It's a
leash, and it takes away the justifiable excuse ("I wasn't near a phone") that essentially gives us the last vestige of
privacy and solitude in our lives.
That is very well said!
Here's to a cell-free life.
People who have a need to get in contact with me generally know one or more of my e-mail addresses. When I explicitly need to be reachable by phone, I stick around one, be it at home, at work, or wherever else, and make sure that those who may want to get in touch and that I am willing to hear from know where I am. Also, anybody who has good enough a reason to "invade" my use of my time at the weirdest moments, can always call back (half) an hour or so later if they didn't get hold of me the first time. And as I wrote already, I will carry a cellphone when I consider that to be useful. But that's a rare occasion for which I refuse to buy my own that I then "cannot get rid of anymore". I strongly prefer e-mail (or, of course, direct face-to-face communication) over using the phone.
You see, the problem is not about needing devices to be reachable. The problem is people assuming that everyone has to carry these devices anyway and thus is - and by illogical but very commonly used extension has to be - "permanently" reachable, which they consider to be A Good Thing. Well, it isn't one.
A lot of people here bitching about cell phones are probably those who have fairly regular locations (normal office hours, then home).
I can't speak for the other bitchers, but this definitely is not the case for me. I'm at home about 11 hours a day (including weekends). 7-8 of those are put aside for sleeping, during 1-2 of them I'm on the net (which in my case locks up my phone). In addition, my working hours are shifted by about 2-3 hours compared to those of the average person, so people will indeed have a hard time catching my on the phone (definitely when at home). Finally, the only people who know my office phone number are 1) immediate collegues, 2) my parents.
Mankind survived for a long time and produced some amazing things without everyone being reachable all the time.
I work for a micro-electonics research institute. One of our many activities actually is making the implementation of ever smarter and feature-rich cellphones and similar devices ever more easy. Even worse, my very own project is about designing for low power from the system level downwards. One could say we're part of the cell phone companies pipe dreams. (Actually, my project worked closely with one of the major cell phone companies in the past, and now another one is very intersted.) All that just to make very clear that I'm not oposed to the technology for the technology's sake. But neither am I in favour of it "just because".
I will personally *never* be caught having my own cell phone. I will carry/use one if the job that I'm doing at that very moment requires that I be reachable while away from any fixed phone system (which happens maybe once per year), but I flat out *refuse* to give in to the "But sir, you have to be reachable, don't you?" pressure. *I* am the one who decides when and where I want to be reachable. And when I've decided that I'm not to be reached, I will implement that very strictly. Now, I know that one can switch off those buggers when one doesn't want to be disturbed, but that is not the same thing: simply by always carrying that thing around, one creates that expection that one be reachable. Maybe not immediately, but definitely within the hour. People then just assume that they can interrupt your life at any moment, because "Hey, what else (s)he's got that cellphone for, afterall?". Then when you diseble it for more than one or two hours on end, they look at you like you're the bad guy/gall who prevented them from doing something "important" such as telling you they ran into Joe or Mary on the way to the bakery. As if that kind of chit-chat can't wait till next time you really see each other. If by then it's still worthy of being told at all, that is.
Also concerning the "but you have to be reachable" craze: Once upon a time my phone company "discovered" that I use the internet a lot when at home. This is over a plain old dial-up modem, so they figured that "he's got to be reachable, so lets enable our nice (and paying!) mailbox service for him". Now there is some poor helpdesk guy over there who probably still has not recovered from what befell him after I found out what they had done and got in touch to get it disabled again. They charge the person who calls you for leaving the message, they charge you *again* for listening to the recording, and then they charge one of you *yet again* when you finnally do get to speak to one another on the phone? Not with me. Not in a million years.
If all that makes me a social outcast, than so be it.
Noone claimed that Linus invented the technique (besides, Linux too has been using it for some time already). The only claim made about him is that he proposed that the way in which Linux uses it be changed. (It would help to actually read the article.)
Sounds nice, but the last thing that I need is even more people telling me not to spam them about diplomas from prestigious non-accredited universities (as if such a things existed...). It's not me who is spamming them, it's some !@#$%^&* idiot *ssh*le out there who is illegally abusing my e-mail address. Grrr...
ACTING like you're smarter than everyone else is socially inept. BEING smarter than everyone else is a given for those to whom it applies. One should not have to hide one's brains in order to survive. Especially not at school, since part of the goal of attending school precisely is to develop them.
I remember being praised by teachers for achieving scores that none of the others were even remotely capable of. There is NO WAY IN HELL that someone like me should have to deliberately turn in wrong answers just so as not to "be different".
Bullying someone because (s)he knows more than you do is the other thing that is socially inept.
That may be a valid reason, but if he cannot come up with better annotations that what he used this time, I move that the underlying memo was not worth the trouble.
Sometimes I get the impression that ESR has painted himself into a corner with these Halloween documents. The first two were absolutely worthwhile, but as time goes on he seems to feel obliged to produce follow-ups at almost regular intervals (advance notice for the trolls: please don't take that literally), whether or not he's got something substantial to comment about. All in all, I think he is doing both himself and a lot of others a considerable disservice with that. When promoting Linux at work, for instance, I do not want to be confronted with "Look at how childish these Linux zealots are. How can we ever entrust our valuable data to software produced by such people." argumuents. Yes such arguments are silly. Yes, they can be debunked. But every minute doing the latter is a minute not spent on promoting Linux.
In any large software system, the truly unique code probably accounts for about 1% of the source.
Excuse me??? Was that measured with a stick that says "0, 1, 2, many"? It has to be, because there is no way that this statement can be true of real code. What's more, if there is a piece of code where all the key stuff makes just up 1% of the total and can be cut out that easily, it is a badly designed piece of bloatware! It's the 99% that should be ripped out, for $deity's sake!
Besides, a good overall design most often is part of the uniqueness and value of really good programs. Never mind that one key algorithm that could be removed by replacing it with a stub, it's often the overall structure that really is worth something to a competitor.
This guy needs to get a life. No, wait... he probably has a life and that is exactly what his problem is. He needs to do some real design and coding for a while. I'm not sure I would hire him for such a job, though, because judging by his "key algorithm" statement he never got past functional decomposition in his "programming 101" course in kindergarten.
I really can't stand MS' My Sh*t idiocy. But in itself the My bit it isn't even the worst part of it, so if they were to support "Joe's Documents", things wouldn't be any better than they are now.
The worst bit is that Microsoft has decided that as a user of their software I must group, store, and retrieve my stuff in classes called documents, pictures, music, movies, and whatnot.
Apparantly MickeySoft thinks no one would ever want to group his or her files based on how their semantics relate to each other. No, no, no! The Only True Way of using your (oops: My) computer is to store all doc files in one folder, all mp3's in another, all...
Just look what happens with Office: you navigate to a directory, double click on a document to fire up Word and then decide to use Word's "File Open" to open another document sitting right next to the first one. Instead of offering you a list of files in the directory where you started out, Word insists on presenting the "My Documents" folder. Which of course is empty most of the time, or at least should be so if you're a somewhat organised mind. After some cursing and a lot of redundant clicking, you want to export a figure from one of the documents and put it in the same directory. But what does Paint do? Sure enough, it takes you to My Pictures, because after all, it is only natural to store a computer flow charts together with the plans for your dream house as well as your holiday fotos. Right?
I frequently make all sorts of changes to documents just to see if an idea might work. I do not want to have to explicitly undo these if I decide to toss the idea later on.
Or, when doing something like merging 2 or more PowerPoint presentations (Oh, the horror!), I always gradually strip down (one of) the source document(s) as work progresses, since this makes it easier to keep track of what still has to be done. I do not want to have to undo all that just to preserve the originals.
And I definitely do not want to have to undo the last 35 modifications out of a list of 45, thereby loosing lost of time to make sure that I do not undo one too many. I want to save the document when I think I have something worthwhile, and I want the ability to go back there with minimal effort.
Basically, my point is that consider 1) myself being clever enough to known what I want to do with my document; 2) the computer being nowhere near clever enough for it to start outsmarting me. This is, actually, the one thing that I dislike most about typical Windows software: the fact that it has a very bad tendency to treat the user like a moron who doesn't know what is best for him/her.
Even if you just succeed in breaking the shell into pieces and due to some magic it does not explode, the pieces will not end up at the original target as designed. First of all, their trajectories and speeds will diverge. Next, shells are designed to do their nasty job in very specific ways (they have care- and purposefully designed geometries, windscreens, armour piercing caps, fuze delays,...). If these things do not arrive as intended, their effect will be greatly reduced and sometimes even nullified. Hell, even a 1 degree change in impact obliquity can make the difference between piercing an armoured plate or bouncing off (for otherwise identical and intact shells).
In terms of speed, the Sun compiler is OK. But in terms of standard compliance and absence of bugs, I found it to be lagging badly. But even so, I use it to port to Sun, because until recently the other alternatives were: g++ 2.95 (sllooowwww) and g++ 3.x for x < 2 (all versions I've tried failed to compile certain multiple inheritance constructs correctly). I believe g++ 3.2 finally fixes the latter issue, but I've also come to the conclusion that I will not change anything until doing so has become unavoidable.
So here's what I do:
use the native compiler on each platform: aCC on HP-UX, CC on Sun, and g++ on Linux (note: I have to use a RedHat 6 machine, because that 2.96 nightmare also has the MI bug!);
use g++ (2.95) as an additional check (parsing only), since no single compiler catches everything and this is one of the things g++ -Wall is relatively good at.
My primary development compiler is HP's aCC, even thouh I have a recent Linux desktop box that can run circles around the aging HP server that I have to use. The reason is very simple: of all C++ compilers that I have used till now, aCC is the best one by far, and g++ has given me the biggest amount of trouble over the years.
Fortran (the WATFIV dialect/system, actually) was my first and only programing language when I came to the univ back in the 1983-85 timeframe. During my first year we even still had to use punched cards. We were supposed to be doing some very small assignments but were each given a certain number of extra "computer points" that we could spend as we saw fit. I don't remember the details, but it went along the lines of: each compilation run costs X points per Y cards; each second of CPU time consumed Z points; each sheet of paper printed costs 1 point; etc.
Me not being satisfied with stupid the kind of little toy programs that my classmates were getting excited about, I opted to burn my points on a "game" program (yes, with punched cards!). The idea was that it first generated a 2D maze and then interpreted any remaining input cards as commands from a human player who had to find his way out. The program would comment on how well the player performed thereby revealing bits of info about the maze. The player was than supposed to rerun it with an other set of comands based on what (s)he had found out. Unfortunately, I only managed to actually run (not just compile) this thing about 5 times before my points were all used up. So I never got all the bugs out. But I still have all the cards!
Having learned from that experience, I used my second year extra points not to program more useless Fortran stuff, but to print out the entire set of manuals for CP and VM/CMS. That gave me a lot of very useful material to find out how OS-es worked "as seen from the outside". And I still have all those listings too!
After that, I opted for CS major and Fortan disappeared from the radar screen except in one class about numerical approximation and the like. That was in somewhere in 1986-88.
All the companies are looking for cures, because if they're the first with a cure for a disease they'll make *billions* of dollars. There are always moer diseases coming.
But the first one to invent a treatment that contains the symptoms as opposed to curing the disease can make many more *billions* and in addition can do so for a very much longer period of time.
On a global scale, even antibiotics (considered "a cure" in most of this discussion) are just a treatment. Curing the world of a disease means to eradicate it. Once you have done that, there's no more money to be made from it. Yes, it has been done, but that's besides the point. The point is that, economically speaking, eradicting is a bad idea if you can make people happy by just treating the symptoms.
Antidepressants are the prime example of all this.
Re:Version number abuse
on
Linux Kernel 3.0?
·
· Score: 2, Interesting
Argh! The first digit in the kernel version number was always meant to indicate the ABI version!
Where did you get that idea? It's simply not true, and never has been. The increase from 1.x to 2.x was to signal the addition of SMP capability. The kernel ABI's change quite regularly during a development series. Usually in such ways as to remain compatible, but compatible does not imply contstant and every so often things are broken (think about some of the problems with some well known binary-only modules for some examples).
Just one datapoint: We just migrated some 60-odd desktop systems to Linux from HP-UX, and are happy campers. Other divisions of our company are now looking into doing the same. Overall, we're a (roughly) 1100 employee company, from which an estimated (by me, here and now) 300 can become Linux users without much problems.
Of course, we operate in the EDA research business (and related areas), so we're atypical and many people around here very much prefer anything UNIX-like over The Other Operating System. But still... Less than two years ago Linux was still a big No-No as far as the head of IT was concerned, even though several unofficial system already existed and the presure to officially support Linux was on already.
Given the structure of the google cookie, it's quite easy: all you need to do, is to replace the ID number with a randomly generated one that has that correct amount of digits. A tiny bit of perl will help. That and some lock file checking, obviously, since it makes no sense to do this while Mozilla is up and running.
With cookies from other sites things can obviously be more difficult or even impossible. There's no generic method that applies to all of them.
Something that does easily apply to all sites, is periodic automatic filtering based on black/white lists (ideally based on regular expressions such as ^ads\..*, something that Mozilla's blocker cannot do at the moment). I must have accumulated a gazillion different profiles in some tracking databases by now. All of them looking like a clueless newbie that surfed the net once for a few hours, visited just a few sites, and then disappeared from the face of the planet.
Mind you, if Mozilla's cookie manager would support regular expressions, I'd most likely throw my script into the dustbin, but since it doesn't...
Sharing it as-is doesn't make a lot of sense. I wrote it to work in my settings, and didn't at all try to be generic. But since it fits into less than 200 lines of Korn shell and a Perl version could probably be done in (a lot) less than half as much, it shouldn't be too hard to reproduce.
I have more stuff than just this google cookie thing. It all got created more for fun than for anything else, but now that I have it, why not use it to annoy some of those cookie monsters out there?
Regarding disclosing the idea: maybe I should have patented it... Including some vague "method and aparatus to circumvent any cookie forgery method and aparatus based on the method and aparatus described in the auther's earlier patent no. 123456789".:-):-)
Seriously, the idea isn't new. I saw it mentioned on/. before by someone else (after I created my script, so that patent could still become mine:-). Besides, if google track the performance of their software, they'll get a tens of hits with IDs that are not in their database on a daily basis from me alone. Tens of reports is peanuts compared to the amount of hits they get overall, but a lot compared to the 0 "ID not found" cases they should be getting if everything went as expected. If I were them, I'd eventually look for a bug somewhere and not being able to find one I'd soon understand what's going on even without reading/. all day. Especially since the ID probably has some internal structure and/or logic that I do not know and thus cannot mimic.
If they devise an alternative that makes this thing fail, they'll just end up in my Mozilla cookie blocking list. So I don't mind.
I even have a setup that replaces my google cookie with a randomly generated one every 6 hours. With Mozilla I could also just block it, but I created this thing back when I was still using NS4, and see no reason to change it.
Slashdot Mirror
That is very well said! Here's to a cell-free life.
Amen to that!
You see, the problem is not about needing devices to be reachable. The problem is people assuming that everyone has to carry these devices anyway and thus is - and by illogical but very commonly used extension has to be - "permanently" reachable, which they consider to be A Good Thing. Well, it isn't one.
I can't speak for the other bitchers, but this definitely is not the case for me. I'm at home about 11 hours a day (including weekends). 7-8 of those are put aside for sleeping, during 1-2 of them I'm on the net (which in my case locks up my phone). In addition, my working hours are shifted by about 2-3 hours compared to those of the average person, so people will indeed have a hard time catching my on the phone (definitely when at home). Finally, the only people who know my office phone number are 1) immediate collegues, 2) my parents. Mankind survived for a long time and produced some amazing things without everyone being reachable all the time.
PS: Sorry about all the typos. I accidently hit the submit button iso the preview one..
No you are not the only one.
I work for a micro-electonics research institute. One of our many activities actually is making the implementation of ever smarter and feature-rich cellphones and similar devices ever more easy. Even worse, my very own project is about designing for low power from the system level downwards. One could say we're part of the cell phone companies pipe dreams. (Actually, my project worked closely with one of the major cell phone companies in the past, and now another one is very intersted.) All that just to make very clear that I'm not oposed to the technology for the technology's sake. But neither am I in favour of it "just because".
I will personally *never* be caught having my own cell phone. I will carry/use one if the job that I'm doing at that very moment requires that I be reachable while away from any fixed phone system (which happens maybe once per year), but I flat out *refuse* to give in to the "But sir, you have to be reachable, don't you?" pressure. *I* am the one who decides when and where I want to be reachable. And when I've decided that I'm not to be reached, I will implement that very strictly. Now, I know that one can switch off those buggers when one doesn't want to be disturbed, but that is not the same thing: simply by always carrying that thing around, one creates that expection that one be reachable. Maybe not immediately, but definitely within the hour. People then just assume that they can interrupt your life at any moment, because "Hey, what else (s)he's got that cellphone for, afterall?". Then when you diseble it for more than one or two hours on end, they look at you like you're the bad guy/gall who prevented them from doing something "important" such as telling you they ran into Joe or Mary on the way to the bakery. As if that kind of chit-chat can't wait till next time you really see each other. If by then it's still worthy of being told at all, that is.
Also concerning the "but you have to be reachable" craze: Once upon a time my phone company "discovered" that I use the internet a lot when at home. This is over a plain old dial-up modem, so they figured that "he's got to be reachable, so lets enable our nice (and paying!) mailbox service for him". Now there is some poor helpdesk guy over there who probably still has not recovered from what befell him after I found out what they had done and got in touch to get it disabled again. They charge the person who calls you for leaving the message, they charge you *again* for listening to the recording, and then they charge one of you *yet again* when you finnally do get to speak to one another on the phone? Not with me. Not in a million years.
If all that makes me a social outcast, than so be it.
Noone claimed that Linus invented the technique (besides, Linux too has been using it for some time already). The only claim made about him is that he proposed that the way in which Linux uses it be changed. (It would help to actually read the article.)
Computer break in is not the issue here. Identity theft is the issue.
Sounds nice, but the last thing that I need is even more people telling me not to spam them about diplomas from prestigious non-accredited universities (as if such a things existed...). It's not me who is spamming them, it's some !@#$%^&* idiot *ssh*le out there who is illegally abusing my e-mail address. Grrr...
ACTING like you're smarter than everyone else is socially inept. BEING smarter than everyone else is a given for those to whom it applies. One should not have to hide one's brains in order to survive. Especially not at school, since part of the goal of attending school precisely is to develop them.
I remember being praised by teachers for achieving scores that none of the others were even remotely capable of. There is NO WAY IN HELL that someone like me should have to deliberately turn in wrong answers just so as not to "be different".
Bullying someone because (s)he knows more than you do is the other thing that is socially inept.
Stroustrup is often quoted for having written: "Library design is language design".
That may be a valid reason, but if he cannot come up with better annotations that what he used this time, I move that the underlying memo was not worth the trouble.
Sometimes I get the impression that ESR has painted himself into a corner with these Halloween documents. The first two were absolutely worthwhile, but as time goes on he seems to feel obliged to produce follow-ups at almost regular intervals (advance notice for the trolls: please don't take that literally), whether or not he's got something substantial to comment about. All in all, I think he is doing both himself and a lot of others a considerable disservice with that. When promoting Linux at work, for instance, I do not want to be confronted with "Look at how childish these Linux zealots are. How can we ever entrust our valuable data to software produced by such people." argumuents. Yes such arguments are silly. Yes, they can be debunked. But every minute doing the latter is a minute not spent on promoting Linux.
Excuse me??? Was that measured with a stick that says "0, 1, 2, many"? It has to be, because there is no way that this statement can be true of real code. What's more, if there is a piece of code where all the key stuff makes just up 1% of the total and can be cut out that easily, it is a badly designed piece of bloatware! It's the 99% that should be ripped out, for $deity's sake!
Besides, a good overall design most often is part of the uniqueness and value of really good programs. Never mind that one key algorithm that could be removed by replacing it with a stub, it's often the overall structure that really is worth something to a competitor.
This guy needs to get a life. No, wait... he probably has a life and that is exactly what his problem is. He needs to do some real design and coding for a while. I'm not sure I would hire him for such a job, though, because judging by his "key algorithm" statement he never got past functional decomposition in his "programming 101" course in kindergarten.
I really can't stand MS' My Sh*t idiocy. But in itself the My bit it isn't even the worst part of it, so if they were to support "Joe's Documents", things wouldn't be any better than they are now.
...
The worst bit is that Microsoft has decided that as a user of their software I must group, store, and retrieve my stuff in classes called documents, pictures, music, movies, and whatnot.
Apparantly MickeySoft thinks no one would ever want to group his or her files based on how their semantics relate to each other. No, no, no! The Only True Way of using your (oops: My) computer is to store all doc files in one folder, all mp3's in another, all
Just look what happens with Office: you navigate to a directory, double click on a document to fire up Word and then decide to use Word's "File Open" to open another document sitting right next to the first one. Instead of offering you a list of files in the directory where you started out, Word insists on presenting the "My Documents" folder. Which of course is empty most of the time, or at least should be so if you're a somewhat organised mind. After some cursing and a lot of redundant clicking, you want to export a figure from one of the documents and put it in the same directory. But what does Paint do? Sure enough, it takes you to My Pictures, because after all, it is only natural to store a computer flow charts together with the plans for your dream house as well as your holiday fotos. Right?
What ARE they thinking???
I frequently make all sorts of changes to documents just to see if an idea might work. I do not want to have to explicitly undo these if I decide to toss the idea later on.
Or, when doing something like merging 2 or more PowerPoint presentations (Oh, the horror!), I always gradually strip down (one of) the source document(s) as work progresses, since this makes it easier to keep track of what still has to be done. I do not want to have to undo all that just to preserve the originals.
And I definitely do not want to have to undo the last 35 modifications out of a list of 45, thereby loosing lost of time to make sure that I do not undo one too many. I want to save the document when I think I have something worthwhile, and I want the ability to go back there with minimal effort.
Basically, my point is that consider 1) myself being clever enough to known what I want to do with my document; 2) the computer being nowhere near clever enough for it to start outsmarting me. This is, actually, the one thing that I dislike most about typical Windows software: the fact that it has a very bad tendency to treat the user like a moron who doesn't know what is best for him/her.
In terms of speed, the Sun compiler is OK. But in terms of standard compliance and absence of bugs, I found it to be lagging badly. But even so, I use it to port to Sun, because until recently the other alternatives were: g++ 2.95 (sllooowwww) and g++ 3.x for x < 2 (all versions I've tried failed to compile certain multiple inheritance constructs correctly). I believe g++ 3.2 finally fixes the latter issue, but I've also come to the conclusion that I will not change anything until doing so has become unavoidable.
So here's what I do:
My primary development compiler is HP's aCC, even thouh I have a recent Linux desktop box that can run circles around the aging HP server that I have to use. The reason is very simple: of all C++ compilers that I have used till now, aCC is the best one by far, and g++ has given me the biggest amount of trouble over the years.
Some moons ago, Linus has been noted saying that the 2.x.100+ numbers were too large for his taste, tough.
Fortran (the WATFIV dialect/system, actually) was my first and only programing language when I came to the univ back in the 1983-85 timeframe. During my first year we even still had to use punched cards. We were supposed to be doing some very small assignments but were each given a certain number of extra "computer points" that we could spend as we saw fit. I don't remember the details, but it went along the lines of: each compilation run costs X points per Y cards; each second of CPU time consumed Z points; each sheet of paper printed costs 1 point; etc.
Me not being satisfied with stupid the kind of little toy programs that my classmates were getting excited about, I opted to burn my points on a "game" program (yes, with punched cards!). The idea was that it first generated a 2D maze and then interpreted any remaining input cards as commands from a human player who had to find his way out. The program would comment on how well the player performed thereby revealing bits of info about the maze. The player was than supposed to rerun it with an other set of comands based on what (s)he had found out. Unfortunately, I only managed to actually run (not just compile) this thing about 5 times before my points were all used up. So I never got all the bugs out. But I still have all the cards!
Having learned from that experience, I used my second year extra points not to program more useless Fortran stuff, but to print out the entire set of manuals for CP and VM/CMS. That gave me a lot of very useful material to find out how OS-es worked "as seen from the outside". And I still have all those listings too!
After that, I opted for CS major and Fortan disappeared from the radar screen except in one class about numerical approximation and the like. That was in somewhere in 1986-88.
I haven't touched it since.
All the companies are looking for cures, because if they're the first with a cure for a disease they'll make *billions* of dollars. There are always moer diseases coming.
But the first one to invent a treatment that contains the symptoms as opposed to curing the disease can make many more *billions* and in addition can do so for a very much longer period of time.
On a global scale, even antibiotics (considered "a cure" in most of this discussion) are just a treatment. Curing the world of a disease means to eradicate it. Once you have done that, there's no more money to be made from it. Yes, it has been done, but that's besides the point. The point is that, economically speaking, eradicting is a bad idea if you can make people happy by just treating the symptoms.
Antidepressants are the prime example of all this.
Argh! The first digit in the kernel version number was always meant to indicate the ABI version!
Where did you get that idea? It's simply not true, and never has been. The increase from 1.x to 2.x was to signal the addition of SMP capability. The kernel ABI's change quite regularly during a development series. Usually in such ways as to remain compatible, but compatible does not imply contstant and every so often things are broken (think about some of the problems with some well known binary-only modules for some examples).
The Kodak one is a indeed single chip one. For more details, see this article at Silicon Strategies.
Just one datapoint: We just migrated some 60-odd desktop systems to Linux from HP-UX, and are happy campers. Other divisions of our company are now looking into doing the same. Overall, we're a (roughly) 1100 employee company, from which an estimated (by me, here and now) 300 can become Linux users without much problems.
Of course, we operate in the EDA research business (and related areas), so we're atypical and many people around here very much prefer anything UNIX-like over The Other Operating System. But still... Less than two years ago Linux was still a big No-No as far as the head of IT was concerned, even though several unofficial system already existed and the presure to officially support Linux was on already.
Given the structure of the google cookie, it's quite easy: all you need to do, is to replace the ID number with a randomly generated one that has that correct amount of digits. A tiny bit of perl will help. That and some lock file checking, obviously, since it makes no sense to do this while Mozilla is up and running.
With cookies from other sites things can obviously be more difficult or even impossible. There's no generic method that applies to all of them.
Something that does easily apply to all sites, is periodic automatic filtering based on black/white lists (ideally based on regular expressions such as ^ads\..*, something that Mozilla's blocker cannot do at the moment). I must have accumulated a gazillion different profiles in some tracking databases by now. All of them looking like a clueless newbie that surfed the net once for a few hours, visited just a few sites, and then disappeared from the face of the planet.
Mind you, if Mozilla's cookie manager would support regular expressions, I'd most likely throw my script into the dustbin, but since it doesn't...
Sharing it as-is doesn't make a lot of sense. I wrote it to work in my settings, and didn't at all try to be generic. But since it fits into less than 200 lines of Korn shell and a Perl version could probably be done in (a lot) less than half as much, it shouldn't be too hard to reproduce.
I have more stuff than just this google cookie thing. It all got created more for fun than for anything else, but now that I have it, why not use it to annoy some of those cookie monsters out there?
:-) :-)
/. before by someone else (after I created my script, so that patent could still become mine :-). Besides, if google track the performance of their software, they'll get a tens of hits with IDs that are not in their database on a daily basis from me alone. Tens of reports is peanuts compared to the amount of hits they get overall, but a lot compared to the 0 "ID not found" cases they should be getting if everything went as expected. If I were them, I'd eventually look for a bug somewhere and not being able to find one I'd soon understand what's going on even without reading /. all day. Especially since the ID probably has some internal structure and/or logic that I do not know and thus cannot mimic.
Regarding disclosing the idea: maybe I should have patented it... Including some vague "method and aparatus to circumvent any cookie forgery method and aparatus based on the method and aparatus described in the auther's earlier patent no. 123456789".
Seriously, the idea isn't new. I saw it mentioned on
If they devise an alternative that makes this thing fail, they'll just end up in my Mozilla cookie blocking list. So I don't mind.
I even have a setup that replaces my google cookie with a randomly generated one every 6 hours. With Mozilla I could also just block it, but I created this thing back when I was still using NS4, and see no reason to change it.