We tried it at my company. For years! We finally shifted to Google Apps, and we're quite happy...
Here are the problems with DIY servers:
Uptime - given that we are not, and have never been, a service provider, our connection was never a datacenter-multiple-redundancies-T3-five-nines-uptime thing. At the time, it was a pair of DSLs with static IPs. Run through an elderly Linux box that acted as the router/NAT and mail server. I don't remember our uptime figures, but it was definitely not perfect.
Speeds - again, not a backbone connection. We had to choose between allowing large attachments and slowing down everyone's browsers.
Maintenance - every time there was a problem with the server, one of us, who knew how it was setup or operated had to break whatever we were doing and spend half a day troubleshooting. Yeah yeah, "have a dedicated IT department"... Makes very little sense for 30 people...
Google gives us each 7 GB, a calendar that we can access anywhere, shared contacts, freakin' phone mail and lots more besides!
Frankly, it's just better value for smaller outfits.
As for losing mails, which do you think is more probably going to lose mails? A hacked-together old box running an ancient copy of sendmail, or Google? Besides, if your mails are really that important, download them using IMAP and back them up as you would anyway! That's what we do...
He resigned, partly over conflicts of interest regarding Android. If he did take up Apple again, presumably he'd have to resign as Executive Chairman of Google for the same reason...
Modest consequences? Flooding of all existing coastal zones is "modest"? Crop failure due to shifted monsoon patterns in the second most populous country in the world is "modest"? Desertification in all of sub-Saharan Africa is "modest"?
In the (geological) past, warming was driven by the sun, and CO2 and methane levels lagged the initial warming, and went into a feedback cycle that caused more warming, until the sun cooled the system down again. Greenhouse effect was secondary, the sun was primary. But greenhouse effect usually accelerated the process.
Now, however, with solar activity at a low, we should be heading for an ice age, but temperatures are going up. And greenhouse effect is going up too. The question here is, "if the sun isn't causing it, what is?", and one extremely likely answer is that greenhouse effect is forcing an accelerated temperature rise, in the same way it did in the geological past. Only, this time, it's doing so without the sun's help.
It also occurs to me to wonder... what would be so BAD about another "Medieval Warm Period", making *practical* arable and habitable places like the Greenland coast, central Canada, and parts of Siberia? Yeah, you might sacrifice a relatively smaller area elsewhere as desert, but wouldn't it be a net gain for human habitability?
Well, for one thing, it would make various other parts of the world a desert - and not necessarily small parts of it either. Much of the Sahara was green during the dawn of the Anthropocene.
For another, it wouldn't be so bad if it stopped at the Goldilocks point, I guess - where we have "just enough" warming for an MWP, but not "too much" so that it leads to desertification of all of Africa. But the problem is, it most probably won't.
It's generally difficult to predict any of these things.
As to species preservation, all well and good, but species come and go all the time; that's the nature of a non-static biosphere.
Seems to my our job is to adapt as needed like any other viable species, not to attempt to freezeframe nature at some theoretically optimal point, lest the nonviable perish. What happens when your freezeframe inevitably collapses and you're stuck with a biosphere that's not *had* to adapt, and is now a large Fail?
Nobody's saying that we shouldn't adapt, but we want changes to be at a gradual pace, so that we can adapt, and the biosphere can adapt. If it changes too fast, it's quite possible that conditions go pear-shaped too fast for us to adapt. If we can slow the change, that is a form of adaptation.
Because, rest assured, the climate will change drastically 4-odd billion years from now (or even a million years from now), but at least it will be at a pace that we can keep up with.
By which time it may be too late to actually do anything about it.
"I have no evidence that my chain smoking hundred cigarettes will give me cancer. I'm studying the issue and will be able to back up that claim after a few decades of putting tar and nicotine into my lungs"
Fair enough, but in that case, I'd point to GPL3 adoption; a 2009 article claims about 50% adoption, at least of active projects on Google Code. That's not bad, I think...
Remember, when you measure success or failure, don't think in binary terms - world domination is not necessary for success.
It would be interesting to see the numbers, wouldn't it? By all accounts, Clang is just about able to compile a Linux kernel - it's going to take it a while to catch up to GCC, and that's after enough things compile with it.
But it's a good thing that it exists - competition is always good, even if only to keep the GCC team on their toes.
Some of the very basic tests I once did show them (MSVC 9.0/Visual Studio 2008 and GCC 4.5-odd) running pretty much neck and neck, without much real advantage to either one or the other. It really comes down to compiler, code and the specific optimization flags one uses. Otherwise, they're both pretty much the same.
Here's someone else's benchmark, showing GCC and MSVC just about the same, and Intel's ICC about 20% or more faster. By your logic, ICC should be the world's favourite compiler, since it's a) the fastest, and b) available on all platforms.
The real world doesn't work like that, though - ICC has other problems...
Mingw32 on Windows is horribly slow in getting developed beyond version 3.4.somethingorother. Also, it can't interoperate with Windows DLLs (made with MSVC) without some hackery. Ergo, it's not so common in Windows, though as others have mentioned, it's more popular than you might expect.
On the other hand, gcc is used on Macs (until recently, it was pretty much exclusive, but now it's lost ground to LLVM, of course), and there's a version for pretty much every embedded platform one would want to use. I think it's safe to say that it's at least one of the top two compilers around.
I doubt that's going to be particularly difficult for 3rd party devs to do - the libavcodec version is only 1400 lines. I'm sure that there's enough reusable code for decoder blocks floating around to assemble an OpenCL version of it...
Eventually, the community decided to move on, and founded MUD Wiki, a Wikia dedicated to the genre.
Exactly! I'd expect to find specific information about obscure Star Trek characters (even those I consider important for some obscure reason) on Memory Alpha, and not in Wikipedia. A link from main Wikipedia to the MUD wiki, explaining that more information is available there seems appropriate. IIRC, such things have been done in other Wiki articles...
Apart from the fact that CUDA/OpenCL/whatever processors are useless for highly I/O dependent tasks, try doing a conditional on any of them and watch your performance drop like crazy...
SIMD has its place, but a multi-core supercomputer is effectively MIMD, and can actually run parallel tasks (or parallely run independent parts of a task).
Thanks for explaining, but honestly, didn't all that give anyone a headache?
Real-world currency works because it's bloody simple to use: give the shopkeeper a few coins or notes, and he gives you an ice-cream. Unless you can simplify it down to something like that, where I don't have to worry about hashes and debt chains, I doubt it's going to really gain traction.
But experimenting with such systems does give us a really good idea of what's actually going on behind those bits of green/blue/brown paper and shiny circular pieces of metal we keep exchanging for goods and services.
Slashdot Mirror
Worse... The one they did build ended up releasing a good helping of Cs 137 and Sr 90 into the local area.
This is an absolute "Wish I had mod points" moment...
I just wish all the extremists on all sides would get this!
You mean Iranian, I believe...
We tried it at my company. For years! We finally shifted to Google Apps, and we're quite happy...
Here are the problems with DIY servers:
Maintenance - every time there was a problem with the server, one of us, who knew how it was setup or operated had to break whatever we were doing and spend half a day troubleshooting. Yeah yeah, "have a dedicated IT department"... Makes very little sense for 30 people...
Frankly, it's just better value for smaller outfits.
As for losing mails, which do you think is more probably going to lose mails? A hacked-together old box running an ancient copy of sendmail, or Google? Besides, if your mails are really that important, download them using IMAP and back them up as you would anyway! That's what we do...
He resigned, partly over conflicts of interest regarding Android. If he did take up Apple again, presumably he'd have to resign as Executive Chairman of Google for the same reason...
Modest consequences? Flooding of all existing coastal zones is "modest"? Crop failure due to shifted monsoon patterns in the second most populous country in the world is "modest"? Desertification in all of sub-Saharan Africa is "modest"?
Which ivory tower do you live in?
Yep yep yep all true.
In the (geological) past, warming was driven by the sun, and CO2 and methane levels lagged the initial warming, and went into a feedback cycle that caused more warming, until the sun cooled the system down again. Greenhouse effect was secondary, the sun was primary. But greenhouse effect usually accelerated the process.
Now, however, with solar activity at a low, we should be heading for an ice age, but temperatures are going up. And greenhouse effect is going up too. The question here is, "if the sun isn't causing it, what is?", and one extremely likely answer is that greenhouse effect is forcing an accelerated temperature rise, in the same way it did in the geological past. Only, this time, it's doing so without the sun's help.
It also occurs to me to wonder... what would be so BAD about another "Medieval Warm Period", making *practical* arable and habitable places like the Greenland coast, central Canada, and parts of Siberia? Yeah, you might sacrifice a relatively smaller area elsewhere as desert, but wouldn't it be a net gain for human habitability?
Well, for one thing, it would make various other parts of the world a desert - and not necessarily small parts of it either. Much of the Sahara was green during the dawn of the Anthropocene.
For another, it wouldn't be so bad if it stopped at the Goldilocks point, I guess - where we have "just enough" warming for an MWP, but not "too much" so that it leads to desertification of all of Africa. But the problem is, it most probably won't.
It's generally difficult to predict any of these things.
As to species preservation, all well and good, but species come and go all the time; that's the nature of a non-static biosphere.
Seems to my our job is to adapt as needed like any other viable species, not to attempt to freezeframe nature at some theoretically optimal point, lest the nonviable perish. What happens when your freezeframe inevitably collapses and you're stuck with a biosphere that's not *had* to adapt, and is now a large Fail?
Nobody's saying that we shouldn't adapt, but we want changes to be at a gradual pace, so that we can adapt, and the biosphere can adapt. If it changes too fast, it's quite possible that conditions go pear-shaped too fast for us to adapt. If we can slow the change, that is a form of adaptation.
Because, rest assured, the climate will change drastically 4-odd billion years from now (or even a million years from now), but at least it will be at a pace that we can keep up with.
By which time it may be too late to actually do anything about it.
"I have no evidence that my chain smoking hundred cigarettes will give me cancer. I'm studying the issue and will be able to back up that claim after a few decades of putting tar and nicotine into my lungs"
Fair enough, but in that case, I'd point to GPL3 adoption; a 2009 article claims about 50% adoption, at least of active projects on Google Code. That's not bad, I think...
Remember, when you measure success or failure, don't think in binary terms - world domination is not necessary for success.
It would be interesting to see the numbers, wouldn't it? By all accounts, Clang is just about able to compile a Linux kernel - it's going to take it a while to catch up to GCC, and that's after enough things compile with it.
But it's a good thing that it exists - competition is always good, even if only to keep the GCC team on their toes.
You mean, apart from developing it?
Some of the very basic tests I once did show them (MSVC 9.0/Visual Studio 2008 and GCC 4.5-odd) running pretty much neck and neck, without much real advantage to either one or the other. It really comes down to compiler, code and the specific optimization flags one uses. Otherwise, they're both pretty much the same.
Here's someone else's benchmark, showing GCC and MSVC just about the same, and Intel's ICC about 20% or more faster. By your logic, ICC should be the world's favourite compiler, since it's a) the fastest, and b) available on all platforms.
The real world doesn't work like that, though - ICC has other problems...
GCC is generally only used on platforms where MSVC isn't available
Pretty much every place except Windows, then...
Mingw32 on Windows is horribly slow in getting developed beyond version 3.4.somethingorother. Also, it can't interoperate with Windows DLLs (made with MSVC) without some hackery. Ergo, it's not so common in Windows, though as others have mentioned, it's more popular than you might expect.
On the other hand, gcc is used on Macs (until recently, it was pretty much exclusive, but now it's lost ground to LLVM, of course), and there's a version for pretty much every embedded platform one would want to use. I think it's safe to say that it's at least one of the top two compilers around.
I doubt that's going to be particularly difficult for 3rd party devs to do - the libavcodec version is only 1400 lines. I'm sure that there's enough reusable code for decoder blocks floating around to assemble an OpenCL version of it...
Right when somebody implements the codec on OpenCL or something, I guess...
Hmm... I'd call GCC pretty successful...
FTFA:
Exactly! I'd expect to find specific information about obscure Star Trek characters (even those I consider important for some obscure reason) on Memory Alpha, and not in Wikipedia. A link from main Wikipedia to the MUD wiki, explaining that more information is available there seems appropriate. IIRC, such things have been done in other Wiki articles...
What's the fuss about?
FTP down, nonexistent or blocked in a client's building. You need to transfer a few hundred megs of data. Rapidshare to the rescue.
Cuda != supercomputers.
Apart from the fact that CUDA/OpenCL/whatever processors are useless for highly I/O dependent tasks, try doing a conditional on any of them and watch your performance drop like crazy...
SIMD has its place, but a multi-core supercomputer is effectively MIMD, and can actually run parallel tasks (or parallely run independent parts of a task).
They really address different segments.
Wait, you're using Uncyclopedia as a citation?
Really, what else is one to think?
I wonder if they actually have a graph display...
Thanks for explaining, but honestly, didn't all that give anyone a headache?
Real-world currency works because it's bloody simple to use: give the shopkeeper a few coins or notes, and he gives you an ice-cream. Unless you can simplify it down to something like that, where I don't have to worry about hashes and debt chains, I doubt it's going to really gain traction.
But experimenting with such systems does give us a really good idea of what's actually going on behind those bits of green/blue/brown paper and shiny circular pieces of metal we keep exchanging for goods and services.
You're thinking Switzerland, not Sweden