One of the features described in Wired's article on this:
http://www.wired.com/news/business/0,1367,39456, 00.html
is that there is a (vaporware) attachment to make it capable of attachment to a projector for doing presentations. For me, this is (nearly) a killer app - no more lugging the laptop onto the plane. Right now I can only ditch the laptop if someone else is presenting too and I can email them slides...now if the Ansyr Primer PDF viewer (http://www.ansyr.com/) works on it (in colour) I'm in.
I agree on the handheld movie front, but handhelds and mp3 are (nearly) a good fit. I would like to reduce the amount of electronics I carry - to *one* power supply, *one* screen, *one* set of rolodex/calendar apps, *one* set of earphones/mic. Unless the mp3 player fits in there somewhere I'll be mighty disappointed. (I carry a Palm Vx and Nokia 7110 everywhere, I'm loath to carry any more)
The mp3 player (IMHO) fits in as part of the headset, not the handset. All the handset should need to do is sequence and stream stuff, eg over bluetooth, and leave mp3 decoding to a dedicated codec. NB that mp3 over bluetooth is not a bad fit as you need to compress the sound for transmission over a limited bandwidth network anyway - so why should the handheld decode anything? In this scenario the handheld mainly acts as a memory device.
WinCE comes nowhere near this ideal (nothing does, yet) and I agree with your main point that handhelds should stick to the KISS principle - mostly. However, handheld (computers) should be expected to supercede other devices which are successful _as_handhelds_ (phones, media players, radio recievers).
Mod trinition up, I wish folk would make the phones that way.
I have a Nokia 7110 and its just dumb: it has all these functions which would be useful while you're on the phone, but which can't be used because its on yer ear. I would say, don't even build an earpiece and mphone on the damn things, if you're not meant to use it that way.
Why not combine a PDA with the Ericsson bluetooth earclip phone headset, for the latest in geek gear. And lastly, even though I think my Pilot is fab, I need a bigger, colour screen. Oh, yes, I do. For streetmaps, and eBooks. When something like that comes out, I'll buy (again)...
Cracking the encryption is only half of the point. SDMI watermarks make the content traceable back to you if'n you post it on the web. And the watermarks are sent to the speaker too.
BTW SDMI watermarks any unwatermarked music with the ID of your player/recorder too (go read the spec!). So the point where music entered the 'network' of SDMI device owners would be traceable too.
I doubt an implementation should even waste CPU checking for the existence of a watermark, just overlaying one each time music is copied onto a device will make anything but 'generation 1' copying horribly noisy.
Working schemes go like this: choose a watermark (a large number). Use it to seed a pseudorandom number generator with a uniform spectrum. Take your stream in short (eg 2s) chunks (possibly in a fourier basis). Alter each chunk using part of your random number sequence (eg by changing the proportions of certain frequencies). Reassemble the stream.
Checking for a given watermark consists of regenerating the random number sequence and performing statistical checks on a sliced up stream to see if you get matches. More than a threshold, and you have the watermark.
The watermark does not depend on the digital nature of the sound, and is robust to D/A conversion, MP3 compression and the like. If you try to overwrite it, all that happens is that the sound degrades a little more, and you will now match *two* watermarks.
The flaw is that these things are supposed to be inaudible, so any transform which is inaudible to humans can be applied, and some of these affect some of the watermark. You can go on to show that there is at least one transform which removes the watermark, in fact one which removes any watermarking scheme, without affecting sound quality. This amounts to discovering a fantastic new lossy compression system, which, BTW, is _hard_. Hence why the SDMI are confident about their watermarks. Also why they want IP rights on anything that breaks them (a condition of the contest).
Complicated? Yes. Usable on small/underpowered hardware? Yes. (a 1k memory requirement is not excessive!). Performance wise? If we look at Western European languages, about 10% of chars are non-ascii, its about 4 times as slow. For non-European languages, its roughly *30* times as slow.
I would really really not thank anyone who made fopen 30 times slower (yes I know I'm assuming the comparison is the slow step here).
DNS has gone for the half-way house - the I18N version is case-sensitive for non-US-ASCII languages.
at some site called slashdot...
http://slashdot.org/article.pl?sid=00/09/13/0634 210&mode=flat
funny how that article got one (abusive) comment and this one got hundreds, they're essentially about the same thing.
Anyway for what its worth - I think all the people thinking about this shit should talk to each other more. Theres not just rpm, deb, and that bsd development, but things happening on the fringes like the rpm-for-cygwin development (http://cygwin-rpm.sourceforge.net/ yay, no more searching through the list of ported software) and loki games' setup tool (http://www.lokigames.com/development/setup.php3)
Okay, this is not an ideal solution, but it is a solution.
Internet
---------------------------------------- firewall
Demilitarized Zone
[ Terminal Server (WTS or an X server) ]
---------------------------------------- firewall
Internal LAN [ client PC goes here ]
Internal users use netscape on the terminal server. This prevents you from leaking information without retyping. However it prevents you from pulling in downloads, and sending email with attachments to customers.
For downloads, open up inbound FTP connections to a fileserver in the DMZ. For outbound emails, warn that emails from the LAN are scanned, and do it. If people want to send a private message, they can use the X or ICA netscape client. This way your users opt in to be scanned when they are deliberately leaking information, because thats what the job requires. Using the X client, all they would have to laboriously retype the information.
Depending on the size of the company, you could scan ALL of these messages by hand, since most outbound mail will be personal or brief.
I didnt say it didnt suck. But it does hang together.
Obviously I was listening to the radio edit of Cages oeuvre;o)
But you're wrong about the watermarks. Think of the human ear of a function we apply to sounds. Watermarking is another such function. The following relationships are advertised as holding for :
For all sounds x in the set X (representing all metallica tracks...), and watermarks w:
ear(noisy(x))=ear(x)
ear(lossy(p,x))=ear(x) [ there exists a lossy protocol p for which this holds
ear(set_watermark(w,x))=ear(x)
get_watermark(set_watermark(w,x))=w
get_watermark(noisy(set_watermark(w,x))=w
Your claim is:
for all protocols p,
get_watermark(lossy(p,set_watermark(w,x)))=w
However let us _define_ protocol 0:
lossy(0,x)=ear(x).
If you know the lossy compression function ear(x) - the function in which each sound indistinguishable by the human ear are mapped on to one encoding - then you can remove the watermark. However you don't need to go that far, you only need to use a lossy compression scheme 'ear-prime' which does not distinguish between differently watermarked copies of the same sound.
Thus my contention is that watermarking has been set up to survive _particular_ lossy schemes as it is theoretically impossible for it to survive _all_ lossy schemes _that_do_not_affect_listening_quality_!!
Hope this clears up my position. Of course if watermarking introduces changes that are actually distinguishable by the human ear then yes you're right, and I can't get rid of them this way. But if there are audible figments nobody will want to listen to SDMI.
Also, if SDMI devices refuse to play tracks which are not watermarked, then again the scheme fails on those devices. However, If you've taken an SDMI track, saved it as MP3, then stripped the watermark, it is no longer possible to determine the first purchaser of the track, ie the pirate. So SDMI watermarking has failed. This leaves the RIAA only one option, which already seems to be happening: devices which do not conform to SDMI for playback will be made illegal.
No, you won't. A decent watermarking scheme could make *every bit* of the files different. Differential crypto on these things will be hard, because they are robust to white noise, and 'inaudible' noise introduced by popular lossy compression schemes like MP3. The only viable attack IMHO is to use a different lossy compression scheme to remove watermarks outright. You dont need to find the watermark for this to work, the attack is general against the concept.
A lot of posts have pointed out that the watermarks work by using sound structures which we would not normally hear (eg subtle time shifts, masked tones, high/low frequencies etc), in order that it is preserved in D/A A/D conversions.
However, removing such structures is/exactly/ how lossy compression works. If we can't hear the watermark, there must be some lossy compression scheme which removes or changes it.
Clearly the watermarking has been tested with the popular schemes (ATRAC, MP3 and so on). But they're not the only possible schemes. It is perfectly possible to come up with a lossy compression scheme which corrupts watermarks, without otherwise affecting the signal.
Why do I believe this? Well, because a compression scheme which does that is exactly what you would use to apply the watermark in the first place....
Its interesting that if we had an 'ideal' lossy compression algorithm, (which had an identical encoding for all sounds we would say sounded identical, and where any change to the encoded form was audible) then it would not be possible to watermark the sound.
BTW I'm interested to see how they manage to watermark John Cage's 4:13.
I agree with the sentiment, but not the words. I hope computing will become ubiquitous to the extent that you will not be able to point at something and say 'that is a computer', because everything is.
Anyway, in the year 2000, there are still far too many people killed in petty conflicts and too many people with inadequate food+water (usually the result of the conflicts, not nature).
Miss Worlds should still ask for world peace, not world computer literacy.
Part of the success of these is that they can be mapped on to well known matrix solving problems...theres enough information in the documents above for you to work out how to write one.
One wrinkle Restil doesnt mention is that the technique is not purely based around link structure. You _seed_ the process with content-ranked pages (hoping the process 'crawls' to the best set independently of the seed), and subsequently you may select the most relevant 'communities' of pages by content ranking. So if you are already in the top 100, say you may be able to content-mangle yourself up the list, but you need good linkages to get in first!
A further criteria used is response time (I strongly suspect Google use this, I got hooked on it when I found that its sites _responded_ rather than hanging as most AltaVista sites did at the time). Again theres publications on this stuff: the shark search algorithm is a spider with this feature.
It seems to me that if you're on the hunt for monopolies Real Networks can't be too far behind. Much as I dislike it (quicktime works much better for me but I know there are linux issues there) a precedent like this would jeapordise any future takeover of Real.
The money from the tools to create real streams must be their biggest asset. Its all very well tring to be free and open, but someone somewhere has to make money somehow if us engineers are gonna get paid.
I actually think this MS activity is fairly innocent, but your reasoning here (that its something like Purls) is all wrong. That facility is provided by a decent implementation of HTTP, to wit dealing with 301 responses as per section 10.3.2 of rfc2616:
10.3.2 301 Moved Permanently
The requested resource has been assigned a new permanent URI and any future references to this resource SHOULD use one of the returned URIs. Clients with link editing capabilities ought to automatically re-link references to the Request-URI to one or more of the new references returned by the server, where possible.
(my emphasis). If MS had set things up so that the URLs were like http://www.ms.com/redir?news_service_1 , and were switching between providers, then yes I would agree with you that this was a valid argument, but thats not what they're at.
As long as they dont mess with *my* bookmarks I don't mind, I've never yet felt the need to use the ones supplied by the browser vendors.
- to the guy who thinks Gutenberg are doing it right; ASCII would be useless for books like Tristram Shandy or Alice in Wonderland which use things other than ASCII. Its also useless for works in languages other than English, textbooks with diagrams, and so on.
- Book readers on my PC are no use; as other people pointed out, if you're using a reference you need to flick between the text and the thing you're doing.
- I hate dead trees. I travel round the country 2 or 3 days a week; I am now down to 1 (1) manual in dead tree format because its an NDA'd text the vendor only supplies on murdered oak; its 700 pages and hurts my shoulder...
In short, give me an eBook reader I can download huge chunks of pdf, or OpenEBooks on to and I will be your friend. My only worry with eBooks is if the reader breaks, then I've lost the text. Paper has an edge here in that it absorbs the food/drink I spill on it and keeps working.
So the added value service *I* want is an online bookshelf I can _upload_ text to; a mymp3.com for eBooks if you will, that records the fact that I've bought the stuff that needs to be bought, but doesnt limit me to commercial texts. That way I won't lose all my stuff when my PC/reader dies.
And then I want someone to scan in all those books that have taken over my house...
Not everything that *could* make life worthwhile has been invented however. Lets go look at Star Trek for a moment. Those matter transporters would be pretty handy in reducing the amount of time taken for doctors to get to emergencies, or for food (from replicators of course) to be sent to areas with famines.
The argument that we already have everything that makes life worthwhile has been made at every point in history. (30 years ago we'd have said, we can put a man on the moon but we can't get a bum off the street...)
Worse, in recent times this same argument is taken to its logical conclusion: the neo-Luddite 'return to a pastoral existence'. A pastoral existence in a world where rabies, smallpox, and TB are rife, could be argued to be no worse than the ills of the world now. But today our life expectancy is verging on 80; in pre-industrial societies life expectancy was closer to 50.
While we're on the subject...the arts are actually highly dependant on technology. Changes in painting techniques were largely driven by the availability of appropriate pigments, discovered over time. Film is pure technology. Large scale theatre can't be done with limelights, or sound systems.
Further, science can even BE art.
http://oposite.stsci.edu/pubinfo/jpeg/M16Full.jp g . 'nuff said.
Reading the article it seemed to me that at least part of the problem here is that CVS is making it difficult to work with multiple repositories linked in at various points in the tree. Wouldnt BitKeeper help here?
http://www.bitmover.com/bitkeeper/bk03.html
Creating a monolithic CVS repository is one answer, but its not the only answer.
I cannot believe the number of folk who are posting 'remember this thing only has a range of 10 metres, eavesdropping isnt an issue'. Bluetooth does not suddenly stop 10 metres from you. Bluetooth receivers must be certified to work at this range, but you can obviously build something much more sensitive.
This was exactly the point Bruce Schneier was making, which a lot of people seem to have missed: if you can pick up transmissions from a monitor from outside a building, just how much easier will it be in a bluetooth environment, where the devices are _intended_ to be transmitters.
I don't work in the industry so I can't say whether these things are any good. epocrates (http://www.epocrates.com/) is software to aid in drug dispensation which runs on the palm, also covers drug interactions. Not quite what you asked but I thought this may interest you.
More relevant is freemed, which of course other folks have suggested. (http://www.freemed.org/)
I'm suprised no-one else mentioned this, since it was invented to cover exactly the situation you're in. In Schneier and Kesley's model, donations are held in escrow, until the publication date. If the funding level you asked for wasnt met, you wouldnt finish the work and money gets refunded; and so on. They wrote a paper on this which describes the process in detail : http://www.counterpane.com/street_performer.pdf
This is no pipe dream, IIRC this has been discussed on/. before in the context of Stephen King's last work.
re the Palm, wasnt there a JavaOne demo of the KVM (the small footprint jvm for the palm etc) which involved it controlling Mindstorm robots? While the KVM (and its source) is downloadable, theres no sign of any Jini implementation for the Palm. The article about that demo is here:
http://developer.java.sun.com/developer/technica lArticles/ConsumerProducts/JavaTanks/PalmP roxy.html
...and while it actually does have all the classes for the demo, if you look closely you'll find the PalmProxy refers out to a half-dozen net.jini packages (not supplied), and rmi packages (ditto). The KVM is pretty limited (no JNI, serialization, lacks coverage of the Palm API) and I'm not convinced the standard Jini/RMI packages would work here.
I did find a palm jini/rmi implementation though, after a little searching:
http://postpc.cs.berkeley.edu/rmilite/
A bit late reading the article... I felt that the main cause of cheats was simply the assumption that you could trust the client (whether its in the information you think they have, or the speed you think they are, etc). You can never trust the client...if you assume this instead, it covers off most classes of cheating (a compromised server is more of a social problem, akin to ringing in auctions). I'm with some of the other respondents who feel that if you have allowed the hack (by design, or by poor design), its within the rules of the game. The clients can be legitimately improved by hacking. If you want to stop this then the protocols need to embody the rules of the game. I think the game designers would benefit from a better understanding of designing for security in general - games have all the same problems as e-commerce. This doesnt just mean encryption - encrypting a bad protocol is just delaying the inevitable hack.
Reading Crypto-Gram (http://www.counterpane.com/crypto-gram.html) would be a start. Check out the articles on the 'Cocaine Auction Protocol' for one way to deal with a hostile world, and on Attack Trees (Dec 99 issue) for reviewing security designs.
Slashdot Mirror
One of the features described in Wired's article on this:, 00.html
http://www.wired.com/news/business/0,1367,39456
is that there is a (vaporware) attachment to make it capable of attachment to a projector for doing presentations. For me, this is (nearly) a killer app - no more lugging the laptop onto the plane. Right now I can only ditch the laptop if someone else is presenting too and I can email them slides...now if the Ansyr Primer PDF viewer (http://www.ansyr.com/) works on it (in colour) I'm in.
I agree on the handheld movie front, but handhelds and mp3 are (nearly) a good fit. I would like to reduce the amount of electronics I carry - to *one* power supply, *one* screen, *one* set of rolodex/calendar apps, *one* set of earphones/mic. Unless the mp3 player fits in there somewhere I'll be mighty disappointed. (I carry a Palm Vx and Nokia 7110 everywhere, I'm loath to carry any more)
The mp3 player (IMHO) fits in as part of the headset, not the handset. All the handset should need to do is sequence and stream stuff, eg over bluetooth, and leave mp3 decoding to a dedicated codec. NB that mp3 over bluetooth is not a bad fit as you need to compress the sound for transmission over a limited bandwidth network anyway - so why should the handheld decode anything? In this scenario the handheld mainly acts as a memory device.
WinCE comes nowhere near this ideal (nothing does, yet) and I agree with your main point that handhelds should stick to the KISS principle - mostly. However, handheld (computers) should be expected to supercede other devices which are successful _as_handhelds_ (phones, media players, radio recievers).
Mod trinition up, I wish folk would make the phones that way.
I have a Nokia 7110 and its just dumb: it has all these functions which would be useful while you're on the phone, but which can't be used because its on yer ear. I would say, don't even build an earpiece and mphone on the damn things, if you're not meant to use it that way.
Why not combine a PDA with the Ericsson bluetooth earclip phone headset, for the latest in geek gear. And lastly, even though I think my Pilot is fab, I need a bigger, colour screen. Oh, yes, I do. For streetmaps, and eBooks. When something like that comes out, I'll buy (again)...
Bzzt! wrong.
Cracking the encryption is only half of the point. SDMI watermarks make the content traceable back to you if'n you post it on the web. And the watermarks are sent to the speaker too.
BTW SDMI watermarks any unwatermarked music with the ID of your player/recorder too (go read the spec!). So the point where music entered the 'network' of SDMI device owners would be traceable too.
I doubt an implementation should even waste CPU checking for the existence of a watermark, just overlaying one each time music is copied onto a device will make anything but 'generation 1' copying horribly noisy.
Watermarking really does work, but is flawed.
Working schemes go like this: choose a watermark (a large number). Use it to seed a pseudorandom number generator with a uniform spectrum. Take your stream in short (eg 2s) chunks (possibly in a fourier basis). Alter each chunk using part of your random number sequence (eg by changing the proportions of certain frequencies). Reassemble the stream.
Checking for a given watermark consists of regenerating the random number sequence and performing statistical checks on a sliced up stream to see if you get matches. More than a threshold, and you have the watermark.
The watermark does not depend on the digital nature of the sound, and is robust to D/A conversion, MP3 compression and the like. If you try to overwrite it, all that happens is that the sound degrades a little more, and you will now match *two* watermarks.
The flaw is that these things are supposed to be inaudible, so any transform which is inaudible to humans can be applied, and some of these affect some of the watermark. You can go on to show that there is at least one transform which removes the watermark, in fact one which removes any watermarking scheme, without affecting sound quality. This amounts to discovering a fantastic new lossy compression system, which, BTW, is _hard_. Hence why the SDMI are confident about their watermarks. Also why they want IP rights on anything that breaks them (a condition of the contest).
For a full discussion of the altenative implementation, see:
6 /part1.html
http://people.netscape.com/ftang/paper/unicode1
Complicated? Yes. Usable on small/underpowered hardware? Yes. (a 1k memory requirement is not excessive!). Performance wise? If we look at Western European languages, about 10% of chars are non-ascii, its about 4 times as slow. For non-European languages, its roughly *30* times as slow.
I would really really not thank anyone who made fopen 30 times slower (yes I know I'm assuming the comparison is the slow step here).
DNS has gone for the half-way house - the I18N version is case-sensitive for non-US-ASCII languages.
at some site called slashdot...4 210&mode=flat
http://slashdot.org/article.pl?sid=00/09/13/063
funny how that article got one (abusive) comment and this one got hundreds, they're essentially about the same thing.
Anyway for what its worth - I think all the people thinking about this shit should talk to each other more. Theres not just rpm, deb, and that bsd development, but things happening on the fringes like the rpm-for-cygwin development (http://cygwin-rpm.sourceforge.net/ yay, no more searching through the list of ported software) and loki games' setup tool (http://www.lokigames.com/development/setup.php3)
Okay, this is not an ideal solution, but it is a solution.
Internet
---------------------------------------- firewall
Demilitarized Zone
[ Terminal Server (WTS or an X server) ]
---------------------------------------- firewall
Internal LAN [ client PC goes here ]
Internal users use netscape on the terminal server. This prevents you from leaking information without retyping. However it prevents you from pulling in downloads, and sending email with attachments to customers.
For downloads, open up inbound FTP connections to a fileserver in the DMZ. For outbound emails, warn that emails from the LAN are scanned, and do it. If people want to send a private message, they can use the X or ICA netscape client. This way your users opt in to be scanned when they are deliberately leaking information, because thats what the job requires. Using the X client, all they would have to laboriously retype the information.
Depending on the size of the company, you could scan ALL of these messages by hand, since most outbound mail will be personal or brief.
I didnt say it didnt suck. But it does hang together.
Obviously I was listening to the radio edit of Cages oeuvre ;o)
But you're wrong about the watermarks. Think of the human ear of a function we apply to sounds. Watermarking is another such function. The following relationships are advertised as holding for :
For all sounds x in the set X (representing all metallica tracks...), and watermarks w:
ear(noisy(x))=ear(x)
ear(lossy(p,x))=ear(x) [ there exists a lossy protocol p for which this holds
ear(set_watermark(w,x))=ear(x)
get_watermark(set_watermark(w,x))=w
get_watermark(noisy(set_watermark(w,x))=w
Your claim is:
for all protocols p,
get_watermark(lossy(p,set_watermark(w,x)))=w
However let us _define_ protocol 0:
lossy(0,x)=ear(x).
Then:
get_watermark(lossy(0,(set_watermark(x,w)))
= get_watermark(ear(set_watermark(w,x))
= get_watermark(x)
= undefined.
If you know the lossy compression function ear(x) - the function in which each sound indistinguishable by the human ear are mapped on to one encoding - then you can remove the watermark. However you don't need to go that far, you only need to use a lossy compression scheme 'ear-prime' which does not distinguish between differently watermarked copies of the same sound.
Thus my contention is that watermarking has been set up to survive _particular_ lossy schemes as it is theoretically impossible for it to survive _all_ lossy schemes _that_do_not_affect_listening_quality_!!
Hope this clears up my position. Of course if watermarking introduces changes that are actually distinguishable by the human ear then yes you're right, and I can't get rid of them this way. But if there are audible figments nobody will want to listen to SDMI.
Also, if SDMI devices refuse to play tracks which are not watermarked, then again the scheme fails on those devices. However, If you've taken an SDMI track, saved it as MP3, then stripped the watermark, it is no longer possible to determine the first purchaser of the track, ie the pirate. So SDMI watermarking has failed. This leaves the RIAA only one option, which already seems to be happening: devices which do not conform to SDMI for playback will be made illegal.
Cheers,
Baz
No, you won't. A decent watermarking scheme could make *every bit* of the files different. Differential crypto on these things will be hard, because they are robust to white noise, and 'inaudible' noise introduced by popular lossy compression schemes like MP3. The only viable attack IMHO is to use a different lossy compression scheme to remove watermarks outright. You dont need to find the watermark for this to work, the attack is general against the concept.
A lot of posts have pointed out that the watermarks work by using sound structures which we would not normally hear (eg subtle time shifts, masked tones, high/low frequencies etc), in order that it is preserved in D/A A/D conversions.
/exactly/ how lossy compression works. If we can't hear the watermark, there must be some lossy compression scheme which removes or changes it.
However, removing such structures is
Clearly the watermarking has been tested with the popular schemes (ATRAC, MP3 and so on). But they're not the only possible schemes. It is perfectly possible to come up with a lossy compression scheme which corrupts watermarks, without otherwise affecting the signal.
Why do I believe this? Well, because a compression scheme which does that is exactly what you would use to apply the watermark in the first place....
Its interesting that if we had an 'ideal' lossy compression algorithm, (which had an identical encoding for all sounds we would say sounded identical, and where any change to the encoded form was audible) then it would not be possible to watermark the sound.
BTW I'm interested to see how they manage to watermark John Cage's 4:13.
-Baz
I agree with the sentiment, but not the words. I hope computing will become ubiquitous to the extent that you will not be able to point at something and say 'that is a computer', because everything is.
Anyway, in the year 2000, there are still far too many people killed in petty conflicts and too many people with inadequate food+water (usually the result of the conflicts, not nature).
Miss Worlds should still ask for world peace, not world computer literacy.
The classic algorithm of this type is called HITS, by J. Kleinberg.
IBM's 'Clever' is an enhancement to 'HITS'.
Part of the success of these is that they can be mapped on to well known matrix solving problems...theres enough information in the documents above for you to work out how to write one.
One wrinkle Restil doesnt mention is that the technique is not purely based around link structure. You _seed_ the process with content-ranked pages (hoping the process 'crawls' to the best set independently of the seed), and subsequently you may select the most relevant 'communities' of pages by content ranking. So if you are already in the top 100, say you may be able to content-mangle yourself up the list, but you need good linkages to get in first!
A further criteria used is response time (I strongly suspect Google use this, I got hooked on it when I found that its sites _responded_ rather than hanging as most AltaVista sites did at the time). Again theres publications on this stuff: the shark search algorithm is a spider with this feature.
It seems to me that if you're on the hunt for monopolies Real Networks can't be too far behind. Much as I dislike it (quicktime works much better for me but I know there are linux issues there) a precedent like this would jeapordise any future takeover of Real.
The money from the tools to create real streams must be their biggest asset. Its all very well tring to be free and open, but someone somewhere has to make money somehow if us engineers are gonna get paid.
I couldn't run netstat, IE had taken all the memory ;o)
Dunno if its got any better but Solaris IE was the only program I used which took more memory than Rose, and that takes some doing...
10.3.2 301 Moved Permanently
The requested resource has been assigned a new permanent URI and any future references to this resource SHOULD use one of the returned URIs. Clients with link editing capabilities ought to automatically re-link references to the Request-URI to one or more of the new references returned by the server, where possible.
(my emphasis). If MS had set things up so that the URLs were like http://www.ms.com/redir?news_service_1 , and were switching between providers, then yes I would agree with you that this was a valid argument, but thats not what they're at.
As long as they dont mess with *my* bookmarks I don't mind, I've never yet felt the need to use the ones supplied by the browser vendors.
- to the guy who thinks Gutenberg are doing it right; ASCII would be useless for books like Tristram Shandy or Alice in Wonderland which use things other than ASCII. Its also useless for works in languages other than English, textbooks with diagrams, and so on.
- Book readers on my PC are no use; as other people pointed out, if you're using a reference you need to flick between the text and the thing you're doing.
- I hate dead trees. I travel round the country 2 or 3 days a week; I am now down to 1 (1) manual in dead tree format because its an NDA'd text the vendor only supplies on murdered oak; its 700 pages and hurts my shoulder...
In short, give me an eBook reader I can download huge chunks of pdf, or OpenEBooks on to and I will be your friend. My only worry with eBooks is if the reader breaks, then I've lost the text. Paper has an edge here in that it absorbs the food/drink I spill on it and keeps working.
So the added value service *I* want is an online bookshelf I can _upload_ text to; a mymp3.com for eBooks if you will, that records the fact that I've bought the stuff that needs to be bought, but doesnt limit me to commercial texts. That way I won't lose all my stuff when my PC/reader dies. And then I want someone to scan in all those books that have taken over my house...
Not everything that *could* make life worthwhile has been invented however. Lets go look at Star Trek for a moment. Those matter transporters would be pretty handy in reducing the amount of time taken for doctors to get to emergencies, or for food (from replicators of course) to be sent to areas with famines.
p g . 'nuff said.
The argument that we already have everything that makes life worthwhile has been made at every point in history. (30 years ago we'd have said, we can put a man on the moon but we can't get a bum off the street...)
Worse, in recent times this same argument is taken to its logical conclusion: the neo-Luddite 'return to a pastoral existence'. A pastoral existence in a world where rabies, smallpox, and TB are rife, could be argued to be no worse than the ills of the world now. But today our life expectancy is verging on 80; in pre-industrial societies life expectancy was closer to 50.
While we're on the subject...the arts are actually highly dependant on technology. Changes in painting techniques were largely driven by the availability of appropriate pigments, discovered over time. Film is pure technology. Large scale theatre can't be done with limelights, or sound systems.
Further, science can even BE art.
http://oposite.stsci.edu/pubinfo/jpeg/M16Full.j
Reading the article it seemed to me that at least part of the problem here is that CVS is making it difficult to work with multiple repositories linked in at various points in the tree. Wouldnt BitKeeper help here?
http://www.bitmover.com/bitkeeper/bk03.html
Creating a monolithic CVS repository is one answer, but its not the only answer.
Didn't spot that one - I browsed at -1 and searched for 'street' and 'performer' before I posted. Mea culpa.
I cannot believe the number of folk who are posting 'remember this thing only has a range of 10 metres, eavesdropping isnt an issue'. Bluetooth does not suddenly stop 10 metres from you. Bluetooth receivers must be certified to work at this range, but you can obviously build something much more sensitive.
This was exactly the point Bruce Schneier was making, which a lot of people seem to have missed: if you can pick up transmissions from a monitor from outside a building, just how much easier will it be in a bluetooth environment, where the devices are _intended_ to be transmitters.
I don't work in the industry so I can't say whether these things are any good. epocrates (http://www.epocrates.com/) is software to aid in drug dispensation which runs on the palm, also covers drug interactions. Not quite what you asked but I thought this may interest you.
More relevant is freemed, which of course other folks have suggested. (http://www.freemed.org/)
I'm suprised no-one else mentioned this, since it was invented to cover exactly the situation you're in. In Schneier and Kesley's model, donations are held in escrow, until the publication date. If the funding level you asked for wasnt met, you wouldnt finish the work and money gets refunded; and so on. They wrote a paper on this which describes the process in detail : http://www.counterpane.com/street_performer.pdf
/. before in the context of Stephen King's last work.
This is no pipe dream, IIRC this has been discussed on
re the Palm, wasnt there a JavaOne demo of the KVM (the small footprint jvm for the palm etc) which involved it controlling Mindstorm robots? While the KVM (and its source) is downloadable, theres no sign of any Jini implementation for the Palm. The article about that demo is here:a lArticles/ConsumerProducts/JavaTanks/PalmP roxy.html
http://developer.java.sun.com/developer/technic
...and while it actually does have all the classes for the demo, if you look closely you'll find the PalmProxy refers out to a half-dozen net.jini packages (not supplied), and rmi packages (ditto). The KVM is pretty limited (no JNI, serialization, lacks coverage of the Palm API) and I'm not convinced the standard Jini/RMI packages would work here.
I did find a palm jini/rmi implementation though, after a little searching:
http://postpc.cs.berkeley.edu/rmilite/
A bit late reading the article...
I felt that the main cause of cheats was simply the assumption that you could trust the client (whether its in the information you think they have, or the speed you think they are, etc). You can never trust the client...if you assume this instead, it covers off most classes of cheating (a compromised server is more of a social problem, akin to ringing in auctions).
I'm with some of the other respondents who feel that if you have allowed the hack (by design, or by poor design), its within the rules of the game. The clients can be legitimately improved by hacking. If you want to stop this then the protocols need to embody the rules of the game.
I think the game designers would benefit from a better understanding of designing for security in general - games have all the same problems as e-commerce. This doesnt just mean encryption - encrypting a bad protocol is just delaying the inevitable hack.
Reading Crypto-Gram (http://www.counterpane.com/crypto-gram.html) would be a start.
Check out the articles on the 'Cocaine Auction Protocol' for one way to deal with a hostile world, and on Attack Trees (Dec 99 issue) for reviewing security designs.
Cheers, Baz