# Neither NOAA nor commercial AM/FM stations can possibly give neighborhood granularity coverage fast enough.
Not true at all. AM News stations do a great job. You may be in an unfortunately poor area for selection.
by time a weather event is verified enough to get into the update cycle, it has probably passed you. NOAA transmitters are pathetically weak and placed in locations where their line of sight coverage is abysmal.
All good reasons why you have 5 different NOAA stations in any given area... Different rotations, different propagation, etc.
but to date there are no portable battery operated televisions capable of receiving a DTV signal.
I've been posting links to portable DTVs for idiots like yourself for a good 5 years now. Get the fuck out of Walmart and look around, dammit. A quick web search will find them.
My point here being that after all of the money spent on DTV, it is within 5 years of being irrelevant thanks to youtube and similar video services and more efficient codecs.
Video (and audio) codecs can't get notably more efficient. MPEG-2 gets surprisingly close to the upper limit of imperceptible lossy compression. No amount of math or hand-waving by non-experts can change the fundamental limit of entropy. All newer codecs can do is make low quality videos not look so artifact-ridden... At HDTV bitrates, they give you slightly more than nothing.
TFA suggests a method of collecting water, for free, from the air without requiring any additional energy input.
It's not "free" in the slightest. It has an extremely high initial cost, and huge operating costs, due to deterioration of solar panels and fatigue of all mechanical parts.
You're suggesting they build a 200km pipe, plus desalination plant, to pump water across miles of hostile desert.
By the above logic, all of that is "free" too... right?
And the desalination plant is definitely "free" because you need precisely that for separating the water from the brine in the air-based system as well.
Assuming TFA's invention isn't very expensive
That's a terrible and baseless assumption.
You also utterly fail to consider VOLUME. Even a tiny pipeline will move a HUGE amount of water, whenever you want. You can't, however, gather much water from the air.
I was expecting to see another machine based on the ammonia absorption cycle. I was pleasantly surprised to see something new.
If you're surprised, you just haven't ever paid attention. The affinity of salt to attract water has been known forever, as has the ability to heat the salt and collect the water.
In fact there was a/. story close to a year ago about a company with a similar product or trying to sell the tech to the US Military for use in Iraq.
The problem with such systems has always been the high salinity of the collected condensate, and the energy hit in purifying it much further, to the level of potability, or use in agriculture.
I am always surprised by how much water gets to the garden.
Humidity is relative. A desert can have much more moisture in the air than a much colder, much more humid area. It's just that, at 50C degrees, the air can hold much more water than it can at 10C degrees. So the same amount of water that makes the desert 15% relative humidity, can result in rain (100% relative humidity) in colder climate.
It's the same thing that allows far more sugar/salt/jello/etc. to dissolve in warm water than cold...
Every company has policies for just about everything they do... If you don't have them, the first thing you need to do is establish policies for IT.
No company of more than 5 people just hires, promotes, and fires people ad-hoc. Neither can IT be expected to support everyone doing everything their own way. If people fail to follow the policy, you tell them where to appeal to have the policy changed, or drop it entirely and require doing it the right way.
Once that's out of the way, the other issue is scheduling. When someone requests your assistance, you tell them when you will be able to get to it... If they insist it's the most important thing in the world, you simply repeat that you don't have the time to do it before then. This is doubly true for those who try to pass jobs off on you 15 minutes before you leave for the day. When faced with missing a deadline because of your schedule, even the worst procrastinators will only fall on their face once before they take the lesson to heart.
If it is, instead, your higher-ups that are the problem, the fix is even simpler. NEVER present the option of putting in extra hours. When new priorities come up, ask which current tasks should be dropped to make time, and don't accept a non-answer. If extra hours become a common occurrence, you simply need to do some contract renegotiation. For a substantial pay raise, I'm happy to work the extra hours requested. If you're not, you may be in the wrong line of work, and need to let your superiors know it's either 9-5 for you, or they need to find someone else... In either case, when faced with additional cash outlay, you might well find that your bosses come to realize that many of those projects aren't so pressing after all.
I want a universal filesystem Which one can you mount on Linux, MacOS and maybe even Windows without precarious hacks, and with journaling, long filenames, and maybe extended attributes?
UFS would be a good candidate, as Linux, BSD, and OS X include support for it, and a couple Windows drivers are available, though I can't comment on their reliability.
No, UFS doesn't have have journaling, but asking for an unseen technical feature is a bit like asking for a specific model of car... There's probably some feature you want, but you can't mentally separate the feature from the "journaling" technology you've heard in the same breath... If you're looking for fault tolerance, or fast recovery without waiting for fsck, that can be done better WITHOUT a journal, as in the case of UFS2 on FreeBSD, and ZFS on FreeBSD and [Open]Solaris. In fact UFS is easily the most reliable filesystem I've ever dealt with, and features like softdeps and UFS2 are improving the performance beyond any other general purpose filesystems out there now.
When removable drives are nearing the terabyte marker, exceeding the reasonable limits for FAT32, and NTFS still isn't open to 3rd parties, UFS seems likely to be the only option.
In fact ATOM uses anywhere from the same amount of power in the best case, to much less than an order of magnitude in the worse...
And even if it were an order of magnitude difference in power consumption as you claim, you're utterly wrong that ARM will outperform x86 solutions. The claim has been made forever by interested parties, and it's never been true. ARM has never been designed for performance equivalent to even the lowest-end x86 CPUs, and ATOM CPUs easily outperform OMAP3 packages, just not as much as they historically always have.
And I must admire how you single-out Geode... Comparing an extremely, EXTREMELY OLD x86 CPU design based on much older tech and vastly larger fab processes (from 10 years ago), with a practically brand-new ARM solution. Just going out of your way to make the comparison as unfair as possible?
Even the Geode has an embarrassingly high power consumption (close to 7W for a complete system, excluding display),
The "excluding display" comment makes it obvious you in-fact know the problem with your claim, but continue to pretend that it doesn't exist...
Cutting the power consumption of the CPU any further than Geode/ATOM has, just doesn't have a market... Not because it's not possible, but simply because other factors begin to dominate. The display is a huge one. Even if you can cut your CPU power consumption by 50%, you're only cutting overall system power consumption by perhaps 10%. The display obviously dominates, and there's no sign of any near-future technology that will substantially reduce that power requirement by a significant amount.
I would never set loose a drive, or my data. I keep it safely locked up inside the case...
Of course you could use another RAID-0 array as your backup.
No kidding. If you were sane, however, you would probably just use a single drive, 4X as large as the drives in your RAID-0 array, and use RSync to frequently copy the changes over.
Basically the greater the data the more valuable it becomes
Not at all. HDTV shows are no more valuable than SD TV shows. Lossless music is no more valuable than MP3s.
It must be noted that what I have just touched on are enterprise solutions which are way beyond the purchase of home systems.
Imagine if you have 100TB drives but they only do sequential transfers at 200MB/sec and are still stuck at about 10milliseconds access time (7200rpm).
Fine with me. rsync rocks. Tape drives, and particularly optical drives (CDs/DVDs/MOs), have FAR WORSE performance characteristics, and they all refuse to die.
Maybe we'll just see Flash take the place of smaller HDDs, and large slow HDDs take the place of tapes and most uses of optical media.
Really with 1TB of HD space there isn't anything you can't have a lot of.
Sure there is.
HDTV takes about 8GBytes/hour. That's just ~120 hours on a a terabyte HDD, or perhaps a month of TV viewing. Now, you're not going to want to save EVERYTHING you view, but you'll probably fill-up a second terabyte drive within a year. Quicker if you're saving blu-ray movies as well.
So, while it isn't as big of a constraint as it once was, we could still use more space...
On the other hand I/O, especially magnetic I/O is the main bottleneck.
So RAID-0 your drives... Instantly double the speed. And if you really can't find a way to use the space, buy a pair of 500GB drives instead of 1x 1TB drive. With the popularity of SATA RAID controllers, most systems can easily enough have 4 HD drives in a RAID set. Are you suggesting that 4x the transfer speed of the best 10kRPM drives still isn't fast enough for you?
Upscaling the video is done slowly through software, even though Overlays surfaces have been around since 1997 with the NVidia Riva 128.
Yes they have. And Flash added support for them... when??? Still less than a year ago. And Flash is a MATURE bit of software that's been used for video for many years now. Firefox with video is still just a beta yet.
top-end Core 2 Duo chips would run circles around your P4, but apparently you'd prefer to pretend they don't exist.
He is making a point about the diminishing value of multiple cores. The fact that the industry has (finally) gotten back up to the point where a single core is clocked faster than a 10 year-old chip doesn't negate the point at all...
And if you can't find a video codec with multiple core support, you're looking in the wrong place. Video decode is one of those embarrassingly-easy things to parallelize,
EXCEPT for H.264, where decoding is highly serialized. The only way to thread it to a significant extent is by skipping some of those steps, and sacrificing quality... That's what the likes of CoreAVC do, hoping nobody notices. That's why Blu-Ray video is divided up into quadrants.
And since threading doesn't work for H.264, why bother? Any other codec is efficient enough that even high bitrate, highdef video can be decoded with a single core on much older machines.
And I'm pretty sure that today's chip designs allow them to be faster per core than your old 3.5GHz P4,
The fact that single-core speeds have now climbed back up certainly does not negate the point he was making. Scalar speed is very important, and exponential diminishing returns result from attempting threading with most common CPU-intensive tasks.
I would think video encoding could be parallelised quite nicely too.
And, for the most part, you'd be wrong. In the worst case, you get about a 10% speed-up even with a well-designed codec.
Actually, after a quick google - avidemux, ffmpeg and mencoder have supported threads for some time.
This is called knowing just enough to be dangerous...
Yes, libavcodec has a "threads" option. It just happens to suck, however. Slice-based threading works wonderfully in a very narrow, controlled scenario, but does next to nothing in most cases.
And for decoding, with H.264 in particular, it's really not possible to get much of a speed-boost with threading... at least not without cheating and scarificing quality by skipping steps, like some commercial decoders seem to do.
Point taken. I mixed-up BPP and Bits/Color in my back of the napkin calculations.
Of course, for all your whining, it's only one very small point which doesn't change the result... ie. Gigabit ethernet is still vastly insufficient. And all my other points remain.
Perhaps the real bread and circuses is all this whining about copyright while your nation fights two wars, has out of control military spending, locks up non-violent drug offenders, arrests medical marijuana growers, denies rights to gays, is in the middle of an economic meltdown, has out of control gun laws, etc etc etc, yet here we are arguing the minutia of copyright law.
Good job. Dismiss his bullshit logic with some even worse false logic of your own.
You're suggesting that if we all drop everything else, we will be able to solve all the major problems in the world, and just work our way down the list... Reality is quite the opposite, really.
You can stop bathing until you've achieved world peace, but the time saved won't gain you world peace, and you'll just go around stinking.
Try this... Don't bother changing the oil in your car. It's not important enough. Just keep going until your car blows up. THEN your car blowing up will be important enough to merit your attention.
Well, the answer is quite obvious then, compress the video
You really want to add a $200+ board to EVERY DEVICE in your home theater?
Computational power is cheaper than HDMI cables anyuwhere you look.
You have no idea the computational requirements of realtime highdef video compression.
And Best Buy isn't "anywhere". Head over to Target and you'll get the cable for $10. Go online and you'll get the cable for $5.
And you don't solve anything... USB and Ethernet cables at Best Buy will continue to be unconscionably expensive... Never mind the "Highdef Video" Ethernet cables they would surely produce.
You might as well advocate flying cars to get around the high mark-up of steering-wheel covers.
Forgive me for not having kept up with the progress of HDMI,
...nor having the most basic knowledge of the topic at hand, correct?
wouldn't it have made infinitely more sense to have simply used gigabit Ethernet for all this?
Oh, my, yes. When transferring 16 Gigabits/sec of uncompressed HD video at high frame rates from your DVD player to your TV, what you really want is a 1 Gigabit/sec standard, designed for unreliable communications over 500 meter distances, using a shared-channel, with LOTS of overhead, and very high computational requirements...
IP addresses don't identify people. They tell routers where to forward packets.
Yeah, in the same way license-plate numbers don't identify people, then just happen to correspond with a name and physical address on file to where bills can be sent, which are them paid by someone... Who know who?
Any reasonably-sized network block has ownership information associated with it. As do the AS numbers used in the BGP announcements, known to those "routers" doing the packet forwarding...
Can we please move beyond this 1980s idea that IP addresses identify people?
Why? Nobody complains that your car, or house, or anything else, identify you, even though it's easily possible to allow someone entirely different to use them in your absence.
The great thing about serial ports as console ports go is they are simple to drive. That means they can be activated very early in the boot process without needing complex dedicated hardware.
USB ports, from the very beginning, were already enabled extremely early in the boot process... If that were not the case, your USB keyboard would be utterly worthless.
Similarly, bootable USB Floppies, CDs, HDDs, etc, are supported by every recent BIOS. And, obviously, if you can load up a USB to serial converter, it's far cheaper, easier, and more featureful, to instead use USB native protocols for such communications.
Slashdot Mirror
Yes, but THAT ISN'T AGAINST THE LAW, so it's not grounds for a warrant, or any other action by the police.
Not true at all. AM News stations do a great job. You may be in an unfortunately poor area for selection.
All good reasons why you have 5 different NOAA stations in any given area... Different rotations, different propagation, etc.
I've been posting links to portable DTVs for idiots like yourself for a good 5 years now. Get the fuck out of Walmart and look around, dammit. A quick web search will find them.
Video (and audio) codecs can't get notably more efficient. MPEG-2 gets surprisingly close to the upper limit of imperceptible lossy compression. No amount of math or hand-waving by non-experts can change the fundamental limit of entropy. All newer codecs can do is make low quality videos not look so artifact-ridden... At HDTV bitrates, they give you slightly more than nothing.
It's not "free" in the slightest. It has an extremely high initial cost, and huge operating costs, due to deterioration of solar panels and fatigue of all mechanical parts.
By the above logic, all of that is "free" too... right?
And the desalination plant is definitely "free" because you need precisely that for separating the water from the brine in the air-based system as well.
That's a terrible and baseless assumption.
You also utterly fail to consider VOLUME. Even a tiny pipeline will move a HUGE amount of water, whenever you want. You can't, however, gather much water from the air.
Less than 200KM to the nearest ocean. Start laying some pipes...
That's very likely more efficient than trying to pull water from the air.
If you're surprised, you just haven't ever paid attention. The affinity of salt to attract water has been known forever, as has the ability to heat the salt and collect the water.
In fact there was a /. story close to a year ago about a company with a similar product or trying to sell the tech to the US Military for use in Iraq.
The problem with such systems has always been the high salinity of the collected condensate, and the energy hit in purifying it much further, to the level of potability, or use in agriculture.
Humidity is relative. A desert can have much more moisture in the air than a much colder, much more humid area. It's just that, at 50C degrees, the air can hold much more water than it can at 10C degrees. So the same amount of water that makes the desert 15% relative humidity, can result in rain (100% relative humidity) in colder climate.
It's the same thing that allows far more sugar/salt/jello/etc. to dissolve in warm water than cold...
The above is the best advice given so far.
Every company has policies for just about everything they do... If you don't have them, the first thing you need to do is establish policies for IT.
No company of more than 5 people just hires, promotes, and fires people ad-hoc. Neither can IT be expected to support everyone doing everything their own way. If people fail to follow the policy, you tell them where to appeal to have the policy changed, or drop it entirely and require doing it the right way.
Once that's out of the way, the other issue is scheduling. When someone requests your assistance, you tell them when you will be able to get to it... If they insist it's the most important thing in the world, you simply repeat that you don't have the time to do it before then. This is doubly true for those who try to pass jobs off on you 15 minutes before you leave for the day. When faced with missing a deadline because of your schedule, even the worst procrastinators will only fall on their face once before they take the lesson to heart.
If it is, instead, your higher-ups that are the problem, the fix is even simpler. NEVER present the option of putting in extra hours. When new priorities come up, ask which current tasks should be dropped to make time, and don't accept a non-answer. If extra hours become a common occurrence, you simply need to do some contract renegotiation. For a substantial pay raise, I'm happy to work the extra hours requested. If you're not, you may be in the wrong line of work, and need to let your superiors know it's either 9-5 for you, or they need to find someone else... In either case, when faced with additional cash outlay, you might well find that your bosses come to realize that many of those projects aren't so pressing after all.
UFS would be a good candidate, as Linux, BSD, and OS X include support for it, and a couple Windows drivers are available, though I can't comment on their reliability.
No, UFS doesn't have have journaling, but asking for an unseen technical feature is a bit like asking for a specific model of car... There's probably some feature you want, but you can't mentally separate the feature from the "journaling" technology you've heard in the same breath... If you're looking for fault tolerance, or fast recovery without waiting for fsck, that can be done better WITHOUT a journal, as in the case of UFS2 on FreeBSD, and ZFS on FreeBSD and [Open]Solaris. In fact UFS is easily the most reliable filesystem I've ever dealt with, and features like softdeps and UFS2 are improving the performance beyond any other general purpose filesystems out there now.
When removable drives are nearing the terabyte marker, exceeding the reasonable limits for FAT32, and NTFS still isn't open to 3rd parties, UFS seems likely to be the only option.
You've got the cause and effect backwards.
There's nothing about cars that FORCES cities to sprawl out. Certainly, there are plenty of car owners in the most dense cities in the world.
Cars are simply most popular in areas where a lot of people WANTED a lot of land / space to themselves, such as the suburbs.
This is completely baseless. See point #1 here: http://bec-systems.com/site/326/intel-atom-vs-ti-omap3
In fact ATOM uses anywhere from the same amount of power in the best case, to much less than an order of magnitude in the worse...
And even if it were an order of magnitude difference in power consumption as you claim, you're utterly wrong that ARM will outperform x86 solutions. The claim has been made forever by interested parties, and it's never been true. ARM has never been designed for performance equivalent to even the lowest-end x86 CPUs, and ATOM CPUs easily outperform OMAP3 packages, just not as much as they historically always have.
And I must admire how you single-out Geode... Comparing an extremely, EXTREMELY OLD x86 CPU design based on much older tech and vastly larger fab processes (from 10 years ago), with a practically brand-new ARM solution. Just going out of your way to make the comparison as unfair as possible?
The "excluding display" comment makes it obvious you in-fact know the problem with your claim, but continue to pretend that it doesn't exist...
Cutting the power consumption of the CPU any further than Geode/ATOM has, just doesn't have a market... Not because it's not possible, but simply because other factors begin to dominate. The display is a huge one. Even if you can cut your CPU power consumption by 50%, you're only cutting overall system power consumption by perhaps 10%. The display obviously dominates, and there's no sign of any near-future technology that will substantially reduce that power requirement by a significant amount.
The question was performance, not redundancy.
I would never set loose a drive, or my data. I keep it safely locked up inside the case...
No kidding. If you were sane, however, you would probably just use a single drive, 4X as large as the drives in your RAID-0 array, and use RSync to frequently copy the changes over.
Not at all. HDTV shows are no more valuable than SD TV shows. Lossless music is no more valuable than MP3s.
Yes, it's called ranting... You're doing it well.
Fine with me. rsync rocks. Tape drives, and particularly optical drives (CDs/DVDs/MOs), have FAR WORSE performance characteristics, and they all refuse to die.
Maybe we'll just see Flash take the place of smaller HDDs, and large slow HDDs take the place of tapes and most uses of optical media.
Sure there is.
HDTV takes about 8GBytes/hour. That's just ~120 hours on a a terabyte HDD, or perhaps a month of TV viewing. Now, you're not going to want to save EVERYTHING you view, but you'll probably fill-up a second terabyte drive within a year. Quicker if you're saving blu-ray movies as well.
So, while it isn't as big of a constraint as it once was, we could still use more space...
So RAID-0 your drives... Instantly double the speed. And if you really can't find a way to use the space, buy a pair of 500GB drives instead of 1x 1TB drive. With the popularity of SATA RAID controllers, most systems can easily enough have 4 HD drives in a RAID set. Are you suggesting that 4x the transfer speed of the best 10kRPM drives still isn't fast enough for you?
Yes they have. And Flash added support for them... when??? Still less than a year ago. And Flash is a MATURE bit of software that's been used for video for many years now. Firefox with video is still just a beta yet.
Number of (english-language) stations I received via analog: 1 (Riverside, CA)
Number of stations with (currently-lower power) digital: 10
He is making a point about the diminishing value of multiple cores. The fact that the industry has (finally) gotten back up to the point where a single core is clocked faster than a 10 year-old chip doesn't negate the point at all...
EXCEPT for H.264, where decoding is highly serialized. The only way to thread it to a significant extent is by skipping some of those steps, and sacrificing quality... That's what the likes of CoreAVC do, hoping nobody notices. That's why Blu-Ray video is divided up into quadrants.
And since threading doesn't work for H.264, why bother? Any other codec is efficient enough that even high bitrate, highdef video can be decoded with a single core on much older machines.
The fact that single-core speeds have now climbed back up certainly does not negate the point he was making. Scalar speed is very important, and exponential diminishing returns result from attempting threading with most common CPU-intensive tasks.
And, for the most part, you'd be wrong. In the worst case, you get about a 10% speed-up even with a well-designed codec.
Actually, after a quick google - avidemux, ffmpeg and mencoder have supported threads for some time.
This is called knowing just enough to be dangerous...
Yes, libavcodec has a "threads" option. It just happens to suck, however. Slice-based threading works wonderfully in a very narrow, controlled scenario, but does next to nothing in most cases.
And for decoding, with H.264 in particular, it's really not possible to get much of a speed-boost with threading... at least not without cheating and scarificing quality by skipping steps, like some commercial decoders seem to do.
Point taken. I mixed-up BPP and Bits/Color in my back of the napkin calculations.
Of course, for all your whining, it's only one very small point which doesn't change the result... ie. Gigabit ethernet is still vastly insufficient. And all my other points remain.
Good job. Dismiss his bullshit logic with some even worse false logic of your own.
You're suggesting that if we all drop everything else, we will be able to solve all the major problems in the world, and just work our way down the list... Reality is quite the opposite, really.
You can stop bathing until you've achieved world peace, but the time saved won't gain you world peace, and you'll just go around stinking.
Try this... Don't bother changing the oil in your car. It's not important enough. Just keep going until your car blows up. THEN your car blowing up will be important enough to merit your attention.
No. 1080 @ 60fps 32bpp ==~ 16GBit/set. I didn't even bother to include audio.
And HDMI can do much MORE than that, which can be utilized by displays some time in the future.
8bpp (uncompressed) would look terrible. That's just 256 colors. Basic VGA? Where did you pull that number from?
Ah, yes... Speaking of pulling numbers from your ass... 60 is very close to 325, right? RIGHT?
You really want to add a $200+ board to EVERY DEVICE in your home theater?
You have no idea the computational requirements of realtime highdef video compression.
And Best Buy isn't "anywhere". Head over to Target and you'll get the cable for $10. Go online and you'll get the cable for $5.
And you don't solve anything... USB and Ethernet cables at Best Buy will continue to be unconscionably expensive... Never mind the "Highdef Video" Ethernet cables they would surely produce.
You might as well advocate flying cars to get around the high mark-up of steering-wheel covers.
Oh, my, yes. When transferring 16 Gigabits/sec of uncompressed HD video at high frame rates from your DVD player to your TV, what you really want is a 1 Gigabit/sec standard, designed for unreliable communications over 500 meter distances, using a shared-channel, with LOTS of overhead, and very high computational requirements...
Your insight is... stunning.
No, they only identify the last publicly accessible router in the chain. The user could well be on the other side of the planet.
VPN tunnels? Proxies? Multihomed ISPs? Nation-wide cellular internet service with static IP addresses? etc.
Yeah, in the same way license-plate numbers don't identify people, then just happen to correspond with a name and physical address on file to where bills can be sent, which are them paid by someone... Who know who?
Any reasonably-sized network block has ownership information associated with it. As do the AS numbers used in the BGP announcements, known to those "routers" doing the packet forwarding...
Why? Nobody complains that your car, or house, or anything else, identify you, even though it's easily possible to allow someone entirely different to use them in your absence.
USB ports, from the very beginning, were already enabled extremely early in the boot process... If that were not the case, your USB keyboard would be utterly worthless.
Similarly, bootable USB Floppies, CDs, HDDs, etc, are supported by every recent BIOS. And, obviously, if you can load up a USB to serial converter, it's far cheaper, easier, and more featureful, to instead use USB native protocols for such communications.