get real support for the base standards (their HTML-generation makes BAD HTML, and their CSS1 support is far from complete) before they go on to `supporting' more or less new standards (like XML, or XSL which is experimental ATM).
If you want Netscape with `proper' DHTML (which Microsoft invented), try Mozilla/Raptor, which supports DOM level 1, 99% CSS1 and well over 50% CSS2.
And your second argument sounds a bit funny to me: Java was too secure, so they made ActiveX instead! Go, go! (If `Java wasn't enough', they could easily have extended it with a few classes. But I guess they don't have HD space for any new classes anymore...)
Perhaps, but at least not new. (Try Debian with dselect.) What worse, you have to make LOTS of scans, and click a lot around. At least with dselect, you know what's happening...
Not wanting to start a distribution war here, I should perhaps notice that Debian has had automatic update for ages. Run dselect, choose `update' and everything (including downloading, installing and configuring) is done automatically for you, at the spot.
If you do a net install, you get the right packages right away, of course.
Since I know close to nothing about RH, I shouldn't say that RH _hasn't_ had a similiar system, but from the comments I've been reading in this article, it certainly sounds that way.
resources like the RealTime BlackHole List make it harder and harder for the spammer to even send or relay spam.
The problem with things such as the RTBL (and address/host filters in general) is that it doesn't catch spammers the first time. Some `plug-pulling' here: I've made a tool that actually looks at the message to determine if it's spam or not. Mail me (remove the obvious part, of course) to get a prerelease version, if you're interested. (It catches 98% or so of my spam, and close to nothing of `real' messages.)
If you have problems with spam, mail me, and I'll send you a (GPLed) preview of my spam-sorter. I'm in the process of making an `how to avoid spam' page, and I'll see if Slashdot accepts it anytime soon;-)
It looks like you're forgetting one big thing (assuming I'm still talking to the same AC here): If you can get a gun, the big and the bad can also get a gun. True, there will always be bad guys, but at least they will require some training to kill.
Another thing is that it is _much_ more difficult to hide a big sword than a gun. At least if you see a guy carrying a sword, you are _warned_, and can run away (running with swords isn't that easy). If people couldn't use guns, I don't think many of us would walk around with swords or clubs. Sorry, I just don't believe in it. Another thing is that it's MUCH easier to kill at long range with a gun. Try even throwing a real spear more than a 10-20 metres, much less actually hitting the target. Do you really think Columbine would have happened if Harris and Klebod ran into the building, armed with clubs? Do you really think hundreds of pupils couldn't have taken them down?
The gun makes _the average guy_ a killer. It looks like you see that as a positive thing -- I'm sorry, I don't. And I'm rather small and weak myself.
Again, it's partly egoism: You'd prefer not being pushed around. Price: Somebody else will probably die. Remember, there can exist _force_ even if there is no guns. Go complain to somebody even bigger.
And if a government is supported by 200 million people, does that mean guns will help that situation? Sorry, I don't really believe in that... The situation will be that the 200 million get guns and _kill_ the 50 million, if they think it's worth it...
In every neural network (at least in traditional networks -- don't think this one would be THAT different), there are two important things: which neurons are connected, and the _weights_ of these neurons. The training only tries to find the optimal weights. After you are satisfied with the results, it's easy just to transfer the weights to a copy of the network (with the same structure). No re-training neccessary at all.
Slashdot Mirror
Caffeine in general, even. Coke contains _both_ caffeine and sugar -- wait, I think I remember my password again! It was `banana'! :-)
Too bad there were so many bad points about this article, as others have pointed out...
/* Steinar */
They've got your user ID, so what? OK, they see what site you are watching, but their point is what you want to _buy_, right? Or more precisely: They want to know what you are likely to be clicking at. If you don't click _anything_, you are likely to keep them in the dark.
/*
/* Steinar
I used to do something I thought was really smart at that time: I just set up an IP chain to block adserver*.doubleclick.net (via some IP ranges). Then, I noticed User Friendly used DoubleClick. Suddenly, I was responsible for Illiad losing money. I turned the filter back on.
:-)
:-)
I think you have three real choices:
1. Don't accept the cookies. Irritating every time, unless you have some auto-option. Go Lynx for this!
2. Accept the cookie, but don't click the banner. They won't be getting any useful info from you.
3. Accept the cookie, and _let them_ build a profile on you. Sure, you will get ads that are more interesting to you -- so what? If you don't want or need the product, don't click... Of course, spam is more irritating -- guard that e-mail address, or get a spam filter (mail me if you're interested in betatesting my own). I thought commercials (ads in this case) actually were supposed to be a Good Thing(TM) for the _user_ as well as the company? If you get ads for products you want, is there anything wrong with that?
OK, I see this might be a bit controversial
/* Steinar */
Have to agree. At least the old page was a bit relaxing. It didn't really stir you up like this one does. Well, I guess I'll get used to it (and HotWired _still_ has worse colors), just like I got used to Freshmeat.
I wonder who they consult for such things? External designers? Was this choice as hard to do as what way PacMan would be facing (left)? And most important of all, is this some kind of marketing stuff? (Santa is red because of Coca-Cola, you know...)
/* Steinar */
Perhaps we can finally get rid of always buying new IBMs at school; not surprised that they lose to Dell. The PCs are OK, but they are _always_ having their own weirdnesses. You have to make holes inside just to get a CD-ROM cable working! Their BIOS is awful -- stick to Award, people. And the machines tend to have all kinds of weird compatibility problems.
Their support seems OK, though -- some stupid guy set the `power switch' on a 486 to 110V (we use 220V), and the machine went OOMPH and up in smoke. They fixed it for free... A 486...
/* Steinar */
A sad thing for all retro gamers, is that you'll have to get a legal copy of a game if you want to play it, even 10 years later. Some very nice people (mostly Amiga folks) have freed their ultra-cool games (now that the Amiga is dead), but you can't force anybody to do it. So I'm afraid you can't just spread the levels around -- their copyright is still valid. However, I myself wouldn't slam you for that -- as long as there's no easy way to get the game, what can you do?
/* Steinar */
Remember that most assumptions are candidates for being wrong :-)
:-(
To make a good cipher like this, you would usually make a misleading code. Not that I've got any better ideas...
BTW, why should _always_ such things be restricted to the U.S. and Canada?
/* Steinar */
At least the main share of the Debian profits goes to SPI (Software in the Public Interest), which is a charity set up to maintain Debian etc. I'm not sure if they donate to the FSF or not, but I wouldn't be surprised if they do.
And yes, Debian is 100% free, at least the stuff in the `main' section. They even have special definitions of what `free' implies -- you can't have anything exclusive to Debian, for instance.
/* Steinar */
The problem with `mainstream hardware' (mice) in this case, was that they killed off the poor guy's hand. He had a bad hand, and had to use a big trackball-like thing instead. :-)
.deb files, on the other hand :-)
And no, _I_ don't like RPM too much. I've installed it twice on my Slackware system (to get some RPM-only stuff), and both times, it messed up my machine completely. The big problem, in fact, is that it _isn't_ centralized -- and there seems to be no real standard for names, versioning etc. At least not in my experience.
/* Steinar */
I think you've addressed most of my points, so I'll just concentrate on the first one (incompabilities).
;-) ). Modularity at work :-)
Sure, installing a new library may break things, but so may installing a new DLL. If a program really needs an old library, there are ways to get around that (LD_LIBRARY_PATH, for instance), even if the libraries have to same version number. Second, in Linux, it's very uncommon to see programs installing their own libraries. This is quite common on NT. If you browse through the other comments on this story, you'll see that people report strange things like `If I install SQL Server, my Visual InterDev menus goes crazy'. In Linux, you won't have to care so much about individual apps. I hear that W2K is going to `solve' such problems, by forcing each program to have all of their (non-system) DLLs in their own folder. Result: You will need a bigger HD.
You are definitely right in that NT is better than 9x. But the last NT machine I saw still needed reboots 4-5 times a day. Not because of a crash, but because the mouse driver faulted all the time. In NT, the user had to restart. In Linux, I do `killall gpm', then `gpm'. Or perhaps even `killall gpm ; rmmod psaux ; insmod psaux ; gpm' (except that I don't have a PS/2 mouse, this was just an example
/* Steinar */
Although you have some valid points, you seem to have overlooked some things. First, managing to keep a system secure _and_ stable at the same time (ie. deciding what `fixes' to install) is a pain for even an experienced user. In Windows, you have dozens of things conflicting with each others (something you notice yourself), and make things terribly fragile. I have yet to see a single such conflict in Linux, after having used lots of different distributions on lots of different machines.
Second, using `mainstream hardware' isn't always feasible. Some PCs simply don't come with mainstream hardware, and you'll just have to live with it. Sure, you should have chosen another (and more expensive) brand, but that's too late. Linux, on the other hand, seems (in my experience) to handle most hardware equally well, but there are of course some exceptions. BTW, a crashing module does _not_ crash the kernel. I've had the FAT filesystem module crash under an experimental kernel -- the filesystems were just no longer accessible, I could simply reload the module and remount the drives, and everything was back up. And there was no bluescreen halting my machine -- all my services etc. went on as usual without requiring me to do a thing.
Third, what's the point of a bugfix if you shouldn't install it?
While it's important to have stable software, you should also keep in mind that the time spent recovering from a crashing program is much lower than that of a crashed machine. You also mention that there will be more `core' applications that can crash -- I don't really see how, as programs are generally much less tangled into each other in Linux. (Take GIMP, for instance -- most plugin communication is done via a simple pipe.)
/* Steinar */
I think the Great Internet Mersenne Prime Search (GIMPS) should be included in this list. We're (I'm a part of it) have a couple thousand users going, and we're one of the few projects who actually have results to show (the three largest primes known to mankind, in just a few years). It has some advantages over some of the others (like very low bandwidth requirements), but it has some disadvantages as well (no fancy GUI, just a working one), but this is not intended as a comparison between those five.
/* Steinar */
This shows only that projects like GIMPS (which IMHO beats distributed.net), distributed.net and SETI@Home has done what part their original intention is: Lead the way for other distributed efforts.
:-)
Without having looked at the article, though, it looks like this will be much less accessible for the masses. GIMPSers (check http://www.mersenne.org if you're interested) have months or even years to complete a single assignments -- this one sounds like it will need the data in time. In other words, only 24/7 (or almost-24/7)-online computers will be able to participate effectively. I wonder how it will cope with having a variable amount of information available as well. OK, off to read the article
/* Steinar */
They claim Windows NT has had C2 security since 3.51. In that case, somebody should fix the C2 standard. Easy example: take the MTF (Microsoft Tape Format) specs -- I believe it's somewhere on Seagate's web site. Throughout the document, there are multiple provisions for "C2-level security". And their password protection? Oh, that's a byte to or from. That's right: FLIP ONE BYTE, and ALL security on the tape goes away... Great! Go NT security!
/* Steinar */
Red tape? Are you referring to the `backup plan' system?
.deb, .tgz or whatever you use for package management.
Should be pretty easy to fix -- if you find that the patch breaks, just install the old RPM,
/* Steinar */
>On the other hand, Linux has so many services started by default that it is a nightmare.
That is a common SuSE/Red Hat problem. Most other distributions (at least not Debian and Slackware) have this problem.
/* Steinar */
>Of course new users are still left to install all 21.
Why?
Just install off the 'Net, and if your distribution can do it (at least dpkg/dselect can do it), then the older packages will never be downloaded at all -- security right away.
If you install from CD, of course, you will have to do a network update. That's approx 10-15 keypresses, for setting up the servers and hitting Update.
/* Steinar */
XML being a key feature??
I think that Microsoft should get real support for the base standards (their HTML-generation makes BAD HTML, and their CSS1 support is far from complete) before they go on to `supporting' more or less new standards (like XML, or XSL which is experimental ATM).
If you want Netscape with `proper' DHTML (which Microsoft invented), try Mozilla/Raptor, which supports DOM level 1, 99% CSS1 and well over 50% CSS2.
And your second argument sounds a bit funny to me: Java was too secure, so they made ActiveX instead! Go, go! (If `Java wasn't enough', they could easily have extended it with a few classes. But I guess they don't have HD space for any new classes anymore...)
/* Steinar */
Perhaps, but at least not new. (Try Debian with dselect.) What worse, you have to make LOTS of scans, and click a lot around. At least with dselect, you know what's happening...
/* Steinar */
Not wanting to start a distribution war here, I should perhaps notice that Debian has had automatic update for ages. Run dselect, choose `update' and everything (including downloading, installing and configuring) is done automatically for you, at the spot.
If you do a net install, you get the right packages right away, of course.
Since I know close to nothing about RH, I shouldn't say that RH _hasn't_ had a similiar system, but from the comments I've been reading in this article, it certainly sounds that way.
/* Steinar */
The problem with things such as the RTBL (and address/host filters in general) is that it doesn't catch spammers the first time. Some `plug-pulling' here: I've made a tool that actually looks at the message to determine if it's spam or not. Mail me (remove the obvious part, of course) to get a prerelease version, if you're interested. (It catches 98% or so of my spam, and close to nothing of `real' messages.)
/* Steinar */
If you have problems with spam, mail me, and I'll send you a (GPLed) preview of my spam-sorter. I'm in the process of making an `how to avoid spam' page, and I'll see if Slashdot accepts it anytime soon ;-)
/* Steinar */
It looks like you're forgetting one big thing (assuming I'm still talking to the same AC here): If you can get a gun, the big and the bad can also get a gun. True, there will always be bad guys, but at least they will require some training to kill.
Another thing is that it is _much_ more difficult to hide a big sword than a gun. At least if you see a guy carrying a sword, you are _warned_, and can run away (running with swords isn't that easy). If people couldn't use guns, I don't think many of us would walk around with swords or clubs. Sorry, I just don't believe in it. Another thing is that it's MUCH easier to kill at long range with a gun. Try even throwing a real spear more than a 10-20 metres, much less actually hitting the target. Do you really think Columbine would have happened if Harris and Klebod ran into the building, armed with clubs? Do you really think hundreds of pupils couldn't have taken them down?
The gun makes _the average guy_ a killer. It looks like you see that as a positive thing -- I'm sorry, I don't. And I'm rather small and weak myself.
/* Steinar */
Again, it's partly egoism: You'd prefer not being pushed around. Price: Somebody else will probably die. Remember, there can exist _force_ even if there is no guns. Go complain to somebody even bigger.
And if a government is supported by 200 million people, does that mean guns will help that situation? Sorry, I don't really believe in that... The situation will be that the 200 million get guns and _kill_ the 50 million, if they think it's worth it...
/* Steinar */
In every neural network (at least in traditional networks -- don't think this one would be THAT different), there are two important things: which neurons are connected, and the _weights_ of these neurons. The training only tries to find the optimal weights. After you are satisfied with the results, it's easy just to transfer the weights to a copy of the network (with the same structure). No re-training neccessary at all.
/* Steinar */