There seems to be a lot of misconceptions about how this thing works. I can only assume the misinformation comes from the wired article which I can not read (slashdotted?).
I googled 'Washington university blood analyzer DxBox'
The research is lead from Washington University, with grants from the Bill & Melinda Gates Foundation, is apparently 'Open technology', and is currently called DxBox (a play on Microsoft xbox since BillG is funding a lot of the work).
No your bullshit. You are basing your conclusions on a stupid wired magazine article that was written for consumer consumption. The device does RNA/DNA amplfications on a credit card sized piece of platic (replacing the refrigerators) as well as flourecense tests you talk about, and a bunch of other test also. I saw the presentation on Washington University TV and it is quite real.
I don't think the descriptions presented here do the device justice. I just happen to catch the PowerPoint presentation from the University of Washington on TV. The device uses chemical pathways printed on a plastic card by a standard ink jet kind of printer. A drop of blood is put in that card, and by 'pumping' the blood around the card it can heat-cycle samples (I think this was some sort of DNA amplifier), It also had membrane 'grids' for spectro type analysis, as well as pass the sample through optical observation stations (for sizing things). It provides 2 different diagnostic tests and results for each test performed (about 6 per card), to insure both results agree.
I am not a medical student by any means, yet in this 1 hour PowerPoint I understood how the device was going to work and why and it was brilliant. The guy explained how it was going to cost less than $200 so poor countries could afford it, and why it worked in 10 minutes v. 10 hours for current technology.
It was hugely funded in part by that evil man who wants to take over the world. What's his name? Oh yeah, Bill & Melinda Gates foundation.
"Now, if Bill G were really interested in changing the world's health... perhaps he'd get on-board with this obvious idea. Who knows. He's got a lot of money."
For some reason, I was bored and just happen to watch the PowerPoint presentation on this thing last week on TV from Washington University. It was quite an impressive device, with many possibilities for multiple uses as part of the design criteria.
And the Bill & Melinda Gates foundation was a huge contributor.
Point me to some IDE edit, compile and debug benchmarks with response time measurements, for windows and Linux that shows what you say and I may reconsider. But in my experience doing exactly that, it is not even close.
If a vm is not good for real time testing, how in the world could it be faster for non real time use? We agree on replacing old servers, I have used it for that purpose before and was happy with the results.
Actually, yes I tried it 30 years ago. Wrote some of the virtual memory paging code in IBM VM/370 - one of the original virtualizing products for multi million dollar mainframes. I also tried it earlier this month, both Microsoft's and VM Ware's. And while I don't have the same intimate knowledge of the code I used to have, I probably have a better understanding of what goes on in virtualization than most people.
But anyhow, your arguments are weak. You say it runs fine. I say it does not run as good as it would native. We are not contradicting each other in any way, but you are putting forth nothing to argue against my statement. And yet you claim it is me that sucks. Eat me asshole.
I was/am in the same boat as you. I noticed a huge drop in memory and concentration ability about 5 years ago. Like you I set off on a journey to find a solution, and what I found was
I don't care what anyone else says, in my experience virtualization sucks for most of the development process.
Years ago (20+?) IBM did studies that statistically proved that sub second response time was critical to productivity. As more and more stuff moves to the web we seem to have forgotten that fact, but it is in my experience very very true. Now I don't care how much money you spend, a virtualized development environment will never be as responsive as a native environment on the same hardware. Fast response time = productivity. You can keep a fast pace on your train of thought. Hit enter and wait 5 seconds for a response and your mind wanders. You lose your concentration.
Virtualization makes sense for replacing legacy hardware servers and for a portion of the development process, Quality Assurance/Testing. But keep the design and code writing native. Take your fastest machine and put your DEV IDE on it. Take your next fastest machine and put your database/web server on it. Or for a better model put them on separate machines, but keep your fastest machine for the IDE. And true, some can do great dev work without an IDE (like with vi/emacs & command line) but that is something for the very experienced, you know the type that has every API call memorized. The rest of us need robustness and the fast context sensitive help of a modern IDE.
Sigh, truly a DBAs idea of what is good enough for developers. MS has indeed had the problem of understanding what a programmer wants.
Everything I have read or seen indicated it wasn't implemented because they just didn't have time to finish it up, not some design limitation. In fact if LINQ is a stepping stone to Microsoft's persistent object strategy and they left it out, then they are pretty much clueless. NHIbernate works as expected, DevExpress XPO works as expected, Bindable LINQ adds it to some extent, and the list goes on. I would be willing to bet some release of LINQ in the near future will properly implement it.
Yeah I understand what you are saying about it being a lazy sequence, but that is a DBAs take on it. Anders was (thank gosh) not a DBA type. He explicitly stated the goal as trying to reduce the mismatch between Code and Databases. As a coder I could care less what the underlying mechanism is, I want to write code with objects and collections, and have the act like objects and collections. If I wanted to write code that talked to a database then what is already in place is fine.
If the intention was a static result set then why would I be able to make changes to an object and persist those changes with a 'submit' statement? I don't think the intentions are static at all. In my code now, I do not even define a query (yes the underlying codebase might). I reference DB.MyTable as a collection - there is nothing in that that conveys static. When I add a row to it, I expect that additional row to reflected in the collection (and the underlying table). I think you nailed the problem on the head, someone with a clue came up with a good idea, but left it to a bunch of database oriented clueless developers to implment it and they spun their impression of how it should be on it.
It had been a while since I had written any code, but I have been doing some again lately, and thought I'd share with you on it.
The problem I have is that LINQ doesn't implement IObservableCollection. You get a set from LINQ and bind it to a listbox for example. Since the properties DO implement INotifyPropertyChanged whenever you make changes they do appear in the UI as expected, but, if you add entities (insertOnSubmit) or delete (deleteOnSubmit) the UI does not get updated. This is the kind of disconnect I am talking about (along with the general disconnect on sortable views etc.) It seems MS half implemented data binding.
I finally did get around to installing SP1 - of course it broke as much as it fixed:( I had to undo the kludges I made before SP1 (now fixed), and then add in new ones for post SP1. MS is just having problems keeping up with the complexity of their own designs.
Yes, I will agree I think WPF is a huge step in the right direction - now if everything else would just catch up to it. It's night and day ahead of flash.
What were some of the tools that you used? My experience with a couple I tried was, like MS, only half implemented.
But how much of your shrink wrapped product was plumbing that should have been built in? Or put another way, how much of what you wrote will be replaced by library code in a future release? Like I said, it's a natural progression, often used code SHOULD be replaced by library code, it is just a matter of where you draw the line between a mature product and one that is having some growing pains.
I'll admit I have not looked in depth at the SP1 beta/previews (is it out for real yet?), and a lot of my frusteration is having to hand code XAML because Blend is such a half finished product. I'm in no hurry, I'll wait until it ferments some more.
"it's not hard to just use ObservableCollection everywhere"
yes, it is hard. Try coupling it to a sortable collection. How many Listviews have you ever seen that you couldn't click a header and sort by that column? I can't think of any. And thus the chaos - when you actually try to tie it all together you MUST write code yourself. And in my experience, the code is not insignificant, it falls apart easily. I am sure they (MS) will work it all out, but at the moment it (WPF/LINQ) fails for real use on anything other than a flashy demo.
There was a lot of new tech in the last couple of.net releases - and unfortunately they are all not in sync at all.
It's no big deal, anytime that much new comes out in so many areas it takes a while to get them synced, but it's a little chaotic now.
Specifically; new GUI paradigm (XAML/WPF/Silverlight) and new Data Access (LINQ) - the standard collections don't have INotifyPropertyChange support across the board, SortedCollections are hit and miss, just in general I have found that interfaces needed for one new component is not well implemented for other new components. Like I said, just a bit of growing pains, but it needs attention.
But I'll agree it has nothing to do with a new language being introduced. I doubt if that will have any affect one way or the other.
After the 'earth at the center of the universe with the sun rotating around it' screw up, the Catholic Church is being a lot more cautious about what they say now. But, I have personally met many non-Catholic Christians who tell me the Bible says that there is no life in the universe except on Earth. Their world is about to crumble. I assume other religions based on the same God, eg; some Jews and Muslims are also about to be in for a shocker.
The ensuing mass panic (I hope) is indeed a threat to national security as the extremist try and save us from any such blasphemy. Of course having a supreme loony in the top job isn't going to help.
beats the hell out of me. I called to get an install. Was told it was the last day they were accepting orders. When I found out who the installing company was I tried 3 others because of the bad reputation of the one I had gotten. All three told me "not accepting any more orders" so I went with the first one. I already have it (and it sucks) but i have been told they are not accepting any more - so I'm not lying about that part - whether or not they actually are I have no idea. 37857 if you care. With the pitiful speed im getting (and I paid for the premium) I can't believe its legal or moral for them to continue selling.
I did not mention it, but I did have T1. Bell south, now AT&T provided a t1 via contract. That contract included a 'one day credit for each hour down time' clause. When I decided to demand credit for the constant downtime they turned the t1 off. I sued them, won the case, and yet they walked away without having to re-activate t1 service or pay credit for the downtime. Like I said, drop $5 in the right hands and get whatever you want around here.
AS my main point was - a lot of people assume there are options, when there is not, and all the stupid comments just proves my point.
And you are doing it at my location? bullshit. you are the one who is lying. You obviously don't install sats because if you did you would know it goes by which bird services your location and you wouldnt make a false blanket statement like that.
fair enough - 2 companies is not a monopoly, but it certainly isnt competition either. It is real obvious the demand exceeds the supply. Neither company is even accepting new installs at the moment. During peak hours, the speed delivered is somewhere around 24.4 even though i am paying for 'up to 1.5k'.
And you mentioned I 'keep using that word' even though I used it once.... hmmm pot calling the kettle black?
Apparently you do not understand the meaning of 'I dont have options'
for reasons unknown to you, I can not move.
Since I don't have money to throw away, starting a broadband company is a rather stupid idea - minimum costs are around $45k perhaps you have that laying around and I would be happy to spend it for you.
I made it a point to say I wasnt whining about my situation, only reminded some of you that just because you have broadband you should not think everyone can have it. Sounds to me like you are the uninformed snob doing the whining.
"Those people in rural area's still have the ability to get high speed internet, such as satellite, direct line of site towers, cellular or even DSL."
People who don't have to deal with are very misinformed about what is available. There is no cellular or towers available. DSL isn't even remotely feasable. And sattelite is so over sold by the 2 monopolies that the speed is OFTEN less than the 24.4 tops dial up that is available from 2 carriers.
Yes, were I live sucks big time. I made the mistake of thinking coverage would eventually be available, but its not. Around here (southern VA, east TN) a $50 dollar bribe to a cop and you can still get away with murder. It's the old west. I dont see things changing any time soon.
But no, I don't expect anyone to do anything to help poor old me out. But just don't go around thinking I have options available, I don't.
It is a shame you did not get to post this earlier where it is more prominent, and hopefully modded up.
I agree. The real question is not evolution v. ID. The real question is 'is God really some sort of being?' and for many 'who had a son?'
I actually like when stuff like this Ben Stein thing happens as it raises the awareness of the few people who can think on there own. What the US and the world in general needs is to bring this whole 'God as a being who writes books' thing to very high level so that it can be treated as witchcraft once was. By the way, I have a lot more faith in there having been actual witches who practice forgotten sciences, than a 'God being'. But anyhow, it takes events like 9/11 and the recent polygamist sect raid in the us to put these religeous fanatics in the spotlight for what they truly are; people using the fear of God to control the weak minded and to justify otherwise unthinkable acts.
In the English language (I assume it is not your native language) ending a sentence with a period is made to emphasize that the statement is as close to fact as possible. Also in the English language (and I assume other languages) any and every written word is considered the opinion of the writer - it is impossible to write a fact. 2 + 2 = 4 is really an opinion, since it implies understanding of the meaning of symbols, and while we all generally agree on the meaning of those symbols, it cannot be a fact.
There are a lot of comments as to how an 'accident could happen'. What I am saying is that I worked on these systems. There are entire departments and consulting contracts dedicated to proper handling of email. It is not some sendmail server running in the corner. There is not 'an IT guy' responsible for backups. These are high paid professionals whose job requires the professional handling of white house correspondence. Things do not disapear because they are classified, they are handled specifically different. Yes, hard drives are destroyed when replaced, but not before a dozen checkmarks on a list.
If there are deleted or missing emails, it was intentional, period.
Slashdot Mirror
There seems to be a lot of misconceptions about how this thing works. I can only assume the misinformation comes from the wired article which I can not read (slashdotted?).
Anyhow here are some links I found on google;
http://uwnews.washington.edu/ni/article.asp?articleID=3045
http://www.rsc.org/publishing/journals/AN/article.asp?doi=b705672a
http://www.rsc.org/delivery/_ArticleLinking/DisplayHTMLArticleforfree.cfm?JournalCode=LC&Year=2008&ManuscriptID=b811158h&Iss=Advance_Article
http://arjournals.annualreviews.org/doi/abs/10.1146/annurev.bioeng.10.061807.160524
I googled 'Washington university blood analyzer DxBox'
The research is lead from Washington University, with grants from the Bill & Melinda Gates Foundation, is apparently 'Open technology', and is currently called DxBox (a play on Microsoft xbox since BillG is funding a lot of the work).
No your bullshit. You are basing your conclusions on a stupid wired magazine article that was written for consumer consumption. The device does RNA/DNA amplfications on a credit card sized piece of platic (replacing the refrigerators) as well as flourecense tests you talk about, and a bunch of other test also. I saw the presentation on Washington University TV and it is quite real.
I don't think the descriptions presented here do the device justice. I just happen to catch the PowerPoint presentation from the University of Washington on TV. The device uses chemical pathways printed on a plastic card by a standard ink jet kind of printer. A drop of blood is put in that card, and by 'pumping' the blood around the card it can heat-cycle samples (I think this was some sort of DNA amplifier), It also had membrane 'grids' for spectro type analysis, as well as pass the sample through optical observation stations (for sizing things). It provides 2 different diagnostic tests and results for each test performed (about 6 per card), to insure both results agree.
I am not a medical student by any means, yet in this 1 hour PowerPoint I understood how the device was going to work and why and it was brilliant. The guy explained how it was going to cost less than $200 so poor countries could afford it, and why it worked in 10 minutes v. 10 hours for current technology.
It was hugely funded in part by that evil man who wants to take over the world. What's his name? Oh yeah, Bill & Melinda Gates foundation.
"Now, if Bill G were really interested in changing the world's health... perhaps he'd get on-board with this obvious idea. Who knows. He's got a lot of money."
For some reason, I was bored and just happen to watch the PowerPoint presentation on this thing last week on TV from Washington University. It was quite an impressive device, with many possibilities for multiple uses as part of the design criteria.
And the Bill & Melinda Gates foundation was a huge contributor.
Point me to some IDE edit, compile and debug benchmarks with response time measurements, for windows and Linux that shows what you say and I may reconsider. But in my experience doing exactly that, it is not even close.
If a vm is not good for real time testing, how in the world could it be faster for non real time use? We agree on replacing old servers, I have used it for that purpose before and was happy with the results.
Actually, yes I tried it 30 years ago. Wrote some of the virtual memory paging code in IBM VM/370 - one of the original virtualizing products for multi million dollar mainframes. I also tried it earlier this month, both Microsoft's and VM Ware's. And while I don't have the same intimate knowledge of the code I used to have, I probably have a better understanding of what goes on in virtualization than most people.
But anyhow, your arguments are weak. You say it runs fine. I say it does not run as good as it would native. We are not contradicting each other in any way, but you are putting forth nothing to argue against my statement. And yet you claim it is me that sucks. Eat me asshole.
I was/am in the same boat as you. I noticed a huge drop in memory and concentration ability about 5 years ago. Like you I set off on a journey to find a solution, and what I found was
What were we talking about?
Oh yeah, I never found anything that helped.
I don't care what anyone else says, in my experience virtualization sucks for most of the development process.
Years ago (20+?) IBM did studies that statistically proved that sub second response time was critical to productivity. As more and more stuff moves to the web we seem to have forgotten that fact, but it is in my experience very very true. Now I don't care how much money you spend, a virtualized development environment will never be as responsive as a native environment on the same hardware. Fast response time = productivity. You can keep a fast pace on your train of thought. Hit enter and wait 5 seconds for a response and your mind wanders. You lose your concentration.
Virtualization makes sense for replacing legacy hardware servers and for a portion of the development process, Quality Assurance/Testing. But keep the design and code writing native. Take your fastest machine and put your DEV IDE on it. Take your next fastest machine and put your database/web server on it. Or for a better model put them on separate machines, but keep your fastest machine for the IDE. And true, some can do great dev work without an IDE (like with vi/emacs & command line) but that is something for the very experienced, you know the type that has every API call memorized. The rest of us need robustness and the fast context sensitive help of a modern IDE.
Sigh, truly a DBAs idea of what is good enough for developers. MS has indeed had the problem of understanding what a programmer wants.
Everything I have read or seen indicated it wasn't implemented because they just didn't have time to finish it up, not some design limitation. In fact if LINQ is a stepping stone to Microsoft's persistent object strategy and they left it out, then they are pretty much clueless. NHIbernate works as expected, DevExpress XPO works as expected, Bindable LINQ adds it to some extent, and the list goes on. I would be willing to bet some release of LINQ in the near future will properly implement it.
Yeah I understand what you are saying about it being a lazy sequence, but that is a DBAs take on it. Anders was (thank gosh) not a DBA type. He explicitly stated the goal as trying to reduce the mismatch between Code and Databases. As a coder I could care less what the underlying mechanism is, I want to write code with objects and collections, and have the act like objects and collections. If I wanted to write code that talked to a database then what is already in place is fine.
If the intention was a static result set then why would I be able to make changes to an object and persist those changes with a 'submit' statement? I don't think the intentions are static at all. In my code now, I do not even define a query (yes the underlying codebase might). I reference DB.MyTable as a collection - there is nothing in that that conveys static. When I add a row to it, I expect that additional row to reflected in the collection (and the underlying table). I think you nailed the problem on the head, someone with a clue came up with a good idea, but left it to a bunch of database oriented clueless developers to implment it and they spun their impression of how it should be on it.
It had been a while since I had written any code, but I have been doing some again lately, and thought I'd share with you on it.
:( I had to undo the kludges I made before SP1 (now fixed), and then add in new ones for post SP1. MS is just having problems keeping up with the complexity of their own designs.
The problem I have is that LINQ doesn't implement IObservableCollection. You get a set from LINQ and bind it to a listbox for example. Since the properties DO implement INotifyPropertyChanged whenever you make changes they do appear in the UI as expected, but, if you add entities (insertOnSubmit) or delete (deleteOnSubmit) the UI does not get updated. This is the kind of disconnect I am talking about (along with the general disconnect on sortable views etc.) It seems MS half implemented data binding.
I finally did get around to installing SP1 - of course it broke as much as it fixed
Yes, I will agree I think WPF is a huge step in the right direction - now if everything else would just catch up to it. It's night and day ahead of flash.
What were some of the tools that you used? My experience with a couple I tried was, like MS, only half implemented.
I'm not saying it can't be done.
But how much of your shrink wrapped product was plumbing that should have been built in? Or put another way, how much of what you wrote will be replaced by library code in a future release? Like I said, it's a natural progression, often used code SHOULD be replaced by library code, it is just a matter of where you draw the line between a mature product and one that is having some growing pains.
I'll admit I have not looked in depth at the SP1 beta/previews (is it out for real yet?), and a lot of my frusteration is having to hand code XAML because Blend is such a half finished product. I'm in no hurry, I'll wait until it ferments some more.
"it's not hard to just use ObservableCollection everywhere"
yes, it is hard. Try coupling it to a sortable collection. How many Listviews have you ever seen that you couldn't click a header and sort by that column? I can't think of any. And thus the chaos - when you actually try to tie it all together you MUST write code yourself. And in my experience, the code is not insignificant, it falls apart easily. I am sure they (MS) will work it all out, but at the moment it (WPF/LINQ) fails for real use on anything other than a flashy demo.
There was a lot of new tech in the last couple of .net releases - and unfortunately they are all not in sync at all.
It's no big deal, anytime that much new comes out in so many areas it takes a while to get them synced, but it's a little chaotic now.
Specifically; new GUI paradigm (XAML/WPF/Silverlight) and new Data Access (LINQ) - the standard collections don't have INotifyPropertyChange support across the board, SortedCollections are hit and miss, just in general I have found that interfaces needed for one new component is not well implemented for other new components. Like I said, just a bit of growing pains, but it needs attention.
But I'll agree it has nothing to do with a new language being introduced. I doubt if that will have any affect one way or the other.
OMG. I just read the wiki. You have got to be kidding me, you think this is a good thing?
Now where is the link to the source code and how can I verify that it is the code that was really running on the machines?
After the 'earth at the center of the universe with the sun rotating around it' screw up, the Catholic Church is being a lot more cautious about what they say now. But, I have personally met many non-Catholic Christians who tell me the Bible says that there is no life in the universe except on Earth. Their world is about to crumble. I assume other religions based on the same God, eg; some Jews and Muslims are also about to be in for a shocker.
The ensuing mass panic (I hope) is indeed a threat to national security as the extremist try and save us from any such blasphemy. Of course having a supreme loony in the top job isn't going to help.
Bingo
Here in the good ole US it is Bell/Att who tells the courts what to do, not the other way around.
beats the hell out of me. I called to get an install. Was told it was the last day they were accepting orders. When I found out who the installing company was I tried 3 others because of the bad reputation of the one I had gotten. All three told me "not accepting any more orders" so I went with the first one. I already have it (and it sucks) but i have been told they are not accepting any more - so I'm not lying about that part - whether or not they actually are I have no idea. 37857 if you care. With the pitiful speed im getting (and I paid for the premium) I can't believe its legal or moral for them to continue selling.
I did not mention it, but I did have T1. Bell south, now AT&T provided a t1 via contract. That contract included a 'one day credit for each hour down time' clause. When I decided to demand credit for the constant downtime they turned the t1 off. I sued them, won the case, and yet they walked away without having to re-activate t1 service or pay credit for the downtime. Like I said, drop $5 in the right hands and get whatever you want around here.
AS my main point was - a lot of people assume there are options, when there is not, and all the stupid comments just proves my point.
And you are doing it at my location? bullshit. you are the one who is lying. You obviously don't install sats because if you did you would know it goes by which bird services your location and you wouldnt make a false blanket statement like that.
fair enough - 2 companies is not a monopoly, but it certainly isnt competition either. It is real obvious the demand exceeds the supply. Neither company is even accepting new installs at the moment. During peak hours, the speed delivered is somewhere around 24.4 even though i am paying for 'up to 1.5k'.
.... hmmm pot calling the kettle black?
And you mentioned I 'keep using that word' even though I used it once
Apparently you do not understand the meaning of 'I dont have options'
for reasons unknown to you, I can not move. Since I don't have money to throw away, starting a broadband company is a rather stupid idea - minimum costs are around $45k perhaps you have that laying around and I would be happy to spend it for you. I made it a point to say I wasnt whining about my situation, only reminded some of you that just because you have broadband you should not think everyone can have it. Sounds to me like you are the uninformed snob doing the whining.
"Those people in rural area's still have the ability to get high speed internet, such as satellite, direct line of site towers, cellular or even DSL."
People who don't have to deal with are very misinformed about what is available. There is no cellular or towers available. DSL isn't even remotely feasable. And sattelite is so over sold by the 2 monopolies that the speed is OFTEN less than the 24.4 tops dial up that is available from 2 carriers.
Yes, were I live sucks big time. I made the mistake of thinking coverage would eventually be available, but its not. Around here (southern VA, east TN) a $50 dollar bribe to a cop and you can still get away with murder. It's the old west. I dont see things changing any time soon.
But no, I don't expect anyone to do anything to help poor old me out. But just don't go around thinking I have options available, I don't.
It is a shame you did not get to post this earlier where it is more prominent, and hopefully modded up.
I agree. The real question is not evolution v. ID. The real question is 'is God really some sort of being?' and for many 'who had a son?'
I actually like when stuff like this Ben Stein thing happens as it raises the awareness of the few people who can think on there own. What the US and the world in general needs is to bring this whole 'God as a being who writes books' thing to very high level so that it can be treated as witchcraft once was. By the way, I have a lot more faith in there having been actual witches who practice forgotten sciences, than a 'God being'. But anyhow, it takes events like 9/11 and the recent polygamist sect raid in the us to put these religeous fanatics in the spotlight for what they truly are; people using the fear of God to control the weak minded and to justify otherwise unthinkable acts.
In the English language (I assume it is not your native language) ending a sentence with a period is made to emphasize that the statement is as close to fact as possible. Also in the English language (and I assume other languages) any and every written word is considered the opinion of the writer - it is impossible to write a fact. 2 + 2 = 4 is really an opinion, since it implies understanding of the meaning of symbols, and while we all generally agree on the meaning of those symbols, it cannot be a fact.
There are a lot of comments as to how an 'accident could happen'. What I am saying is that I worked on these systems. There are entire departments and consulting contracts dedicated to proper handling of email. It is not some sendmail server running in the corner. There is not 'an IT guy' responsible for backups. These are high paid professionals whose job requires the professional handling of white house correspondence. Things do not disapear because they are classified, they are handled specifically different. Yes, hard drives are destroyed when replaced, but not before a dozen checkmarks on a list.
If there are deleted or missing emails, it was intentional, period.