Actually, it supports JSSS - JavaScript Style Sheets, which was Netscape's stylesheet language proposal. Unfortunately for them (and, rather fortunately for us) it got rejected, partly because nobody else would implement it.
So when CSS was accepted, they had to quickly hack together a way to get it into the browser, but since a complete rewrite of the stylesheet engine was out of the question, they were forced instead to write a CSS -> JSSS translator. Much of the problems with NS4's CSS support is that it isn't a 1:1 mapping with JSSS.
On the contary, ISO's, CCD's and bin/cue's are very common, since rips tend to be missing things, be unpatchable and not exactly of the highest quality (er, yeah, like I'm gonna run your dodgy bat file which runs an.exe to do what a.reg would do, and have all the data crunched into some lossy format for the rip to meet some scene guidelines).
Disk images, on the other hand, are easy to make, aren't much bigger than a rip to anyone except dialup users, and are not missing components.
Copy protection schemes like this have been around for ages in the style of SafeDisk; some copiers can cope and some software tools can cope with the dodgy ECC stuff the disks use, but quite often they also come with cracked executables in the image filesystem, so even if your copy is bad, the crack will work just fine.
Probably the most effective protection I can remember is Operation Flashpoint's Fade; it kind of turned poorly cracked versions into demo versions which degraded over time. It gave those who liked it an extra reason to buy it, and gave those who could take it or leave it a chance to play the game. Whether they made enough back off it to pay for it's development is another matter entirely.
Quake 3-style protection is even better, of course; you need a valid key to play online, so no key, no game; you can hardly keygen it, or crack most servers. Not really applicable to single player games though.
It's a Gecko (well, a lizard at least) eating the Microsoft butterfly, no need to panic, moderators.
dcstimm: Blame your high userid (I didn't realise they'd hit half a million, heh) and moderators not bothering to check links. Although tbh I would have prefered it without the butterflies:)
Well, the patent doesn't cover this, since its limited to email where the forwarding server is a third party key => value (old email => new email) database, and where it's the client that decides to use it.
It goes something like this:
Take a current idea that's obvious to anyone.
Patent it with 18 extra claims that limit scope.
Profit!
This is of course slightly different to BT's hyperlink patent, which is more along the lines of:
Make some dodgy system and patent it in case it might be useful some day.
Get upset when you notice you could have generalised the patent and covered something that could make you money if you could only get rid of some claims.
Try to convince court that because of (2) the patent should cover the more general case anyway.
Profit!
Personally, I think the patent offices should have a moderation system. I'd vote this one down (-1, File an RFC).
Eventually, over billions or even trillions of years, every possible combination of molecular structures will be created from these random interactions.
Don't underestimate the total possible combinations.
A good example is your display; say you use a mere 64*64 pixels, and it's only 2 colour. You have a total of 2 ** (64 * 64) possible combinations, which is 4.383 * 10^1234. That's 4383[... another 1231 digits...].
Let's say you can do a trillion permutations a second, it'll take you 3.311 * 10^1214 years to exhaust them. You'd either have to be incredibly lucky or there'd have to be a huge number of potential "good" permutations for you to find one before the universe could no longer power your 64*64*2 display randomizer.
You have a whole extra dimension, more than one type of atom, and a lot more space than a mere 64*64 grid.
And Regex in PHP? its a nightmare, two different types, neither near as powerful as Perl
You've got PCRE and POSIX regexp, what else do you need? How is PCRE nowhere "near as powerful as Perl"?
and why are function names in general ten_freaking_miles_long()!?
Because there are over 1,000 of them all sharing the same namespace. What's especially irritating is they don't all follow the same naming convention.
If you're going to learn a new language anyway why not learn Perl 6?
Because it's not very interesting. Learning new languages helps you gain multiple perspectives on problems, sticking with the same language sticks you in a rut where you only have a few views of how to solve a problem, none of which are necessarily optimal.
Personally, I'm looking at SmallTalk after Ruby, or maybe Lisp.
Damn fine until you realize you can't block popups or have tabs. But then again -- maybe I am the only one who does not liked popups and thinks 1 window is cleaner than 15 windows.
I use Opera, but MyIE really impressed me as a GUI for IE's engine. Tabs, popup filters, etc; it's just a shame you can't use Gecko in it:)
F12 and R will turn them off, F12 A will turn them back on.
Down arrow indeed.. tsk:)
Turning off plugins also comes highly recommended for killing Flash banners (F12 p, toggle), and disabling gif anims (F12 g, toggle) makes the rest much less irritating.
Well, given that GPU's are highly targeted at consumers, I can really see non-realtime GPU-rendering for the lower end of the market (the basic consumers who make wallpapers and the like), while the higher end market sticks with the far more computationally expensive CPU-bound raytracing, probably mixing in GPU based stuff on less complex scenes (plenty of CG is just fancy texture mapping and a bit of warping/shading, after all).
Seriously, by the time most people have nv30+-level GPU's, they'll have an enormous amount of rendering power that's quite comparable in most cases with raytracing. If you can render a scene on the GPU in a few seconds and have it look almost identical to a raytraced image that takes an hour, which do you think the average user will choose?:)
But my problems weren't a difficulty learning the course material; quite the opposite, the problem was the fact that in the 2 years I was there there was maybe 3 months of actual work.
The problem was that the course was filled mainly with total newbies who truely didn't know anything, so the course was targeted at them. Never mind that I'd just got an A in my A-level Computing, I had to start from scratch, "learning to program in visual basic" (woo, event based programming with minimal looks at flow control and subroutines), learning C (three hours spent going over "Hello World", I shit you not, and that's without even breaking it down to assm, then two semesters of groups writing near-identical tsv readers and making 10 minute presentations (this was for 33% of the marks, 16% doing a presentation, and 16% marking a presentation out of A..D)), and so on.
This, believe it or not, is supposed to be one of the better Uni's for computing in the region. And I dare say for the majority of students it was just right (seeing as half of them still didn't know what a variable was after the third semester). In the end I left in disgust. Boring students, boring course, and £3500 debt p.a. did not make for a happy time.
Er, yeah, anyway, you can talk about how to help the lower end strugglers which presumably happen quite a lot in good Uni's, but what about all the crappy Uni's where the strugglers tend to be the bored to death top end ones?
Well, sort of. I'm mainly thinking in terms of using the GPU to render the scenes in hardware rather than just using a software renderer. Since this seems to be the direction cards are moving in (that is, hardware rendered scenes competing directly with traditional raytracers due to all the shader stuff and higher bit depths), bottlenecks like this will become more of an issue.
Although having said that, I doubt even hardware accelerated rasterisers will be pushing 10MB/s of video data out in most cases, so..:)
OK, scenario: I use my expensive GFX card to play Unreal Tournament. I don't just want to grab screenshots. I want to actually grab sequential frames in REAL-TIME from my game while I'm playing to create movies.
Actually, my scenario is more like:
I use my expensive GFX card to render shots for my incredibly innovative but poorly funded sci-fi flick. I want to grab each frame in perfect detail so it can be post-processed. The easiest and cheapest way to do this is to have the renderer save each frame as it's computed. Real-time is not an issue, just like it's not an issue with a raytracer or whatever.
Using the GFX card to play AND capture at the same time is just not feasible, not to mention unwise (read: stability issues).
It better become feasable if companies are going to want renderfarms based on the nv30/40/whatever. Having two seperate machines per renderer would be pretty.. dodgy:)
No, not really. You've got a lot more than just pushing pixels around to worry about for it to be that simple.
4x the pixels doesn't mean 4x the geometry, or 4x the AI, or 4x the bandwidth requirements (ok, the final image is 4x bigger, but the textures and geometry you're pushing aren't also 4x bigger).
Look at the benchmarks for the 9700 for Q3; 1024*768 = 203FPS. 1600*1200 = 180.6FPS. The nearly three times as many pixels of the higher resolution resulted in a loss of just 10% of the framerate.
Uh, except 3-400FPS in RTCW in a given resolution means you can double the resolution and get.. 1-200FPS. Handy when the next generation of monitors come out that let you use resolutions of 4000x4000 or whatever. It also means you can make more detailed maps, and have busier scenes; instead of being limited to 2-3 characters in view at any one time, you can do a Doom/Serious Sam with the same detail level. Or you can tone up the LOD and have higher resolution textures further away and reduce mipmap artifacts. Or you can seriously concider adding bumpmaps everywhere, or more detailed volumetric self shadowing and *still* not drop below monitor refresh for the most detailed scenes.
And even IF you've still got tonnes of power left, this is what multitasking is for; if my system can push RTCW along at 400FPS, I can leave some expensive background task running and still have perfectly smooth gameplay.
Um, no, you've missed the point. This isn't about using a GFX card as a device to capture external data, it's to grab the images it's rendering into system memory so you can use it to, say, render some CG in a movie. You don't want to use a seperate device to grab the images your expensive cinematic quality GFX card produces when you could just dump it directly into the device it's running on.
Don't forget, the engine's having to do a lot more than current FPS's, because the lighting is *all* dynamic, bump mapped, self shadowed, etc. Id's working on removing or reducing the differences between objects in the world and the world itself, and removing things like lightmaps.
Basically this means that Doom III will be able to do things like have doors with lit rooms behind them cast light onto unlit rooms and for the lighting to actually change as the door opens while everything in the room lights up from it and casts shadows on themselves and everything else. But stills with lighting similar to what you'd get with a normal lightmap will take much more power to render, and hence need to have a lower level of detail.
Basically it's trading freedom and clever lighting effects for polygons and performance. Eventually this will pay off, but don't expect the average screenshot to necessarily look better than anything else. Expect it to MOVE better in the right maps, though:)
Probably also worth pointing out that, with Opera at least, the client still downloads the banners, and you can use the author/user mode toggle to turn it on/off should you feel the need to click on a couple of banners to support a site.
In this way, I probably click on more banners than I ever did when I just ignored them, AND the sites I visit still get their impressions.
I have a user CSS file which does this. It also works in Opera (and any other browser with reasonable CSS 2 selector support; i.e. not IE;), and is quite effective.
There's a bit of JS you can link to which makes it work.
Alternatively you can use absolute positioning and hide the fixed from IE using one of the many tricks/selectors it doesn't support (assuming this is compatible with your use of fixed).
Or you can do what the W3C do and ignore it. Nice how IE6 SP1 *still* doesn't fix this.
Slashdot Mirror
It does, uh, slightly.
Actually, it supports JSSS - JavaScript Style Sheets, which was Netscape's stylesheet language proposal. Unfortunately for them (and, rather fortunately for us) it got rejected, partly because nobody else would implement it.
So when CSS was accepted, they had to quickly hack together a way to get it into the browser, but since a complete rewrite of the stylesheet engine was out of the question, they were forced instead to write a CSS -> JSSS translator. Much of the problems with NS4's CSS support is that it isn't a 1:1 mapping with JSSS.
Death to NS4!
I find P2P largely useless for anything aside the odd mp3. Usenet's where it's at. It's not eating 400GB/day for the good of it's health ;)
On the contary, ISO's, CCD's and bin/cue's are very common, since rips tend to be missing things, be unpatchable and not exactly of the highest quality (er, yeah, like I'm gonna run your dodgy bat file which runs an .exe to do what a .reg would do, and have all the data crunched into some lossy format for the rip to meet some scene guidelines).
Disk images, on the other hand, are easy to make, aren't much bigger than a rip to anyone except dialup users, and are not missing components.
Copy protection schemes like this have been around for ages in the style of SafeDisk; some copiers can cope and some software tools can cope with the dodgy ECC stuff the disks use, but quite often they also come with cracked executables in the image filesystem, so even if your copy is bad, the crack will work just fine.
Probably the most effective protection I can remember is Operation Flashpoint's Fade; it kind of turned poorly cracked versions into demo versions which degraded over time. It gave those who liked it an extra reason to buy it, and gave those who could take it or leave it a chance to play the game. Whether they made enough back off it to pay for it's development is another matter entirely.
Quake 3-style protection is even better, of course; you need a valid key to play online, so no key, no game; you can hardly keygen it, or crack most servers. Not really applicable to single player games though.
There's at least one keygen, so unless they also match against all known good keys it'll be little more than an irritation.
Heh, I saw this on TV a couple of months ago. Way to go /. :)
It's a Gecko (well, a lizard at least) eating the Microsoft butterfly, no need to panic, moderators.
:)
dcstimm: Blame your high userid (I didn't realise they'd hit half a million, heh) and moderators not bothering to check links. Although tbh I would have prefered it without the butterflies
Or maybe you just look up the key and send to an alternative email registered on the keyring :)
It goes something like this:
This is of course slightly different to BT's hyperlink patent, which is more along the lines of:
Personally, I think the patent offices should have a moderation system. I'd vote this one down (-1, File an RFC).
Don't underestimate the total possible combinations.
A good example is your display; say you use a mere 64*64 pixels, and it's only 2 colour. You have a total of 2 ** (64 * 64) possible combinations, which is 4.383 * 10^1234. That's 4383[... another 1231 digits
Let's say you can do a trillion permutations a second, it'll take you 3.311 * 10^1214 years to exhaust them. You'd either have to be incredibly lucky or there'd have to be a huge number of potential "good" permutations for you to find one before the universe could no longer power your 64*64*2 display randomizer.
You have a whole extra dimension, more than one type of atom, and a lot more space than a mere 64*64 grid.
Still think a trillion years is enough?
You've got PCRE and POSIX regexp, what else do you need? How is PCRE nowhere "near as powerful as Perl"?
Because there are over 1,000 of them all sharing the same namespace. What's especially irritating is they don't all follow the same naming convention.
Because it's not very interesting. Learning new languages helps you gain multiple perspectives on problems, sticking with the same language sticks you in a rut where you only have a few views of how to solve a problem, none of which are necessarily optimal.
Personally, I'm looking at SmallTalk after Ruby, or maybe Lisp.
No, I just made a typo, sorry.
6.04 here.. wasn't that a security update?
Ask them
I use Opera, but MyIE really impressed me as a GUI for IE's engine. Tabs, popup filters, etc; it's just a shame you can't use Gecko in it
F12 and R will turn them off, F12 A will turn them back on.
:)
Down arrow indeed.. tsk
Turning off plugins also comes highly recommended for killing Flash banners (F12 p, toggle), and disabling gif anims (F12 g, toggle) makes the rest much less irritating.
http://voi.aagh.net/code/anti-banner.css kills most of the rest.
Well, given that GPU's are highly targeted at consumers, I can really see non-realtime GPU-rendering for the lower end of the market (the basic consumers who make wallpapers and the like), while the higher end market sticks with the far more computationally expensive CPU-bound raytracing, probably mixing in GPU based stuff on less complex scenes (plenty of CG is just fancy texture mapping and a bit of warping/shading, after all).
:)
Seriously, by the time most people have nv30+-level GPU's, they'll have an enormous amount of rendering power that's quite comparable in most cases with raytracing. If you can render a scene on the GPU in a few seconds and have it look almost identical to a raytraced image that takes an hour, which do you think the average user will choose?
Worth keeping an eye on, anyway.
But my problems weren't a difficulty learning the course material; quite the opposite, the problem was the fact that in the 2 years I was there there was maybe 3 months of actual work.
The problem was that the course was filled mainly with total newbies who truely didn't know anything, so the course was targeted at them. Never mind that I'd just got an A in my A-level Computing, I had to start from scratch, "learning to program in visual basic" (woo, event based programming with minimal looks at flow control and subroutines), learning C (three hours spent going over "Hello World", I shit you not, and that's without even breaking it down to assm, then two semesters of groups writing near-identical tsv readers and making 10 minute presentations (this was for 33% of the marks, 16% doing a presentation, and 16% marking a presentation out of A..D)), and so on.
This, believe it or not, is supposed to be one of the better Uni's for computing in the region. And I dare say for the majority of students it was just right (seeing as half of them still didn't know what a variable was after the third semester). In the end I left in disgust. Boring students, boring course, and £3500 debt p.a. did not make for a happy time.
Er, yeah, anyway, you can talk about how to help the lower end strugglers which presumably happen quite a lot in good Uni's, but what about all the crappy Uni's where the strugglers tend to be the bored to death top end ones?
Well, sort of. I'm mainly thinking in terms of using the GPU to render the scenes in hardware rather than just using a software renderer. Since this seems to be the direction cards are moving in (that is, hardware rendered scenes competing directly with traditional raytracers due to all the shader stuff and higher bit depths), bottlenecks like this will become more of an issue.
:)
Although having said that, I doubt even hardware accelerated rasterisers will be pushing 10MB/s of video data out in most cases, so..
Actually, my scenario is more like:
I use my expensive GFX card to render shots for my incredibly innovative but poorly funded sci-fi flick. I want to grab each frame in perfect detail so it can be post-processed. The easiest and cheapest way to do this is to have the renderer save each frame as it's computed. Real-time is not an issue, just like it's not an issue with a raytracer or whatever.
It better become feasable if companies are going to want renderfarms based on the nv30/40/whatever. Having two seperate machines per renderer would be pretty.. dodgy
No, not really. You've got a lot more than just pushing pixels around to worry about for it to be that simple.
4x the pixels doesn't mean 4x the geometry, or 4x the AI, or 4x the bandwidth requirements (ok, the final image is 4x bigger, but the textures and geometry you're pushing aren't also 4x bigger).
Look at the benchmarks for the 9700 for Q3; 1024*768 = 203FPS. 1600*1200 = 180.6FPS. The nearly three times as many pixels of the higher resolution resulted in a loss of just 10% of the framerate.
Uh, except 3-400FPS in RTCW in a given resolution means you can double the resolution and get.. 1-200FPS. Handy when the next generation of monitors come out that let you use resolutions of 4000x4000 or whatever. It also means you can make more detailed maps, and have busier scenes; instead of being limited to 2-3 characters in view at any one time, you can do a Doom/Serious Sam with the same detail level. Or you can tone up the LOD and have higher resolution textures further away and reduce mipmap artifacts. Or you can seriously concider adding bumpmaps everywhere, or more detailed volumetric self shadowing and *still* not drop below monitor refresh for the most detailed scenes.
And even IF you've still got tonnes of power left, this is what multitasking is for; if my system can push RTCW along at 400FPS, I can leave some expensive background task running and still have perfectly smooth gameplay.
Um, no, you've missed the point. This isn't about using a GFX card as a device to capture external data, it's to grab the images it's rendering into system memory so you can use it to, say, render some CG in a movie. You don't want to use a seperate device to grab the images your expensive cinematic quality GFX card produces when you could just dump it directly into the device it's running on.
The idea is to use the [GV]PU to render your production quality images, so grabbing the rendered image directly off the card is exactly what you want.
Don't forget, the engine's having to do a lot more than current FPS's, because the lighting is *all* dynamic, bump mapped, self shadowed, etc. Id's working on removing or reducing the differences between objects in the world and the world itself, and removing things like lightmaps.
:)
Basically this means that Doom III will be able to do things like have doors with lit rooms behind them cast light onto unlit rooms and for the lighting to actually change as the door opens while everything in the room lights up from it and casts shadows on themselves and everything else. But stills with lighting similar to what you'd get with a normal lightmap will take much more power to render, and hence need to have a lower level of detail.
Basically it's trading freedom and clever lighting effects for polygons and performance. Eventually this will pay off, but don't expect the average screenshot to necessarily look better than anything else. Expect it to MOVE better in the right maps, though
Probably also worth pointing out that, with Opera at least, the client still downloads the banners, and you can use the author/user mode toggle to turn it on/off should you feel the need to click on a couple of banners to support a site.
In this way, I probably click on more banners than I ever did when I just ignored them, AND the sites I visit still get their impressions.
I have a user CSS file which does this. It also works in Opera (and any other browser with reasonable CSS 2 selector support; i.e. not IE ;), and is quite effective.
http://voi.aagh.net/code/anti-banner.css
Let me know if you use it.
There's a bit of JS you can link to which makes it work.
Alternatively you can use absolute positioning and hide the fixed from IE using one of the many tricks/selectors it doesn't support (assuming this is compatible with your use of fixed).
Or you can do what the W3C do and ignore it. Nice how IE6 SP1 *still* doesn't fix this.