Gosh, now I wonder what kind of certain source would generate PostScript that was so broken that a simple filter would be unable to do a 2-up transformation on it.
'Broken'? You are missing the point. PostScript is intended as a programming language for telling the printer where to put ink on the page; if it does that then it is not 'broken'. A tool which formats PS files as 2up might make certain assumptions about the format of the file, but if those assumptions turn out to be wrong then the tool is broken, not the PS file.
To make an analogy: suppose you ran the PostScript through a 'grep' program, which dumped core because it couldn't handle lines longer than 80 characters.
Write a stupid story, you get thousands of hits from Slashdot, boosting the figures you give to advertisers. If advertisers pay per click rather than per view this won't work though.
ZDNET ought to have custom advertising for stories such as this, advertising which is aimed at the Slashdot hordes.
Re:I don't understand the question.
on
Solving Chess?
·
· Score: 2
After all, if you're looking for a strategy to win every time if you are (white|black), this is quite different than a sequence of moves. The aforementioned grandmasters could probably provide a lot of insight into this.
What I mean by a 'strategy' is a way to work out which move to play for any possible game position. Picking a move at random is one possible strategy, not a very good one. Another is to number all your pieces and always move the lowest-numbered piece as far as possible towards the bottom left corner of the board, or if that piece can't move or is captured, try the next piece and so on. Another rather poor strategy, but it will tell you exactly what move to play in any circumstance (deterministic).
Most computer chess programs will use a strategy that looks ahead to a certain depth and picks the move that ends up with the most favourable-looking board position after a few moves, allowing for the fact that the other player is trying to spoil things. A perfect strategy would look ahead without limit and work out the best move leading to checkmate, or at least avoiding being checkmated.
Re:I don't understand the question.
on
Solving Chess?
·
· Score: 2
Chess would probably be subject to the same problem, since both players would know all possible outcomes and the paths leading to them, they would try to steer it such that their chances of winning are the highest. Problem is that both want to win and that they will therefore in the end settle for a draw, because that is the highest either one will be able to get.
You don't know that. That might happen, but it's also possible that white can always force a win, or maybe black can always force a win. Without exhaustively checking all possible games it's hard to tell which.
Ofcourse this carries the assumption that the game cannot be determined from the outset by the first move that has been done.
It can be determined in this manner, since you would know exactly what strategy the players are using - the best possible.
It would be cool if you could submit snippets of Perl code to process comments for you; then you could get things like 'truncate all -1 posts over 15 lines' without needing to pester Rob to implement them. The problem with this is security and the scope for DoS attacks. Perl has a module called Safe to run code in sandboxes, and it's easy to timeout things if they run for too long, but even so it's a scary idea.
Java has something similar called 'servlets' (untrusted code to run on the server), but I don't think it ever took off.
Re:I don't understand the question.
on
Solving Chess?
·
· Score: 3
What exactly does it mean to "solve" the game of chess? After all, there are widely published sequences of chess moves which lead to a given outcome.
There are published sequences of moves for both players. But there is not a published strategy for white which will always win, no matter what moves black plays. This would of course be an impossibly enormous task, at least using conventional computers. (Work out the number of possibilities, even approximately, after twenty moves.)
AFAIK 'solving' the game of chess means finding a perfect strategy for both players, a strategy which would have to work out all the possible cominations in advance and plays the best move. If both players did this and neither made a mistake, what would the outcome be?
There is already a perfect strategy for each player in noughts and crosses (tic-tac-toe); AFAIK it leads to a draw. (Some people say that the first player can always win, but this seems to be an urban myth.)
I almost always use red links (purple visited ones) because I think links should stand out.
Because you think that? What do your users think? If they wanted different-coloured links for some reason they could change the settings in their browser. And if you want links to stand out, fine, you can set them that way in your browser.
Of course, I'm just trolling here. But I think web designers often forget that it's not about what the designer prefers, it's what the users prefer - and as far as possible, you should let the users make their own choices. This is why CSS is good - your style sheet could specify red links, but the user could have an 'important' part of his style sheet to override it.
People make these GIFs and then make them transparent, assuming that my browser will have a white background. So for users with a black background, often the GIF looks really crappy, or in the case of diagrams, often totally invisible!
Arguably this is the fault of your browser. It ought to do something sensible when a transparent image doesn't stand out against the background. Don't forget that in most cases, transparent images are helpful when you want to switch background colours; you don't get white rectangles messing up your tasteful lime green page.
Most scripting languages like Perl or Python need to be compiled into bytecode, and then executed. The overhead of compiling the program every time it is executed can be quite a burden, especially for applications like CGI programs where the program is started very frequently but runs only for a short time.
Even if you can cache the compiled bytecode (as Python does), there is still the overhead of starting/usr/bin/perl for each hit. Again, this becomes significant once you start getting lots of hits.
The best thing you can do is use something like Apache's mod_perl which avoids the overhead of starting a new perl process for each page, and also the overhead of compiling the program each time. Similar features exist for all popular scripting languages and Web servers.
Once you've done that, you may find that there are other startup costs you can factor out. For example, does each hit make a new connection to the database? Depending on your application you may be able to get away with some sort of local cache, either on disk or cached in memory in some way.
Finally, you can make your site faster to _use_ and save the user's time if you make your pages appear gradually. If you're churning out a long, complex page, try to make it appear as it is computed, so the user can start reading immediately. This is partly a matter of choosing layout that web browsers can render incrementally, and partly of how you write your program. (Slashdot's comment display, at least in Minimalist Mode, is very bad at this. In Netscape 4.x you have to wait for the whole 100Kbyte or so before seeing any of it.)
I don't know if the situation has changed since I last checked, but IIRC the licence for YaST forbids you to sell 'data carriers' containing the software for a fee. So what I said is correct - you cannot resell copies of SuSE.
Caldera recently freed their installer - or I think they did - so it might now be possible to sell copies of OpenLinux as you can RedHat, Mandrake and so on. You'd have to strip out a lot of the third-party stuff first though.
The way things are now you can't lock users into your distribution because anyone can take it, sell it, and support it without paying you a cent.
That's the way things are with Red Hat, Mandrake, Slackware, Corel, and so on. But it's not the way things are with Caldera and SuSE; they already stop people from copying and reselling their distributions.
I think you left out MP3 from your list of successful formats...
'Stream control' is a red herring. If you can listen to it, you can record it. There are sound drivers for Windows and for Linux that let you save the output stream to disk, so you can make perfect digital recordings of anything coming from RealPlayer and other 'secure' players.
You can apply the anthropic principle, or something like that, to science fiction. You might think it is unbelievable that the Star Trek universe contains so many things which just happen to make good (or bad) television. But consider that there are an infinite number of possible sci-fi universes, but only a select few get made into books or TV series. Therefore the fact that you, the viewer, are watching the programme is the reason why semi-inverted Hoffenschauer tachyon field anomalies exist. If they didn't exist, the writers wouldn't be able to crank out a plot, so no episode.
Linux vs Minix is yet another example of 'worse is better'. Start with something that isn't pretty but does the job, and then improve it (modularize, make portable, etc) once you have built up some momentum.
If Minix had been free when it was released in 1985, people might today be using Minix instead of Linux. But that's another of those 'what if' questions which you can never prove either way.
Okay, so they've changed the copying conditions. When are they going to fix the spelling, from MINIX to Minix?
In general, you cannot fix security problems by trying to make people use a particular client. People trying to bring down the servers can easily learn how to pretend to be the new 'secure' client.
What's the point in Tarantella, now we have cool things like Workspot (on Slashdot a few days ago) which are a free solution for delivering Unix apps over the Web?
No, the fun in Dune 2 was building a long line of concrete blocks from your base to the enemy base, then placing rocket turrets in the middle of the enemy base. Watch the enemy vehicles destroy their own buildings trying to get at your turrets. If you build a block of four turrets and keep repairing them, they should be almost invulnerable.
Slashdot Mirror
'Broken'? You are missing the point. PostScript is intended as a programming language for telling the printer where to put ink on the page; if it does that then it is not 'broken'. A tool which formats PS files as 2up might make certain assumptions about the format of the file, but if those assumptions turn out to be wrong then the tool is broken, not the PS file.
To make an analogy: suppose you ran the PostScript through a 'grep' program, which dumped core because it couldn't handle lines longer than 80 characters.
Write a stupid story, you get thousands of hits from Slashdot, boosting the figures you give to advertisers. If advertisers pay per click rather than per view this won't work though.
ZDNET ought to have custom advertising for stories such as this, advertising which is aimed at the Slashdot hordes.
What I mean by a 'strategy' is a way to work out which move to play for any possible game position. Picking a move at random is one possible strategy, not a very good one. Another is to number all your pieces and always move the lowest-numbered piece as far as possible towards the bottom left corner of the board, or if that piece can't move or is captured, try the next piece and so on. Another rather poor strategy, but it will tell you exactly what move to play in any circumstance (deterministic).
Most computer chess programs will use a strategy that looks ahead to a certain depth and picks the move that ends up with the most favourable-looking board position after a few moves, allowing for the fact that the other player is trying to spoil things. A perfect strategy would look ahead without limit and work out the best move leading to checkmate, or at least avoiding being checkmated.
You don't know that. That might happen, but it's also possible that white can always force a win, or maybe black can always force a win. Without exhaustively checking all possible games it's hard to tell which.
It can be determined in this manner, since you would know exactly what strategy the players are using - the best possible.
It would be cool if you could submit snippets of Perl code to process comments for you; then you could get things like 'truncate all -1 posts over 15 lines' without needing to pester Rob to implement them. The problem with this is security and the scope for DoS attacks. Perl has a module called Safe to run code in sandboxes, and it's easy to timeout things if they run for too long, but even so it's a scary idea.
Java has something similar called 'servlets' (untrusted code to run on the server), but I don't think it ever took off.
There are published sequences of moves for both players. But there is not a published strategy for white which will always win, no matter what moves black plays. This would of course be an impossibly enormous task, at least using conventional computers. (Work out the number of possibilities, even approximately, after twenty moves.)
AFAIK 'solving' the game of chess means finding a perfect strategy for both players, a strategy which would have to work out all the possible cominations in advance and plays the best move. If both players did this and neither made a mistake, what would the outcome be?
There is already a perfect strategy for each player in noughts and crosses (tic-tac-toe); AFAIK it leads to a draw. (Some people say that the first player can always win, but this seems to be an urban myth.)
Because you think that? What do your users think? If they wanted different-coloured links for some reason they could change the settings in their browser. And if you want links to stand out, fine, you can set them that way in your browser.
Of course, I'm just trolling here. But I think web designers often forget that it's not about what the designer prefers, it's what the users prefer - and as far as possible, you should let the users make their own choices. This is why CSS is good - your style sheet could specify red links, but the user could have an 'important' part of his style sheet to override it.
Arguably this is the fault of your browser. It ought to do something sensible when a transparent image doesn't stand out against the background. Don't forget that in most cases, transparent images are helpful when you want to switch background colours; you don't get white rectangles messing up your tasteful lime green page.
We already have a common Internet-based language, it's called Perl :-)
Yeah, Netscape 4.x is a bit stupid, but it doesn't help that Slashdot sticks things in a table for no reason at all.
Most scripting languages like Perl or Python need to be compiled into bytecode, and then executed. The overhead of compiling the program every time it is executed can be quite a burden, especially for applications like CGI programs where the program is started very frequently but runs only for a short time.
/usr/bin/perl for each hit. Again, this becomes significant once you start getting lots of hits.
Even if you can cache the compiled bytecode (as Python does), there is still the overhead of starting
The best thing you can do is use something like Apache's mod_perl which avoids the overhead of starting a new perl process for each page, and also the overhead of compiling the program each time. Similar features exist for all popular scripting languages and Web servers.
Once you've done that, you may find that there are other startup costs you can factor out. For example, does each hit make a new connection to the database? Depending on your application you may be able to get away with some sort of local cache, either on disk or cached in memory in some way.
Finally, you can make your site faster to _use_ and save the user's time if you make your pages appear gradually. If you're churning out a long, complex page, try to make it appear as it is computed, so the user can start reading immediately. This is partly a matter of choosing layout that web browsers can render incrementally, and partly of how you write your program. (Slashdot's comment display, at least in Minimalist Mode, is very bad at this. In Netscape 4.x you have to wait for the whole 100Kbyte or so before seeing any of it.)
I don't know if the situation has changed since I last checked, but IIRC the licence for YaST forbids you to sell 'data carriers' containing the software for a fee. So what I said is correct - you cannot resell copies of SuSE.
Caldera recently freed their installer - or I think they did - so it might now be possible to sell copies of OpenLinux as you can RedHat, Mandrake and so on. You'd have to strip out a lot of the third-party stuff first though.
That's the way things are with Red Hat, Mandrake, Slackware, Corel, and so on. But it's not the way things are with Caldera and SuSE; they already stop people from copying and reselling their distributions.
It's hard disk, not hard drive, dammit!
I think you left out MP3 from your list of successful formats...
'Stream control' is a red herring. If you can
listen to it, you can record it. There are sound
drivers for Windows and for Linux that let you
save the output stream to disk, so you can make
perfect digital recordings of anything coming
from RealPlayer and other 'secure' players.
You can apply the anthropic principle, or
something like that, to science fiction. You
might think it is unbelievable that the Star Trek
universe contains so many things which just
happen to make good (or bad) television. But
consider that there are an infinite number of
possible sci-fi universes, but only a select few
get made into books or TV series. Therefore the
fact that you, the viewer, are watching the
programme is the reason why semi-inverted
Hoffenschauer tachyon field anomalies exist. If
they didn't exist, the writers wouldn't be able
to crank out a plot, so no episode.
Does that make any sort of sense?
Linux vs Minix is yet another example of 'worse is better'. Start with something that isn't pretty but does the job, and then improve it (modularize, make portable, etc) once you have built up some momentum.
If Minix had been free when it was released in 1985, people might today be using Minix instead of Linux. But that's another of those 'what if' questions which you can never prove either way.
Okay, so they've changed the copying conditions. When are they going to fix the spelling, from MINIX to Minix?
In general, you cannot fix security problems by trying to make people use a particular client. People trying to bring down the servers can easily learn how to pretend to be the new 'secure' client.
I thought the DMCA had already been passed. Will the protesters be asking for it to be repealed, or what?
Still, anything that provides a counterweight to lobbying by the copyright industry is welcome.
So how is Tarantella any better, apart from the SCO brand name?
What happened to perlfs? I'm sure that would have been a fairly painless way to do things, if there is a Napster library for Perl.
But does it run Linux?
What's the point in Tarantella, now we have cool things like Workspot (on Slashdot a few days ago) which are a free solution for delivering Unix apps over the Web?
You could call it 'Win2.001k'. Then rearrange that into 'Wink 2.001', and tada! we have version numbers again.
No, the fun in Dune 2 was building a long line of concrete blocks from your base to the enemy base, then placing rocket turrets in the middle of the enemy base. Watch the enemy vehicles destroy their own buildings trying to get at your turrets. If you build a block of four turrets and keep repairing them, they should be almost invulnerable.