They *still* aren't supported properly on the most used web browser (IE prompts me to save them to disk rather than open them and display them, perhaps this is fixed in Me, I don't know. I've also had problems with the alpha channels in IE, as well as the gamma correction stuff - they displayed very dark). Obviously this is typical MS sloth at work, and has nothing to do with PNG itself..
I've never managed to get PNG files to approach JPG's in terms of compactness. Obviously JPG is lossy, so it can always have an advantage - but the fact of the matter is, for the vast majority of image distribution that takes place, a certain amount of information loss is acceptable (JPG's look perfectly good to 99% of people anyway, so why would these people want to download bigger files in PNG format?) (Most people probably aren't even aware that JPG's are lossy, hence the large of amounts of ultra-crappy looking porn on newsgroups, from people saving and resaving images just to STUPID things like adding a huge blue border around an image.)
Software for creating/reading them (end-user software as well as API's) still isn't as common and widespread as it *should* be. These things take time, yes, but they are slowed by factors like popular browsers not supporting PNG's properly anyway.
Don't get me wrong though, I would like to see PNG becoming the dominant format. But PNG seems to forever be stuck in the "early adopter" phase..
OK.. so I may have been exaggerating a little wrt the times I quoted.. but as far as I'm concerned it's still waaaay too slow for a machine of spec I mentioned. It is capable of a lot more than it gets out with Windows/MS Word etc... PS I'm using Win98 (for work reasons, I can't use Win2k) which is obviously quite different also.
In general I agree with you, the interfaces within Windows and MS software are generally well-thought out (sourcesafe has some really stupid things in it though, but it does seem to be in general an app that MS has no real interest in improving.) This does not make them innovative though. Designing a decent UI requires a lot of careful thought and design, normally some prototyping and testing etc, and is generally pretty hard work (anyone who has actually tried to make a good UI would probably know that.)
But that, as I said, is not innovation. If I think of the notion of UI innovation, I think of *new* ways of doing things. MS might put careful thought into putting the sort of options on the context menu that you would expect there - but that is not innovation, innovation would imply that they actually came up with the idea of context menus. Similarly, MS handles the whole clipboard concept very well, better than anyone else that I know, but they certainly did not come up with the concept of the OS clipboard.
I guess that's where we're differing here.
Personalized menus (which you guessed right, I hate 'em:) I don't know if they came up with, it's possible, and as for hardware such as joysticks I wasn't really commenting on. My gut feeling says that the sidewinder was probably bought by MS, not created. I don't have facts to back that up, but I do vaguely remember something like that.
16 color bitmaps are ugly and archaic, and quite frankly I simply expect more from my hardware and software than that, which considering I have a GeForce2 and PIII667 with 512MB RAM, I don't think is an unreasonable expectation - we're not living in the 80's anymore, computers can handle a little more than that..
This doesn't have nothing to do with functionality either - in fact, for an application we're developing for a client we need to place images of various things you can place in the world (e.g. ships, tanks etc) and with anything less than at least 256 colours, it's extremely difficult to tell what they are (aside from being butt ugly.) Sure, we can design some abstract 16-color iconic images that represent the items you're placing in the world - but that is clearly less intuitive to the user than if he/she could instantly see on the icon what he was placing. Possible considered alternatives include writing our own toolbar classes, or using a different interface altogether, maybe an owner-draw list in a modeless dialog (which would take up much more space on the screen.)
Phew, someone else that feels the way I do. I use a 667 MHz Pentium III with 512 MB of RAM and a GeForce2 at work, and Windows is horribly slow on it. It can't even keep up with my keypresses in applications like Visual Studio and Microsoft Word (MS Word is particularly bad.) It often does absolutely *nothing* for a good twenty seconds or so while I'm typing before keypresses are responded to, normally in bursts.) It's terrible. Clicking on the Windows Explorer shortcut I wait a good 10 to 40 seconds for flipping Windows Explorer to start. If I was Joe User and I'd forked over all that money myself I would be most PO and demand some sort of refund or at the very least an explanation for why my system is so unresponsive.. but I guess Joe User doesn't have the computer knowledge and doesn't know any better.
Are the types hardcoded or is it a generic 'plug-in' type model? E.g. if I wanted to create a.3ds model previewer that worked with nautilus, could I? Just asking.. I've wanted a file manager that worked like this from the first time I used ACDSee.. actually set out to develop one myself at some stage, but never took it very far.
Call me when you've read more than the nickle version of computer interface history and we might have an intelligent discussion
You want to back up your implied knowledge on computer interface history here with some facts + references? You're quick to say that MS is innovating, but you don't give any examples. Please, I invite you to show us the list of UI innovations that MS came up with themselves. For a company that spends $5bil a year on R&D, show us what it's produced. And please, give us examples of things that *are* actually original, *not* copied concepts like the start menu. I personally can't think of a single thing in Windows UI design that hadn't been done before on X WM's, OS2, Mac's, Xerox, or outlined in various twenty year old UI design articles. MFC toolbars can't even handle more than 16 color bitmaps, I for one do not associate that sort of crap with a company that pushes any UI envelops.
As far as I can tell re gnome and kde, they seem to be trying to copy many of the things I've always regarded as the *worst* aspects of windows. For example, the gnome file manager. I thought it was great news when I heard they were using midnight commander as the base, ms is a well-thought out, useful file manager. I was disgusted when I learnt that they'd removed most of what was *good* about mc, and degraded it into some sort of Windows Explorer clone, complete with many of the absolutely bad things about WE. Windows Explorer is not my idea of a target worth aiming at.
So I'm not disagreeing with you that the gnome/kde people copy ideas a lot (sometimes for the worse, they seem to copy ideas sometimes without first asking themselves if they *should*). But to claim that MS is springing forth some non-negligible number of innovations that they're copying borders on ridiculous.
Name 10 original UI interface ideas that MS has come up with, that MS didn't themselves copy from someone else. Hint: the start menu is *not* one of them.
"Whistler will have the option to only run signed applications. You can turn this off"
This 'feature' is probably currently largely aimed at corporate desktop users for medium to large corporations as a means of controlling what software employees install and run. So in practice probably only your administrator is going to be able to turn it off.
"If you're having trouble with gnome placing windows off the screen, maybe you have a virtual desktop that is bigger than your visible desktop"
No, that was most certainly not the case, I've been using X more than long enough to not be confused by the virtual desktop. I always configure the virtual to be the same as the actual resolution anyway. I'm talking about plain straightforward window creation. And it wasn't the only enlightenment version I tried that did that.. it behaved like that for months..
But like I say, my last install is RH 6.1, so I'll give RH7 a try..
OK, I found the GIMP on the menu, under RedHat/graphics, of all places. Seems like a silly place to me. The first place a user would look for graphics applications would be under the root "Graphics" submenu, and that is where it should be. Users shouldn't have to worry about "OK was that a gnome or a KDE application", and shouldn't have to peruse several different "graphics" menus to find a graphics app. Gosh.. netscape is under "red hat" too.. I was not aware that Gimp and netscape are RedHat applications! (sarcasm..)
Agreed.. I've gotten pretty annoyed at some of the supposed-to-be-user-friendly window managers on Linux. And the worst part of it is, horribly enough, not *big* things but *little* things. Lots of *little things* are just plain wrong or stupid. And it's the little things that are so easy to fix.
I've just re-set-up my Linux box at home (it's still RH 6.1 so admittedly it's old) and I fired up KDE, and a few things bothered me within a few minutes: (1) No netscape available either on the KDE menu or the Gnome submenu (these days a web browser is such a basic, commonly used thing, what were they thinking?). (2) No GIMP available on either the KDE menu or the Gnome submenu (one of Linux's finest points, and it's not on the menu?) (3) Right-click on a file on the desktop to delete something: (3a) No keyboard shortcuts on the Yes/No message box!!! In windows I always just press 'Y' but I had to use the mouse for KDE. Windows might be crap as hell, but at least you can use the keyboard for basically everything. (3b) I don't know where to turn off the "ask me if I'm sure" delete option. In windows I can turn it off, I don't like it, it's one of the lousiest aspects of Windows file manager design (to perpetually ask "are you sure, are you really really totally sure" to everything, but at least most of the annoying message boxes can be turned off.)
Also the default window manager in Gnome is stupid enough to place windows by default so that *half of the window is off the screen*. So whenever a new window pops up, you have to manually drag it back onto the display. I'm sorry, that's just plain stupid.
And like I said - all of these things are *little things*.. I mean, I think both KDE and Gnome are really impressive, the *big things* are all in place. So there is little excuse for not having all the little things in place.. I'm going to install the latest RH sometime soon, and I just know I'm going to be extremely annoyed if those things aren't fixed yet.. I mean, it's been more than long enough now.
True.. good point. Nonetheless, this doesn't automatically throw physics into the category of metaphysics.. there is still a notable difference between the two.
"thus putting science on a solid basis of empirical observation and experimentation"
Which, as it turns out, seeks to obtain the meaning, nature and source of life.
That is exactly physics - the search for "truth". Are you saying that the terms "meaning of life" or "nature of life" automatically imply some sort of non-truth or metaphysics? I'm afraid I can't see why that would be the case at all.
That's human nature, and thats why Gnutella does and always will suck
Yes, that's probably why Napster never really caught on, right? Nobody was making their files available for downloading, they only wanted to get them..
Yeah right. Almost everybody I know loves to share files, using any possible network technology.. LAN/SMB is an extremely popular way to transfer files (especially large ones like movies, or large collections of mp3's)
Anyway, it only takes a small percentage of people to be "altruistic" for files to spread very quickly, since many people will copy a file from one "altruistic" person..
"One small question, when did serial port change into hot PnP?"
Not trolling.. otherwise I would have posted anon, and not included "OT" in the title.. anyway, I still don't see why you should have to restart your entire OS to plug in a serial mouse. Even in the days of DOS/TSR's you could hot-plug a serial mouse and just load the driver.. I'm sure Win2K could do the same, especially since they seem to go out of their way to detect when you unplug the thing, they might as well check if you plug one in. (It is an MS plug-n-play mouse BTW.. )
I suppose the issue would be performance problems related to polling the ports, I don't know, doesn't seem to bother them to poll the network status for cable unplugged..
"Your mouse has moved. Windows NT must be restarted for this change to take effect."
You joke, but I've noticed something horribly similar enough; if I unplug my serial mouse in Windows2000, it's smart enough to detect this, and the mouse is disabled and the mouse cursor disappears. (I was unplugging to switch my modem and mouse around, the modem was configured on the other port, so I figured it would be easier to switch them around physically.)
But when plugging the mouse back in, Win2k didn't seem as smart anymore.. so restart:) And that's supposed to be a server operating system.. "attention all employees, the server will be brought down at 14:00 today, as I have to plug a mouse in"..
It sounds like the software is some sort of audio pattern-recognition system backed with a large database of songs to recognize. There is then no reason why the software wouldn't be able to check that you don't have the entire song, making fair use more unlikely already. But that's probably irrelevant anyway...
If they detected less than (say) 70% of a song, they could just as well leave it, rather than have to worry about fair use cases. However, any serious music pirate site would most likely (1) include a large number of.mp3 files, and (2) include the entire songs. Detecting these characteristics will be able to tell you with probably 99.99% certainty that a site is guilty of piracy. Not proof, of course - but more than enough evidence to prove guilt in court.
This is probably the only type of site that this software is really going to be used to stop anyway, so if your thesis or whatever happens to contain 30 seconds of one or two songs, I wouldn't worry, the record industry couldn't be bothered. Much like Microsoft who mainly go after large software piracy organizations.
Anyway, the software isn't going to *be* a lawyer/judge. If you were caught by this software with some mp3's, I'm sure you'll still have the opportunity to defend yourself - and if you were in the right, you'd be OK.
So I suspect your point may be moot.
The majority of mp3 music piracy probably happens in (1) LAN-based SMB file-sharing in dorms etc, (2) napster, (3) ratio'd ftp's. The web is generally a stupid place to look for pirated mp3's. Napster is probably the only one of the abovementioned that "they"'ll eventually be able to integrate this detection into. Most mp3 piracy will therefore still go undetected anyway. Plus, protocols like gnutella will no doubt continue to improve and will probably eventually include encryption as some sort of standard. It's a technology one-up-manship battle that the RIAA cannot win - they are already behind, and they are much slower than the technology piracy "underground", and they are legally entitled to do everything 'properly'.
Might this software constitute some sort of 'unreasonable search' or invasion of privacy.. ? Hmm.. probably not if it's scouring websites.. but if it sits on routers checking passing emails etc (like carnivore) it might..
Sure it sucks, but there will be abuse with any system that gets implemented. Some systems will just allow for a smaller amount of abuse than others; but it's stupid to blindly reject any system that gets proposed just because "it might be abused".
Of course it'll be abused. But having domains like.xxx and.kids will make it a lot easier to control, and cause a lot less confusion and abuse than the current system. I'm sure most people will stick to intended uses of the domain names; why throw out the idea just because a tiny minority of people might not? That's just plain stupid. Overall, life would most likely be much easier for everyone if we had.kids and.xxx. The small amount of abuse might actually become manageable... I mean, a name like fuck.kids is going to stand out a lot more amongst the list of.kids domains than a name like fuckkids.com would stand out amongst the list of fuck*.com domaings.
"I could just as easily put my home page up at www.porn-free.xxx with no pornographic content whatsoever and be filtered for no other reason than my TLD happens be.xxx "
I can't for the life of me think of a reason why you might want to though. Some real-life analogies might be putting up a sign in front of your totally-clean candy store "hardcore porn available here".
One of the whole ideas behind having the.xxx domain is making things easy to filter. In real life, if you're out shopping with your kids, you *know* you're not going to walk into a store that has a huge sign outside saying "XXX adult toys". Same with.xxx. So if you put up your no-porn home page on a.xxx domain, you're just being stupid, and it's your own problem if you get filtered - it cannot by any means be considered an unfair filtering. That would be like complaining that people are discriminating against your candy store with the porn sign out front.
Slashdot Mirror
Also remember Intel was also the target of an antitrust investigation at the same time as MS .. it was just less well publicised ..
.. why so many /. articles get reposted more than once? :)
PNG's are good, BUT:
Software for creating/reading them (end-user software as well as API's) still isn't as common and widespread as it *should* be. These things take time, yes, but they are slowed by factors like popular browsers not supporting PNG's properly anyway.
Don't get me wrong though, I would like to see PNG becoming the dominant format. But PNG seems to forever be stuck in the "early adopter" phase ..
OK .. so I may have been exaggerating a little wrt the times I quoted .. but as far as I'm concerned it's still waaaay too slow for a machine of spec I mentioned. It is capable of a lot more than it gets out with Windows/MS Word etc ... PS I'm using Win98 (for work reasons, I can't use Win2k) which is obviously quite different also.
In general I agree with you, the interfaces within Windows and MS software are generally well-thought out (sourcesafe has some really stupid things in it though, but it does seem to be in general an app that MS has no real interest in improving.) This does not make them innovative though. Designing a decent UI requires a lot of careful thought and design, normally some prototyping and testing etc, and is generally pretty hard work (anyone who has actually tried to make a good UI would probably know that.)
But that, as I said, is not innovation. If I think of the notion of UI innovation, I think of *new* ways of doing things. MS might put careful thought into putting the sort of options on the context menu that you would expect there - but that is not innovation, innovation would imply that they actually came up with the idea of context menus. Similarly, MS handles the whole clipboard concept very well, better than anyone else that I know, but they certainly did not come up with the concept of the OS clipboard.
I guess that's where we're differing here.
Personalized menus (which you guessed right, I hate 'em :) I don't know if they came up with, it's possible, and as for hardware such as joysticks I wasn't really commenting on. My gut feeling says that the sidewinder was probably bought by MS, not created. I don't have facts to back that up, but I do vaguely remember something like that.
16 color bitmaps are ugly and archaic, and quite frankly I simply expect more from my hardware and software than that, which considering I have a GeForce2 and PIII667 with 512MB RAM, I don't think is an unreasonable expectation - we're not living in the 80's anymore, computers can handle a little more than that ..
This doesn't have nothing to do with functionality either - in fact, for an application we're developing for a client we need to place images of various things you can place in the world (e.g. ships, tanks etc) and with anything less than at least 256 colours, it's extremely difficult to tell what they are (aside from being butt ugly.) Sure, we can design some abstract 16-color iconic images that represent the items you're placing in the world - but that is clearly less intuitive to the user than if he/she could instantly see on the icon what he was placing. Possible considered alternatives include writing our own toolbar classes, or using a different interface altogether, maybe an owner-draw list in a modeless dialog (which would take up much more space on the screen.)
What for? Nobody here claimed that they had come up with anything, certainly I didn't.
There were such claims regarding MS though.
Phew, someone else that feels the way I do. I use a 667 MHz Pentium III with 512 MB of RAM and a GeForce2 at work, and Windows is horribly slow on it. It can't even keep up with my keypresses in applications like Visual Studio and Microsoft Word (MS Word is particularly bad.) It often does absolutely *nothing* for a good twenty seconds or so while I'm typing before keypresses are responded to, normally in bursts.) It's terrible. Clicking on the Windows Explorer shortcut I wait a good 10 to 40 seconds for flipping Windows Explorer to start. If I was Joe User and I'd forked over all that money myself I would be most PO and demand some sort of refund or at the very least an explanation for why my system is so unresponsive .. but I guess Joe User doesn't have the computer knowledge and doesn't know any better.
Are the types hardcoded or is it a generic 'plug-in' type model? E.g. if I wanted to create a .3ds model previewer that worked with nautilus, could I? Just asking .. I've wanted a file manager that worked like this from the first time I used ACDSee .. actually set out to develop one myself at some stage, but never took it very far.
Call me when you've read more than the nickle version of computer interface history and we might have an intelligent discussion
You want to back up your implied knowledge on computer interface history here with some facts + references? You're quick to say that MS is innovating, but you don't give any examples. Please, I invite you to show us the list of UI innovations that MS came up with themselves. For a company that spends $5bil a year on R&D, show us what it's produced. And please, give us examples of things that *are* actually original, *not* copied concepts like the start menu. I personally can't think of a single thing in Windows UI design that hadn't been done before on X WM's, OS2, Mac's, Xerox, or outlined in various twenty year old UI design articles. MFC toolbars can't even handle more than 16 color bitmaps, I for one do not associate that sort of crap with a company that pushes any UI envelops.
As far as I can tell re gnome and kde, they seem to be trying to copy many of the things I've always regarded as the *worst* aspects of windows. For example, the gnome file manager. I thought it was great news when I heard they were using midnight commander as the base, ms is a well-thought out, useful file manager. I was disgusted when I learnt that they'd removed most of what was *good* about mc, and degraded it into some sort of Windows Explorer clone, complete with many of the absolutely bad things about WE. Windows Explorer is not my idea of a target worth aiming at.
So I'm not disagreeing with you that the gnome/kde people copy ideas a lot (sometimes for the worse, they seem to copy ideas sometimes without first asking themselves if they *should*). But to claim that MS is springing forth some non-negligible number of innovations that they're copying borders on ridiculous.
Name 10 original UI interface ideas that MS has come up with, that MS didn't themselves copy from someone else. Hint: the start menu is *not* one of them.
"Whistler will have the option to only run signed applications. You can turn this off"
This 'feature' is probably currently largely aimed at corporate desktop users for medium to large corporations as a means of controlling what software employees install and run. So in practice probably only your administrator is going to be able to turn it off.
"If you're having trouble with gnome placing windows off the screen, maybe you have a virtual desktop that is bigger than your visible desktop"
No, that was most certainly not the case, I've been using X more than long enough to not be confused by the virtual desktop. I always configure the virtual to be the same as the actual resolution anyway. I'm talking about plain straightforward window creation. And it wasn't the only enlightenment version I tried that did that .. it behaved like that for months ..
But like I say, my last install is RH 6.1, so I'll give RH7 a try ..
OK, I found the GIMP on the menu, under RedHat/graphics, of all places. Seems like a silly place to me. The first place a user would look for graphics applications would be under the root "Graphics" submenu, and that is where it should be. Users shouldn't have to worry about "OK was that a gnome or a KDE application", and shouldn't have to peruse several different "graphics" menus to find a graphics app. Gosh .. netscape is under "red hat" too .. I was not aware that Gimp and netscape are RedHat applications! (sarcasm ..)
Agreed .. I've gotten pretty annoyed at some of the supposed-to-be-user-friendly window managers on Linux. And the worst part of it is, horribly enough, not *big* things but *little* things. Lots of *little things* are just plain wrong or stupid. And it's the little things that are so easy to fix.
I've just re-set-up my Linux box at home (it's still RH 6.1 so admittedly it's old) and I fired up KDE, and a few things bothered me within a few minutes: (1) No netscape available either on the KDE menu or the Gnome submenu (these days a web browser is such a basic, commonly used thing, what were they thinking?). (2) No GIMP available on either the KDE menu or the Gnome submenu (one of Linux's finest points, and it's not on the menu?) (3) Right-click on a file on the desktop to delete something: (3a) No keyboard shortcuts on the Yes/No message box!!! In windows I always just press 'Y' but I had to use the mouse for KDE. Windows might be crap as hell, but at least you can use the keyboard for basically everything. (3b) I don't know where to turn off the "ask me if I'm sure" delete option. In windows I can turn it off, I don't like it, it's one of the lousiest aspects of Windows file manager design (to perpetually ask "are you sure, are you really really totally sure" to everything, but at least most of the annoying message boxes can be turned off.)
Also the default window manager in Gnome is stupid enough to place windows by default so that *half of the window is off the screen*. So whenever a new window pops up, you have to manually drag it back onto the display. I'm sorry, that's just plain stupid.
And like I said - all of these things are *little things* .. I mean, I think both KDE and Gnome are really impressive, the *big things* are all in place. So there is little excuse for not having all the little things in place .. I'm going to install the latest RH sometime soon, and I just know I'm going to be extremely annoyed if those things aren't fixed yet .. I mean, it's been more than long enough now.
No, I never meant to imply that you were claiming physics were metaphysics .. I understand your point completely .. sorry if it seemed like it.
We can never know if a theory is "true" or not
True .. good point. Nonetheless, this doesn't automatically throw physics into the category of metaphysics .. there is still a notable difference between the two.
"thus putting science on a solid basis of empirical observation and experimentation"
Which, as it turns out, seeks to obtain the meaning, nature and source of life.
That is exactly physics - the search for "truth". Are you saying that the terms "meaning of life" or "nature of life" automatically imply some sort of non-truth or metaphysics? I'm afraid I can't see why that would be the case at all.
Hmm .. isn't "hat trick" also used in cricket for something? Not sure, but it rings a bell of some sort ..
That's human nature, and thats why Gnutella does and always will suck
Yes, that's probably why Napster never really caught on, right? Nobody was making their files available for downloading, they only wanted to get them ..
Yeah right. Almost everybody I know loves to share files, using any possible network technology .. LAN/SMB is an extremely popular way to transfer files (especially large ones like movies, or large collections of mp3's)
Anyway, it only takes a small percentage of people to be "altruistic" for files to spread very quickly, since many people will copy a file from one "altruistic" person ..
I think something's up, It appears I have just moderated my own post .. I thought that wasn't allowed .. or have the rules changed or something?
"One small question, when did serial port change into hot PnP?"
Not trolling .. otherwise I would have posted anon, and not included "OT" in the title .. anyway, I still don't see why you should have to restart your entire OS to plug in a serial mouse. Even in the days of DOS/TSR's you could hot-plug a serial mouse and just load the driver .. I'm sure Win2K could do the same, especially since they seem to go out of their way to detect when you unplug the thing, they might as well check if you plug one in. (It is an MS plug-n-play mouse BTW .. )
I suppose the issue would be performance problems related to polling the ports, I don't know, doesn't seem to bother them to poll the network status for cable unplugged ..
"Your mouse has moved. Windows NT must be restarted for this change to take effect."
You joke, but I've noticed something horribly similar enough; if I unplug my serial mouse in Windows2000, it's smart enough to detect this, and the mouse is disabled and the mouse cursor disappears. (I was unplugging to switch my modem and mouse around, the modem was configured on the other port, so I figured it would be easier to switch them around physically.)
But when plugging the mouse back in, Win2k didn't seem as smart anymore .. so restart :) And that's supposed to be a server operating system .. "attention all employees, the server will be brought down at 14:00 today, as I have to plug a mouse in" ..
It sounds like the software is some sort of audio pattern-recognition system backed with a large database of songs to recognize. There is then no reason why the software wouldn't be able to check that you don't have the entire song, making fair use more unlikely already. But that's probably irrelevant anyway ...
If they detected less than (say) 70% of a song, they could just as well leave it, rather than have to worry about fair use cases. However, any serious music pirate site would most likely (1) include a large number of .mp3 files, and (2) include the entire songs. Detecting these characteristics will be able to tell you with probably 99.99% certainty that a site is guilty of piracy. Not proof, of course - but more than enough evidence to prove guilt in court.
This is probably the only type of site that this software is really going to be used to stop anyway, so if your thesis or whatever happens to contain 30 seconds of one or two songs, I wouldn't worry, the record industry couldn't be bothered. Much like Microsoft who mainly go after large software piracy organizations.
Anyway, the software isn't going to *be* a lawyer/judge. If you were caught by this software with some mp3's, I'm sure you'll still have the opportunity to defend yourself - and if you were in the right, you'd be OK.
So I suspect your point may be moot.
The majority of mp3 music piracy probably happens in (1) LAN-based SMB file-sharing in dorms etc, (2) napster, (3) ratio'd ftp's. The web is generally a stupid place to look for pirated mp3's. Napster is probably the only one of the abovementioned that "they"'ll eventually be able to integrate this detection into. Most mp3 piracy will therefore still go undetected anyway. Plus, protocols like gnutella will no doubt continue to improve and will probably eventually include encryption as some sort of standard. It's a technology one-up-manship battle that the RIAA cannot win - they are already behind, and they are much slower than the technology piracy "underground", and they are legally entitled to do everything 'properly'.
Might this software constitute some sort of 'unreasonable search' or invasion of privacy .. ? Hmm .. probably not if it's scouring websites .. but if it sits on routers checking passing emails etc (like carnivore) it might ..
Sure it sucks, but there will be abuse with any system that gets implemented. Some systems will just allow for a smaller amount of abuse than others; but it's stupid to blindly reject any system that gets proposed just because "it might be abused".
Of course it'll be abused. But having domains like .xxx and .kids will make it a lot easier to control, and cause a lot less confusion and abuse than the current system. I'm sure most people will stick to intended uses of the domain names; why throw out the idea just because a tiny minority of people might not? That's just plain stupid. Overall, life would most likely be much easier for everyone if we had .kids and .xxx. The small amount of abuse might actually become manageable ... I mean, a name like fuck.kids is going to stand out a lot more amongst the list of .kids domains than a name like fuckkids.com would stand out amongst the list of fuck*.com domaings.
"I could just as easily put my home page up at www.porn-free.xxx with no pornographic content whatsoever and be filtered for no other reason than my TLD happens be .xxx "
I can't for the life of me think of a reason why you might want to though. Some real-life analogies might be putting up a sign in front of your totally-clean candy store "hardcore porn available here".
One of the whole ideas behind having the .xxx domain is making things easy to filter. In real life, if you're out shopping with your kids, you *know* you're not going to walk into a store that has a huge sign outside saying "XXX adult toys". Same with .xxx. So if you put up your no-porn home page on a .xxx domain, you're just being stupid, and it's your own problem if you get filtered - it cannot by any means be considered an unfair filtering. That would be like complaining that people are discriminating against your candy store with the porn sign out front.