They're dangerous inefficient. Letting barely trained people drive polluting hunks of metal at great speed past pedestrians is a dumb idea.
If people paid taxes equal to the cost of urban driving then large cities could have good enough transit systems that people wouldn't need to drive.
Highway driving is fairly expensive. A car with a few people eats up a lot more fuel than those people's seat on even a farily luxurious bus.
Few people actually need cars and with a bit of decent city planning even less would.
Then let people drive outside of cities where they're less likely to drive drunk (people drive home from bars, etc, usually not from state to state, while drunk) and less likely to run off the road in a populated area.
If someone wants a car to drive on private property, or on interstate highways, or in similar situations, then let them. But if they want a car to drive in the city because they can't be bothered with transit, then screw 'em. (I only know a few drivers who I think are good enough to deserve a license for city driving. The rest of them... It'd be like giving guns to kids after a one-week anger management course and a few target-practice sessions.)
Bleh. I've heard this before, that it's important to protect our heritage, and our traditions.
Why?
Think of it like a free market economy, but with culture. If your culture is less attractive than others (stupid restrictive traditions, lame rituals, etc) then your children will likely not follow it, if they have a choice.
The whole push to preserve culture is based on giving people less choice. Limiting the outside influence they can be exposed to, so that they'll do the only think they know.
You can see it in culture, where for instance there are limits on the ammount of foreign TV shows and local artists are subsidised not based on skill, but on quota. You can also see it in religion, where people forbid teaching of other religions or non-religious viewpoints (think Kansas) so that the children will think the way their parents want them to.
It's not like preventing others from speaking makes your viewpoint any more compelling, but if someone is exposed to one viewpoint from everyone around them while they grow up, often they will believe it just because that's all they know.
Culture works in a similar way. You can get someone to practice your culture by smothering other cultures, but if your culture offered something of value, then people would pick it without being forced or deprived of options. And if they don't, then it's probably because your culture is nothing but a collection of useless rituals and stupid beliefs that hold no objective value.
I don't think that saying GNU utillities are based on AT&T Unix utils is incorrect. The code may be completely from scratch, but the names, functions, and usage is similar.
The code is often the smaller part of making a new util. Design, both in identifying the problem, and in deciding on the implementation, are as much of a problem as the actual coding.
This isn't to run down the GNU utils, they are intended to be very similar to Unix tools. If they were vastly different, they wouldn't be as useful.
(Why can't we have a thread that mentions GNU without someone using the term fanatic? They have stated a goal and are working towards that goal, not getting distracted in the meme of the moment isn't fanaticism.)
If you think about it, a keyword list, and a name list, like the white and yellow pages, are the ways to go.
This was worked out long ago by the phone companies. The actual address of a company is (mostly) insignificant and if you make it significant and easy, there aren't enough to go around.
One phone number is mostly the same as another, with some exceptions. The differentiator is the name of the company and their type of business. Trademark law already covers companies naming themselves after competitors trying to steal their business, so whatever system we develop doesn't need to deal with this. All 'we' need to develop is a way for Apple Records and Apple Computer to share the WWW without being unweildly.
Probably the best having a registry that stores the name of a company and a few (two or three) keywords about them, then you simply type in "Apple:Music" and the registry finds all entries with 'Apple' in the name and 'Music' (or synonyms) in the keywords, then automatically your browser replaces what you typed with the url, which except for bookmarking purposes is useless. (Like memorizing an often called phone number)
Then, not having to rely on having the shortest, easiest url, companies could take applecomputer.com or applerecords.com or OrganicAppleOrchardsAssociation.com, wanting it to be descriptive and hard to mistake, similar to a trademarked logo.
Trademark law should be ammended to say that having "Clue" (etc) trademarked in one context, not only does not give protection in other contexts (This is already in there), but specificaly does not give any rights in context-free situations, like clue.com.
> 2. yes. only the second key appears to be replacable, but since there is NO warning that a test has failed the first, the first question still applies.
Perhaps only the second is trivially replaceable, (ie, simply hexedit the string to a bunch of zeros)...
> 4. The key "locations" were never "known" prior to sp5.
This is not known. I'm sure people poked through MS's DLLs before. And the key did have the characteristic form of a key, as in, that string of bytes did look like a key. What we didn't know until recently was the symbol names that went with the data. (But the actual location and value of the data is *trivial* to rip out of an executable.)
> 5. In exported versions its 40 bit. Not a terribly difficult crack.
The strongest encryption you are allowed to install is 40b. But there are fewer restrictions on the use of encryption in such a way that arbitrary data can't be sent. One example of this is digital signing, such as it's used in the CryptoAPI. Especially when the intended use of the digital signing is to keep people from installing strong crypto in violation of those very export laws. 6. Yes you can replace it with a program, but with the first key to authenticate me as i spoof widgets.com, I can replace it without your knowledge or approval.
No, there's nothing in here that lets you install anything on someone's computer without their approval. They still need to initiate the install (by running the installer, or a trojan of some sort). What the key does is allow the CryptoAPI to check if the crypto package you install is signed by MS, or the NSA, the only two entities (in this case) who are supposed to sign valid crypto components for Windows.
You *could*, with a trojan, replace that key, or simply crack the cryptoAPI to not check, and then allow the installation of unsigned crypto modules. But, this would be silly and redundant to use in an attack.
If you could gain access to someone's computer (via trojan, physical access, etc) then you could simply install a keyboard monitor, or some such, and read all of their data directly, bypassing any crypto. You wouldn't need to hack the crypto module, install broken/weak crypto, then crack it later. And if you did this, the people they tried to communicate wouldn't understand it, because the crypto algorithms wouldn't match.
No, the only '3v1l h4x0r' trick to do here is simply crack *YOUR* CryptoAPI (and let your friends do the same) so that you can easily encrypt data with a strong algorithm and large key. To do this you'd need to either convince MS or the NSA to sign your strong crypto, or remove the check in the CryptoAPI, much the same as crackers everywhere remove shareware expiration checks, etc.
How would they find these unconnected sites to do an analysis of them?
I for instance have a two websites I use to distribute private files and such to friends. But, to the best of my knowledge neither of these has links to it.
They need to recruit a few ISPs, examine all the pages on them (after stripping usernames) to see how many webpages don't have links to or from them , then use that figure to adjust their estimate of the total number of pages.
(They need to get an estimate of the number of pages they can't find by spidering, and the only way to do that is go to the source...)
Personal settings re other people's posts. Perhaps John Katz bugs me, then I could give his posts a -1 on top of any other points he may get, or if I really liked his usual posts, I could give them a +1 or +2 so that they'd float to the top. This would be sort of like/ignore on IRC, it'd let us filter out those who personally bug us without having to inflict our moderation biases on the community.
(On a practical note, few users would use this, and only vs a few other users, so it's not a matter of storing 80k x 80k of settings.)
Instead of having 'Funny +1', perhaps we could seperate the descriptors and the points. Some funny posts are inapproptiately funny and some point out a good point with humor, so for funny to always be a +1 isn't always right. Also, if keywords were seperate it'd be easier to display a list of all the keywords for a message (funny, insightful, etc) which would aid in a filter for inc/excluding posts by keyword.
Also, don't think that the whole point of this moderation is to censor people whose opinions are unpopular. It's also to chop down the number of messages that we have to read. Preferably by selecting those messages which are 'best' by our and community standards.
We also need a way for messages to influence threads they're in. If someone posts a message and then the reply to it is +5, then the original post is probably relevant. This would work better by moderating parent posts up, but not down. (Don't hide a parent because the reply is -1, but do show a 0 post if it generated high scoring replies.)
Seriously, if you wish to spend tons of time making something that someone can make a closed-source version of, fine, but don't expect sane people to help.
The proper people to rewrite the system in such a way that they can make a closed-source version of it are the actualy comapnies wishing to do so.
Well, as long as you understand that your closed-source goal is laughable and only a few freaks want it, then fine. Just shut up about the GPL virus. It's that way because we want it that way so take a flying leap.
Translation: "I want to steal your code, if you stop me, I'll whine."
It's free for everyone unless they try to close the source. Can't you get that through your pea-brain? The only restriction is that you MUST release it as open-source and in such a way that future versions are also open-source.
Your right to take code, rewrite a bit, close it, and claim it as your own is one that I'm perfectly willing to take away from you grubby bastards.
Nope. No restrictions at all, for users. If you don't touch the source code, you don't get bothered by the license. The only people restricted are those rewriting the app, and they're only restricted in that they can't close the source.
Users not only are not harmed by the license, but they are unaware of it.
I've got _Beneath Apple Dos_ which I used to write a package of programs to transfer files to and from disk images. (For us emulator people)
I can't part with it because I still use it now and then, but if you need any information from it, I could try to OCR, or retype parts of it.
If anyone has _Beneath Apple Prodos_, I would like to have access to allow me to write in Prodos support for the apps.
(The allow reading/writing files to the disk images, reading/writing dos images, formatting (making) images and/or adding dos images, translating files (cr -> cr/lf, basic bytecode to text, etc) )
If anyone wants a copy, or wants to help development, contact me. I'll throw a GPL on it and we can update it.
Actually, I understand the code just has to be made available, either on disk, or on a website, etc.
But, this is to keep people from using the code in their project and denying users the benefits of an opensource application while using someone else's code to write it. It's a good thing.
And the GPL doesn't prevent FreeBSD people from using the GNU utils, as long as they provide the source, etc. They will have problems using parts of the source code if they insist on the BSDL, but if they don't insist on every app having a BSDL license, they're fine.
You're not really dictating the terms of the sublicenses, you're offering a contract to someone provided that they dictate the terms of their license in such a way.
Each person is only the copyright owner of the work they produce and as such is only able to dictate the terms of usage of that work. But, they can offer a contract which gives the right to use the work provided the joint author licenses it in a certain way.
The GPL is Stallman's greatest work. The compiler and the utils, etc are good, but it's only the GPL which made them as usable and perpetually free as they are. He is as much the father of what we think of as Linux as Torvalds is. Stallman made it possible for Torvalds to write Linux in a free environment.
And this is why we fight back with tools designed to keep the OS open for the users, licenses like the GPL, tactics like keeping up with proprietary extensions, etc.
If we sit and watch, people will take what we collectively have made and use it for their own purposes. This is not 'bad' but it can hurt us if we don't watch out and if we let it happen, we'd deserve it.
We need to fight tooth and nail against anyone who uses laws against us, by trying to trademarks names of our products, or patent processes that they use. And we should use every means at our disposal to fight these predators. Use what little legal protection we have (without being able to afford lawyers) like the GPL, mass lobbying of politicians, etc. We should also use other methods, boycotting companies like this, bringing out products similar to theirs specifically to remove their markets where they are attacking us, etc.
As long as our goal is always to keep the software free and open for the users then we should feel justified about throwing their own predatory tactics back at them.
But, probably the best weapon is simply making them depend on us. For example, unless Redhat decided to rewrite all of GNU/Linux without the GPL they need us. Companies that use the OS also need us in the same way that other companies need MS. And the companies that use Linux have a vested interest in keeping it free for users because that also means free for them. We need to set up these symbiotic relationships whenever possible and destroy with any means availble any parasitic relationships.
Not at all. Open means open. It doesn't mean anything else. A door can be open but the area behind it restricted. That doesn't mean the door isn't open, it just means you can't go there.
The doom/hexen/etc source is open in that you can see it. You can't call it your own and sell it, but if you really have to whine when someone won't let you profit from their product, tough shit.
And if you're any good, you can simply see how they did it and write your own game if it's that important.
Greater income to a few artists means less income to the rest and he doubted that those few megastars made up for the creativity lost when the lesser known artists quit.
The assertion is that the recording industry supports the recording industry and a few star artists, not all artists. Thus, the industry will propose law to help itself, not to help the artists. Being that we think the art is good and the industry is irrelevant, we should advocate laws that support the artists.
With DAT copying, you could have a good underground music network, much like with MP3s. If the taxes actually went to the artists whose work was copied instead of the recording industry then we could legitimatize this underground copying because it wouldn't deprive artists of their money.
If you think about it, there are probably more pirated copies of small artists' music than the stars, in relation to the total number of copies sold. This scheme would benefit the small artists because they're the ones who don't have contracts with the recording industry that would put their music in all the stores.
Yes, the computer world is colorblind, and also age, sex, looks, etc, aren't important. But you need to be able to afford a computer and have a minimum of schooling in how to use it.
This is a racial issue in that there are more poor blacks and hispanics than poor whites.
But, dealing with this from a racial angle is completely wrong. Not only does helping someone because of their race pretty well equate with holding others back because of their race. (Would it be racism to give blacks a 50% off coupon for a computer? How about to do it for whites?)
What needs to be done is to target the situation people are in. If the poor can't get good schooling then chances are that they won't be eligible for any decent jobs, or better schooling, etc, and won't know what to do with a computer if the government requires that they be given one.
This is *not* a racial issue. I went to school with plenty of white kids who were destined to live McLives when I transfered into the inner city for my final year of high school.
We need to reach these children somehow. Offering them real choices and the education to make the most of their chances. It's not fair to say that the poor have as good of a chance at winning a scholarship when they have to grow up in a slum and are sick all the time and it's not enough to simply raise welfare payments hoping that it will miraculously make these children live an easier life.
I think advocates do themselves a disservice by making issues like this racial. I would be behind anti-poverty movements, but the minute someone brings race into it, like "Helping poor *black* families" I rebel completely. Treating one race differently from another is what got us into this mess in the first place and repreating that mistake won't help anyone.
So yes, technology is completely color blind. Computers don't care who hits their keys and the users at the other end won't know who you are unless you tell them. They are economically discriminatory because you need to buy them and they're fairly expensive to a family barely getting by as is.
There is a small racial issue here in that there are many poor blacks, but this just clouds the real issue by ignoring all the poor of other races.
They picked INFO because HTML wasn't as widespread at the time and man pages don't allow hyperlinks.
man pages are good, but they're best suited for documenting a specific command, as in "What's that switch for ls to show all files..." When you don't know what the command is called, you don't really want to guess, check a bunch of 'see also' references by loading a new pages, etc..
This is where a HTMLish document is preferable. You select something, check it out, go back, etc. It's suited to putting all of the linux documentation in a big tree and letting you start and browse to what you need. Much more suitable for finding something you don't know the name of.
But, GNU used INFO because there wasn't anything better at the time.
I too think INFO blows. HTML is easier, has more viewers, is very platform independent.
I haven't used DocBook but unless it comes with an HTML version of the documents, it won't be as useful.
I do a lot of my linux work via telnet from windows machines, or in win32 with DJGPP (from work) and unless I want to open a new telnet window just to view docs, I need to have them in an easy to view format, like HTML. And HTML is nicely suited to local or remote viewing.
Yes. Of a sort. It wouldn't appeal to everyone in the same way that pencil sketches often don't appeal, no matter how good they actually are, to people used to full color oil paintings.
Mel's program would be sort of like the hidden art of Bev Dolittle (?) where there's a simple nature scene and hidden in it, if you know where to look, are ten wolves, or a man on horseback, etc. Or maybe it's be like some piece of kinetic sculpture that appears to be out of an MC Escher book.
But yes, he wrote a program in ways other people have problems understanding, and it worked faster and better because of this, so yes, there is art in that.
Not that I'd want it hanging on my wall, for two reasons. One, I usually like big colorful things, regardless of actual artistic merit (ie, sat. pics of the area, N.G. Maps of the world, etc) and two, while it may have been artistic, that just means that everything Mel did was probably that way and this was just the minor job that was later given to someone untrained enough to have to wrestle with it yet enlightened enough to be able to grok it with study. This is probably as artistic to Mel as the canvas Rembrandt cleaned his brushes on.
Almost anything we do can be done quickly and haphazardly, or carefully and elegantly. What makes one elegant work art and another not? Functionality.
Most people are pompous asses and refuse to call something art if it has the slightest use. Can you use a dead sheep in a tank except as a footrest (or sex object for programmers.:) No, so it's art. Can you use a well-designed bridge? Yes, so it's 'just' a bridge.
Code is one of these functional items. Except perhaps for tiny snippets in a textbook, we expect code to work. The most beautiful 'looking' code is totally worthless if it doesn't work, and well, and people would praise line-noise if it was an efficient kernel when run in intercal (not that this is likely...)
Contrast this to 'art forms'. FLW's architecture had many groundbreaking elements, but was also a failure in many ways. If it was a program we'd see it as a nicely indented example in COBOL (worthless) instead of 'Art'. This is because programming isn't seen as artistic and is judged primarily on getting the job done. Only if code is as fast and robust as reasonable, do we start looking for elegance.
In the 'art' world, elegant failures are still successful. In the real world, failures are failures, elegant or not.
So, art is something which has no use. People have a hard time seeing art in anything which actually does something useful, and are also willing to overlook nearly any failings in something called art.
And the "near primes" that PGP uses have an astronomically small chance of being non-prime.
Well, it's more that the 'near primes' that PGP uses have an astronomically small chance of having 'small' factors. The number may or may not be prime, but it isn't obviously unprime. (It's not even for instance.)
And generally if the number is non-prime, PGP encryption just plain won't work.
Not quite. A prime number is no different from a non-prime except that it has less factors. PGP can't tell if a given number is prime except by trying to factor it, otherwise it'd simply test all the numbers it generated and this whole 'near prime' thing wouldn't even be talked about.
We can't tell how far ahead of the rest of the world the NSA is. Most of their mathmeticians aren't actually working 9-5 at cracking codes, they're engaged in general research, so they've got many more people working at advancing the field than anyone else. Other mathmeticians have to pay the bills and then work on crypto in their offtime. Also, it's hard to look at the general level of math knowledge and tell how far ahead someone is. We do know that the NSA advised (IBM?) to change DES for a then-unknown reason which was only discovered years later. But, do we know when this was discovered then by the NSA? Perhaps they worked it out years earlier. But perhaps this was discovered mainly by someone curious as to why the S-Box design was dictated by the NSA. If unguided, perhaps it would have been fifty years before that was discovered.
We don't really know what state of the art is for the NSA because they tend not to publish. They also hire a lot of civilian mathmeticians without saying precisely why. Are they perhaps taking the one who seem closest to actually figuring something important out and leaving the non-crypto oriented and the less bright (and the contientous objectors who won't do government work.) to advance common knowledge?
Perhaps prime-based crypto is better to use than one of the newer non-prime algorithms simply because of the ammount of peer review, but we don't know if we're 2% or 95% of the way to being able to trivially factor large numbers, trying to guess where the NSA is seems even harder.
Well, after reading Abrash's _Graphics Programming Black Book_, I think Mel is alive in some form. Carmack touched that for a while with his infamous free FP (For the record, the texturing used fixed point, but they used a float instruction 'for free' to correct for the innacuracies every 16 pixels.) which pretty well became *the* killer game and changed a generation of CPUs. (The AMD k6 was faster than the P5 except for FP, if not for Quake changing the FP is too slow for games mentality, the K6 would have been a kickass gamers machine. As is, we had to wait till the K7 for an x86 compatible to kill the pentium lead.)
And Abrash himself is no mean optimizer. But, my favorite story out of the book was a guy who wrote a program to compile 'life' screens into enough assembly code to compile the next generation and beat the next fastest code by a huge margin. That sounds a bit like a 'Mel' program, esp since Abrash said it took him a few days to understand how it even worked.
Sure, MS Office will never be svelte, but there will always be people who optimize. People who know enough about the machine to write code that flies. Eventually computer speed/storage increases will start to slow down, even a bit, and then we'll see people looking to the old days for hints.
It seems to me that the sheep are the people who listen to the whole thing, in order, because that's the way they're meant to do it.
I listen to music because I want to hear music I enjoy. If 80% of a CD doesn't interest me, why should I listen to it just because it's 'part of the composition'?
The argument is partially valid. Art looks better when matted or framed. The frames/matte could be compared to the extra songs on a CD. But, nobody would say that the frame is as important as the art, just that it helps set it off.
It comes down to, do I know my own tastes, or does some musician know them better than me?
You may find that you can't arrange music in as enjoyable a way as the artists can, but I know my own tastes well enough to program music I want to listen to.
Thus, I'll buy singles of the songs I like, or maybe a package deal on great albums, but once I have them, I'll delete the crap that I don't like. This composition thing is for sheep who can't decide what they like.
Slashdot Mirror
Yes.
They're dangerous inefficient. Letting barely trained people drive polluting hunks of metal at great speed past pedestrians is a dumb idea.
If people paid taxes equal to the cost of urban driving then large cities could have good enough transit systems that people wouldn't need to drive.
Highway driving is fairly expensive. A car with a few people eats up a lot more fuel than those people's seat on even a farily luxurious bus.
Few people actually need cars and with a bit of decent city planning even less would.
Then let people drive outside of cities where they're less likely to drive drunk (people drive home from bars, etc, usually not from state to state, while drunk) and less likely to run off the road in a populated area.
If someone wants a car to drive on private property, or on interstate highways, or in similar situations, then let them. But if they want a car to drive in the city because they can't be bothered with transit, then screw 'em. (I only know a few drivers who I think are good enough to deserve a license for city driving. The rest of them... It'd be like giving guns to kids after a one-week anger management course and a few target-practice sessions.)
Bleh. I've heard this before, that it's important to protect our heritage, and our traditions.
Why?
Think of it like a free market economy, but with culture. If your culture is less attractive than others (stupid restrictive traditions, lame rituals, etc) then your children will likely not follow it, if they have a choice.
The whole push to preserve culture is based on giving people less choice. Limiting the outside influence they can be exposed to, so that they'll do the only think they know.
You can see it in culture, where for instance there are limits on the ammount of foreign TV shows and local artists are subsidised not based on skill, but on quota. You can also see it in religion, where people forbid teaching of other religions or non-religious viewpoints (think Kansas) so that the children will think the way their parents want them to.
It's not like preventing others from speaking makes your viewpoint any more compelling, but if someone is exposed to one viewpoint from everyone around them while they grow up, often they will believe it just because that's all they know.
Culture works in a similar way. You can get someone to practice your culture by smothering other cultures, but if your culture offered something of value, then people would pick it without being forced or deprived of options. And if they don't, then it's probably because your culture is nothing but a collection of useless rituals and stupid beliefs that hold no objective value.
I don't think that saying GNU utillities are based on AT&T Unix utils is incorrect. The code may be completely from scratch, but the names, functions, and usage is similar.
The code is often the smaller part of making a new util. Design, both in identifying the problem, and in deciding on the implementation, are as much of a problem as the actual coding.
This isn't to run down the GNU utils, they are intended to be very similar to Unix tools. If they were vastly different, they wouldn't be as useful.
(Why can't we have a thread that mentions GNU without someone using the term fanatic? They have stated a goal and are working towards that goal, not getting distracted in the meme of the moment isn't fanaticism.)
If you think about it, a keyword list, and a name list, like the white and yellow pages, are the ways to go.
This was worked out long ago by the phone companies. The actual address of a company is (mostly) insignificant and if you make it significant and easy, there aren't enough to go around.
One phone number is mostly the same as another, with some exceptions. The differentiator is the name of the company and their type of business. Trademark law already covers companies naming themselves after competitors trying to steal their business, so whatever system we develop doesn't need to deal with this. All 'we' need to develop is a way for Apple Records and Apple Computer to share the WWW without being unweildly.
Probably the best having a registry that stores the name of a company and a few (two or three) keywords about them, then you simply type in "Apple:Music" and the registry finds all entries with 'Apple' in the name and 'Music' (or synonyms) in the keywords, then automatically your browser replaces what you typed with the url, which except for bookmarking purposes is useless. (Like memorizing an often called phone number)
Then, not having to rely on having the shortest, easiest url, companies could take applecomputer.com or applerecords.com or OrganicAppleOrchardsAssociation.com, wanting it to be descriptive and hard to mistake, similar to a trademarked logo.
Trademark law should be ammended to say that having "Clue" (etc) trademarked in one context, not only does not give protection in other contexts (This is already in there), but specificaly does not give any rights in context-free situations, like clue.com.
Perhaps only the second is trivially replaceable, (ie, simply hexedit the string to a bunch of zeros)...
> 4. The key "locations" were never "known" prior to sp5.
This is not known. I'm sure people poked through MS's DLLs before. And the key did have the characteristic form of a key, as in, that string of bytes did look like a key. What we didn't know until recently was the symbol names that went with the data. (But the actual location and value of the data is *trivial* to rip out of an executable.)
> 5. In exported versions its 40 bit. Not a terribly difficult crack.
The strongest encryption you are allowed to install is 40b. But there are fewer restrictions on the use of encryption in such a way that arbitrary data can't be sent. One example of this is digital signing, such as it's used in the CryptoAPI. Especially when the intended use of the digital signing is to keep people from installing strong crypto in violation of those very export laws. 6. Yes you can replace it with a program, but with the first key to authenticate me as i spoof widgets.com, I can replace it without your knowledge or approval.
No, there's nothing in here that lets you install anything on someone's computer without their approval. They still need to initiate the install (by running the installer, or a trojan of some sort). What the key does is allow the CryptoAPI to check if the crypto package you install is signed by MS, or the NSA, the only two entities (in this case) who are supposed to sign valid crypto components for Windows.
You *could*, with a trojan, replace that key, or simply crack the cryptoAPI to not check, and then allow the installation of unsigned crypto modules. But, this would be silly and redundant to use in an attack.
If you could gain access to someone's computer (via trojan, physical access, etc) then you could simply install a keyboard monitor, or some such, and read all of their data directly, bypassing any crypto. You wouldn't need to hack the crypto module, install broken/weak crypto, then crack it later. And if you did this, the people they tried to communicate wouldn't understand it, because the crypto algorithms wouldn't match.
No, the only '3v1l h4x0r' trick to do here is simply crack *YOUR* CryptoAPI (and let your friends do the same) so that you can easily encrypt data with a strong algorithm and large key. To do this you'd need to either convince MS or the NSA to sign your strong crypto, or remove the check in the CryptoAPI, much the same as crackers everywhere remove shareware expiration checks, etc.
How would they find these unconnected sites to do an analysis of them?
I for instance have a two websites I use to distribute private files and such to friends. But, to the best of my knowledge neither of these has links to it.
They need to recruit a few ISPs, examine all the pages on them (after stripping usernames) to see how many webpages don't have links to or from them , then use that figure to adjust their estimate of the total number of pages.
(They need to get an estimate of the number of pages they can't find by spidering, and the only way to do that is go to the source...)
A few ideas...
Personal settings re other people's posts. Perhaps John Katz bugs me, then I could give his posts a -1 on top of any other points he may get, or if I really liked his usual posts, I could give them a +1 or +2 so that they'd float to the top. This would be sort of like
(On a practical note, few users would use this, and only vs a few other users, so it's not a matter of storing 80k x 80k of settings.)
Instead of having 'Funny +1', perhaps we could seperate the descriptors and the points. Some funny posts are inapproptiately funny and some point out a good point with humor, so for funny to always be a +1 isn't always right. Also, if keywords were seperate it'd be easier to display a list of all the keywords for a message (funny, insightful, etc) which would aid in a filter for inc/excluding posts by keyword.
Also, don't think that the whole point of this moderation is to censor people whose opinions are unpopular. It's also to chop down the number of messages that we have to read. Preferably by selecting those messages which are 'best' by our and community standards.
We also need a way for messages to influence threads they're in. If someone posts a message and then the reply to it is +5, then the original post is probably relevant. This would work better by moderating parent posts up, but not down. (Don't hide a parent because the reply is -1, but do show a 0 post if it generated high scoring replies.)
And we should care because why?
Seriously, if you wish to spend tons of time making something that someone can make a closed-source version of, fine, but don't expect sane people to help.
The proper people to rewrite the system in such a way that they can make a closed-source version of it are the actualy comapnies wishing to do so.
Well, as long as you understand that your closed-source goal is laughable and only a few freaks want it, then fine. Just shut up about the GPL virus. It's that way because we want it that way so take a flying leap.
Translation: "I want to steal your code, if you stop me, I'll whine."
It's free for everyone unless they try to close the source. Can't you get that through your pea-brain? The only restriction is that you MUST release it as open-source and in such a way that future versions are also open-source.
Your right to take code, rewrite a bit, close it, and claim it as your own is one that I'm perfectly willing to take away from you grubby bastards.
Nope. No restrictions at all, for users. If you don't touch the source code, you don't get bothered by the license. The only people restricted are those rewriting the app, and they're only restricted in that they can't close the source.
Users not only are not harmed by the license, but they are unaware of it.
I've got _Beneath Apple Dos_ which I used to write a package of programs to transfer files to and from disk images. (For us emulator people)
I can't part with it because I still use it now and then, but if you need any information from it, I could try to OCR, or retype parts of it.
If anyone has _Beneath Apple Prodos_, I would like to have access to allow me to write in Prodos support for the apps.
(The allow reading/writing files to the disk images, reading/writing dos images, formatting (making) images and/or adding dos images, translating files (cr -> cr/lf, basic bytecode to text, etc) )
If anyone wants a copy, or wants to help development, contact me. I'll throw a GPL on it and we can update it.
And this is intentional.
Actually, I understand the code just has to be made available, either on disk, or on a website, etc.
But, this is to keep people from using the code in their project and denying users the benefits of an opensource application while using someone else's code to write it. It's a good thing.
And the GPL doesn't prevent FreeBSD people from using the GNU utils, as long as they provide the source, etc. They will have problems using parts of the source code if they insist on the BSDL, but if they don't insist on every app having a BSDL license, they're fine.
You're not really dictating the terms of the sublicenses, you're offering a contract to someone provided that they dictate the terms of their license in such a way.
Each person is only the copyright owner of the work they produce and as such is only able to dictate the terms of usage of that work. But, they can offer a contract which gives the right to use the work provided the joint author licenses it in a certain way.
The GPL is Stallman's greatest work. The compiler and the utils, etc are good, but it's only the GPL which made them as usable and perpetually free as they are. He is as much the father of what we think of as Linux as Torvalds is. Stallman made it possible for Torvalds to write Linux in a free environment.
And this is why we fight back with tools designed to keep the OS open for the users, licenses like the GPL, tactics like keeping up with proprietary extensions, etc.
If we sit and watch, people will take what we collectively have made and use it for their own purposes. This is not 'bad' but it can hurt us if we don't watch out and if we let it happen, we'd deserve it.
We need to fight tooth and nail against anyone who uses laws against us, by trying to trademarks names of our products, or patent processes that they use. And we should use every means at our disposal to fight these predators. Use what little legal protection we have (without being able to afford lawyers) like the GPL, mass lobbying of politicians, etc. We should also use other methods, boycotting companies like this, bringing out products similar to theirs specifically to remove their markets where they are attacking us, etc.
As long as our goal is always to keep the software free and open for the users then we should feel justified about throwing their own predatory tactics back at them.
But, probably the best weapon is simply making them depend on us. For example, unless Redhat decided to rewrite all of GNU/Linux without the GPL they need us. Companies that use the OS also need us in the same way that other companies need MS. And the companies that use Linux have a vested interest in keeping it free for users because that also means free for them. We need to set up these symbiotic relationships whenever possible and destroy with any means availble any parasitic relationships.
Sure, but can the CPUs handle it? That's the part of the system that's being used the most.
Not at all. Open means open. It doesn't mean anything else. A door can be open but the area behind it restricted. That doesn't mean the door isn't open, it just means you can't go there.
The doom/hexen/etc source is open in that you can see it. You can't call it your own and sell it, but if you really have to whine when someone won't let you profit from their product, tough shit.
And if you're any good, you can simply see how they did it and write your own game if it's that important.
Greater income to a few artists means less income to the rest and he doubted that those few megastars made up for the creativity lost when the lesser known artists quit.
The assertion is that the recording industry supports the recording industry and a few star artists, not all artists. Thus, the industry will propose law to help itself, not to help the artists. Being that we think the art is good and the industry is irrelevant, we should advocate laws that support the artists.
With DAT copying, you could have a good underground music network, much like with MP3s. If the taxes actually went to the artists whose work was copied instead of the recording industry then we could legitimatize this underground copying because it wouldn't deprive artists of their money.
If you think about it, there are probably more pirated copies of small artists' music than the stars, in relation to the total number of copies sold. This scheme would benefit the small artists because they're the ones who don't have contracts with the recording industry that would put their music in all the stores.
Yes, the computer world is colorblind, and also age, sex, looks, etc, aren't important. But you need to be able to afford a computer and have a minimum of schooling in how to use it.
This is a racial issue in that there are more poor blacks and hispanics than poor whites.
But, dealing with this from a racial angle is completely wrong. Not only does helping someone because of their race pretty well equate with holding others back because of their race. (Would it be racism to give blacks a 50% off coupon for a computer? How about to do it for whites?)
What needs to be done is to target the situation people are in. If the poor can't get good schooling then chances are that they won't be eligible for any decent jobs, or better schooling, etc, and won't know what to do with a computer if the government requires that they be given one.
This is *not* a racial issue. I went to school with plenty of white kids who were destined to live McLives when I transfered into the inner city for my final year of high school.
We need to reach these children somehow. Offering them real choices and the education to make the most of their chances. It's not fair to say that the poor have as good of a chance at winning a scholarship when they have to grow up in a slum and are sick all the time and it's not enough to simply raise welfare payments hoping that it will miraculously make these children live an easier life.
I think advocates do themselves a disservice by making issues like this racial. I would be behind anti-poverty movements, but the minute someone brings race into it, like "Helping poor *black* families" I rebel completely. Treating one race differently from another is what got us into this mess in the first place and repreating that mistake won't help anyone.
So yes, technology is completely color blind. Computers don't care who hits their keys and the users at the other end won't know who you are unless you tell them. They are economically discriminatory because you need to buy them and they're fairly expensive to a family barely getting by as is.
There is a small racial issue here in that there are many poor blacks, but this just clouds the real issue by ignoring all the poor of other races.
They picked INFO because HTML wasn't as widespread at the time and man pages don't allow hyperlinks.
man pages are good, but they're best suited for documenting a specific command, as in "What's that switch for ls to show all files..." When you don't know what the command is called, you don't really want to guess, check a bunch of 'see also' references by loading a new pages, etc..
This is where a HTMLish document is preferable. You select something, check it out, go back, etc. It's suited to putting all of the linux documentation in a big tree and letting you start and browse to what you need. Much more suitable for finding something you don't know the name of.
But, GNU used INFO because there wasn't anything better at the time.
I too think INFO blows. HTML is easier, has more viewers, is very platform independent.
I haven't used DocBook but unless it comes with an HTML version of the documents, it won't be as useful.
I do a lot of my linux work via telnet from windows machines, or in win32 with DJGPP (from work) and unless I want to open a new telnet window just to view docs, I need to have them in an easy to view format, like HTML. And HTML is nicely suited to local or remote viewing.
Yes. Of a sort. It wouldn't appeal to everyone in the same way that pencil sketches often don't appeal, no matter how good they actually are, to people used to full color oil paintings.
Mel's program would be sort of like the hidden art of Bev Dolittle (?) where there's a simple nature scene and hidden in it, if you know where to look, are ten wolves, or a man on horseback, etc. Or maybe it's be like some piece of kinetic sculpture that appears to be out of an MC Escher book.
But yes, he wrote a program in ways other people have problems understanding, and it worked faster and better because of this, so yes, there is art in that.
Not that I'd want it hanging on my wall, for two reasons. One, I usually like big colorful things, regardless of actual artistic merit (ie, sat. pics of the area, N.G. Maps of the world, etc) and two, while it may have been artistic, that just means that everything Mel did was probably that way and this was just the minor job that was later given to someone untrained enough to have to wrestle with it yet enlightened enough to be able to grok it with study. This is probably as artistic to Mel as the canvas Rembrandt cleaned his brushes on.
Almost anything we do can be done quickly and haphazardly, or carefully and elegantly. What makes one elegant work art and another not? Functionality.
:) No, so it's art. Can you use a well-designed bridge? Yes, so it's 'just' a bridge.
Most people are pompous asses and refuse to call something art if it has the slightest use. Can you use a dead sheep in a tank except as a footrest (or sex object for programmers.
Code is one of these functional items. Except perhaps for tiny snippets in a textbook, we expect code to work. The most beautiful 'looking' code is totally worthless if it doesn't work, and well, and people would praise line-noise if it was an efficient kernel when run in intercal (not that this is likely...)
Contrast this to 'art forms'. FLW's architecture had many groundbreaking elements, but was also a failure in many ways. If it was a program we'd see it as a nicely indented example in COBOL (worthless) instead of 'Art'. This is because programming isn't seen as artistic and is judged primarily on getting the job done. Only if code is as fast and robust as reasonable, do we start looking for elegance.
In the 'art' world, elegant failures are still successful. In the real world, failures are failures, elegant or not.
So, art is something which has no use. People have a hard time seeing art in anything which actually does something useful, and are also willing to overlook nearly any failings in something called art.
Well, it's more that the 'near primes' that PGP uses have an astronomically small chance of having 'small' factors. The number may or may not be prime, but it isn't obviously unprime. (It's not even for instance.)
And generally if the number is non-prime, PGP encryption just plain won't work.
Not quite. A prime number is no different from a non-prime except that it has less factors. PGP can't tell if a given number is prime except by trying to factor it, otherwise it'd simply test all the numbers it generated and this whole 'near prime' thing wouldn't even be talked about.
We can't tell how far ahead of the rest of the world the NSA is. Most of their mathmeticians aren't actually working 9-5 at cracking codes, they're engaged in general research, so they've got many more people working at advancing the field than anyone else. Other mathmeticians have to pay the bills and then work on crypto in their offtime. Also, it's hard to look at the general level of math knowledge and tell how far ahead someone is. We do know that the NSA advised (IBM?) to change DES for a then-unknown reason which was only discovered years later. But, do we know when this was discovered then by the NSA? Perhaps they worked it out years earlier. But perhaps this was discovered mainly by someone curious as to why the S-Box design was dictated by the NSA. If unguided, perhaps it would have been fifty years before that was discovered.
We don't really know what state of the art is for the NSA because they tend not to publish. They also hire a lot of civilian mathmeticians without saying precisely why. Are they perhaps taking the one who seem closest to actually figuring something important out and leaving the non-crypto oriented and the less bright (and the contientous objectors who won't do government work.) to advance common knowledge?
Perhaps prime-based crypto is better to use than one of the newer non-prime algorithms simply because of the ammount of peer review, but we don't know if we're 2% or 95% of the way to being able to trivially factor large numbers, trying to guess where the NSA is seems even harder.
Well, after reading Abrash's _Graphics Programming Black Book_, I think Mel is alive in some form. Carmack touched that for a while with his infamous free FP (For the record, the texturing used fixed point, but they used a float instruction 'for free' to correct for the innacuracies every 16 pixels.) which pretty well became *the* killer game and changed a generation of CPUs. (The AMD k6 was faster than the P5 except for FP, if not for Quake changing the FP is too slow for games mentality, the K6 would have been a kickass gamers machine. As is, we had to wait till the K7 for an x86 compatible to kill the pentium lead.)
And Abrash himself is no mean optimizer. But, my favorite story out of the book was a guy who wrote a program to compile 'life' screens into enough assembly code to compile the next generation and beat the next fastest code by a huge margin. That sounds a bit like a 'Mel' program, esp since Abrash said it took him a few days to understand how it even worked.
Sure, MS Office will never be svelte, but there will always be people who optimize. People who know enough about the machine to write code that flies. Eventually computer speed/storage increases will start to slow down, even a bit, and then we'll see people looking to the old days for hints.
Did you post it as HTML or 'Plain Old Text'?
Look at the selector box when you post again, that's probably your problem.
It seems to me that the sheep are the people who listen to the whole thing, in order, because that's the way they're meant to do it.
I listen to music because I want to hear music I enjoy. If 80% of a CD doesn't interest me, why should I listen to it just because it's 'part of the composition'?
The argument is partially valid. Art looks better when matted or framed. The frames/matte could be compared to the extra songs on a CD. But, nobody would say that the frame is as important as the art, just that it helps set it off.
It comes down to, do I know my own tastes, or does some musician know them better than me?
You may find that you can't arrange music in as enjoyable a way as the artists can, but I know my own tastes well enough to program music I want to listen to.
Thus, I'll buy singles of the songs I like, or maybe a package deal on great albums, but once I have them, I'll delete the crap that I don't like. This composition thing is for sheep who can't decide what they like.