I'm experiencing a little cognitive dissonance here. I think I was thrashed recently in a BSD forum for making such assertions.
I've been running Java 5 on a Mac Mini at work and on a Sempron Pro 2600 FC3 box at home, and frankly I don't see the Mac Mini as that much slower, other than what I would expect for a desktop HD vs. a laptop HD.
I think I'm seeing a lot of rationalization from iNTEL fanboys in this thread.
Lousy MSWindows-only "free" DICOM viewer
on
Film to X-rays?
·
· Score: 1
Yeah there are FOSS tools in the works various places, but the "free" viewer is MSWindows-only.
I've watched judges righting down serial numbers in the margins of the sign-in books in US elections. I don't remember if I realized at the time I was supposed to tell them not to do that.
The serial numbers, when used, are supposed to be for matching the stub to the ballot, so that the ballot and the stub can be delivered by independent paths. But the serial numbers are _not_ supposed to be tracked, and are even supposed to be in somewhat random order so that election judges can't just remember that Joe Brown voted the 57th ballot and then count up on the serial numbers later. If the numbers are tracked, then you are right. Anonimity has been breached.
Getting the administration process correct on paper ballots does take a little more than ordinary skill, but it is visible, and the ordinary person has a hope of understanding it.
Ordinary voting judges can count ballots and stubs if the ballots and stubs are paper. If they are electronic, it takes at least a geek with moderate skill to check the counting program and the count.
Checking the program requires analyzing the source, including every library linked, and then making sure every byte of the object matches the source. And when the analyst is done, the best we can assert is that, if there is a pattern-matching trap in the counting program, it was constructed by someone with more skill than the analyst (assuming the analyst is not in on the game).
Watching the count of paper ballots requires only ordinary people of ordinary skill (who can stay awake). (Chads are evidence that paper ballots can suffer from design defects, but the effects are somewhat more random.)
You can't rely on a statistical test of the count, because statistic quality assurance relies on randomness. Deliberately faked results are not random, and can be made dependent on precinct, on time of day, on all sorts of obfuscation techniques.
I guess you'd need at least three different sets of programs, each written a different group, preferably of different party affiliation, to get anywhere close to being able to monitor the count. Does that save time and money over paper ballots? Is it going to be more reliable? Can Republicans program? How about Democrats? How about fundamentalist political supporters of the religion of the far rock? (Random political organization of known bias, there.) What happens when the biased political organization that can't program (or fudges their results) brings suit against the ones which are accurate?
Every step of the process has the same defects -- the balloting, the judging, the collection at precints (Failing to collect at precincts just weakens the system.), the delivery to the local authorities, the counting by the local authorities (and the checking the count of ballots against the count of stubs), the storage of ballots, everything. Physical paper requires human action, but can be performed, and _monitored_ by humans of ordinary skill. Each political party can send monitors to observe every step, and those moniters don't have to be highly trained. (Just courteours at some basic level.)
Monitoring electronic balloting requires somewhat extraordinary skill. Some aspects of the monitoring may even require hardware that ordinary people don't have. (Want something to do a radio sweep of the ballot place all during ballots?)
I know computer technology is attractive. It can be applied, actually, but the paper trail must remain. (And the ballots themselves should be as simple as possible, to reduce opportunity for hidden radio transmitters.)
that Microsoft is giving to some African communities and governments.
Give them a little free MS software, train them how to use (to the extent it's possible) and hope they learn how to make money with it to buy upgrades at some time in the future.
I think the reason the system has become disconnected from the people is that the media has been happy to step in between, but that's a different thread.
How does the voter see the electronic ballot and the government see the same electronic ballot without giving the software the chance to fink the voter to some nameless official on a vendetta?
If it's not the same, how does the voter know his vote really counted?
How does the voter get any assurance that the government tally matches the collected votes?
You can build, I suppose, a public key document that would reveal the vote and not the voter, except to the person who owns the key for the vote, but if you have n votes, how can any observer prove that the other n-1 are not faked?
One time passwords as serial numbers on the ballot? Then you have the same problem again, how does the government prove the OTP is valid without reading it? And even if that can actually be solved, how does the average voter assure him or herself that the government's software is actually always jumping through all the hoops just to protect the voter's privacy when there are nameless officials with potential motivation to not bother?
Any way of solving one of the problems opens a hole somewhere else.
How does _any_ encryption allow the average voter to monitor the voting process?
The average voter can see the anonymous ballot go into the box with a lot of other ballots. The average voter can see the anonymous stub go in the other box when such stubs are used (and they ought to be required everywhere). The average voter can look around and observe whether there is anyone looking over his or her shoulder at the voting booth.
The simple act of pushing the votes onto a wire or into a database before the vote has been detached from the voter and judged completely undermines the ability of the average voter to observe the voting process for either accuracy or anonymity. No amount of encryption or other mathematical games can fix that.
Perhaps your next argument is that only geeks with the mathematical skills to understand holomorphic encryption (and the network and radio monitoring equipment to watch the polling station) should be allowed to vote? Would that make it difficult enough?
It's called PKI. From what little the article said, it sounds like they used it.
But the real problem is not the connection, it's the disconnection. The government can claim they don't make a record of who voted how, but there's no way for the voter to check. Also, there is a way for the voter to be sure the vote is recorded, but no way to be sure that a vote recorded is a vote counted.
With paper ballots, the only way to connect a voter with a ballot would be leftover fingerprints, or (becoming viable now, perhaps) wireless monitoring by someone who can see who enters the booth when. By providing separable anonymous stubs, there is also a way to be reasonably sure that every vote gets counted.
You know, some webapps have parts the run on the client. This basically clarifies that, while code can run on the server without being distributed, code cannot run on the client without being distributed.
They're a consumer electronics distributor (TV, radio, cables, etc). The only "IP" these guys have is their customer and supplier lists.
That's the one they fight tooth and nail to protect, right?
I guess I should RTFA, but I've been in enough companies that try to use FOSS without realizing that getting answers and help has to be done differently.
It happens. The companies get excited about the low cost of entry and fail to allocate time for the mail lists, much less for reporting bugs. Heaven help the employee that suggests actually fixing one.
Perhaps the whole root of the problem is that, fearing to let their (future) IP out of the bag, they prevented their techs from going to the community with questions.
Is anyone asked when her website is indexed ? Why should this be any different with books?
I'm sure you remembered right after you posted that about the robots.txt file and the ROBOTS meta-headers. Yeah, something similar could be put in books (for instance, right under the copyright declaration) starting one or two printing runs from now, but what about all the books that have been published to date?
Besides, the copyright declaration really is already enough of a notice that any copying activity legally requires contacting the copyright holder. Rather than additional notices that the copyright holder really means it, a notice specifically granting the right to systematically index or copy for public archival purposes would be more appropriate. (Sort of like the GPL for software, donchanoh.)
Mind you, when corporations started buying up copyrights and wangled Congress into extending copyright into a permanent easement, they eroded some of their claims to consideration for this kind of use. I'd like to see some legal action broadening fair use when copyright is held by a corporation or other quasi-public entity, so that corporations would have to specifically opt out of public indexes. Corporate ownership of copyright is something of an abuse of the commons, so the copyright should be more limited in scope.
My worst nightmare -- that someone sane starts fixing the problems at MicroSoft. How would there be any room left to compete?
Joke.
The only way to really fix Microsoft is to split it into two corporations each for every product line, and open all APIs with no anti-GPL license restrictions. And use the ill-gotten gains Gates, Balmer, et. al. have accumulated to fund start-ups to company with the baby-Softs. And open the evolution of the APIs under the control of a joint committee of the EFF and representatives of the several Linux and BSD distributions.
Casual reading makes it sound like the examiner is either incompetent or has been bribed by someone trying to raise the open source communities legal costs.
Admittedly, that's casual reading.
Maybe the lawyer assumed the examiner would be more familiar with the history of Linux and the examiner assume that he shouldn't be for some reason?
Guess I'd have to read a bit more carefully to be sure.
I suppose that it shows that Apple is investing _some_ resources into this thing.
Lousy update also started up QT on reboot. I had forgotten I was logged back into the admin account, so if Gretchen had any malware on her front page for her new video, the malware may have had a chance to execute.
One thing the Mac OS really needs is a way to turn off apps for admin accounts. (You can do that for non-amin accounts.) At least all the auto-execute gadgetry should be shut off by default when you click the "allow this account to admin the mac" button.
They mentioned that you'd need to carry 44 to 84 pounds, and the discussion sounds like that's the weight of the pendulum. Or maybe the pendulum plus the generator plus the frame. If the generator is, say, only ten pounds, then the 44 to 84 could include your tent, the device(s) being powered, lunch, collected rocks, etc.
Slashdot Mirror
I'm experiencing a little cognitive dissonance here. I think I was thrashed recently in a BSD forum for making such assertions.
I've been running Java 5 on a Mac Mini at work and on a Sempron Pro 2600 FC3 box at home, and frankly I don't see the Mac Mini as that much slower, other than what I would expect for a desktop HD vs. a laptop HD.
I think I'm seeing a lot of rationalization from iNTEL fanboys in this thread.
Yeah there are FOSS tools in the works various places, but the "free" viewer is MSWindows-only.
I've watched judges righting down serial numbers in the margins of the sign-in books in US elections. I don't remember if I realized at the time I was supposed to tell them not to do that.
The serial numbers, when used, are supposed to be for matching the stub to the ballot, so that the ballot and the stub can be delivered by independent paths. But the serial numbers are _not_ supposed to be tracked, and are even supposed to be in somewhat random order so that election judges can't just remember that Joe Brown voted the 57th ballot and then count up on the serial numbers later. If the numbers are tracked, then you are right. Anonimity has been breached.
Getting the administration process correct on paper ballots does take a little more than ordinary skill, but it is visible, and the ordinary person has a hope of understanding it.
and that is neither an insult or a compliment.
Ordinary voting judges can count ballots and stubs if the ballots and stubs are paper. If they are electronic, it takes at least a geek with moderate skill to check the counting program and the count.
Checking the program requires analyzing the source, including every library linked, and then making sure every byte of the object matches the source. And when the analyst is done, the best we can assert is that, if there is a pattern-matching trap in the counting program, it was constructed by someone with more skill than the analyst (assuming the analyst is not in on the game).
Watching the count of paper ballots requires only ordinary people of ordinary skill (who can stay awake). (Chads are evidence that paper ballots can suffer from design defects, but the effects are somewhat more random.)
You can't rely on a statistical test of the count, because statistic quality assurance relies on randomness. Deliberately faked results are not random, and can be made dependent on precinct, on time of day, on all sorts of obfuscation techniques.
I guess you'd need at least three different sets of programs, each written a different group, preferably of different party affiliation, to get anywhere close to being able to monitor the count. Does that save time and money over paper ballots? Is it going to be more reliable? Can Republicans program? How about Democrats? How about fundamentalist political supporters of the religion of the far rock? (Random political organization of known bias, there.) What happens when the biased political organization that can't program (or fudges their results) brings suit against the ones which are accurate?
Every step of the process has the same defects -- the balloting, the judging, the collection at precints (Failing to collect at precincts just weakens the system.), the delivery to the local authorities, the counting by the local authorities (and the checking the count of ballots against the count of stubs), the storage of ballots, everything. Physical paper requires human action, but can be performed, and _monitored_ by humans of ordinary skill. Each political party can send monitors to observe every step, and those moniters don't have to be highly trained. (Just courteours at some basic level.)
Monitoring electronic balloting requires somewhat extraordinary skill. Some aspects of the monitoring may even require hardware that ordinary people don't have. (Want something to do a radio sweep of the ballot place all during ballots?)
I know computer technology is attractive. It can be applied, actually, but the paper trail must remain. (And the ballots themselves should be as simple as possible, to reduce opportunity for hidden radio transmitters.)
that Microsoft is giving to some African communities and governments.
Give them a little free MS software, train them how to use (to the extent it's possible) and hope they learn how to make money with it to buy upgrades at some time in the future.
That's part of the reason the tap runs dry.
Big question -- Is Mil^Hcrosoft milking the system?
Agreed that PKI does not cut it.
I think the reason the system has become disconnected from the people is that the media has been happy to step in between, but that's a different thread.
If _who_ uses their algorithm carefully?
Remember, we are talking about end-users who are average voters, not geeks.
How does the voter see the electronic ballot and the government see the same electronic ballot without giving the software the chance to fink the voter to some nameless official on a vendetta?
If it's not the same, how does the voter know his vote really counted?
How does the voter get any assurance that the government tally matches the collected votes?
You can build, I suppose, a public key document that would reveal the vote and not the voter, except to the person who owns the key for the vote, but if you have n votes, how can any observer prove that the other n-1 are not faked?
One time passwords as serial numbers on the ballot? Then you have the same problem again, how does the government prove the OTP is valid without reading it? And even if that can actually be solved, how does the average voter assure him or herself that the government's software is actually always jumping through all the hoops just to protect the voter's privacy when there are nameless officials with potential motivation to not bother?
Any way of solving one of the problems opens a hole somewhere else.
Seriously missing. Seriously MS-only.
Seriously you're going to need these to access your bank account in a couple of years.
Conspiracy? Microsoft has Verisign in their back pocket or somethin?
Didn't somebody mention a _contractual_ limitation that one of the vendors acknowledged?
Seriously.
How does _any_ encryption allow the average voter to monitor the voting process?
The average voter can see the anonymous ballot go into the box with a lot of other ballots. The average voter can see the anonymous stub go in the other box when such stubs are used (and they ought to be required everywhere). The average voter can look around and observe whether there is anyone looking over his or her shoulder at the voting booth.
The simple act of pushing the votes onto a wire or into a database before the vote has been detached from the voter and judged completely undermines the ability of the average voter to observe the voting process for either accuracy or anonymity. No amount of encryption or other mathematical games can fix that.
Perhaps your next argument is that only geeks with the mathematical skills to understand holomorphic encryption (and the network and radio monitoring equipment to watch the polling station) should be allowed to vote? Would that make it difficult enough?
Oh, and every time the latest, best hashing algorithm gets defeated, you have to give the voters new ID cards.
It's called PKI. From what little the article said, it sounds like they used it.
But the real problem is not the connection, it's the disconnection. The government can claim they don't make a record of who voted how, but there's no way for the voter to check. Also, there is a way for the voter to be sure the vote is recorded, but no way to be sure that a vote recorded is a vote counted.
With paper ballots, the only way to connect a voter with a ballot would be leftover fingerprints, or (becoming viable now, perhaps) wireless monitoring by someone who can see who enters the booth when. By providing separable anonymous stubs, there is also a way to be reasonably sure that every vote gets counted.
You know, some webapps have parts the run on the client. This basically clarifies that, while code can run on the server without being distributed, code cannot run on the client without being distributed.
(As someone pointed out up above.)
That's the one they fight tooth and nail to protect, right?
I guess I should RTFA, but I've been in enough companies that try to use FOSS without realizing that getting answers and help has to be done differently.
It happens. The companies get excited about the low cost of entry and fail to allocate time for the mail lists, much less for reporting bugs. Heaven help the employee that suggests actually fixing one.
are not wordprocessing, spreadsheet, or even presentation junk.
The most important apps in an office are the ones that you fill out and pass around paperwork with.
sad, but true
Perhaps the whole root of the problem is that, fearing to let their (future) IP out of the bag, they prevented their techs from going to the community with questions.
Have you checked that the problem with your company's timecard page isn't that your company's certificate is incorrectly built?
We have that problem, but not enough motivation to fix it yet. Safari is also useable on the page.
I'm sure you remembered right after you posted that about the robots.txt file and the ROBOTS meta-headers. Yeah, something similar could be put in books (for instance, right under the copyright declaration) starting one or two printing runs from now, but what about all the books that have been published to date?
Besides, the copyright declaration really is already enough of a notice that any copying activity legally requires contacting the copyright holder. Rather than additional notices that the copyright holder really means it, a notice specifically granting the right to systematically index or copy for public archival purposes would be more appropriate. (Sort of like the GPL for software, donchanoh.)
Mind you, when corporations started buying up copyrights and wangled Congress into extending copyright into a permanent easement, they eroded some of their claims to consideration for this kind of use. I'd like to see some legal action broadening fair use when copyright is held by a corporation or other quasi-public entity, so that corporations would have to specifically opt out of public indexes. Corporate ownership of copyright is something of an abuse of the commons, so the copyright should be more limited in scope.
Perhaps the suit is to make Google ask first, instead of allowing Google to proceed unless authors specifically opt out.
Yes, it may be good advertising, but an author should have a right to opt in, not a right to opt out. Google should be asking first.
My worst nightmare -- that someone sane starts fixing the problems at MicroSoft. How would there be any room left to compete?
Joke.
The only way to really fix Microsoft is to split it into two corporations each for every product line, and open all APIs with no anti-GPL license restrictions. And use the ill-gotten gains Gates, Balmer, et. al. have accumulated to fund start-ups to company with the baby-Softs. And open the evolution of the APIs under the control of a joint committee of the EFF and representatives of the several Linux and BSD distributions.
It ain't gonna happen.
Did I read that right?
5%
And they think they call the shots? Have a right to define Novell's future direction? Think it means something that they don't like being ignored?
5% could hurt, I suppose, but it really makes them look like jerks. I wonder if they are buddies with somebody at SCO.
Casual reading makes it sound like the examiner is either incompetent or has been bribed by someone trying to raise the open source communities legal costs.
Admittedly, that's casual reading.
Maybe the lawyer assumed the examiner would be more familiar with the history of Linux and the examiner assume that he shouldn't be for some reason?
Guess I'd have to read a bit more carefully to be sure.
gets me an iTunes phone driver.
I suppose that it shows that Apple is investing _some_ resources into this thing.
Lousy update also started up QT on reboot. I had forgotten I was logged back into the admin account, so if Gretchen had any malware on her front page for her new video, the malware may have had a chance to execute.
One thing the Mac OS really needs is a way to turn off apps for admin accounts. (You can do that for non-amin accounts.) At least all the auto-execute gadgetry should be shut off by default when you click the "allow this account to admin the mac" button.
They mentioned that you'd need to carry 44 to 84 pounds, and the discussion sounds like that's the weight of the pendulum. Or maybe the pendulum plus the generator plus the frame. If the generator is, say, only ten pounds, then the 44 to 84 could include your tent, the device(s) being powered, lunch, collected rocks, etc.
Am I reading this wrong?