Sorry, I typoed decoding for encoding, but I'd be surprised if youtube doesn't do decoding too, behind the scenes, for various things (scaling down a video comes to mind).
Exactly. Google would have to implement hardware decoding of theora, and then roll that out to all of its servers. We're talking about a huge investment there, so, lamentably, that's a non-starter. My guess is google is also using H264 for their highest resolution format, which means transcoding to theora would also introduce significant generation loss, whereas downcoding H264 to H264 is doable with far fewer artifacts.
Saying the issue was bandwidth was bogus, but how often are companies straight with customers about their internal reasons for doing or not doing something?
The thing that peeves me about silverlight is that MS gets to brag about having a "multi-platform" tool, when in reality, apple and novell do the work porting the code to for osx and linux, respectively. Having a semi-open spec which doesn't rule out third party implementations just isn't the same as having a tool that is officially released on multiple platforms like flash is.
The main thing as far as I can tell (besides from having a few proprietary drivers built in) is that ubuntu server has newer versions of programs than debian stable does (but is still "stable"). If you need the latest version of mysql or php for something, for example, you might consider ubuntu server over compiling it yourself. But server software like that doesn't change as rapidly as desktop stuff, so it's not much of an advantage. Hence why very few people use it. Another big thing is support. My hosting company "supports" ubuntu, but not debian (we decided we'd support ourselves). And some vendors sell with ubuntu support contracts. But the short answer is there is no compelling reason you are missing.
Seriously. CS asks questions of the form will X accomplish Y? Software Engineering asks questions of the form "what is the best way X to accomplish y"?
Your description is of an airplane going from just below stall speed to an airspeed that could damage the aircraft before the pilot is able to regain control. I don't think a "high-speed dive" could explain that. Once you're in a dive it should be pretty easy to get the nose up.
I think you're missing that stall speed is much higher at altitude than at sea level. So even though a plane like that might have a canonical stall speed on the order of ~150 mph for landing with full flaps down, the stall speed is much higher at 35k feet. Add in a little disorientation or any number of other possible factors and its entirely plausible.
Going into something like a flat spin may be a possibility; I have no idea on how vulnerable airliners are to unrecoverable spins.
The general rule of thumb is the bigger the aircraft, the more vulnerable it is. Small prop planes are usually recoverable from spins. Large, fast planes avoid them.
That sounds plausible, except the last I heard there weren't any radio transmissions during the episode. Wouldn't you expect at least 1 of 3 pilots to be screaming into the headset?
Screaming into the headset? Yes. On the radio? No. You have to press a button in order to talk on the radio. The headsets pilots have on are only used for intra-cabin communications.
I have a question unrelated to this fuss. How exactly does G1 work, and why is it expected to be "better" (no credit for answers that just say the current java GC is crap).
According to economic theory, in most markets you get two market leaders--e.g. Coke and Pepsi, Bud and Miller, Ford and GM--and I don't see FOSS as any different.
[ Citation needed ] -- which economic theory is that?
That's what the F-35 is for (still needing manned aircraft for awhile). The F-22s the US has will keep giving it an edge against the countries who are sold F-35s (not to mention the export version will be somewhat stripped down, of course). There's really no need for more F-22s, UAVs or no.
Slashdot Mirror
Sorry, I typoed decoding for encoding, but I'd be surprised if youtube doesn't do decoding too, behind the scenes, for various things (scaling down a video comes to mind).
Exactly. Google would have to implement hardware decoding of theora, and then roll that out to all of its servers. We're talking about a huge investment there, so, lamentably, that's a non-starter. My guess is google is also using H264 for their highest resolution format, which means transcoding to theora would also introduce significant generation loss, whereas downcoding H264 to H264 is doable with far fewer artifacts.
Saying the issue was bandwidth was bogus, but how often are companies straight with customers about their internal reasons for doing or not doing something?
The thing that peeves me about silverlight is that MS gets to brag about having a "multi-platform" tool, when in reality, apple and novell do the work porting the code to for osx and linux, respectively. Having a semi-open spec which doesn't rule out third party implementations just isn't the same as having a tool that is officially released on multiple platforms like flash is.
parent is not flamebait - its insightful.
Is it not possible that apple is getting ready to rebrand ZFS as iFiles or something?
The main thing as far as I can tell (besides from having a few proprietary drivers built in) is that ubuntu server has newer versions of programs than debian stable does (but is still "stable"). If you need the latest version of mysql or php for something, for example, you might consider ubuntu server over compiling it yourself. But server software like that doesn't change as rapidly as desktop stuff, so it's not much of an advantage. Hence why very few people use it. Another big thing is support. My hosting company "supports" ubuntu, but not debian (we decided we'd support ourselves). And some vendors sell with ubuntu support contracts. But the short answer is there is no compelling reason you are missing.
Seriously. CS asks questions of the form will X accomplish Y? Software Engineering asks questions of the form "what is the best way X to accomplish y"?
Nothing to see here, move along.
[citation needed]
Your description is of an airplane going from just below stall speed to an airspeed that could damage the aircraft before the pilot is able to regain control. I don't think a "high-speed dive" could explain that. Once you're in a dive it should be pretty easy to get the nose up.
I think you're missing that stall speed is much higher at altitude than at sea level. So even though a plane like that might have a canonical stall speed on the order of ~150 mph for landing with full flaps down, the stall speed is much higher at 35k feet. Add in a little disorientation or any number of other possible factors and its entirely plausible.
Going into something like a flat spin may be a possibility; I have no idea on how vulnerable airliners are to unrecoverable spins.
The general rule of thumb is the bigger the aircraft, the more vulnerable it is. Small prop planes are usually recoverable from spins. Large, fast planes avoid them.
That sounds plausible, except the last I heard there weren't any radio transmissions during the episode. Wouldn't you expect at least 1 of 3 pilots to be screaming into the headset?
Screaming into the headset? Yes. On the radio? No. You have to press a button in order to talk on the radio. The headsets pilots have on are only used for intra-cabin communications.
See google, it's not that hard, so stop crying about standards and get your act together. It's easier to criticize than to create.
Was that supposed to be hilarious? Because it was.
I have a question unrelated to this fuss. How exactly does G1 work, and why is it expected to be "better" (no credit for answers that just say the current java GC is crap).
Oh, snap!
since when does solaris have a microkernel?
GNU HURD is no longer 2-5 years away. Even the debian port has stopped being remotely useful.
Interestingly, I see absolutely no bluetooth icon on my Kubuntu 8.10 machine (can't risk upgrading yet), until I turn it on (via the hardware switch).
On many if not most laptops, bluetooth is implemented as a USB device. Try a sudo lsusb -v | grep -i bluetooth. I've got something like this:
Bus 003 Device 002: ID 13d3:3249 IMC Networks
Device Descriptor:
bLength 18
bDescriptorType 1
bcdUSB 2.00
bDeviceClass 224 Wireless
bDeviceSubClass 1 Radio Frequency
bDeviceProtocol 1 Bluetooth
[snip]
huh? what about TIA/EIA-568-B ?
Yeah, I'm sure their email servers would let that through. And who the hell modded that informative?
According to economic theory, in most markets you get two market leaders--e.g. Coke and Pepsi, Bud and Miller, Ford and GM--and I don't see FOSS as any different.
[ Citation needed ] -- which economic theory is that?
wow, look at the size of this thread. all about a point that is addressed in TFA -- on the first page no less.
That's what the F-35 is for (still needing manned aircraft for awhile). The F-22s the US has will keep giving it an edge against the countries who are sold F-35s (not to mention the export version will be somewhat stripped down, of course). There's really no need for more F-22s, UAVs or no.
Surely you're joking? stunnel costs an exec() per connection.
...4-way redundant power supply...
What on earth is that for? I understand wanting 2, or 3 just to be extra safe, but 4? How do you lose 3 of 4 without noticing?
He meant comparable to. Not as an example of global warning. You, sir, are an idiot for jumping to conclusions.
Please mod racist troll parent down. Thx.