By focusing on the PEOPLE USING THE TOOLS, you get to the root of the problem.
the US government is attempting to focus on terrorists. Do you think they are not?
Remember - if terrorists followed laws, we wouldn't have to worry about them.
if private use of back-door-less encryption is made illegal in the US, then the government could more easily
look inside of emails to see that they do not contain terrorist instructions
or focus on whomever is violating this law.
Now, I'm sure you would object to such a law, but it works to accomplish the goal you've outlined. You have not given much reasoning to convince someone who thinks this is a good idea that it is not.
I'm sick of all the knee-jerk anti-US government sentiment I see here. It wasn't too long ago that everybody here was calling for massive "email disobedience" to overwhelm Echelon and/or Carnivore... do you think this is a good in today's climate? should all Slashdotter's right now be trying to swamp the NSA and FBI servers with bogus terrorist message references in routine emails to their friends? If you really believe what you are saying, I'd say you think we should...
you're getting all sorts of plaudits for what you wrote, but it's a piece of crap. you clearly support the majority opinion on slashdot, that's why the slashbots modded you up, but I'm not clear on what exactly is your point. Aircraft, plastic explosives, and several of the other "inanimate objects" on your list are currently heavily regulated, precisely because they are believed by legislative majorities to be unsafe if used improperly. What are you saying?
Are you saying these things should have no regulation?
or are you saying that encryption should be regulated the way these things are?
or are you saying that everything is just fine the way it is with a mix of regulated and unregulated.
I ask because you didn't actually say anything at all as it applies to reality. "Starting down the road of outlawing inanimate objects that can be used for multiple purposes"... is exactly where we've been for hundreds of years, and I kind of like living here so I'm finding it a very satisfying experience. Sure, I don't agree with all regulations, but I can't figure out what you are proposing...
you would find plenty of places where buf is
passed in as an argument so that 'n' is not locally known.
a passed in buffer is exactly where it is dangerous. code that is implemented as a sprintf is generally code that is concerned with user i/o, == chance for a malicious exploit. otherwise, what the hell is sprintf being used for? who would implement something in terms of sprintf? but that is an unimportant point.
in 99% of such code, and I mean that 99% because I clean up code like this all the time, it is simple to programmatically determine the buffer size from the caller's declaration, and programmatically correct it by adding an n argument. you don't need to solve the general case; programmer's each have their own style and they follow it pretty reliably and generally not too many programmers will be involved in coding an app. [hint: don't do it programmatically in C; try perl]
and yes, you do need to test it, but guess what? the only bugs you will encounter will be from actual buffer overruns. though the code may have worked before because the stack was "free":) you will now sleep the sleep of the angels realizing that you've fixed it.
I speak from experience. people are always so scared of cleanup because it could theoretically be bad. and it could. theoretically. but in practice it is straightforward.
i think that both you and he and the unix powers-that-be think about this problem the wrong way. the source is available: the answer is disimplement sprintf. period. if it didn't exist, you would never call it.
the only thing that would break is other programs you build which use it, but even a program that called it 5000 times would be easy to clean up.
solutions that incrementally improve the world are so easy, but so few see the beauty in that.
find your fun where you can, for sure;
and, you may not find it fun to document anything.
but if you did find it fun, you would also discover that the time you spent on the project would be more productive and highly leveraged precisely because other people could contribute more.
then that submitter ought to first a) know WHAT the developers' original goal was so
that he can demonstrate that he b) is informed enough to know why his new method will
work better. The developers aren't going to trust
Great! I completely agree with your points. Thus, I am sure that you make sure that
projects that you work on publish clear and concise technical documentation which outlines both where the project is going, and details how the architecture is approaches the goals. Thus potential contributors will better know where to make their changes. As, I am equally sure,
projects you work on, when rejecting submissions, send back constructive comments that point out where in the architecture a set of changes should have been made so the submitter can make an acceptable set of changes.
Following is the list of equipment that the Red Cross needs for its field workers
and expanded Emergency Operations Centers. It also needs certified Citrix
engineers and Microsoft-certified consultants.
Apparently, some weaker bridges HAVE collapsed or sustained damage
you've written several confused posts now: you're on the right track but you've got the different concepts all muddled. it is not the strength of the bridge, nor is it the combined energy of the marchers, though those things do play a role.
What is important to the collapse is the undamped resonance of the bridge which stores the energy as it is added to the system.
in the early days, Microsoft was kept afloat by sales of SoftCard, a piece of hardware that added a Zilog Z80 to an Apple II so that MosTek 6502-based machine could run CP/M software designed for S100 machines. The product was successful enough that Apples were the largest installed base of CP/M machines.
so, Microsoft can credibly said to have been a hardware company at that time. BTW, the SoftCard was the brainchild of Paul Allen, Bill Gates's partner who never gets enough credit since it was he who was the responsible for Microsoft's existence and early successes.
This is not flamebait. I'm using the word "left-wing" as shorthand for the position that "abortion is ok because a fetus is not a human, and therefore fetal tissue is ok to experiment with". This viewpoint is more prevalent among liberals than conservatives in my experience, and among scientists. But if you disagree, you'll understand what I'm saying just the same now that I've explained it, and you won't need to object on those grounds: my left equals liberal (in the American sense), and I think this is a liberal position.
One common argument in favor of a zygote's not being a life is the argument that it is just a ball of cells. This same argument is carried further to early stage fetuses before the nervous system is developed to the extent of anything resembling consciousness. [some take it even further. I am trying to summarize a variety of arguments quickly here and have probably not captured every nuance. Please do not quibble unless you think there is some way of phrasing it that is especially useful]
Recently, findings were published from a study which entailed injecting live human fetal cells into a developing monkey fetus brain. The experiment was a success in that the human cells developed fully and were incorporated into the monkey brain... huh? When the researchers were killing these monkeys, did they give any thought to the notion that they were killing a being with partial but fully developed human nervous system?
Am I taking sides in this issue? Well, you decide after I tell you where in the middle I stand. Between what isn't a human life and what is a human life there is a vast grey area. Clearly we need to draw a line somewhere, but wherever we draw a line we are going to be able to find seeming "inconsistencies". But draw a line we must. I am in favor of drawing a more conservative line that errs on the side of preserving more of what "might be" humans, because I think devaluing humanity is a slippery slope. Is this an inconsisten position? Not more than any other. But is it a "costly" position in terms of "humanity"?
We know that there are plenty of scientists among us who would be perfectly willing to experiment on human adults or children in the name of science. Certainly we'd get the best results that way, and the cost of a small number of botched experiments would be more than made up for by the millions of lives improved and saved with our new knowledge. If we experimented on volunteers, what's the diff? Most/many scientists give at least lip service to the supposed ethical problem they see with experimenting on actual human subjects. Well, limits on fetal research or stem cell research are simply a small extension to the "keep off the grass" area. The cost is less knowledge about human biology, only more slowly developing cures to defects and frailties. But defects and frailties are part of what makes us human. What we have in common with our ancestors is that we are mortal. We live, we love, we die. (interesting: I'm applying the leftwing/romanticized/artsy view of humanity, liberal arts if you will, as opposed to the cold calculations of cost-benefit... now who is the hypocrite?) What if we could eradicate death... should we? Really?
What I find disturbing is the insanely egotistical drive for prolonging the lives of those close to us that this medical research represents. If prolonging and improving human life is your goal, well dig deep and save the children of Africa. If prolonging and improving your life is the goal, I have trouble joining in. Or maybe it's the "Nazi-scientist's" pursuit of knowledge for its own sake without regard to the humanity of the subjects that disturbs me. Or maybe these scientists are just buried in their research and don't even want to think about the issues, and it bothers me that they draw a conclusion without much thinking? Or maybe there is some merit to my suspicion that politics plays a role and if it's "conservative" they hate it and if it's "liberal" they like it, for what else could explain the way the two sides seem to line up?
I don't expect you to instantly come around to my postions here, but I hope you walk away realizing that there is more to think about here than "oh, the other side just doesn't get it". I, for one, think I've shown that I get a lot more of it than you do.
I defy anyone to explain to me how (as W would have it) it can be okay to finance research on
human stem cell lines that were created before a certain date (date of W's speech?), and verboten
to finance research on stem cells created after that date.
If scientists can live with a ban on experimenting on humans, they ought to be able to live with an only slightly more liberal definition of what is a human. Different people have different opinions and we reach middle ground in the political arena. I'd guess that Bush doesn't think he knows all the answers either, but realizes there are solid pros and cons and powerful political forces on both sides, and his decision was a compromise--generally, the ability to compromise is extolled as a virtue, you will recall.
This is the definitive message in that thread, where Hans Reiser shows that he is "kinda dumb" sometimes, and as a result he can't be trusted deciding issues about shipping his code. Read that and the next three messages and you will see that Christoph Hellwig just makes more sense when talking about Hans's own code, though I think Christoph could have been more polite.
plot synopsis [SPOILER!]:) Hans says "there is no reason for RedHat to build Reiser in debug mode" but Christoph points out that in debug mode it detects some extra errors and halts. Hans seems to think it's better for it to keep running with erroneous data. That ain't a guy I want writing my fscking filesystem.
BTW, I have nothing against Hans or ReiserFS: I just learned this 10 seconds ago, and I've been running ReiserFS with good results under the Manstroke distro. I had already planned to stop using it but that was because it is not integrated with the latest RedHats so it was limiting my options for upgrading or testing.
Also, earlier in that thread, Hans indicates that he doesn't test with Red Hat himself, and he accuses them of just shoveling code onto CDROMs. This strikes me as a juvenile attitude if he at all wants people to run his code. Hate RedHat all you want... but, it will still have the #1 market share in the US and you ought to check that your code works with it if you want your code to acheive stardom in the US. Maybe he doesn't care--that's cool too--but then he shouldn't care if people who run RedHat are a little uneasy with ReiserFS.
Re:are you building a gaming laptop?
on
The New Athlons
·
· Score: 1
of course I read to the end, that's where it talks about Durons. Gamers want Athlons.
sure... but if you read the thread from where I started it (not deeply nested at all, BTW), the discussion I started was about execptions being useful for "normal" flow of control, not just for "errors". that is a point that seems lost on you and everyone else since you never say, "hey, good idea. I've never noticed that before..." it is an important point so I will keep harping on it till somebody gets it.
tuples were proposed as an alternative but they don't accomplish the goal. you'll have to forgive me for attempting to stick to my own thread:)
as to your advocacy of tuples in general, sure, but you need to tighten up your argument by talking about data types. when is a tuple common enough to warrant it's own type (a class or a struct)? should a tuple of a particular type be assignable to a struct of similar type? should a tuple of uniform type be assignable to an array or that type? etc.
are you building a gaming laptop?
on
The New Athlons
·
· Score: 2
these are mobile chips and a Duron... I thought gamers needed top speed... do you read the articles?
So do we get to see this computer beat another chess champion?
probably not: i just hacked into it for a minute and told it "there is a winning strategy for tic tac toe, find it"... heh heh... the prompt never came back. c u l8r asci!
Thinking about it more, I realize this whole -tuple method is a completely broken hack for solving the problem. Exceptions are the One True Way to handle these cases.
Here, you can see the Right Thing is to write code naturally and functionally:
that is a billion times better (I put infinitely first, but got an overflow:) than this:
bool, float factorial(float x) {
// and by now if you can't see what
// is so fscked up about this approach
// you shouldn't be writing code
// forget about compilers...:)
}
Now, you're probably thinking "yeah, but overflow is an exception already" but my point is that looping over an array of strings and looking them up in a dictionary and doing something with the result should be treated just the same way: functionally till there is a not found, and it is so much more natural to write it as an exception rather than put that ugly if in there every time.
This was my whole point in the beginning: stop thinking of exceptions as "errors" and think of them as normal control flow and build compilers that can handle it.
Slashdot Mirror
the US government is attempting to focus on terrorists. Do you think they are not?
Remember - if terrorists followed laws, we wouldn't have to worry about them.
if private use of back-door-less encryption is made illegal in the US, then the government could more easily
- look inside of emails to see that they do not contain terrorist instructions
- or focus on whomever is violating this law.
Now, I'm sure you would object to such a law, but it works to accomplish the goal you've outlined. You have not given much reasoning to convince someone who thinks this is a good idea that it is not.I'm sick of all the knee-jerk anti-US government sentiment I see here. It wasn't too long ago that everybody here was calling for massive "email disobedience" to overwhelm Echelon and/or Carnivore... do you think this is a good in today's climate? should all Slashdotter's right now be trying to swamp the NSA and FBI servers with bogus terrorist message references in routine emails to their friends? If you really believe what you are saying, I'd say you think we should...
- Are you saying these things should have no regulation?
- or are you saying that encryption should be regulated the way these things are?
- or are you saying that everything is just fine the way it is with a mix of regulated and unregulated.
I ask because you didn't actually say anything at all as it applies to reality. "Starting down the road of outlawing inanimate objects that can be used for multiple purposes"... is exactly where we've been for hundreds of years, and I kind of like living here so I'm finding it a very satisfying experience. Sure, I don't agree with all regulations, but I can't figure out what you are proposing...except, he's saying he wants the blue pill?
passed in as an argument so that 'n' is not locally known.
I speak from experience. people are always so scared of cleanup because it could theoretically be bad. and it could. theoretically. but in practice it is straightforward.
the only thing that would break is other programs you build which use it, but even a program that called it 5000 times would be easy to clean up.
solutions that incrementally improve the world are so easy, but so few see the beauty in that.
samba
very sloppy, no wonder he got caught
you misunderstand: required notification about the GPL license does not give credit to the author.
FinePrint pdfFactory also uses a virtual printer paradigm to "print" to a pdf file, and it is also very slick.
find your fun where you can, for sure;
and, you may not find it fun to document anything.
but if you did find it fun, you would also discover that the time you spent on the project would be more productive and highly leveraged precisely because other people could contribute more.
Great! I completely agree with your points. Thus, I am sure that you make sure that
Following is the list of equipment that the Red Cross needs for its field workers
and expanded Emergency Operations Centers. It also needs certified Citrix
engineers and Microsoft-certified consultants.
a <drool>Beowulf Cluster!</drool>
I think you're being slashdotted :)
I would guess that they meant the energy of 75,000 tons of children jumping up and down.
you've written several confused posts now: you're on the right track but you've got the different concepts all muddled. it is not the strength of the bridge, nor is it the combined energy of the marchers, though those things do play a role.
What is important to the collapse is the undamped resonance of the bridge which stores the energy as it is added to the system.
so, Microsoft can credibly said to have been a hardware company at that time. BTW, the SoftCard was the brainchild of Paul Allen, Bill Gates's partner who never gets enough credit since it was he who was the responsible for Microsoft's existence and early successes.
One common argument in favor of a zygote's not being a life is the argument that it is just a ball of cells. This same argument is carried further to early stage fetuses before the nervous system is developed to the extent of anything resembling consciousness. [some take it even further. I am trying to summarize a variety of arguments quickly here and have probably not captured every nuance. Please do not quibble unless you think there is some way of phrasing it that is especially useful]
Recently, findings were published from a study which entailed injecting live human fetal cells into a developing monkey fetus brain. The experiment was a success in that the human cells developed fully and were incorporated into the monkey brain... huh? When the researchers were killing these monkeys, did they give any thought to the notion that they were killing a being with partial but fully developed human nervous system?
Am I taking sides in this issue? Well, you decide after I tell you where in the middle I stand. Between what isn't a human life and what is a human life there is a vast grey area. Clearly we need to draw a line somewhere, but wherever we draw a line we are going to be able to find seeming "inconsistencies". But draw a line we must. I am in favor of drawing a more conservative line that errs on the side of preserving more of what "might be" humans, because I think devaluing humanity is a slippery slope. Is this an inconsisten position? Not more than any other. But is it a "costly" position in terms of "humanity"?
We know that there are plenty of scientists among us who would be perfectly willing to experiment on human adults or children in the name of science. Certainly we'd get the best results that way, and the cost of a small number of botched experiments would be more than made up for by the millions of lives improved and saved with our new knowledge. If we experimented on volunteers, what's the diff? Most/many scientists give at least lip service to the supposed ethical problem they see with experimenting on actual human subjects. Well, limits on fetal research or stem cell research are simply a small extension to the "keep off the grass" area. The cost is less knowledge about human biology, only more slowly developing cures to defects and frailties. But defects and frailties are part of what makes us human. What we have in common with our ancestors is that we are mortal. We live, we love, we die. (interesting: I'm applying the leftwing/romanticized/artsy view of humanity, liberal arts if you will, as opposed to the cold calculations of cost-benefit... now who is the hypocrite?) What if we could eradicate death... should we? Really?
What I find disturbing is the insanely egotistical drive for prolonging the lives of those close to us that this medical research represents. If prolonging and improving human life is your goal, well dig deep and save the children of Africa. If prolonging and improving your life is the goal, I have trouble joining in. Or maybe it's the "Nazi-scientist's" pursuit of knowledge for its own sake without regard to the humanity of the subjects that disturbs me. Or maybe these scientists are just buried in their research and don't even want to think about the issues, and it bothers me that they draw a conclusion without much thinking? Or maybe there is some merit to my suspicion that politics plays a role and if it's "conservative" they hate it and if it's "liberal" they like it, for what else could explain the way the two sides seem to line up?
I don't expect you to instantly come around to my postions here, but I hope you walk away realizing that there is more to think about here than "oh, the other side just doesn't get it". I, for one, think I've shown that I get a lot more of it than you do.
I defy anyone to explain to me how (as W would have it) it can be okay to finance research on human stem cell lines that were created before a certain date (date of W's speech?), and verboten to finance research on stem cells created after that date.
If scientists can live with a ban on experimenting on humans, they ought to be able to live with an only slightly more liberal definition of what is a human. Different people have different opinions and we reach middle ground in the political arena. I'd guess that Bush doesn't think he knows all the answers either, but realizes there are solid pros and cons and powerful political forces on both sides, and his decision was a compromise--generally, the ability to compromise is extolled as a virtue, you will recall.
no, they are sending them from The Final Destination... that's the problem ;)
plot synopsis [SPOILER!] :) Hans says "there is no reason for RedHat to build Reiser in debug mode" but Christoph points out that in debug mode it detects some extra errors and halts. Hans seems to think it's better for it to keep running with erroneous data. That ain't a guy I want writing my fscking filesystem.
BTW, I have nothing against Hans or ReiserFS: I just learned this 10 seconds ago, and I've been running ReiserFS with good results under the Manstroke distro. I had already planned to stop using it but that was because it is not integrated with the latest RedHats so it was limiting my options for upgrading or testing.
Also, earlier in that thread, Hans indicates that he doesn't test with Red Hat himself, and he accuses them of just shoveling code onto CDROMs. This strikes me as a juvenile attitude if he at all wants people to run his code. Hate RedHat all you want... but, it will still have the #1 market share in the US and you ought to check that your code works with it if you want your code to acheive stardom in the US. Maybe he doesn't care--that's cool too--but then he shouldn't care if people who run RedHat are a little uneasy with ReiserFS.
of course I read to the end, that's where it talks about Durons. Gamers want Athlons.
tuples were proposed as an alternative but they don't accomplish the goal. you'll have to forgive me for attempting to stick to my own thread :)
as to your advocacy of tuples in general, sure, but you need to tighten up your argument by talking about data types. when is a tuple common enough to warrant it's own type (a class or a struct)? should a tuple of a particular type be assignable to a struct of similar type? should a tuple of uniform type be assignable to an array or that type? etc.
these are mobile chips and a Duron... I thought gamers needed top speed... do you read the articles?
I tested my other link when I posted it...
newly correct link
probably not: i just hacked into it for a minute and told it "there is a winning strategy for tic tac toe, find it"... heh heh... the prompt never came back. c u l8r asci!
peace.
Here, you can see the Right Thing is to write code naturally and functionally:
float factorial (float x) {
return x * factorial(x-1)
catch overflow throw overflow;
}
that is a billion times better (I put infinitely first, but got an overflow :) than this:
bool, float factorial(float x) { :)
// and by now if you can't see what
// is so fscked up about this approach
// you shouldn't be writing code
// forget about compilers...
}
Now, you're probably thinking "yeah, but overflow is an exception already" but my point is that looping over an array of strings and looking them up in a dictionary and doing something with the result should be treated just the same way: functionally till there is a not found, and it is so much more natural to write it as an exception rather than put that ugly if in there every time.
This was my whole point in the beginning: stop thinking of exceptions as "errors" and think of them as normal control flow and build compilers that can handle it.