17. Can SmartPaper(TM) be tampered with electronically?
No. Due to the way signs are constructed, they can not be tampered with by magnets, static electricity, PDAs, cell phones, or other electronic devices.
So I suppose he rather had something like a strong coil?
True. I like Reiser's first idea though - giving credit to the developers in splash screens, screensavers, etc. But changing the license - no. As others on the newsforge board said too, a Free Software license does not restrict the user in any way other than forcing people to keep the license. If people want more credit for their products, they could easily change their product so it shows the names in more places (as long as that doesn't lead to nag screens...).
Maybe Hans Reiser is writing this article because, of course, his software has no higher-level splash screens etc., and credit is thus not too visible, but I think he could reach most distro makers and ask them to, for example, give him more credit in the documentation, and they might happily do it. But I think that enforcing credits through the license is unnecessary. Most software includes credits in many ways, and many people who use free software to improve their own free software do give credit. In the case of ReiserFS, for example, Hans does not gain anything from his name being included in every software building on it. The people he seems to be trying to reach won't read the licenses anyway (he seems to be looking for more of a marketing/'general public' approach).
So, it would of course be nice (and not much of a problem) to have, say, more splash screens/screensavers/hints in the docs or whatever showing credit (without annoying the user of course). I would take his message as "hey, people, let's promote our own names in a way that's a little more visible", and I'd be fine with that. But a license change? Not justified, imho.
SpamAssassin works pretty well for me too, since my college's mail server has it running and I can filter everything to a folder on my IMAP account. My spam stats today, since March 5th: 564 Spams, 27 not caught, i.e. 4.8% and somewhere about 11 Spams a day. This is not a Bayes-enabled SpamAssassin, and it runs on the standard spam threshold of 5.0 which I can not change.
Note that I hardly published my address anywhere; my biggest trouble are mailing list archives that don't obfuscate or remove the addresses.
But while it's nice that the crap doesn't flood my inbox anymore, I still/get/ those mails. And I find that bad enough -- unacceptable, really. The only good spam is no spam at all...
It's actually quite interesting how in the comics, whenever they explain that you have to press the number 0, they use that "old-school" slashed-zero... reminds me of my Commodore days...
Oh, and I enjoyed the Computer Trap story, when they are in the Museum hiding from the terrorists, and discover a phone... so they go ahead and use their spiffy Model-100 and an acoustic coupler (gasp) to set off an emergency message... and I'm thinking, why didn't they just place a normal phone call... funny stuff.
Yeah, for practical reasons the legacy stuff will still just work and be fine... but it's really nice to think of a system that's redesigned from scratch with well-specfied, open, up-to-date interfaces...
But anyway, that isn't gonna happen. Technology just evolves, and anyone designing such a new hardware for the mass market would fail spectacularly (not so in, say, some special server surroundings where the context is just different). But the good point of the article is, at some point there's a time when old stuff should be just thrown overboard instead of kept. Hey, even MS realized that when finally dropping (most of) the DOS legacy...
No... quantum computing will not allow you to factor in constant time or anything, or by "assuming the answer", get back the factors. Your idea seems to be "let's take the result and then calculate backwards" so to speak. But that won't work. Now if we could create two superpositions of "all numbers between 0 and sqrt(c)" (to put it in an easy way), calculate the product and then find a way to filter out all results equal to c (which seems to be what you're looking for), then we'd of course be able to simply measure the factors. But the problem is that you can't "filter out" only the results you want to look at. You might be able to slightly increase the likelihood of measuring the 'correct c' and therefore getting correct factors. That's (very simply put) what Shor's algorithm is doing - it only manages to increase the likelihood to measure the right result and therefore retrieve correct factors.
Note that I'm grossly oversimplifying...
Another example is trying to solve 3CNF-SAT - figure out whether a formula in 3CNF can be satisfied - in O(1). Classically, it's an NP-complete problem with exponential complexity. Now the naïve attempt would be to create a superposition of all possible inputs, filter out only those that yield "true" as a result, and then measure the "filtered" superposition to get a solution. Same problem; you can't really filter out the "true" results, you can only make it slightly more likely to measure a "1" as a result and therefore retrieve a solution for the input. You'd still need to repeat that for a couple of times, only less often as in the classical case - but still not in O(1), or even O(n).
So no, quantum computing is not that much of a magic solve-everything-instantly machine... e.g. Grover's algorithm to find an element in an unsorted list will not bring you from classical O(n) to O(1), but rather O(sqrt(n)).
But then again, maybe you're just trolling:)
Anyway, I found this paper here very interesting: it's called "Quantum Computing for Non-Physicists".
I'd just like to review the 'jokes' so far... call it Flamebait, if you want to, but it just annoys me and I have to say it...
RFC 3514
This one was nice. Obvious, but nice... as usual, the RFC people are doing a good job. Too bad Slashdot ruined it with the first April Fools dupe...
Gentoo on RPM
Well, good idea of the Gentoo people, but waaay too obvious... imho a good AF joke is one you believe to be true for at least a couple of minutes until you've looked at it very closely. But OK, this was only number 2, so it was still nice... now AF story bloat yet.
Whitespace programming language
Hm. As has been pointed out, it's not new... and even more obvious than the previous one (and pretty much boring too). I think then CPAN people know how to really make a good AF joke...
Microsoft + Security
Hey... better at least. Nicely combining a true story with a joke story... though the joke was not very believable either. Also, it's number four already, and we're only half-way through the day...
The Register's story
OK. By itself, not all that bad... but not too overwhelming either. And, on Slashdot, we're now at Number 5 and counting.
The dupe.
OK, increasingly stupid... but then again, maybe the best joke today on Slashdot:)
Enlightenment 1.0
Aawwww, come on. Enough. It hurts. And again, blunt joke. Latency between reading and noting the joke:.01 ms. Including the dupe, we're at seven now... and the day is still not over. I think I'll stop reloading the web site until I'm sure April 1 is over in all time zones...
Yep... as I've said severaltimes already... it's just like last year, when after the fourth joke everyone was annoyed. Make it fewer jokes... but make 'em good instead.
Somehow I don't find stories too good if there's a green box saying "Note: this is an April Fools joke" at the top... but I guess that just proves how many slashdotters actually follow the links and read the stories;-))
But Editors, pleease remember: do not drown us in hundreds of April Fools jokes... last year it was getting kinda annoying. And we already have three stories up now... please make it fewer jokes, but make 'em good.
That being said, I liked what my University'scomputer center did... we have a wi-fi net running on campus, on today they announced Germany-wide wireless access via the new Satellite network... very detailed prank, they put up the homepage, created the corresponding mail addresses (sukath-admin@...), and wrote a whole bunch of pages on that site I linked to... funny FAQs etc. Good job!
And thinking about that, I'd like to say: Editors, please do not overact tomorrow! Last year was funny for the first two 4-1 stories, but after that it just got annoying. Please limit yourself to maybe two or three April Fool stories, but make 'em good instead...
And you might want to take a look at their CD register. Click on "Datenbank abfragen" to find out about problems with specific CDs (or Un-CDs) users registered.
Hmm... yes, the font rendering in X might be good. But what the Linux world really is missing is a centralized, standard font system for all applications. I can certainly enjoy nice on-screen fonts. But try writing a document using the app of your choice, and then printing it. OpenOffice is on the right way (at least it manages to use TrueType fonts to print correct PostScript documents). But currenty it's a real pain to be able to pick any font on the system by its unique name, and then go and use this same font for on-screen rendering and printing. While it might be possible to make this work, it's a pain to do that. Just think about the mess of directories on your system that contain fonts. On my Debian Testing box, there's/usr/X11R6/lib/X11/fonts, and there's/usr/share/fonts; and to be able to use my TrueType fonts in OpenOffice, I have them all in $HOME/.openoffice/1.0.1/user/fonts. A software using these TT fonts to print will most likely have to embed the font data. It can not refer to the font name in PostScript as it can in X. Where is that a problem? Last time I tried, when I wanted to use a vector graphics program (sketch, kontour,...) and write some text using my installed fonts, I would be disappointed as soon as it came to printing. So either it didn't work, or it was just way too much of a hassle to make work.
That being said, here comes the disclaimer:-) I might be wrong on how printing and X fonts interoperate... but I would definitely be interested in a way to set up proper font handling (as good as it gets) that does not cause as much confusion.
Oh, and then on Debian there's defoma... what does that do? For me (<- dumb user in this respect;-)) it seems to make it all even more complicated...
Well yes, it is overrated. Here are my guesses why:
The US Administration has created a big part of the "publicity" for this event
The 9/11 context -- whether or not it actually plays a role for the Iraq crisis -- creates more attention
The media are eagerly repeating publicly every tiny rumor concerning this war, again this is related to the 9/11 context I guess. 9/11 was alsowhen media first started with those "War on Terror"/"War on Iraq"/"Attack on..." etc. brand names, having seemingly discovered a 'war==profit' rule. Who knows, maybe the fact that a huge part of media networks belong to Rupert Murdoch (strongly pro-War in this case) might have a little bit to do with that too.
'Innocent' (whatever that means) Iraqi civilians will inevitably die in this war. But is that really the same as saying that no 'innocent' Iraqi people will die if Mr Hussein is left to his own device and in power?
Problem here is, you are trading lives by saying "it's better to kill 1,000 than letting 100,000 be killed." Where is the limit? If you kill n innocent people to prevent the killing of m innocent people, for which ratio n/m do you still call your war "just"? Where do you draw the line?
The answer: you simply can't. Noone has the right to trade innocent lives. To give an example, the German constitution states in its very first article, "human dignity is inviolable." There are similar statements in most free nations' constitutions.
I personally can not support a war that is (at least partly) justified on such a tradeoff.
I definitely agree with you. But I have really been wondering: Throughout the years, the US has often displayed such an arrogant attitude. But the Americans always have such confidence that they are doing the one Right and Good Thing, and it is almost impossible to change their minds, and difficult enough to even try and create a non-biased, open discussion.
I wonder what the reasons for this "character trait" are. For one, it's probably due to the strong, sometimes blind, patriotism... but there might be more to it than just that -- historic background,...?
(Yes, you could give me "it's the economic interests, the 'good will' is just pretent" -- but while that might be true for some of the politicians, it's unlikely to be true for the majority of Americans, who still follow their President into war.)
Slashdot Mirror
From the Gyricon link you gave:
So I suppose he rather had something like a strong coil?
True. I like Reiser's first idea though - giving credit to the developers in splash screens, screensavers, etc. But changing the license - no. As others on the newsforge board said too, a Free Software license does not restrict the user in any way other than forcing people to keep the license. If people want more credit for their products, they could easily change their product so it shows the names in more places (as long as that doesn't lead to nag screens...).
Maybe Hans Reiser is writing this article because, of course, his software has no higher-level splash screens etc., and credit is thus not too visible, but I think he could reach most distro makers and ask them to, for example, give him more credit in the documentation, and they might happily do it. But I think that enforcing credits through the license is unnecessary. Most software includes credits in many ways, and many people who use free software to improve their own free software do give credit. In the case of ReiserFS, for example, Hans does not gain anything from his name being included in every software building on it. The people he seems to be trying to reach won't read the licenses anyway (he seems to be looking for more of a marketing/'general public' approach).
So, it would of course be nice (and not much of a problem) to have, say, more splash screens/screensavers/hints in the docs or whatever showing credit (without annoying the user of course). I would take his message as "hey, people, let's promote our own names in a way that's a little more visible", and I'd be fine with that. But a license change? Not justified, imho.
I liked the fact that the car of the passer-by is a Ford Prefect :)
Interesting... pretty much the same thing Microsoft says... except for a very different definition of "worst." ;-)
SpamAssassin works pretty well for me too, since my college's mail server has it running and I can filter everything to a folder on my IMAP account. My spam stats today, since March 5th: 564 Spams, 27 not caught, i.e. 4.8% and somewhere about 11 Spams a day. This is not a Bayes-enabled SpamAssassin, and it runs on the standard spam threshold of 5.0 which I can not change.
/get/ those mails. And I find that bad enough -- unacceptable, really. The only good spam is no spam at all...
Note that I hardly published my address anywhere; my biggest trouble are mailing list archives that don't obfuscate or remove the addresses.
But while it's nice that the crap doesn't flood my inbox anymore, I still
It's actually quite interesting how in the comics, whenever they explain that you have to press the number 0, they use that "old-school" slashed-zero... reminds me of my Commodore days...
Oh, and I enjoyed the Computer Trap story, when they are in the Museum hiding from the terrorists, and discover a phone... so they go ahead and use their spiffy Model-100 and an acoustic coupler (gasp) to set off an emergency message... and I'm thinking, why didn't they just place a normal phone call... funny stuff.
Hm, das ist dann wohl der erste deutsche Thread (Entschuldigung, 'Faden') auf Slashdot? :-)
Yeah, for practical reasons the legacy stuff will still just work and be fine... but it's really nice to think of a system that's redesigned from scratch with well-specfied, open, up-to-date interfaces...
But anyway, that isn't gonna happen. Technology just evolves, and anyone designing such a new hardware for the mass market would fail spectacularly (not so in, say, some special server surroundings where the context is just different). But the good point of the article is, at some point there's a time when old stuff should be just thrown overboard instead of kept. Hey, even MS realized that when finally dropping (most of) the DOS legacy...
No... quantum computing will not allow you to factor in constant time or anything, or by "assuming the answer", get back the factors. Your idea seems to be "let's take the result and then calculate backwards" so to speak. But that won't work. Now if we could create two superpositions of "all numbers between 0 and sqrt(c)" (to put it in an easy way), calculate the product and then find a way to filter out all results equal to c (which seems to be what you're looking for), then we'd of course be able to simply measure the factors. But the problem is that you can't "filter out" only the results you want to look at. You might be able to slightly increase the likelihood of measuring the 'correct c' and therefore getting correct factors. That's (very simply put) what Shor's algorithm is doing - it only manages to increase the likelihood to measure the right result and therefore retrieve correct factors.
Note that I'm grossly oversimplifying...
Another example is trying to solve 3CNF-SAT - figure out whether a formula in 3CNF can be satisfied - in O(1). Classically, it's an NP-complete problem with exponential complexity. Now the naïve attempt would be to create a superposition of all possible inputs, filter out only those that yield "true" as a result, and then measure the "filtered" superposition to get a solution. Same problem; you can't really filter out the "true" results, you can only make it slightly more likely to measure a "1" as a result and therefore retrieve a solution for the input. You'd still need to repeat that for a couple of times, only less often as in the classical case - but still not in O(1), or even O(n).
So no, quantum computing is not that much of a magic solve-everything-instantly machine... e.g. Grover's algorithm to find an element in an unsorted list will not bring you from classical O(n) to O(1), but rather O(sqrt(n)).
But then again, maybe you're just trolling :)
Anyway, I found this paper here very interesting: it's called "Quantum Computing for Non-Physicists".
Simple answer: because it's an Alpha release. I would expect that to change, and surely it will be easy to (de)activate in a stable release.
When it comes around the next time, should we call it Quad Damage?
I'd just like to review the 'jokes' so far... call it Flamebait, if you want to, but it just annoys me and I have to say it...
This one was nice. Obvious, but nice... as usual, the RFC people are doing a good job. Too bad Slashdot ruined it with the first April Fools dupe...
Well, good idea of the Gentoo people, but waaay too obvious... imho a good AF joke is one you believe to be true for at least a couple of minutes until you've looked at it very closely. But OK, this was only number 2, so it was still nice... now AF story bloat yet.
Hm. As has been pointed out, it's not new... and even more obvious than the previous one (and pretty much boring too). I think then CPAN people know how to really make a good AF joke...
Hey... better at least. Nicely combining a true story with a joke story... though the joke was not very believable either. Also, it's number four already, and we're only half-way through the day...
OK. By itself, not all that bad... but not too overwhelming either. And, on Slashdot, we're now at Number 5 and counting.
OK, increasingly stupid... but then again, maybe the best joke today on Slashdot
Aawwww, come on. Enough. It hurts. And again, blunt joke. Latency between reading and noting the joke:
For some quality 4-1 jokes, see here (German), the above-mentioned cpan.org, or even the Freshmeat one which isn't so bad. This ain't so bad either. Kuro5hin points to this interesting link.
Can you /. editors pleease try to come up with a single good hoax and dump the rest? That would be nice.
</rant>Yep... as I've said several times already... it's just like last year, when after the fourth joke everyone was annoyed. Make it fewer jokes... but make 'em good instead.
Please.
As for RFC 1149, it's also worth noting there was an implementation too...
Somehow I don't find stories too good if there's a green box saying "Note: this is an April Fools joke" at the top... but I guess that just proves how many slashdotters actually follow the links and read the stories ;-))
But Editors, pleease remember: do not drown us in hundreds of April Fools jokes... last year it was getting kinda annoying. And we already have three stories up now... please make it fewer jokes, but make 'em good.
That being said, I liked what my University's computer center did... we have a wi-fi net running on campus, on today they announced Germany-wide wireless access via the new Satellite network... very detailed prank, they put up the homepage, created the corresponding mail addresses (sukath-admin@...), and wrote a whole bunch of pages on that site I linked to... funny FAQs etc. Good job!
Hmm yes, I thought so for a second too...
And thinking about that, I'd like to say: Editors, please do not overact tomorrow! Last year was funny for the first two 4-1 stories, but after that it just got annoying. Please limit yourself to maybe two or three April Fool stories, but make 'em good instead...
Thank you.
And you might want to take a look at their CD register. Click on "Datenbank abfragen" to find out about problems with specific CDs (or Un-CDs) users registered.
Hmm... yes, the font rendering in X might be good. But what the Linux world really is missing is a centralized, standard font system for all applications. I can certainly enjoy nice on-screen fonts. But try writing a document using the app of your choice, and then printing it. OpenOffice is on the right way (at least it manages to use TrueType fonts to print correct PostScript documents). But currenty it's a real pain to be able to pick any font on the system by its unique name, and then go and use this same font for on-screen rendering and printing. While it might be possible to make this work, it's a pain to do that. Just think about the mess of directories on your system that contain fonts. On my Debian Testing box, there's /usr/X11R6/lib/X11/fonts, and there's /usr/share/fonts; and to be able to use my TrueType fonts in OpenOffice, I have them all in $HOME/.openoffice/1.0.1/user/fonts. A software using these TT fonts to print will most likely have to embed the font data. It can not refer to the font name in PostScript as it can in X. Where is that a problem? Last time I tried, when I wanted to use a vector graphics program (sketch, kontour, ...) and write some text using my installed fonts, I would be disappointed as soon as it came to printing. So either it didn't work, or it was just way too much of a hassle to make work.
That being said, here comes the disclaimer :-) I might be wrong on how printing and X fonts interoperate... but I would definitely be interested in a way to set up proper font handling (as good as it gets) that does not cause as much confusion.
Oh, and then on Debian there's defoma... what does that do? For me (<- dumb user in this respect ;-)) it seems to make it all even more complicated...
gkrellm (see bottom of that page), for example, has (at least for the name) been inspired by the movie Forbidden Planet.
Do you know any other SF-inspired software that /.ers use a lot?
Nononono. .fr is short for .freedom. ;-)
Well yes, it is overrated. Here are my guesses why:
Just some guesses.
Good point. By not acting, we are also trading lives; thus, no matter how we react (war or not), we are implicitly making this trade-off.
So what follows is simply that you can not argument one way or the other with a trade-off argument. It is no justification for any action.
Although this has definitely a rather philosphical nature now...
Problem here is, you are trading lives by saying "it's better to kill 1,000 than letting 100,000 be killed." Where is the limit? If you kill n innocent people to prevent the killing of m innocent people, for which ratio n/m do you still call your war "just"? Where do you draw the line?
The answer: you simply can't. Noone has the right to trade innocent lives. To give an example, the German constitution states in its very first article, "human dignity is inviolable." There are similar statements in most free nations' constitutions.
I personally can not support a war that is (at least partly) justified on such a tradeoff.
I definitely agree with you. But I have really been wondering: Throughout the years, the US has often displayed such an arrogant attitude. But the Americans always have such confidence that they are doing the one Right and Good Thing, and it is almost impossible to change their minds, and difficult enough to even try and create a non-biased, open discussion.
...?
I wonder what the reasons for this "character trait" are. For one, it's probably due to the strong, sometimes blind, patriotism... but there might be more to it than just that -- historic background,
(Yes, you could give me "it's the economic interests, the 'good will' is just pretent" -- but while that might be true for some of the politicians, it's unlikely to be true for the majority of Americans, who still follow their President into war.)
Can someone give some 'psychoanalysis'?
Just remember what Bush said to the Iraqi on Tuesday evening: "And it will be no excuse to say, 'I was just following orders.' "
Mind that this also applies to you.