It isn't always true that old machines burn lots of power, not PCs anyway. 386 and 486 processors run much cooler than the Pentium or its successors, and it looks to me as if other parts like memory also use less power. Old hard disks are much thirstier than their modern equivalents but that doesn't seem to make much of a difference.
For example, I have a PS/2 Model 55SX which powers 16 megs of RAM (including eight SIMMs on an expansion card), a 1987-model 3c523 Ethernet card, XGA-2 accelerated video and a 120Mbyte ESDI disk all from a 90W power supply. It's because none of these older components need much power - the only heatsink is on the video card.
What I'd like to know is why the patent office can make this decision in the first place. The law is made by governments (or EU directives implemented by national governments) and then it is the patent office's job to implement those decisions. Not vice-versa.
BTW: citizens of EU countries, sign the petition if you haven't already.
But if there isn't already a category called 'linux' in the standard IE5.5 bookmarks, Microsoft will just ignore any hits in that category, knowing that they must have come from Slashdot readers.
If you want to mess up the 'tracking' you need to hit URLs that actually are in the bookmarks.
It's kind of like the first version of Windows 98: you install the software.. and then it asks you for the agreement and key. It's not like you have any choice by then, right?
You do have a choice - reboot in Safe Mode and edit the registry to add a serial number by hand. And I think set the RegDone key to 1.
CPUs are getting faster but the demands placed on CPUs are growing at least as fast. If you're taking part in a videoconference, you need to compress and stream out your own image at the same time as streaming in and displaying images from the other participants. And if you can manage that, try again when the screen resolution and refresh rate have reached television-like levels.
Of course in the long term everything will be fast enough that clever schedulers won't be so necessary. But you could say the same of optimizing compilers, transparently-compressed network links, or almost any performance improvement invented over the past thirty years. Most of them are still with us. Even if the need will eventually disappear, there will be at least a five-year window when people will need a better RT scheduler (is my rough guess).
The tainting mode _forces_ you to check things. And you don't have to 'look for' formatting strings. That's the wrong approach - you should have a criterion for allowing things through, not one for weeding things out. For example, have a regex listing all the characters you know for sure will not cause a problem. There might be a couple of harmless characters you've forgotten, causing the program to reject input unnecessarily, but that's better than trying to list all the 'bad' characters and forgetting a couple (as you surely will at some point).
When you say 'improper use of sprintf()', do you mean things like:
/* get 'str' from user */
sprintf(buffer, str);
That would mean the user could put %d or whatever into the format string and corrupt the stack. The proper way to do things would be like printf("%s", str). Is that what you mean?
Yes, it is annoying that.bash_history gets created 644. I don't know why bash doesn't create it with a sensible mode to start with. (Although even here you can argue that this is illusory security - anybody could run 'ps' every second and get a complete log of the commands you run.)
I don't think that's a reason to change default permissions for all files in a home directory; the answer is to fix bash. I don't see how.newsrc or.bashrc is particularly sensitive. If your account could be compromised by somebody who knew the contents of.bashrc, making it go-rwx is just security through obscurity. (If you do have secret stuff like database passwords, put them in a separate file and read them in.)
I was thinking more of NFS, where files will get sent unencrypted over the network anyway. I agree, on a system where users log in locally, which isn't likely to be cracked or physically broken into, chmod go-rwx might be effective.
I didn't advocate ditching permissions on home directories, just changing the default. If users want obnoxious permissions, that's their choice, but the default should be something sensible.
And the British show Scrapheap Challenge (which may or may not be based on an American programme) is co-presented by Robert Llewellyn (Kryten). It's similar to Robot Wars in some ways, having teams which have to build machines and then compete. (Scavenge for junk and use it to build a land yacht or whatever within ten hours.)
So what will the remaining Dwarf cast be presenting?
English used to have a politeness indicator - whether you address someone as 'thou' (familiar) or 'you' (polite or plural). However 'you' got used so much that it came to take over both meanings, and 'thou' became obsolete.
In object-oriented Perl you already can choose whether to put the verb at the beginning or the end (at least a little). For example, if you're using the IO::Handle module and you want to make standard output unbuffered:
Can't you tunnel your VPN traffic over ssh or something? Tell ssh to forward port 50 on the local machine to port 50 on some remote machine, and the remote machine then continues the VPNing.
In computing, anything that is installed and working is a 'legacy system'. Order a shiny new 512-processor compute farm, set it up - and as soon as it starts running its first job, it's a legacy.
I've never been that keen on God-games like Dune 2, Transport Tycoon, Warcraft and so on that operate in 'real time'. They often end up being a test of how frantically you can click the mouse, rather than of real skill or strategy. What's especially stupid is that 'real time' ticks by many times faster than you'd expect. If a building takes 'three months' to build, why is the player not allowed even three seconds of thinking time without 'days' or 'weeks' of game time being wasted?
It might be better to have a game where you can pause, think about things, and order various actions. The actions only start happening once you unpause the game. In this way speed of mouse clicking would not be a factor. OTOH such a game would probably be very dull and would not have the 'real-time strategy' element that makes eg Dune 2 so fun to play. I used to play Transport Tycoon by pausing and unpausing all the time, but then I decided life was too short to play in such an anal fashion.
It would be interesting to _encourage_ the development of scripts for Command&Conquer type games. It might make the games more fun to play. If I have to sit in front of a Windows box and repetitively click on things I get annoyed. On Unix I'd probably write a script to automate it and get on with something more interesting. There are tedious elements in strategy games too, why not let users automate those?
I'm puzzled about your comment on FTP. Why is anonymous FTP somehow not as bad as FTP with passwords?
If the only way to FTP files off the system is non-anonymously, people have to send passwords in plain text. Anonymous FTP doesn't have the risk of giving away passwords, the worst that can happen is that somebody exploits a buffer overrun (or similar) in the FTP server - and that server is chrooted, right?
I agree that FTP is generally worth disabling, but why pick on anonymous, when it's the non-anonymous kind that lets sniffers get hold of users' passwords? (You could have separate FTP and login passwords, but that's not the default setup.)
Slashdot is interested in getting people worked up because when people are worked up, they post a larger number of interesting comments. In theory.
As for Miguel's talk, I went to his talk at the Linux Expo in London a couple of weeks ago, and it was definitely titled 'Unix Sucks'. I imagine the talk at Ottawa was the same. (Unix sucks because there is no code reuse and not enough consistency between different programs, according to Miguel.)
As far as I can tell, we have *three* component architectures. GNOME's Bonobo based on CORBA, Mozilla's components based on COM, and KDE's KOM based on... I'm not sure.
There's going to be a lot of duplication of effort if components for any one of these three won't work on either of the others. Maybe the GNOME and KDE people don't get along well enough to make a unified component architecture, or maybe there just isn't the will to do it. But perhaps one of them could work together with the Mozilla stuff?
Then there is Wine. A modified version of Wine to let you use all those Windows components out there in native Unix apps would be useful, but it would raise the number of component systems to four.
Isn't the point of component models to _avoid_ rewriting the same thing?
Slashdot Mirror
It isn't always true that old machines burn lots of power, not PCs anyway. 386 and 486 processors run much cooler than the Pentium or its successors, and it looks to me as if other parts like memory also use less power. Old hard disks are much thirstier than their modern equivalents but that doesn't seem to make much of a difference.
For example, I have a PS/2 Model 55SX which powers 16 megs of RAM (including eight SIMMs on an expansion card), a 1987-model 3c523 Ethernet card, XGA-2 accelerated video and a 120Mbyte ESDI disk all from a 90W power supply. It's because none of these older components need much power - the only heatsink is on the video card.
What I'd like to know is why the patent office can make this decision in the first place. The law is made by governments (or EU directives implemented by national governments) and then it is the patent office's job to implement those decisions. Not vice-versa.
BTW: citizens of EU countries, sign the petition if you haven't already.
But if there isn't already a category called 'linux' in the standard IE5.5 bookmarks, Microsoft will just ignore any hits in that category, knowing that they must have come from Slashdot readers.
If you want to mess up the 'tracking' you need to hit URLs that actually are in the bookmarks.
Does Microsoft have a category called 'linux'? If not, it wouldn't affect their statistics much.
You do have a choice - reboot in Safe Mode and edit the registry to add a serial number by hand. And I think set the RegDone key to 1.
CPUs are getting faster but the demands placed on CPUs are growing at least as fast. If you're taking part in a videoconference, you need to compress and stream out your own image at the same time as streaming in and displaying images from the other participants. And if you can manage that, try again when the screen resolution and refresh rate have reached television-like levels.
Of course in the long term everything will be fast enough that clever schedulers won't be so necessary. But you could say the same of optimizing compilers, transparently-compressed network links, or almost any performance improvement invented over the past thirty years. Most of them are still with us. Even if the need will eventually disappear, there will be at least a five-year window when people will need a better RT scheduler (is my rough guess).
The tainting mode _forces_ you to check things. And you don't have to 'look for' formatting strings. That's the wrong approach - you should have a criterion for allowing things through, not one for weeding things out. For example, have a regex listing all the characters you know for sure will not cause a problem. There might be a couple of harmless characters you've forgotten, causing the program to reject input unnecessarily, but that's better than trying to list all the 'bad' characters and forgetting a couple (as you surely will at some point).
When you say 'improper use of sprintf()', do you mean things like:
/* get 'str' from user */
sprintf(buffer, str);
That would mean the user could put %d or whatever into the format string and corrupt the stack. The proper way to do things would be like printf("%s", str). Is that what you mean?
The $1M in damages is probably dwarfed by MS's legal costs.
Yes, it is annoying that .bash_history gets created 644. I don't know why bash doesn't create it with a sensible mode to start with. (Although even here you can argue that this is illusory security - anybody could run 'ps' every second and get a complete log of the commands you run.)
.newsrc or .bashrc is particularly sensitive. If your account could be compromised by somebody who knew the contents of .bashrc, making it go-rwx is just security through obscurity. (If you do have secret stuff like database passwords, put them in a separate file and read them in.)
I don't think that's a reason to change default permissions for all files in a home directory; the answer is to fix bash. I don't see how
I was thinking more of NFS, where files will get sent unencrypted over the network anyway. I agree, on a system where users log in locally, which isn't likely to be cracked or physically broken into, chmod go-rwx might be effective.
I didn't advocate ditching permissions on home directories, just changing the default. If users want obnoxious permissions, that's their choice, but the default should be something sensible.
Go to petition.eurolinux.org and register your opposition to this kind of crap spreading to Europe.
And the British show Scrapheap Challenge (which may or may not be based on an American programme) is co-presented by Robert Llewellyn (Kryten). It's similar to Robot Wars in some ways, having teams which have to build machines and then compete. (Scavenge for junk and use it to build a land yacht or whatever within ten hours.)
So what will the remaining Dwarf cast be presenting?
Yes, use spaces to separate groups of numbers, or use underscores like Perl (and Ada IIRC). For example,
good: 12345 or 12 345 or 12_345
bad: 12,345 or 12.345
English used to have a politeness indicator - whether you address someone as 'thou' (familiar) or 'you' (polite or plural). However 'you' got used so much that it came to take over both meanings, and 'thou' became obsolete.
:-)
It's a kind of politeness inflation
In object-oriented Perl you already can choose whether to put the verb at the beginning or the end (at least a little). For example, if you're using the IO::Handle module and you want to make standard output unbuffered:
STDOUT->autoflush();or
autoflush STDOUT;
What Web browser does it use? Navigator? Mozilla?
Can't you tunnel your VPN traffic over ssh or something? Tell ssh to forward port 50 on the local machine to port 50 on some remote machine, and the remote machine then continues the VPNing.
In computing, anything that is installed and working is a 'legacy system'. Order a shiny new 512-processor compute farm, set it up - and as soon as it starts running its first job, it's a legacy.
You still don't have any incentive to pay. What difference will your $1 make among thousands of readers?
I've never been that keen on God-games like Dune 2, Transport Tycoon, Warcraft and so on that operate in 'real time'. They often end up being a test of how frantically you can click the mouse, rather than of real skill or strategy. What's especially stupid is that 'real time' ticks by many times faster than you'd expect. If a building takes 'three months' to build, why is the player not allowed even three seconds of thinking time without 'days' or 'weeks' of game time being wasted?
It might be better to have a game where you can pause, think about things, and order various actions. The actions only start happening once you unpause the game. In this way speed of mouse clicking would not be a factor. OTOH such a game would probably be very dull and would not have the 'real-time strategy' element that makes eg Dune 2 so fun to play. I used to play Transport Tycoon by pausing and unpausing all the time, but then I decided life was too short to play in such an anal fashion.
It would be interesting to _encourage_ the development of scripts for Command&Conquer type games. It might make the games more fun to play. If I have to sit in front of a Windows box and repetitively click on things I get annoyed. On Unix I'd probably write a script to automate it and get on with something more interesting. There are tedious elements in strategy games too, why not let users automate those?
I'm puzzled about your comment on FTP. Why is anonymous FTP somehow not as bad as FTP with passwords?
If the only way to FTP files off the system is non-anonymously, people have to send passwords in plain text. Anonymous FTP doesn't have the risk of giving away passwords, the worst that can happen is that somebody exploits a buffer overrun (or similar) in the FTP server - and that server is chrooted, right?
I agree that FTP is generally worth disabling, but why pick on anonymous, when it's the non-anonymous kind that lets sniffers get hold of users' passwords? (You could have separate FTP and login passwords, but that's not the default setup.)
Yes, 2D totally sucks ass!
(Screw you guys, I'm going to eat.)
Slashdot is interested in getting people worked up because when people are worked up, they post a larger number of interesting comments. In theory.
As for Miguel's talk, I went to his talk at the Linux Expo in London a couple of weeks ago, and it was definitely titled 'Unix Sucks'. I imagine the talk at Ottawa was the same. (Unix sucks because there is no code reuse and not enough consistency between different programs, according to Miguel.)
As far as I can tell, we have *three* component architectures. GNOME's Bonobo based on CORBA, Mozilla's components based on COM, and KDE's KOM based on... I'm not sure.
There's going to be a lot of duplication of effort if components for any one of these three won't work on either of the others. Maybe the GNOME and KDE people don't get along well enough to make a unified component architecture, or maybe there just isn't the will to do it. But perhaps one of them could work together with the Mozilla stuff?
Then there is Wine. A modified version of Wine to let you use all those Windows components out there in native Unix apps would be useful, but it would raise the number of component systems to four.
Isn't the point of component models to _avoid_ rewriting the same thing?