That's the annoying thing - my patients ask for my email address all the time.
I decline, explaining email isn't appropriately secure for health communications... and get the "Come on, get into the 21st century, you luddite!" response. Particularly from the Blackberry crowd.
But everyone posting at slashdot is "I wouldn't go to anyone who isn't super safe."
It works on Mac and Windows. My understanding is that it sets up an automatic encrypted peer-to-peer mirror of folders between machines you designate. So for me, I have it mirror folders on my personal laptop, my work desktop, and my brother's desktop (which is the automated-internet-equivalent of what someone said in an earlier post, "mail a DVD-R to my mom"). If anything crashes, we've got two more copies of it.
Only drawback, doesn't work for Linux yet; likely never will work since they were recently acquired by Microsoft. But there's probably a way to Samba-serve a work-around for that...
Above poster has it totally right -- more exposure for commercials, and broadcaster could track the numbers (+email address for BitTorrent tracker registration, for example). Plus, you could update the commercials to keep them current... or even tailor them to the specific downloader.
This seems like such an obvious business model, I don't understand why someone doesn't do this... if this were available for free, huge numbers of people doing this already on BitTorrent would likely switch.
Then - you wouldn't have to pay for cable, you wouldn't have to pay for a DVR, you'd just have to have a (fairly speedy) internet connection and a computer somehow attached to your TV. Total simplicity.
Cringley predicts this (probably other did as well). Will be interesting to see the execution/format Google decides to pursue. Should pit them against Amazon and Ebay, huh?
I wonder - can this thing be accessed on your home wireless network without physically manipulating the device itself (ie. like a wake-on-LAN computer)?
To me it seems that a good use of this thing would be to just leave it constantly in your briefcase/sack/backpack/car, and push stuff(files, media, podcast, etc) to it from your home wireless network automatically. Then you just pick up your briefcase/sack/backpack/car on your way out the door...
Doesn't this concept seem like it would be a great weapon for Verizon/SBC against Comcast. Last I heard, V/SBC are
1) Racing to roll-out TV services to compete against Comcast&friends.
2) Owners/sellers of lots of moderately-high-speed Internet connections via DSL.
3) Big enough to negotiate with television studios for distribution rights.
4) Use some kind of BT/Tivo-type end-device to ultimately distribute the content to the end-user.
It could be like Comcast's On-Demand, but with thousands of shows. I'd probably pay money for that type of thing, if it were well-executed.
What about the combination of choice #2 and #3? The advantage of email is that you can have access to it at any computer that is connected to the Internet. The advantage of drive-based, is that you can have access to it at any (non-Internet connected) computer that you pop the key-drive into.
So I guess that would be an auto-script that copies files to the key drive every time it's plugged in, and also mails a copy of those files to Gmail. So then your "backup activity" would consist of plugging in your key drive every once in awhile. Which is good, because that's easy, and your not likely to forget to do it.
I highly doubt most hospital/health care facilities have "upgraded" to Win98, much less Win2K. So much legacy software, it's pretty near impossible to upgrade these big organizations and still get work done. This won't be an issue for a couple of years yet...
Gould always does this - ie. because a scientific problem is difficult, one must quiver at "Nature's mystical powers". And, by extension, stop doing any well-controlled experiments. Because, after all, reductionist experimentation is hogwash.
I agree that this line of argument is basically self-inflating puffery, "I told you - humans are complex," who the hell is going to argue that! We're all humans(except for some of my ex's); we love being told we're complex.
I do think that too much crap is going to be foisted on us now as misguided biologists and math/comp guys get together and over-determine a model system of gene interactions with too many degrees of freedom. Then when we question if their work is meaningless drivel that can't make scientific predictions, they say "complexity! splicing factors! difficult problem! mystical! blasphemer! off with your head!"
Bottom line: Thinking hard about what Gould has to say in not a good way to do science. His stuff maybe good for making people feel warm and fuzzy - and if the smart and/or attractive ones pursue careers in science as a result, then maybe Gould is good for science.
Dawkins rocks. (Okay, so I'm a bored mol. biologist/genejock. Now I'll get back to work.).
I receive my MD in a few months, and let me tell you - I can't imagine big American hospital medical care without a Palm. You have no idea how many times I've sat in the OR, reading the list of medications that a patient has filled out about themselves, and have to use Epocrates (literally the best free program EVER written) to sort out what's going on based on phonetics. Sounds crazy, huh? Trust me, it's an improvement.
Anyways, I think the fact we (doctor-types) get labeled as "always want to do it the old way" is a sign of intelligence. Systems for med info can't tolerate the kind of crap you have to put up with your standard freshmeat development cycles(ie. paper and handwriting may suck, but it works). However, I think once you have an exmaple of a breakthrough and useful and reliable UI, like epocrates and some others, there pretty much >80% incorporation across the resident housestaff at the big hospitals.
Don't forget the network effect when it comes to patient tracking software because of the Palm-to-Palm beaming. Once some programmer-doodz figure out a quick checkbox way to initiate patient records, you'll get widespread acceptance. It really is only the patient record setup time that is the bottleneck now, as I see it. I would kill to have an accurate Palm-based patient census. We're carrying ~10-50 patients per, and right now it's the lowest level team member's (ie. intern) job to type an MS Word file EVERY 3AM with all the updated info.
The genomic loci used for the current national database are public information.
Given this info, here's something to consider (Please bear with me... I'm a scientist, not a writer):
To assemble the perfect hypothetical database to search DNA "from the scene" against, you would have to test everyone first, and put their allele set at each loci into a database. To render DNA that is "at the scene" useless vis-a-vis the database, one does not need to replace it with someone else's DNA at every position, but just cover your 13 loci with everyone else's alleles (ie. all possible alleles at each of the known 13 loci - fairly straightforward pooled-sample PCR - and, incidentally, is a bit easier than isolating a specific person's & erasing your alleles).
The same combinatorial logic that makes the identification so specific leads to a rapid collapse if there are more than one allele found at each position (I hope you can see how this works - instead of just allele A or allele B, you find alleles A, B, C, D, and E - obviously fake, but also useless for identification purposes. Sort of looking for a needle in a pile of needles instead of a haystack). On a practical level, this type of "obscuration defense" can easily be gotten around by testing different loci in any individual two-point comparison(assuming you only "obscure" the reference allele set). However, the reference set of loci that determines the national database (ie. the search set) will thus be rendered useless. This makes this trick nice because it stakes out somewhat higher moral grounds. It's a way to maintain the usefulness of DNA evidence in specific cases, but cause a national database to collapse.
Slashdot Mirror
Umm... Clear 4G = unlimited. For computers. Using it now.
That's the annoying thing - my patients ask for my email address all the time.
I decline, explaining email isn't appropriately secure for health communications... and get the "Come on, get into the 21st century, you luddite!" response. Particularly from the Blackberry crowd.
But everyone posting at slashdot is "I wouldn't go to anyone who isn't super safe."
Sadly, most patients aren't like that...
Foldershare
It works on Mac and Windows. My understanding is that it sets up an automatic encrypted peer-to-peer mirror of folders between machines you designate. So for me, I have it mirror folders on my personal laptop, my work desktop, and my brother's desktop (which is the automated-internet-equivalent of what someone said in an earlier post, "mail a DVD-R to my mom"). If anything crashes, we've got two more copies of it.
Only drawback, doesn't work for Linux yet; likely never will work since they were recently acquired by Microsoft. But there's probably a way to Samba-serve a work-around for that...
Above poster has it totally right -- more exposure for commercials, and broadcaster could track the numbers (+email address for BitTorrent tracker registration, for example). Plus, you could update the commercials to keep them current... or even tailor them to the specific downloader.
This seems like such an obvious business model, I don't understand why someone doesn't do this... if this were available for free, huge numbers of people doing this already on BitTorrent would likely switch.
Then - you wouldn't have to pay for cable, you wouldn't have to pay for a DVR, you'd just have to have a (fairly speedy) internet connection and a computer somehow attached to your TV. Total simplicity.
Argh! Comment submission doesn't start till Nov 2nd -- to maximize the Slashdot effect, the editors probably should re-post this article then.
(although, let's face it... with the frequency of article dupes, they probably will re-post it as part of the normal Slashdot practice)...
I particularly like the text-entry feature...
Alternative Info at this Site
Cringley predicts this (probably other did as well). Will be interesting to see the execution/format Google decides to pursue. Should pit them against Amazon and Ebay, huh?
I wonder - can this thing be accessed on your home wireless network without physically manipulating the device itself (ie. like a wake-on-LAN computer)?
To me it seems that a good use of this thing would be to just leave it constantly in your briefcase/sack/backpack/car, and push stuff(files, media, podcast, etc) to it from your home wireless network automatically. Then you just pick up your briefcase/sack/backpack/car on your way out the door...
I think you were thinking of this Google Tidbits article (source) from a little while ago.
Doesn't this concept seem like it would be a great weapon for Verizon/SBC against Comcast. Last I heard, V/SBC are
1) Racing to roll-out TV services to compete against Comcast&friends.
2) Owners/sellers of lots of moderately-high-speed Internet connections via DSL.
3) Big enough to negotiate with television studios for distribution rights.
4) Use some kind of BT/Tivo-type end-device to ultimately distribute the content to the end-user.
It could be like Comcast's On-Demand, but with thousands of shows. I'd probably pay money for that type of thing, if it were well-executed.
Compared to my workplace where 80 hour weeks are universally recognized as signs of laziness.
And to think before the changes we routinely worked 110-120/wk.
"No-one forcing them to work there." misses the point of the obvious non-monetary benefit to the allure of working there.
What about the combination of choice #2 and #3? The advantage of email is that you can have access to it at any computer that is connected to the Internet. The advantage of drive-based, is that you can have access to it at any (non-Internet connected) computer that you pop the key-drive into.
So I guess that would be an auto-script that copies files to the key drive every time it's plugged in, and also mails a copy of those files to Gmail. So then your "backup activity" would consist of plugging in your key drive every once in awhile. Which is good, because that's easy, and your not likely to forget to do it.
For Example, Google Groups
I highly doubt most hospital/health care facilities have "upgraded" to Win98, much less Win2K. So much legacy software, it's pretty near impossible to upgrade these big organizations and still get work done. This won't be an issue for a couple of years yet...
The abstract, and some comments.
Consider Rosetta (RSTA) with ~175 employees vs. VA (LNUX).
0 .html
$620m, not too shabby.
http://www.thestandard.com/article/0,1902,24434,0
Gould always does this - ie. because a scientific problem is difficult, one must quiver at "Nature's mystical powers". And, by extension, stop doing any well-controlled experiments. Because, after all, reductionist experimentation is hogwash.
I agree that this line of argument is basically self-inflating puffery, "I told you - humans are complex," who the hell is going to argue that! We're all humans(except for some of my ex's); we love being told we're complex.
I do think that too much crap is going to be foisted on us now as misguided biologists and math/comp guys get together and over-determine a model system of gene interactions with too many degrees of freedom. Then when we question if their work is meaningless drivel that can't make scientific predictions, they say "complexity! splicing factors! difficult problem! mystical! blasphemer! off with your head!"
Bottom line: Thinking hard about what Gould has to say in not a good way to do science. His stuff maybe good for making people feel warm and fuzzy - and if the smart and/or attractive ones pursue careers in science as a result, then maybe Gould is good for science.
Dawkins rocks.
(Okay, so I'm a bored mol. biologist/genejock. Now I'll get back to work.).
I receive my MD in a few months, and let me tell you - I can't imagine big American hospital medical care without a Palm. You have no idea how many times I've sat in the OR, reading the list of medications that a patient has filled out about themselves, and have to use Epocrates (literally the best free program EVER written) to sort out what's going on based on phonetics. Sounds crazy, huh? Trust me, it's an improvement.
Anyways, I think the fact we (doctor-types) get labeled as "always want to do it the old way" is a sign of intelligence. Systems for med info can't tolerate the kind of crap you have to put up with your standard freshmeat development cycles(ie. paper and handwriting may suck, but it works). However, I think once you have an exmaple of a breakthrough and useful and reliable UI, like epocrates and some others, there pretty much >80% incorporation across the resident housestaff at the big hospitals.
Don't forget the network effect when it comes to patient tracking software because of the Palm-to-Palm beaming. Once some programmer-doodz figure out a quick checkbox way to initiate patient records, you'll get widespread acceptance. It really is only the patient record setup time that is the bottleneck now, as I see it. I would kill to have an accurate Palm-based patient census. We're carrying ~10-50 patients per, and right now it's the lowest level team member's (ie. intern) job to type an MS Word file EVERY 3AM with all the updated info.
Given this info, here's something to consider (Please bear with me... I'm a scientist, not a writer):
To assemble the perfect hypothetical database to search DNA "from the scene" against, you would have to test everyone first, and put their allele set at each loci into a database. To render DNA that is "at the scene" useless vis-a-vis the database, one does not need to replace it with someone else's DNA at every position, but just cover your 13 loci with everyone else's alleles (ie. all possible alleles at each of the known 13 loci - fairly straightforward pooled-sample PCR - and, incidentally, is a bit easier than isolating a specific person's & erasing your alleles).
The same combinatorial logic that makes the identification so specific leads to a rapid collapse if there are more than one allele found at each position (I hope you can see how this works - instead of just allele A or allele B, you find alleles A, B, C, D, and E - obviously fake, but also useless for identification purposes. Sort of looking for a needle in a pile of needles instead of a haystack). On a practical level, this type of "obscuration defense" can easily be gotten around by testing different loci in any individual two-point comparison(assuming you only "obscure" the reference allele set). However, the reference set of loci that determines the national database (ie. the search set) will thus be rendered useless. This makes this trick nice because it stakes out somewhat higher moral grounds. It's a way to maintain the usefulness of DNA evidence in specific cases, but cause a national database to collapse.