You know, other than the obvious slippery slope that is involved, I'd be extremely satisfied with a reasonable (ie: big) on-peak cap, if off-peak usage were either unmetered or significantly greater. But that's not how it seems to be playing out, at least in the States.
If it were, it would add an incentive for Netflix to implement just such a scheme. But without a distinction between peak/off-peak, there is no reason for them to do anything differently than they already are: Whether they watch a show at 6PM or 4AM, their customers are equally fucked by bandwidth caps.
I remember a time, many moons ago, where land-line telephone usage was handled similarly: During the day, long distance calls were expensive. At night (after 9, IIRC, usually) it was somewhat less expensive.
Of course, back then, the telephone network was still pretty heavily regulated. And, perhaps the Intarwebs should be, as well, if this stuff doesn't get sorted out in a sensible way on its own. (I'll be burned at the stake for even mentioning regulation, but I mentioned it anyway.)
I'm all for using the network more efficiently. Rewarding folks for using CDN-delivered services, and/or doing so at off-peak times would seem like a win for everyone.
Hell: Rewarding folks for choosing cheaper BitTorrent peers instead of preferring to use whoever is simply the fastest (even if they're on another continent!) would do a lot to improve the efficiency of things, but neither geography nor network topology seem to be a factor in peer selection.
Even Apple had to cave when it came to MP3 (they wouldn't sell it, but the iPod had to play it).
Your history is backwards:
The iPod first launched on October 23, 2001, and it played MP3s just fine at that time.
The iTunes Music Store opened on April 28, 2003. Prior to this, Apple didn't sell content for the devices.
Apple never "caved," they simply built an MP3 player which happened to be successful. Later on, they started selling content for it (in the arguably more space-efficient AAC format).
There's a bunch of old panoramic maps at the Library of Congress which are available in JPEG2000. The amount detail preserved in these offerings is astounding (at least to me).
I mention this only because, as already pointed out, nobody much bothers with the format and, and good examples can be hard to come by. And old maps are cool.
Windows has gone backward with audio hardware support for entertainment with Vista and 7, and has done a very thorough job of making things as dumb as possible, by crippling Directsound.
It used to be sophisticated, and support all kinds of hardware acceleration for reverberation, filtering, occlusion, and surround sound (both actual and simulated) with just about any number of speakers.
Now, it's up to developers to implement these functions in software, and the CPU does the heavy lifting (which admittedly isn't very heavy these days, but I buy dedicated sound hardware for a reason).
It has pretty much killed the market for sound cards, and the development of new consumer-accessible DSP tech, since they all do the same thing these days, no matter how fancy or capable the underlying hardware is: As far as software is concerned, they all just play back audio.
Unless you count clever hacks like Creative's Alchemy, that is. But by the time we start comparing platform-specific audio hacks between Linux and Windows the discussion has turned pretty well moot.
Right, because the sound library my applications use is exactly the kind of thing I should be thinking about as a user. Why the hell should I have to care what audio library I'm using? Why is that still an issue on Linux? It is 2011 for Chrissake.
Indeed, it is 2011. Linux distributions could just emulate Windows 7/Vista by eliminating all fancy sound card functions altogether, while moving backward in terms of hardware functionality and at the same time and presenting a simple (but very limited) API for software to deal with.
Would that be preferable?
As a current Windows 7 user (at least on the desktop), who has from time to time done reasonably serious audio work under both Windows and Linux (and FreeBSD for that matter), I'd like to say that I'd prefer choice.
And given the fact that any new "web startup" is at least 90% likely to fail, I think it's a perfect match.
If it never takes off, which is very likely on average, then he'll never be out more than $14.99/mo. If it does take off, he'll be sent scrambling for a more robust solution, but it's likely that he'll have a few weeks/months/years to get ready for that, and will then have the funds to back something more serious.
If that's "most" of them, as you say, then I've never had to cure "most" infections because they never happened to begin with.
TFA is about a program, running on the local computer, which proclaims quite persuasively to be part of Windows. Changing themes will do fuck-all to help folks see the difference.
If the malware takes control of the PC (which it does, in the context of the FA), then having a single, locally-attached backup disk isn't necessarily a good answer: It can destroy/disrupt the backup just as easily as it can anything else on that PC.
A well-thought-out rotation of backup media would help, but that's no good because it involves humans who simply won't do it.
This wouldn't be a problem, so much, with good online storage: Even Dropbox does a good job of keeping old copies of your data intact for a period of time. I simply want the concept extended to an entire disk, with metadata intact, to enable a bare metal recovery.
This, combined with extra, out-of-band human verification (SMS?) for when you Really, Really want to destroy backup data, would work well against malware.
(And, yeah: I did use System Restore eventually. I consider it to be a last resort, though, simply because I am ignorant as to the extent of its workings and I am prejudiced against system-level programs which do not provide meaningful feedback as to what they're doing.)
While I believe your advice is well-intentioned, it's really no good.
This only works if the malware isn't using existing Windows widgets for its displays.
If I were I Windows programmer (I'm not) and I were writing malware (good heavens!), I'd use the native toolkit for all of my dealings...just like most other software does. It's easier, that way.
And then: Changing themes, for properly-implemented malware, would also change the look of that ill program to match.
RAID would do nothing to protect against the thing described in TFA.
RAID only protects against hardware failure, and even then only if the failure is actually detected instead of just silently munging data.
This is not to say that RAID is not useful: It can be a performance boost in some applications. It can provide a clever way to combine many smaller disks into one larger volume, which can also be useful in some instances. To be sure, some of the things RAID does do can be very cool for a lot of different reasons.
But RAID is not a backup. It never was, and it never will be.
I just cleaned this off of a computer two days ago.
It set some registry entries values meant for maximum fuckery, marked every file on the disk that it could access as being hidden (thus even "dir" from a command line would result in "File not found,") and nuked the contents of the start menu, and did some other mean stuff.
Malwarebytes removed it but left the registry broken (which is arguably correct behavior). I changed the registry entries by hand, and I restored the start menu from an earlier copy.
After that, things were happy...except for a lingering, and possibly unrelated, issue with links from Google being redirected to spam. This turned out to be an infected Windows DLL, which "sfc/scannow" couldn't/didn't bother to fix. I was just about to give up on the machine for a happy time of nuke/reinstall, and another half-dozen hours of putting the machine back how it was... but then I tried combofix and the redirect problem went away, too.
All said: While I am a little richer having fixed these problems, money is poor compensation for this sort of pain.
I welcome the day when an affordable online service* can do incremental backups that can be used for a simple, bare-metal restore. Bandwidth isn't the issue anymore, and spinning storage is cheap; where is it?
*: Yes, online. If it's offline, that means that folks will have to think about it on a regular basis, and it won't be done.
Turns out one company had microfilmed one libraries collection - and the others had all bought the microfilm and trashed their paper copies and nobody had ever actually verified that the microfilm represented a complete run.
Well now, that's a perfectly reasonable explanation.
Yes, but remember, if it fails in the market its another example of Apples shitty track record with developing standard connectors and it is becomes a huge success and the Mac Book pros cause a a bunch of equipment manufacturers to build devices for it, then its Intel's technology and Apple had nothing to do with bringing it to market or helping to make it successful.
USB, incidentally, has similar development heritage to Thunderbolt:
It was initially designed by Intel, and essentially ignored until Apple started using it. After Apple rolled it out, a market for USB peripherals began to grow, and only then did it begin to be built-out on PCs.
Fast forward a bunch of years and it is essentially ubiquitous (for better or for worse), and folks have completely forgotten whether it is an Apple thing or an Intel thing. All we know is that it is a standard thing, and that's the important part.
I suspect that if Thunderbolt succeeds (and I frankly hope that it does) that its heritage to be similarly forgotten.
I've been seeing bitcoin mentioned here and there for a few weeks now.
With this FA, I've been introduced to the concept of "mining" Bitcoin. (It seems I'm a few months late, perhaps -- and yes, you can get off my lawn.)
Which, I must say, is interesting -- if people are willing to pay for it.
But in my own preliminary experience, I will generate two 10,000ths of a bitcoin per hour on my Intel Core2 Quad Q6600. (I found it interesting that all 4 cores were appropriately maxed out with in-browser Java, and that the system still seemed as responsive as always.)
But that's for my years-old CPU, which everyone seems to agree is the wrong way to mine Bitcoin. And while I can harness my GPU(s) to do the work considerably faster, given appropriate kit, here's something I've been so far completely unable to figure out:
What in the fuck are these cycles being used for? Is there some problem being solved? Is it just a measure of masochistic tolerance? What's going on here?
Now, if only they can ban those annoying red lights that some cars have when they're driving, you know the ones that make it hard to tell if the driver has applied the brakes.
What -- you mean, rear fog lights? Some countries require cars to have them, and some imports (mostly of European origin) retain the function. The brake lights are separate.
And in times where they're intended to be used (heavy fog, rain, snow, or other times when visibility is limited), they work fine: They help show the driver behind you that you are, indeed, present. The rest of the time, they're meant to be off.
Of course, most stupid Americans (of which I am one) don't have any idea what they are, or whether they're on or off, or likely even where the switch controlling them is located. It's entirely likely that the only reason they ever got switched on in the first place was that the sod bumped the switch with their knee while wrangling two big McDonalds bags and drink carrier, while chatting on the phone with a dog on their lap and trying to fix their hair using the makeup mirror that is conveniently located at the top of the windshield between the visors.
All kidding aside: Yes, it's annoying, but abolishment of this useful safety feature isn't the answer; education is. Getting these ignorant folks to read the manual that came with their $80k Range Rover would be a good start, and making them standard equipment would help awareness even more.
News flash: P2P is not efficient of bandwidth compared to a good, geographically-diverse CDN, as Netflix uses. For example: BitTorrent clients have little, if any, preference for those peers who are close. However, a well-developed CDN exists just for this purpose.
TFA isn't about Netflix complaining about their bandwidth utilization; it's about the fact that they do use a shitload of it. AFAICT, they're quite pleased with their bandwidth usage.
If you just want your kids to be able to efficiently watch the same shows over and over, then local caching is in order, not P2P. You've got the wrong solution to the problem.
As a counterpoint: My family watches a lot of Netflix.
Most of it is in the form of streamed TV series.
I'm a Top Gear fan (but can't stand their Ameican presentations), so I tend to watch either that or, sometimes, documentaries. (All of the old and not-so-old UK Top Gear episodes are on Netflix with excellent quality.)
The boy likes teenager-oriented shows, usually involving singing or socialite aspects that I have not and will never understand (I'm on Slashdot, after all). They either have all of them, or none of them -- he can park and watch and episode or two without any commercials at all.
And the wife likes mystery/detective dramas. (Rinse, repeat.)
Back when I had an antenna on a small tower (substitute cable or u-verse or satellite as required), it was nice to watch NCIS when it aired. Later on, DVRs happened, and it wasn't so important to watch it when it was on.
Fast forward a bit, and rewind a season or so, and it's far more pleasurable to just stream these shows with Netflix without commercials at all, to such an extent that I've gone from three-at-a-time to one-at-a-time on my account and dropped paid-for cable/satellite broadcasts completely.
It took it a minute to boot, but it workes in Firefox on my Droid 2 Global.
Great, thanks.
Since you've pointed that out, we'll never see Firefox on the iPhone, since Firefox can run arbitrary code.
(For the sarcasm-impaired: I jest, but only a little.)
FWIW, it failed completely with the Dolphin browser on my OG Droid with CM 7.something. So no Linux-within-browser-within-Linux for me on my phone for now. (I am heartbroken.)
Doesn't /etc/passwd do an acceptable job of this?
You know, other than the obvious slippery slope that is involved, I'd be extremely satisfied with a reasonable (ie: big) on-peak cap, if off-peak usage were either unmetered or significantly greater. But that's not how it seems to be playing out, at least in the States.
If it were, it would add an incentive for Netflix to implement just such a scheme. But without a distinction between peak/off-peak, there is no reason for them to do anything differently than they already are: Whether they watch a show at 6PM or 4AM, their customers are equally fucked by bandwidth caps.
I remember a time, many moons ago, where land-line telephone usage was handled similarly: During the day, long distance calls were expensive. At night (after 9, IIRC, usually) it was somewhat less expensive.
Of course, back then, the telephone network was still pretty heavily regulated. And, perhaps the Intarwebs should be, as well, if this stuff doesn't get sorted out in a sensible way on its own. (I'll be burned at the stake for even mentioning regulation, but I mentioned it anyway.)
I'm all for using the network more efficiently. Rewarding folks for using CDN-delivered services, and/or doing so at off-peak times would seem like a win for everyone.
Hell: Rewarding folks for choosing cheaper BitTorrent peers instead of preferring to use whoever is simply the fastest (even if they're on another continent!) would do a lot to improve the efficiency of things, but neither geography nor network topology seem to be a factor in peer selection.
Why should I care what UIDs my applications run as?
More like Pepsi vs. New Coke.
Your history is backwards:
The iPod first launched on October 23, 2001, and it played MP3s just fine at that time.
The iTunes Music Store opened on April 28, 2003. Prior to this, Apple didn't sell content for the devices.
Apple never "caved," they simply built an MP3 player which happened to be successful. Later on, they started selling content for it (in the arguably more space-efficient AAC format).
JPEG2000 can work quite well for images, as well.
There's a bunch of old panoramic maps at the Library of Congress which are available in JPEG2000. The amount detail preserved in these offerings is astounding (at least to me).
I mention this only because, as already pointed out, nobody much bothers with the format and, and good examples can be hard to come by. And old maps are cool.
Windows has gone backward with audio hardware support for entertainment with Vista and 7, and has done a very thorough job of making things as dumb as possible, by crippling Directsound.
It used to be sophisticated, and support all kinds of hardware acceleration for reverberation, filtering, occlusion, and surround sound (both actual and simulated) with just about any number of speakers.
Now, it's up to developers to implement these functions in software, and the CPU does the heavy lifting (which admittedly isn't very heavy these days, but I buy dedicated sound hardware for a reason).
It has pretty much killed the market for sound cards, and the development of new consumer-accessible DSP tech, since they all do the same thing these days, no matter how fancy or capable the underlying hardware is: As far as software is concerned, they all just play back audio.
Unless you count clever hacks like Creative's Alchemy, that is. But by the time we start comparing platform-specific audio hacks between Linux and Windows the discussion has turned pretty well moot.
Really? Microsoft has no say at all?
You act as if you've never seen someone sell something expensive before. The buyer usually has quite a bit of say in how things play out...
Indeed, it is 2011. Linux distributions could just emulate Windows 7/Vista by eliminating all fancy sound card functions altogether, while moving backward in terms of hardware functionality and at the same time and presenting a simple (but very limited) API for software to deal with.
Would that be preferable?
As a current Windows 7 user (at least on the desktop), who has from time to time done reasonably serious audio work under both Windows and Linux (and FreeBSD for that matter), I'd like to say that I'd prefer choice.
YMMV.
There are also AAAA batteries. A common 9V alkaline battery is full of them.
For whatever that's worth.
(On the end of the spectrum, F-sized batteries are also reasonably common, and are one size bigger than D.)
And given the fact that any new "web startup" is at least 90% likely to fail, I think it's a perfect match.
If it never takes off, which is very likely on average, then he'll never be out more than $14.99/mo. If it does take off, he'll be sent scrambling for a more robust solution, but it's likely that he'll have a few weeks/months/years to get ready for that, and will then have the funds to back something more serious.
*shrug*
If that's "most" of them, as you say, then I've never had to cure "most" infections because they never happened to begin with.
TFA is about a program, running on the local computer, which proclaims quite persuasively to be part of Windows. Changing themes will do fuck-all to help folks see the difference.
If the malware takes control of the PC (which it does, in the context of the FA), then having a single, locally-attached backup disk isn't necessarily a good answer: It can destroy/disrupt the backup just as easily as it can anything else on that PC.
A well-thought-out rotation of backup media would help, but that's no good because it involves humans who simply won't do it.
This wouldn't be a problem, so much, with good online storage: Even Dropbox does a good job of keeping old copies of your data intact for a period of time. I simply want the concept extended to an entire disk, with metadata intact, to enable a bare metal recovery.
This, combined with extra, out-of-band human verification (SMS?) for when you Really, Really want to destroy backup data, would work well against malware.
(And, yeah: I did use System Restore eventually. I consider it to be a last resort, though, simply because I am ignorant as to the extent of its workings and I am prejudiced against system-level programs which do not provide meaningful feedback as to what they're doing.)
While I believe your advice is well-intentioned, it's really no good.
This only works if the malware isn't using existing Windows widgets for its displays.
If I were I Windows programmer (I'm not) and I were writing malware (good heavens!), I'd use the native toolkit for all of my dealings...just like most other software does. It's easier, that way.
And then: Changing themes, for properly-implemented malware, would also change the look of that ill program to match.
Seconded, and furthered:
RAID would do nothing to protect against the thing described in TFA.
RAID only protects against hardware failure, and even then only if the failure is actually detected instead of just silently munging data.
This is not to say that RAID is not useful: It can be a performance boost in some applications. It can provide a clever way to combine many smaller disks into one larger volume, which can also be useful in some instances. To be sure, some of the things RAID does do can be very cool for a lot of different reasons.
But RAID is not a backup. It never was, and it never will be.
I just cleaned this off of a computer two days ago.
It set some registry entries values meant for maximum fuckery, marked every file on the disk that it could access as being hidden (thus even "dir" from a command line would result in "File not found,") and nuked the contents of the start menu, and did some other mean stuff.
Malwarebytes removed it but left the registry broken (which is arguably correct behavior). I changed the registry entries by hand, and I restored the start menu from an earlier copy.
After that, things were happy...except for a lingering, and possibly unrelated, issue with links from Google being redirected to spam. This turned out to be an infected Windows DLL, which "sfc /scannow" couldn't/didn't bother to fix. I was just about to give up on the machine for a happy time of nuke/reinstall, and another half-dozen hours of putting the machine back how it was... but then I tried combofix and the redirect problem went away, too.
All said: While I am a little richer having fixed these problems, money is poor compensation for this sort of pain.
I welcome the day when an affordable online service* can do incremental backups that can be used for a simple, bare-metal restore. Bandwidth isn't the issue anymore, and spinning storage is cheap; where is it?
*: Yes, online. If it's offline, that means that folks will have to think about it on a regular basis, and it won't be done.
Well now, that's a perfectly reasonable explanation.
But I, for one, blame the Illuminati.
USB, incidentally, has similar development heritage to Thunderbolt:
It was initially designed by Intel, and essentially ignored until Apple started using it. After Apple rolled it out, a market for USB peripherals began to grow, and only then did it begin to be built-out on PCs.
Fast forward a bunch of years and it is essentially ubiquitous (for better or for worse), and folks have completely forgotten whether it is an Apple thing or an Intel thing. All we know is that it is a standard thing, and that's the important part.
I suspect that if Thunderbolt succeeds (and I frankly hope that it does) that its heritage to be similarly forgotten.
I've been seeing bitcoin mentioned here and there for a few weeks now.
With this FA, I've been introduced to the concept of "mining" Bitcoin. (It seems I'm a few months late, perhaps -- and yes, you can get off my lawn.)
Which, I must say, is interesting -- if people are willing to pay for it.
But in my own preliminary experience, I will generate two 10,000ths of a bitcoin per hour on my Intel Core2 Quad Q6600. (I found it interesting that all 4 cores were appropriately maxed out with in-browser Java, and that the system still seemed as responsive as always.)
But that's for my years-old CPU, which everyone seems to agree is the wrong way to mine Bitcoin. And while I can harness my GPU(s) to do the work considerably faster, given appropriate kit, here's something I've been so far completely unable to figure out:
What in the fuck are these cycles being used for? Is there some problem being solved? Is it just a measure of masochistic tolerance? What's going on here?
What -- you mean, rear fog lights? Some countries require cars to have them, and some imports (mostly of European origin) retain the function. The brake lights are separate.
And in times where they're intended to be used (heavy fog, rain, snow, or other times when visibility is limited), they work fine: They help show the driver behind you that you are, indeed, present. The rest of the time, they're meant to be off.
Of course, most stupid Americans (of which I am one) don't have any idea what they are, or whether they're on or off, or likely even where the switch controlling them is located. It's entirely likely that the only reason they ever got switched on in the first place was that the sod bumped the switch with their knee while wrangling two big McDonalds bags and drink carrier, while chatting on the phone with a dog on their lap and trying to fix their hair using the makeup mirror that is conveniently located at the top of the windshield between the visors.
All kidding aside: Yes, it's annoying, but abolishment of this useful safety feature isn't the answer; education is. Getting these ignorant folks to read the manual that came with their $80k Range Rover would be a good start, and making them standard equipment would help awareness even more.
News flash: P2P is not efficient of bandwidth compared to a good, geographically-diverse CDN, as Netflix uses. For example: BitTorrent clients have little, if any, preference for those peers who are close. However, a well-developed CDN exists just for this purpose.
TFA isn't about Netflix complaining about their bandwidth utilization; it's about the fact that they do use a shitload of it. AFAICT, they're quite pleased with their bandwidth usage.
If you just want your kids to be able to efficiently watch the same shows over and over, then local caching is in order, not P2P. You've got the wrong solution to the problem.
As a counterpoint: My family watches a lot of Netflix.
Most of it is in the form of streamed TV series.
I'm a Top Gear fan (but can't stand their Ameican presentations), so I tend to watch either that or, sometimes, documentaries. (All of the old and not-so-old UK Top Gear episodes are on Netflix with excellent quality.)
The boy likes teenager-oriented shows, usually involving singing or socialite aspects that I have not and will never understand (I'm on Slashdot, after all). They either have all of them, or none of them -- he can park and watch and episode or two without any commercials at all.
And the wife likes mystery/detective dramas. (Rinse, repeat.)
Back when I had an antenna on a small tower (substitute cable or u-verse or satellite as required), it was nice to watch NCIS when it aired. Later on, DVRs happened, and it wasn't so important to watch it when it was on.
Fast forward a bit, and rewind a season or so, and it's far more pleasurable to just stream these shows with Netflix without commercials at all, to such an extent that I've gone from three-at-a-time to one-at-a-time on my account and dropped paid-for cable/satellite broadcasts completely.
Perhaps you're doing it wrong. :)
You pay sales tax in California when you buy food items like orange juice?
Great, thanks.
Since you've pointed that out, we'll never see Firefox on the iPhone, since Firefox can run arbitrary code.
(For the sarcasm-impaired: I jest, but only a little.)
FWIW, it failed completely with the Dolphin browser on my OG Droid with CM 7.something. So no Linux-within-browser-within-Linux for me on my phone for now. (I am heartbroken.)
Seriously?