also...if yuo DO need the order back, since the number of chunks is finite, yuo just add some extra data on the end to give the order of the chunks as extra data...it wont change the size of the compressed data THAT much. the larger the size(number of chunks) in fact, the less the extra order chunk will actually make a difference. so, for a 1GB file, 1k chunks, that is 10^6 chunks, that is 20bits * 10^6 chunks that is 20 extra MB.
although for 100:1 compression that is twice the compressed data, the total compression is STILL 30:1 which is QUITE a bit better than 10:1...again the bigger the data (optimize, actually for chunk size, # of chunks) the more the compression.
this could be used for MOVIES!! (yay)...take a while to compress the stuff yuo want, but then the tranfer is fast, and uncompression is fast. the only downside of this method is that yuo can't stream (SORRY!) but otherwise, for providing full movies to ppl, this would work really well.
well...so much for all that technobabble on their site, and it obviously being some sort of hoax, scam, something to impress mommy, etc. or whatever yuo want to call it.
the point is, that it ALL depends on what the application is. for example, sometimes yuo don't care about the order of the data yuo get back, yuo just want to the same discrete chunks yuo started with. and yuo sometimes also don't care how long it takes to COMPRESS, as long as uncompression is fast. for example, in the case of hardware testing, yuo usually want to test a bunch of input to make sure yuo have the correct output. if yuo are testing this capability, then yuo DONT care how long it takes to compress, yuo only need to do it once, what yuo care about is the QUICK transmission of the compressed data and uncompressing it as fast as possible.
so what yuo do in this case, is divide the data into chunks of the test codes, and try to solve the TSP of the test codes as coordinates of an N-dimensional space where N=# of bits in code. although solving TSP for a large number of codes is impractical, yuo can use, say, a kohonen self organizing neural net to approximate. and at every step of the way, yuo would have data that can be compressed via a running length compression if yuo express the differences between the test codes (that is what the TSP is for). the longer yuo wait, the better the compression is. at least it is only O(x) and not anymore (not including actual the updating time for each node, which is neglible anyway compared to the actual solving of the problem).
getting it back is trivial, just add all the differences one by one to get the test codes back in some order and feed them to the hardware to be tested. works, well, doesnt it??? its more than 100:1 if yuo wait long enough. and it is very practical for the application at hand. however, it is clearly impractical for every day use. so...can this BS company claim what they have done is real?
SURE! if they dont mention the details of what they are doing...
unfortunately, despite my usual line of linux advocacy, i must say one thing: as purely consumer and end-user operating system, windows suffices. in fact, since dave "doesn't mess with the insides of his computer", he is truly a consumer, and therefore more likely to do something wrong especially if he is not used to it. if he decides to upgrade to linux, then in THAT case, i doubt he will find anyone as ready to help him as his friendly tech support guy....so in his case...i suggest he stays with what he has.
apparently ff IS stale, or they wouldn't be making a new adaption to MMORPG. the thing is, before, the newer final fantasy games drove sales, but hopefully (for them) not like this one. since multiplayer doesnt exist on console without making yuor screen a fraction of its already small size, which made multiplayer NOT very enticing. now...yuo can use the web going capability of the PS/2 to play REAL multiplayer...YAY!
is that on *nix, i have to save a file before i run it, unlike in windows, where i simply click on the attachment and it runs. their comment on linux users being more sophisticated was obviously well taken by the/. crowd, but in reality, the same exact idiots exist onl inux, as much as we would like to believe it. the diff is that when yuo DO click on an attachment, it DOESNT automagically get executed.
for once i am disappointed. instead of commenting on how much better the operating system is,/. users decide instead that their ego is more important and therefore its probably more useful to post on how "sophisticated" we are all...well, we ARE...but isnt that rather egotistical???
also, just another thought: even with all the stuff that was mentioned...couldn't the integration of data (a server with a rigid protocol of some sort) be run on TOP of the kernel. so, use the linux kernel, and instead of spending precious money (400 million) and time (it'll take a while) developing a new kernel, use what exists and instead spend your time on writing a server that allows for programs to communicate better that runs on top of imbedded linux.
well, first of all, they seem to be hiding something that the story is not telling us. all the things that they say an OS can do, so can linux and windows and just about every other mature OS on the market today.
if they were developing a new OS, they need to do one thing: the thing that makes the Palm different from the Newton: all data is integrated into the OS, making it very easy to communicate between different programs. that is what most OSs lack. it is very difficult to make programs communicate with each other, except maybe with FIFO files that take up extra space. it is not really possible for one program to write in the memory of another. sure that is 3V1L for desktop and server OSs b/c they are VERY vulnerable, but its perfect for an OS that is supposed to communicate with appliances.
Now, if you have been thinking to mod me down because I did not mention grep on unix..wait right there! grep is a great tool, but it still has its extents. you cannot do everything with grep. programs can use each other's data (making it very useful for doing complicated tasks via simple programs) but they cannot in the full sense of the word intercommunicate. most importantly, grep is essentially a one way connection.
usefullness and functionality and speed of use and easiness AFTER you've learned it? we shall have to see. it seems that THAT would be the plugins job..so i guess someone could make a plugin that could have all the emacs keycodes. or maybe i'm just dreaming
i think i can generalize this to: if there exists a functionality of a program that the program uses...use it! generators are just a special case of this, because their whole functionality revolves around what they are made of...but many people make the mistake of designing something based on some OTHER library, when what they are making just extends the functionality of the library itself...so the question is: why not use what you have made???? editors are another group of programs that fall into this category.
yuo are correct ofcourse in saying that ogg vorbis is a better system. however, i don't think that discussion of ogg vorbis is quite relavent in terms of "winners" and "losers". ogg is great stuff, no doubt, it beats mp3, yes...but if i want some music, and i don't want to buy it, then all i have for the moment is mp3! unfortunate but true.
and it is obviously useless to encode mp3 files into ogg. and if BOUGHT the music? well then...why would i be using a lower quality sound format anyway? i use the original CDs. many ppl rip their music just to boast a 10 GB mp3 (or ogg) collection, but CDs are of better quality than both mp3 and ogg. plus, when i use a cd, there is less CPU usage then decoding ogg or mp3. the only mp3z i have are those that I download...why would i waste space on mp3z that I ripped from higher quality music?
making an evolving fpga try to imitate current computers may be hopelessly pointless, because we do not know how much our view of computing may slow down the processes it is handling. however, if the fpga was used to evolve a simple (at first) turing machine, it could essentially do everything that computers can currently do. if, however, this turing machine was optimized, (the modules would be separate, having different fitness functions, but in sync, namely the operation of one would depend on another), we dont know how efficient the machine would become.
maybe the hal in the article (not 2001, at least yet) really IS a reality soon...
everyone seems to be posting how doctors, police, etc have to work since they are public services, but that jewish colleagues cover them. well, that seems very nice, but not very fair. on yom kippur and rosh hashanah, jews don't get the day off, and end up taking a personal day so they can do whatever it is that floats their boat.
so why should they work for christmas?
in israel, nothing is closed on christmas, but a lot of people go out and have a nice dinner or do something special for christmas, simply because its a holiday everywhere else and the businesses in every OTHER country are closed.
maybe its because christmas is no longer about celebrating the birth of christ and more about capitalist consumerism taking over the world and making everyone spend lots of $$ on presents...
but have yuo used, say, a given DISTRIBUTION'S tech support. if yuo have, yuo are lying, because they are paid to do their job and hence dont tell yuo to RFM, and if yuo havent?...
well, if any FREE support measures up in ANY way to paid support, then the free support must be rEALLY DAMNED GOOD!
NOT QUITE!!! sure its one of the reactions in one...but if they created a supernova...we wouldnt need canada.com (of all places) to report....the shock wave would deliver the news a little faster!
that its their nature to diss each other's products because they are biased, and that redhat / lineo overreacted to the m$ propaganda...but think of what would happen if they DIDN't respond to them. everyone would keep believing m$, and we would all be bought out...
what makes you think that people will suddenly start saying that "i am fed up", etc...just because linux is out? NO! it has been going on for years, and ppl at this point (those that are still using windows and not linux extensively) have never seen anything better. the sad thing is, they consider microsoft a reliable source, so they will believe the m$ BS that they spit out constantly.
sure, m$ is scared shitless of linux, because an open source OS that they can't buy was the only thing they didn't predict; but getting people to switch to linux is a slow process...yuo can't force them to switch unless you start using microsoft marketing strategies.
having done some technical writing myself, i now know the difference between software developers and technical writers. software developers make the software (DUH), but also know MORE than anyone else about what the project / software is about. however they usually know jack shit or close to nothing about writing and eloquent expression. because of this, they hire technical writers. the problem is the fact that technical writers write better (DUH, again), but also usually dont know THAT much about what they are writing, and the developers end up unsatisfied with the result.
writers should work in conjunction with software developers to document their work. yuo cant have a bunch of guys contribute to the cause...
finally, developers usually find the docs the MOST ANNOYING part of the process of completing their project / producing their software. therefore...we can't really reach a golden mean...and creating an organization of sorts to document things is NOT incredibly useful, IMHO.
however, the rating system might remedy this. the developers might not be pleased and be too lazy to write their own docs, but can write the written stuff.
eh, whatever: hate to judge prematurely...so good luck guys!
Slashdot Mirror
also...if yuo DO need the order back, since the number of chunks is finite, yuo just add some extra data on the end to give the order of the chunks as extra data...it wont change the size of the compressed data THAT much. the larger the size(number of chunks) in fact, the less the extra order chunk will actually make a difference. so, for a 1GB file, 1k chunks, that is 10^6 chunks, that is 20bits * 10^6 chunks that is 20 extra MB.
although for 100:1 compression that is twice the compressed data, the total compression is STILL 30:1 which is QUITE a bit better than 10:1...again the bigger the data (optimize, actually for chunk size, # of chunks) the more the compression.
this could be used for MOVIES!! (yay)...take a while to compress the stuff yuo want, but then the tranfer is fast, and uncompression is fast. the only downside of this method is that yuo can't stream (SORRY!) but otherwise, for providing full movies to ppl, this would work really well.
QED
well...so much for all that technobabble on their site, and it obviously being some sort of hoax, scam, something to impress mommy, etc. or whatever yuo want to call it.
the point is, that it ALL depends on what the application is. for example, sometimes yuo don't care about the order of the data yuo get back, yuo just want to the same discrete chunks yuo started with. and yuo sometimes also don't care how long it takes to COMPRESS, as long as uncompression is fast. for example, in the case of hardware testing, yuo usually want to test a bunch of input to make sure yuo have the correct output. if yuo are testing this capability, then yuo DONT care how long it takes to compress, yuo only need to do it once, what yuo care about is the QUICK transmission of the compressed data and uncompressing it as fast as possible.
so what yuo do in this case, is divide the data into chunks of the test codes, and try to solve the TSP of the test codes as coordinates of an N-dimensional space where N=# of bits in code. although solving TSP for a large number of codes is impractical, yuo can use, say, a kohonen self organizing neural net to approximate. and at every step of the way, yuo would have data that can be compressed via a running length compression if yuo express the differences between the test codes (that is what the TSP is for). the longer yuo wait, the better the compression is. at least it is only O(x) and not anymore (not including actual the updating time for each node, which is neglible anyway compared to the actual solving of the problem).
getting it back is trivial, just add all the differences one by one to get the test codes back in some order and feed them to the hardware to be tested. works, well, doesnt it??? its more than 100:1 if yuo wait long enough. and it is very practical for the application at hand. however, it is clearly impractical for every day use. so...can this BS company claim what they have done is real?
SURE! if they dont mention the details of what they are doing...
QED
unfortunately, despite my usual line of linux advocacy, i must say one thing: as purely consumer and end-user operating system, windows suffices. in fact, since dave "doesn't mess with the insides of his computer", he is truly a consumer, and therefore more likely to do something wrong especially if he is not used to it. if he decides to upgrade to linux, then in THAT case, i doubt he will find anyone as ready to help him as his friendly tech support guy....so in his case...i suggest he stays with what he has.
QED
apparently ff IS stale, or they wouldn't be making a new adaption to MMORPG. the thing is, before, the newer final fantasy games drove sales, but hopefully (for them) not like this one. since multiplayer doesnt exist on console without making yuor screen a fraction of its already small size, which made multiplayer NOT very enticing. now...yuo can use the web going capability of the PS/2 to play REAL multiplayer...YAY!
QED
is that on *nix, i have to save a file before i run it, unlike in windows, where i simply click on the attachment and it runs. their comment on linux users being more sophisticated was obviously well taken by the /. crowd, but in reality, the same exact idiots exist onl inux, as much as we would like to believe it. the diff is that when yuo DO click on an attachment, it DOESNT automagically get executed.
/. users decide instead that their ego is more important and therefore its probably more useful to post on how "sophisticated" we are all...well, we ARE...but isnt that rather egotistical???
for once i am disappointed. instead of commenting on how much better the operating system is,
QED
also, just another thought: even with all the stuff that was mentioned...couldn't the integration of data (a server with a rigid protocol of some sort) be run on TOP of the kernel. so, use the linux kernel, and instead of spending precious money (400 million) and time (it'll take a while) developing a new kernel, use what exists and instead spend your time on writing a server that allows for programs to communicate better that runs on top of imbedded linux.
QED
well, first of all, they seem to be hiding something that the story is not telling us. all the things that they say an OS can do, so can linux and windows and just about every other mature OS on the market today.
if they were developing a new OS, they need to do one thing: the thing that makes the Palm different from the Newton: all data is integrated into the OS, making it very easy to communicate between different programs. that is what most OSs lack. it is very difficult to make programs communicate with each other, except maybe with FIFO files that take up extra space. it is not really possible for one program to write in the memory of another. sure that is 3V1L for desktop and server OSs b/c they are VERY vulnerable, but its perfect for an OS that is supposed to communicate with appliances.
Now, if you have been thinking to mod me down because I did not mention grep on unix..wait right there! grep is a great tool, but it still has its extents. you cannot do everything with grep. programs can use each other's data (making it very useful for doing complicated tasks via simple programs) but they cannot in the full sense of the word intercommunicate. most importantly, grep is essentially a one way connection.
QED
IDE, eh?
M-x mode-xxx
QED
learning curve steeper? maybe...
usefullness and functionality and speed of use and easiness AFTER you've learned it? we shall have to see. it seems that THAT would be the plugins job..so i guess someone could make a plugin that could have all the emacs keycodes. or maybe i'm just dreaming
QED
even if its NOT a generator of any sort.
i think i can generalize this to: if there exists a functionality of a program that the program uses...use it! generators are just a special case of this, because their whole functionality revolves around what they are made of...but many people make the mistake of designing something based on some OTHER library, when what they are making just extends the functionality of the library itself...so the question is: why not use what you have made???? editors are another group of programs that fall into this category.
QED
yuo are correct ofcourse in saying that ogg vorbis is a better system. however, i don't think that discussion of ogg vorbis is quite relavent in terms of "winners" and "losers". ogg is great stuff, no doubt, it beats mp3, yes...but if i want some music, and i don't want to buy it, then all i have for the moment is mp3! unfortunate but true.
and it is obviously useless to encode mp3 files into ogg. and if BOUGHT the music? well then...why would i be using a lower quality sound format anyway? i use the original CDs. many ppl rip their music just to boast a 10 GB mp3 (or ogg) collection, but CDs are of better quality than both mp3 and ogg. plus, when i use a cd, there is less CPU usage then decoding ogg or mp3. the only mp3z i have are those that I download...why would i waste space on mp3z that I ripped from higher quality music?
QED
i am in a rather similar situation, but i am not going to rant about it...i started early (full-time paying job, that is)
the only thing i am wondering, if colleges are just as skeptical as this guy's IT manager is? oh well, i guess its their loss in the end.
QED
making an evolving fpga try to imitate current computers may be hopelessly pointless, because we do not know how much our view of computing may slow down the processes it is handling. however, if the fpga was used to evolve a simple (at first) turing machine, it could essentially do everything that computers can currently do. if, however, this turing machine was optimized, (the modules would be separate, having different fitness functions, but in sync, namely the operation of one would depend on another), we dont know how efficient the machine would become.
maybe the hal in the article (not 2001, at least yet) really IS a reality soon...
QED
see sig...apparently most ppl believe it to be so...
QED
this is because everyone who finally got tired of windows and moved to linux says: WHAT the hell??? where is ms word, etc..(you get the point).
instead we should be looking at what microsoft is: every one of their apps is a clone of some other product: can you say: LOTUS, WORD-PERFECT...etc
so, who is the one who is following?
QED
should relax and do something fun!
everyone seems to be posting how doctors, police, etc have to work since they are public services, but that jewish colleagues cover them. well, that seems very nice, but not very fair. on yom kippur and rosh hashanah, jews don't get the day off, and end up taking a personal day so they can do whatever it is that floats their boat.
so why should they work for christmas?
in israel, nothing is closed on christmas, but a lot of people go out and have a nice dinner or do something special for christmas, simply because its a holiday everywhere else and the businesses in every OTHER country are closed.
maybe its because christmas is no longer about celebrating the birth of christ and more about capitalist consumerism taking over the world and making everyone spend lots of $$ on presents...
QED
that was an excellent troll
it started out rather serious....trolls like
that are great! they make me so ANGRY >:o!!!
QED
"and this is the most secure version of windows ever!"
QED
but have yuo used, say, a given DISTRIBUTION'S tech support. if yuo have, yuo are lying, because they are paid to do their job and hence dont tell yuo to RFM, and if yuo havent?...
well, if any FREE support measures up in ANY way to paid support, then the free support must be rEALLY DAMNED GOOD!
QED
where did yuo actually GET this data...although everyone on /. will OBVIOUSLY say this is wrong (this is clearly flamebait)
/. readers to rebuke mindlessly!
also, i wonder what school(s) were yuo speaking of?
in most high tech schools linux is not only known of , but used!
please stop posting flamebait for
QED
now we know who is regular canada.com reader! we spotted the canadans(sic)!!!
QED
NOT QUITE!!! sure its one of the reactions in one...but if they created a supernova...we wouldnt need canada.com (of all places) to report....the shock wave would deliver the news a little faster!
QED
that its their nature to diss each other's products because they are biased, and that redhat / lineo overreacted to the m$ propaganda...but think of what would happen if they DIDN't respond to them. everyone would keep believing m$, and we would all be bought out...
QED
what makes you think that people will suddenly start saying that "i am fed up", etc...just because linux is out? NO! it has been going on for years, and ppl at this point (those that are still using windows and not linux extensively) have never seen anything better. the sad thing is, they consider microsoft a reliable source, so they will believe the m$ BS that they spit out constantly.
sure, m$ is scared shitless of linux, because an open source OS that they can't buy was the only thing they didn't predict; but getting people to switch to linux is a slow process...yuo can't force them to switch unless you start using microsoft marketing strategies.
QED
having done some technical writing myself, i now know the difference between software developers and technical writers. software developers make the software (DUH), but also know MORE than anyone else about what the project / software is about. however they usually know jack shit or close to nothing about writing and eloquent expression. because of this, they hire technical writers. the problem is the fact that technical writers write better (DUH, again), but also usually dont know THAT much about what they are writing, and the developers end up unsatisfied with the result.
writers should work in conjunction with software developers to document their work. yuo cant have a bunch of guys contribute to the cause...
finally, developers usually find the docs the MOST ANNOYING part of the process of completing their project / producing their software. therefore...we can't really reach a golden mean...and creating an organization of sorts to document things is NOT incredibly useful, IMHO.
however, the rating system might remedy this. the developers might not be pleased and be too lazy to write their own docs, but can write the written stuff.
eh, whatever: hate to judge prematurely...so good luck guys!
QED