I'm more questioning the whole process of drug development and approval as it applies to diseases that are always fatal. I think the modern western approach has gone too far along the "it must be perfect for approval" path with the attached very high cost without showing that it has eliminated or even greatly reduced risk vs a more relaxed approval approach. Sure for a drug like Viagra perhaps the bar should be set high as the gain from the drug is fairly minimal in comparison to the potential problems but if you are talking about eliminating Parkinsons or non-Hodgkin's lymphoma then the bar should probably be lower as the potential gains are that it likely helps everyone it doesn't kill (who would have died shortly of the disease anyways).
Basically the reason so many cures are stuck at that early stage is that drug companies don't see a potential for profit because the system has become so cumbersome. I'm not normally a free-marketeer but perhaps the balance should be tipped back to the drug companies a bit for some classes of drugs?
Since the disease leads to paralysis then death how safe does it have to be to be effective? If the cure kills 5% of the people that take it I would think that will be less than the 10 year delay in getting a "perfect" cure out of the lab and through FDA testing.
It's $60/month for BB data plans with tethering and they pull from one shared 5GB/month pool so actually a worse deal than Verizon's plan. Though I have to say I wish we had the option to pay for more data. We have a DR solution for our remote sites that involves a router that can pair with our phones, and even at the stupid high prices they are charging it would probably be worth it. Actually, I just figure out that for most of our offices they could just rotate which phone got plugged into the router each day so we should be fine on data cap as it's 5GB per employee at the office.
No, genetic testing of the virus is EXPENSIVE and generally not needed at this point as we are so early in the normal flu season that it is obvious on the face that the majority of cases are related to H1N1.
Yes, on the show tonight they had someone who had the same issue with I believe it was a Highlander that did not have the mats that were implicated in the recall.
Seriously? People are freaked out by redlining their vehicle? Wow! Yeah, that would be an almost weekly occurrence for me, downshift (in an automatic!) and then stomp the gas to merge. Of course my first car was a manual so I actually understand how the whole clutch, downshift, accelerate thing works.
ABS wasn't even an option on most cards until the mid 90's and many of the systems available before ~2000 were hydraulic (my dad brought me one from a customer who was having rusting issues with the inner valve in a system they made for GM).
(FLOPS/W)/latency For non embarrassingly parallel jobs it won't matter how efficiently you can compute if you can't communicate the results between nodes.
From the numbers I gave in reply to a sibling post it's closer to 2-3lbs per sq ft where I live, but that's still several times the concentration of the atmospheric carbon. The problem is that arable landmass is a fairly small percentage of the earths surface. I'm not saying using biomass sequestration isn't useful, it's just not going to fight heightened CO2 levels on its own (not to suggest you were saying it is).
Or use a pair of them like the Sun Unified storage cluster using the 7310/7410. Of course Sun charges a fairly hefty fee for what you get (I got 72x450GB 15k drives in my EVA6400 for what they charge for the same storage is SATA and mine included 5 years of support).
We have a strict Dev/Test/Prod/DR model due to both internal and external compliance requirements so for every app we need at least 4 OS images. In the past that meant 4 servers, now it's simply 4 VM's. Also for n-tier apps we can provide multiple front end servers behind the load balancer for reliability, again without needing additional lightly utilized servers (per environment!) All of this means that my power and hardware budgets are decreasing even as I provide more useful services to my clients, and as I said we were able to not buildout a new datacenter space which would have cost ~$1M.
For us on the midend storage vmotion provides all the benefits that storage virtualization promised without the drawbacks. Not sure you are going to run a Fortune10 enterprise that way but it sure is nice knowing I can migrate to another SAN without downtime or expensive migration tools or services.
Huh? There are XPembedded based thin terms with 128MB flash and 512MB ram. They are basically Citrix/RDP clients that can work with winprinters and load a local IE session. They boot in seconds and can be locked down even tighter than the most well locked workstation. They also sell for $200 or less. If you don't need IE or printer support there are even lower powered solutions running a variety of OS's.
For most workload profiles vSphere has a 3-5% overhead penalty, given all the advantages it provides I'd say it's well worth it. Things like per machine snapshots aren't really something you can easily graft into an uber OS (how does it know what to snapshot for a given application)?
Slashdot Mirror
I'm more questioning the whole process of drug development and approval as it applies to diseases that are always fatal. I think the modern western approach has gone too far along the "it must be perfect for approval" path with the attached very high cost without showing that it has eliminated or even greatly reduced risk vs a more relaxed approval approach. Sure for a drug like Viagra perhaps the bar should be set high as the gain from the drug is fairly minimal in comparison to the potential problems but if you are talking about eliminating Parkinsons or non-Hodgkin's lymphoma then the bar should probably be lower as the potential gains are that it likely helps everyone it doesn't kill (who would have died shortly of the disease anyways).
Basically the reason so many cures are stuck at that early stage is that drug companies don't see a potential for profit because the system has become so cumbersome. I'm not normally a free-marketeer but perhaps the balance should be tipped back to the drug companies a bit for some classes of drugs?
Since the disease leads to paralysis then death how safe does it have to be to be effective? If the cure kills 5% of the people that take it I would think that will be less than the 10 year delay in getting a "perfect" cure out of the lab and through FDA testing.
Huh, how is T-Mobile not credible?
It's $60/month for BB data plans with tethering and they pull from one shared 5GB/month pool so actually a worse deal than Verizon's plan. Though I have to say I wish we had the option to pay for more data. We have a DR solution for our remote sites that involves a router that can pair with our phones, and even at the stupid high prices they are charging it would probably be worth it. Actually, I just figure out that for most of our offices they could just rotate which phone got plugged into the router each day so we should be fine on data cap as it's 5GB per employee at the office.
You must have a pre Rev-A card, Rev-A get about 1.9-2.1 Mbps, older cards were capped at ~950Kbps max.
Dude, it's a town of 7,000 people, the old ladies sitting on the porches will be sufficient to enforce any ban.
No, genetic testing of the virus is EXPENSIVE and generally not needed at this point as we are so early in the normal flu season that it is obvious on the face that the majority of cases are related to H1N1.
Hmm, I know of a company that started out as a specialist in aftermarket support for DG boxes, wonder how hard they have looked for replacement parts.
We have a winner. Earliest FPS with location detection I can remember.
They'll only potentially save you if they are GFCI, standard breakers will let you complete the circuit quite long enough to fry you.
Yes, on the show tonight they had someone who had the same issue with I believe it was a Highlander that did not have the mats that were implicated in the recall.
Seriously? People are freaked out by redlining their vehicle? Wow! Yeah, that would be an almost weekly occurrence for me, downshift (in an automatic!) and then stomp the gas to merge. Of course my first car was a manual so I actually understand how the whole clutch, downshift, accelerate thing works.
ABS wasn't even an option on most cards until the mid 90's and many of the systems available before ~2000 were hydraulic (my dad brought me one from a customer who was having rusting issues with the inner valve in a system they made for GM).
Could be worse, could be a Merc with the damn jog dial.
(FLOPS/W)/latency
For non embarrassingly parallel jobs it won't matter how efficiently you can compute if you can't communicate the results between nodes.
From the numbers I gave in reply to a sibling post it's closer to 2-3lbs per sq ft where I live, but that's still several times the concentration of the atmospheric carbon. The problem is that arable landmass is a fairly small percentage of the earths surface. I'm not saying using biomass sequestration isn't useful, it's just not going to fight heightened CO2 levels on its own (not to suggest you were saying it is).
Or use a pair of them like the Sun Unified storage cluster using the 7310/7410. Of course Sun charges a fairly hefty fee for what you get (I got 72x450GB 15k drives in my EVA6400 for what they charge for the same storage is SATA and mine included 5 years of support).
We have a strict Dev/Test/Prod/DR model due to both internal and external compliance requirements so for every app we need at least 4 OS images. In the past that meant 4 servers, now it's simply 4 VM's. Also for n-tier apps we can provide multiple front end servers behind the load balancer for reliability, again without needing additional lightly utilized servers (per environment!) All of this means that my power and hardware budgets are decreasing even as I provide more useful services to my clients, and as I said we were able to not buildout a new datacenter space which would have cost ~$1M.
For us on the midend storage vmotion provides all the benefits that storage virtualization promised without the drawbacks. Not sure you are going to run a Fortune10 enterprise that way but it sure is nice knowing I can migrate to another SAN without downtime or expensive migration tools or services.
Huh? There are XPembedded based thin terms with 128MB flash and 512MB ram. They are basically Citrix/RDP clients that can work with winprinters and load a local IE session. They boot in seconds and can be locked down even tighter than the most well locked workstation. They also sell for $200 or less. If you don't need IE or printer support there are even lower powered solutions running a variety of OS's.
For most workload profiles vSphere has a 3-5% overhead penalty, given all the advantages it provides I'd say it's well worth it. Things like per machine snapshots aren't really something you can easily graft into an uber OS (how does it know what to snapshot for a given application)?
It saved us from having to do a $1M datacenter upgrade so yeah, I'd say it benefited us.
Lots of it. According to this paper 10.2 ± 2.8 kg m^2 in the upper 1-m depth.
Well, they sort of already are part of an umbrella corp or keiretsu, they are part of the Toyota Group.
Because then people would buy more cars from their former competitors who are not so focused on sustainability?