Even if you have a RS-232 port on the motherboard, there's no way to unstick a wedged machine, like you can do by sending a break on the serial console of most Sun machines.
I find the reset switch on the front of the case generally works as well.
Its called a converter. USB is a high-speed, multinode reimplementation of serial ports. It has bandwidth to spare and can incorporate all of the nuances of RS232/422 communication with the right adapters.
Most of my problems are with the decrepid software that comes with elderly scientific equipment, not the ports.
Re:Does this ver. solve the WinXP security "featur
on
Samba 3.0.0 Released
·
· Score: 1
It does a hell of a lot more. I have more software that doesn't work under XP than I ever did under 2000.
And hey, who can't love the fisher-price dialog system. You have no need to go in and change a setting that you know where it goes. There is a ritual now by which you painstaking step through a set of droolproof dialogs, enter the setting you wanted 4 steps in, and then have step 7 negate them.
Re:NT4 support? Err , what about 2000, XP?
on
Samba 3.0.0 Released
·
· Score: 2, Insightful
In a word yes.
That's exactly what I did 3 years ago when M$ started playing games with Active Directory, and I still had a network full of 98 and NT boxes. We set up a new domain, and moved all the file and print services to it.
Now that we have aged out all of the decrepid hardware and standardized on 2k, ActiveDirectory is a good idea. But that is 3 years, and a $100,000 in hardware later.
Having trust support would have saved me from having to hike to all the machines and add them to the new domain. I can imagine with NT entering M$'s discard pile, there are quite a few NT shops that are looking for a drop in replacement.
Enter the dragon...
Re:Open source top 5 best contributions
on
Samba 3.0.0 Released
·
· Score: 0, Offtopic
I'd replace postgres with MySQL, Apache with either Python/Tcl/Perl.
Scripting language power a lot more under the hood of your Linux box than Apache.
Legal code and computer code have a lot in common. They are incomprehensible to the common man, riddled with logical errors, and open to exploitation an abuse by hacks.
Ignoring the fact that its impossible to bring 200 million people on sailing vessels in the 18th and 19th centuries...
Actually the problem wasn't that 200 million people started the journey to America. It's that a fraction of them actually survived the trip. A slave ship packed solid with about 400 people was a) well documented and b) VERY common. Your window of the 18th and 19th century is also artificially small. The slave trade began as early as the 1500's.
And of course, not everyone made it as far as the ships. Many died of disease, dehydration, starvation, or getting slapped around on the way to the ship or waiting at the harbor.
As gory and nasty as the movie Amistad was, it was historically accurate. The slave trade was that damn gory and nasty.
If it helps anyone sleep better at night, it was Africans that sold fellow Africans as slaves.
I just take comfort that during all this, my ancestors where starving in Ireland, or in-between invasions in Poland.
Thank you for your excellent comments. I'm still convinced that most involved in this discussion don't understand some of the logistics involved. You have demonstrated that under the right conditions it can work well and cost effectively.
While I hear you on the "Windows" part, we do live in a world of sheep that know no better. But I digress...
The PDF is every bit as corruptible as any other medium. You would do better to simply maintain an electronic log with an cryptographic cypher in each line, an auto-incrementing line number, a date stamp, and a serial number for the voting machine. The cryptographic hash would generate a cipher that uses the time, log entry number, and the votes recorded. Basically all of this information are plugged into factors of a math equation, the result is the hash. Theoretially, only that combination of factors can produce that hash.
When the validating software reads the log entry, it also re-generates the hash and makes sure the numbers jive. If they don't it is a possible sign of tampering, and the records are escalated up to a technician who can investigate further.
In order to be secure, there has to be several other factors that are not store in the log, but are known to both the auditor and the voting machine. This presents it's own issues, but it does raise the level of expertise required to fudge numbers.
Have you ever, in your life, run a helpdesk? Printer ARE the spawn of Satan. They are the single most error prone piece of hardware that exists. They jam. They go bezerk spewing control codes like some arameic chant. They lock up, accepting data but doing nothing with it. And then you have the problem of keeping paper, toner, and/or ink stocked in the unit.
Thermal reciept printers are an order of magnetude more reliable, but the paper they print on is not. The print fades over time, and the paper decomposes too rapidly for any kind of permanent record. Did I mention that a good thermal printer runs between $700-$2000 dollars?
Now, does solid and reliable printer technology exists? Yes. But it is far more expensive than the current budgets for electronic voting allow for. You would actually do better to devise a cheap and reliable electronic (or even mechanical) WORM media, that embeds a timestamp and cryptographic hash into every log entry for verification.
Or better yet, use MySQL or Postgres as the basis for "USSQL". For what most branches of government spend on Oracle licenses, we could have our own DB development department, staffed with the best minds money can by, with money to spare for extra mainframes.
Slashdot Mirror
I find the reset switch on the front of the case generally works as well.
Most of my problems are with the decrepid software that comes with elderly scientific equipment, not the ports.
We are just quibbling over whether said converters are soldered to the motherboard or are replaceable external units.
No, you just buy an LCD that has analog inputs. I have one such unit on my desk (and am posting on Slashdot through it.)
Sure. As if postgres can mystically operate over multiple servers at once, and makes a server immune to hardware failure.
That's as bad as the MySQL guys saying "You don't NEED server-side includes or foriegn keys."
Then, Gnome Vs. KDE
Now its MySQL Vs. Postgres
At least we are evolving from text editors and eye-candy to relational databases.
(Cough) replication (Cough)
And hey, who can't love the fisher-price dialog system. You have no need to go in and change a setting that you know where it goes. There is a ritual now by which you painstaking step through a set of droolproof dialogs, enter the setting you wanted 4 steps in, and then have step 7 negate them.
That's exactly what I did 3 years ago when M$ started playing games with Active Directory, and I still had a network full of 98 and NT boxes. We set up a new domain, and moved all the file and print services to it.
Now that we have aged out all of the decrepid hardware and standardized on 2k, ActiveDirectory is a good idea. But that is 3 years, and a $100,000 in hardware later.
Having trust support would have saved me from having to hike to all the machines and add them to the new domain. I can imagine with NT entering M$'s discard pile, there are quite a few NT shops that are looking for a drop in replacement.
Enter the dragon...
Scripting language power a lot more under the hood of your Linux box than Apache.
Don't forget about robbing children. All they need to do now is kick a begger. Or better yet, sue some guy who is dying of cancer or something...
Have you actually read an act of congress? Talk about security through obscurity!
In Soviet Russia, Kazaa searches YOU!
Legal code and computer code have a lot in common. They are incomprehensible to the common man, riddled with logical errors, and open to exploitation an abuse by hacks.
We are at war with Oceana. We have always been at war with Oceana. Eurasia is our ally...
The RIAA were just sucking files, and nobody likes a mooch.
I don't care if they have to reindex all of reality, I'm sick of these clitches.
Actually the problem wasn't that 200 million people started the journey to America. It's that a fraction of them actually survived the trip. A slave ship packed solid with about 400 people was a) well documented and b) VERY common. Your window of the 18th and 19th century is also artificially small. The slave trade began as early as the 1500's.
And of course, not everyone made it as far as the ships. Many died of disease, dehydration, starvation, or getting slapped around on the way to the ship or waiting at the harbor.
As gory and nasty as the movie Amistad was, it was historically accurate. The slave trade was that damn gory and nasty.
If it helps anyone sleep better at night, it was Africans that sold fellow Africans as slaves.
I just take comfort that during all this, my ancestors where starving in Ireland, or in-between invasions in Poland.
Thank you for your excellent comments. I'm still convinced that most involved in this discussion don't understand some of the logistics involved. You have demonstrated that under the right conditions it can work well and cost effectively.
Or at least document the lack of documentation of the lack of a procedure.
The PDF is every bit as corruptible as any other medium. You would do better to simply maintain an electronic log with an cryptographic cypher in each line, an auto-incrementing line number, a date stamp, and a serial number for the voting machine. The cryptographic hash would generate a cipher that uses the time, log entry number, and the votes recorded. Basically all of this information are plugged into factors of a math equation, the result is the hash. Theoretially, only that combination of factors can produce that hash.
When the validating software reads the log entry, it also re-generates the hash and makes sure the numbers jive. If they don't it is a possible sign of tampering, and the records are escalated up to a technician who can investigate further.
In order to be secure, there has to be several other factors that are not store in the log, but are known to both the auditor and the voting machine. This presents it's own issues, but it does raise the level of expertise required to fudge numbers.
Good printers would cost more than the rest of the voting machine. I actually get a little more in dept in this comment.
Thermal reciept printers are an order of magnetude more reliable, but the paper they print on is not. The print fades over time, and the paper decomposes too rapidly for any kind of permanent record. Did I mention that a good thermal printer runs between $700-$2000 dollars?
Now, does solid and reliable printer technology exists? Yes. But it is far more expensive than the current budgets for electronic voting allow for. You would actually do better to devise a cheap and reliable electronic (or even mechanical) WORM media, that embeds a timestamp and cryptographic hash into every log entry for verification.
To the voting people this is a project to be pulled off as quickly and cheaply as possible. They are the low bidder, after all.
Or better yet, use MySQL or Postgres as the basis for "USSQL". For what most branches of government spend on Oracle licenses, we could have our own DB development department, staffed with the best minds money can by, with money to spare for extra mainframes.