Wow. Welcome to 1973. What's it like back there, caveman?
Dude! You had a computer playing 44kHz 16-bit stereo sound back in 1973? Color me impressed.
Seriously - what is wrong with typing '80' instead of moving your hand off the keyboard over onto the mouse, moving the mouse pointer to some designated region on the screen, clicking on a small slider icon and dragging it in one direction or the other?
Of course, if you are one of those people who spend more time with your hand on the mouse than on the keyboard, it may well be a waste of time to try and explain why the GUI isn't the be-all and end-all of HCI. (:
Re:IT'S 11/9 NOT 9/11 IT'S 11/9 NOT 9/11 IT's 11/9
on
Collateral Damage
·
· Score: 1
The other reason that it's sometimes referred to as "nine-eleven" is the similarity it has with the emergency telephone number "9-1-1".
This is a rhetorical question, but... can we please stop with the pseudo-dramatic little "9/11 == 9-1-1" congealing spoonful of chicken soup for the soul? It's a country song waiting to be written, if not already recorded.
Tell me, if Al'Quaeda's day of judgment had happened to be the 4th of October, would we be making much of the fact that it was "10-4 day, get it? The police use that call on their CB radios. So appropriate!"?
In summary, I feel it fitting to refer back to one of this nation's most important events in history as "September 11th", or "nine-eleven", or even "nine-one-one".
And I find it exceedingly cheesy. There. Glad to have got that off my chest. Please continue, pay me no mind.[*]
[*] Though on second thought I could use a new mind - this one doesn't always perform to spec.
This was because SunOS had a dog-slow filesystem; even today,/tmp is usually backed by ram. Linux (and
probably BSD) has a fast enough filesystem that this isn't an issue
There's also the small matter of write-back caching - any modern OS should cache writes aggressively (or at least should have the option), such that short-lived temp files (you know, the ones whose speed matters most) usually never reach the platter before being deleted anyway.
I've seen this before, why is it that everyone says "boxen"? Whos started this?
It's by analogy to vaxen, itself of course an analog of oxen, which is legit.
I think boxen was a massively parallel coinage. I'm pretty sure I started using it before seeing anyone else's usage, but I hardly think I'm influential enough to have been the originator. (Same story with Pentia meaning more than one Intel 586.)
Does anyone really believe that music consumers "backup" thier discs to mp3 for purely "personal" use? Let's at least be adult enough not to sugar coat this: we want to get around Cactus Data Shield because we want to "share" [or steal] music.
The majority, almost certainly, but not me. I really do use cdparanoia and oggenc purely for my own convenience. I've got 10 GB of ogg files on my hard disk here at work, every last octet of which was ripped from CDs I have at home [well a few are still at work, from being ripped]. Nobody has access to this drive except me.
I appreciate what you're saying - I bet there are more people on/. who use p2p to infringe copyrights than there are people like me - but we do exist. Not that I've given you any evidence to back up my claim of personal audio copyright integrity. (:
Hey - all of us agree, CDS and similar measures suck kidney stones through Fallopian tubes, but if Big Media were to make it clear that they would only use technological measures "against" their customers and never call down the lawyers for piracy/DRM issues, I figure that's almost a fair deal. I just hate it when they call down the lawyers.
Unfortunately we know they would never agree to tie their legal hands like that....
Assuming what you get in the mail is on a paper medium. What if it was a floppy disk, or a flash memory stick?
I suspect that computer CDROM players will become "smart" and eventually this copy protection will be thwarted.
I suppose this is (-1, Redundant), but of course the thing the music industry either hasn't noticed yet or keeps ignoring is that not every player has to circumvent the protection; it's enough for one electrical engineer to get something through a S/PDIF cable and an mp3 encoder... and then game over. People already illegally download music, all the time, so what is to stop people from downloading something they already own on CD (which in my book is absolutely permissible - morally, if not legally)?
However, it would also open the door for the "good guys", allowing small production companies to use open codecs to play their stuff in an unrestricted fashion on the same hardware that the big houses are using to enforce their DRM nonsense.
Your faith in the harmlessness of DRM in the face of open honest competition is touching. (:
DVDs are not required to be encrypted or region-coded. CDs are not required to be CRC-crippled. The Big Money is not forcing DRM down every content producer's throat.
Back on topic, I guess we're talking about people being forced to use MPEG-4 because that's what the hardware will support. Well, I suppose a VM-based player architecture would help there, but honestly, that's what flash rom is for.
And flash rom has the advantage that if you need to crack obnoxious features like DRM, you only have to do it once, not once per disc.
For example, try buying an Ogg Vorbis hardware player today.
Good point, but I still believe the support will come. Various reasons not to use vorbis are falling apart, and when Monty finishes with his integer-only decoder with performance comparable to mp3 (two reasons not to do embedded vorbis, currently), most of the pieces will be in place. There is still brand recognition, to be sure....
I'm assuming that the only way the MPEG-4 Forum can recoup the expense of developing and publishing their spec is through licensing it, which I think is true.
Perhaps, but that only matters if the MPEG-4 Forum needs to recoup their costs, which is another way of saying if the MPEG-4 Forum is not funded in some other way. What I'm saying is that there are other ways to fund R&D than the promise of patent royalties.
The X Consortium is a well-known example of an entity which made no money at all from any sort of license fees, but only existed to develop technology to benefit its member companies. It is undeniable than having everyone implement the X Window System (15-year-old tech and still very much alive, but that's a different thread) was a better solution than everyone having his own private windowing system like NeWS or NeXTstep, as cool as those might have been (studlycaps and all).
Of course, many parties benefitted from X11 without paying for X Consortium memberships. However, contrary to popular belief, IP economics is not a zero-sum game. The existence of freeloaders does not detract from the value to the paying sponsors - at least not the primary value, which is having a robust, cross-platform, widely supported standard for workstation graphics. In fact the freeloaders actually helped achieve this!
I see no reason the various parties who benefit from good media codecs can't form a similar consortium, sponsored the same way. (Whether they would or not is another question, but they could.)
(Now don't get me wrong, I'm not accusing you of being arrogant!)
On the other hand, as I understand it, the Ogg file format already supports having video streams embedded in it (i.e. it wouldn't have to be Tarkin-codec video). I've been ITCHING to see somebody add Ogg format support to a video encoder like ffmpeg (and I suspect MPlayer would have support for playback within days afterwards).
Indeed, ogg is just a container format - but think of how confusing that would be! If you show anyone a "*.ogg" file today, they assume it's Vorbis-coded audio. (Assuming they have any clue what it is at all, of course.) With your proposal, it could be any sort of multimedia file.
Not that there's anything wrong with a little confusion, but people would have to start learning to say "ogg vorbis file" instead of just "ogg file". (:
Hardware manufacturers would get together and establish an Open Player Platform or something, which essentially standardizes the instruction set and capabilities of the player and the image format of the codec executable.
That is absolutely the best way to introduce all sorts of exciting digital rights management schemes. Remember, this is what some "copy-protected CDs" do today - they autoload a player binary under Windows.
Imagine - the player bytecode would checksum your hardware player's ROM to make sure it is an "authorized AOLTW playback device", then try to connect to the web. If your media station isn't plugged into your phone line it'll refuse to play. Then it will make the necessary pay-per-play arrangements between your credit card company and AOLTW, and finally you get your movie playback. (Oh, and btw, no freeze-frame - since you chose not to pay the extra 15% for that privilege.)
I prefer smart (and openly-implementable) players and dumb media, thanks very much.
But, to address your point, consider a world in which there was no protection of intellectual property. Let's say it cost $1 million to develop MPEG-4. (I have no idea what it cost. That's just an example.) The people who developed have no protection; if they release their specification, anybody who wants to could develop a product based on it. If they don't release their spec, anybody who wants to could reverse-engineer it. The MPEG-4 Forums loses $1 million.
The problem with your argument is that you assume that the only people who profit from MPEG-4 are the people who license the patent.
Fact is, anyone who wants to provide streaming video - or for that matter any digital video - profits from said R&D. Digital camera companies, DVD publishers, digital cable providers, pr0n sites... lots of people stand to be able to provide their customers with an improved service when someone develops an improved codec.
It is in the digital video industry's best interest to sponsor R&D of new compression technology - by way of working groups, consortiums, research grants, whatever. Also, for many or most of these parties, having one's own proprietary codec is of no use at all - they need open standards to drive consumer acceptance, which means they need to cooperate, on some level. (Again, consortiums are the usual vehicle for this.)
Side note: this is similar to the Microsoft fallacy of "GPL software is bad for business". Their key assumption is that the harm GPL software does to the software-producing industry somehow outweighs the benefit to the whole rest of the software-consuming industry. That's a pretty arrogant assumption, if you think about it. (Now don't get me wrong, I'm not accusing you of being arrogant!)
I still don't understand how they thought that data from programs running under different kernels was automatically incomaptible.
Yeah you did (: - you and I both know it's because they're ill-informed and they just assume stuff. Or perhaps they were talking about mixing hardware architectures - something which if you're careful with htonl() and friends you can still do just fine on Linux. (For obvious reasons you can't do it on OSX!)
The OSX nodes can be left running with a simple renice to make it more user responsive. Whether or not this can be done with a Beowulf cluster, I have no idea.
Shouldn't be necessary - you should already be running your batch jobs at a nice value of say 5 or 10. After all, if the machine is otherwise idle, nice processes still get the same amount of CPU.
From what I understand, Maya is a popular Mac program, and it takes quite a while to do the crunching for that program.
Perhaps now... but it only came out for the mac like a year ago.
As for the number crunching - Maya has two modes of operation, interactive use and rendering. The interactive part can make good use of a beefy system, but I'm not sure it's even multithreaded, much less clusterable.
The render part is what you use a cluster for. And for that, you don't need Macs - Alias|Wavefront supports their render engine on IRIX, NT and Linux (and perhaps other platforms by now). Sure, they don't support distributed rendering out of the box except for IRIX, but any company serious enough about doing large renders will develop or purchase a system to handle the cluster glue.
If you buy anything by credit card face-to-face, the EULA can be assumed to be binding, as you have signed the charge slip, and some crafty lawyer could probably twist that into having signed the EULA.
Huh? What I signed was a contract between me, the merchant, and the Greenwood Trust Company. What on Earth does that have to do with any supposed contract between me and the software house?
I know, I know, never underestimate the power of the legal eagles, but what you suggest would be ridiculous.
Incidentally, I imagine click-through would be upheld in court. Part of that is my innate cynicism, of course. Part is the fact that if a court were shown the consequences of following common sense (collapse of current off-the-shelf software sales method; requirement for PKI / certs for every e-commerce consumer), they would be afraid to rock this particular boat.
But the clincher: I recently noticed that even the US government is now using click-through contracts. My taxes are still simple enough that I can file via phone, and I noticed that the process includes a step where you "sign" a statement by punching in a 5-or-so-digit code.
As far as I'm concerned, the trouble is our dear kernel trying to get those 4Mb pages found in Pentium and not in Athlons. Now, why everybody call this a BUG????? I think this is just a "NOT supported feature".
No, it's a bug. 4Mb pages are a supported feature on Athlons - but we now know that the CPU has a small problem invalidating such pages when AGP activity is going on, resulting in memory corruption, which in turn can cause crashes or lockups.
What was the mean-time-to-patch for all major operating systems for the P5 FP bug? A month or two?
Bad example - you can't work around that class of bug in the OS. Best thing you can do is get Intel to ship you a replacement - which I guess they did, on request.
You should have picked the Pentium F00F bug from a few years back. That's one you can work around in software. Ingo Molnar produced a working Linux patch over the span of a weekend, and it was shipped with a new kernel rev (2.0.35 or so) a couple days later. Microsoft didn't fix it at all in NT4 - presumably they did in Win2k, if they felt the Pentium-MMX 233 and below was still worth supporting by then.
None of which had much to do with Intel, except that they were very cooperative distributing intel about the bug and suggested workaround to the Linux people.
So, I ask again, what is the point of having 4MB pages?
They are only used for kernel memory, which is never paged out. Memory for user processes and kernel modules always uses 4k pages. (I say always, but of course I mean always on the i386 architecture. The Alpha for example uses 8k pages natively.)
You see, Alan Cox and friends are planning to make Linux "recognizes" Athlons, and do a special case - the "nomem" thingy - on it.
But if the Stepping A5 of AthlonXP/MP core has the bug licked, then, the "nomem" thingy by Alan Cox et al may be a step backward.
You underestimate kernel hackers. They'll definitely check steppings. If Linux ends up downgrading all Athlons to 4k pages, just email the kernel list and demonstrate / assert / prove that certain Athlons are OK. Alan or Dave Jones will patch the kernel correctly in no time, I'd put money on it.
Imagine if all porn sites would store their pictures in.gif format
GIF is lossless? Only if your source image uses 256 or fewer colors. Since GIF only supports an 8-bit colormap, it's horribly lossy for any sort of photo-quality image. JPEG may blur your edges but at least it can handle 24-bit color.
The notion that I would pay to carry around a telephone that would spout advertisements based on my location is worse than absurd -- it's insulting.
Just used up my mod points. Figures. That was +1 Interesting.
Re:I would prefer the other way around
on
Debian NetBSD
·
· Score: 2
LinuxBSD would be funny--nothing GNU except for the toolchain and a few other utilities. I don't actually see a real advantage of a system like that over a standard Linux distro, but the idea amuses me:)
David Parsons created a distro a few years ago with pretty much exactly this in mind. Parsons on the GNU utilities: "Unfortunately, these tools come with a few albatrosses around their neck: FSF bloat, FSF philosophy, and Richard Stallman." One of his goals was to minimise the amount of GPL code in his distro.
Slashdot Mirror
Dude! You had a computer playing 44kHz 16-bit stereo sound back in 1973? Color me impressed.
Seriously - what is wrong with typing '80' instead of moving your hand off the keyboard over onto the mouse, moving the mouse pointer to some designated region on the screen, clicking on a small slider icon and dragging it in one direction or the other?
Of course, if you are one of those people who spend more time with your hand on the mouse than on the keyboard, it may well be a waste of time to try and explain why the GUI isn't the be-all and end-all of HCI. (:
This is a rhetorical question, but ... can we please stop with the pseudo-dramatic little "9/11 == 9-1-1" congealing spoonful of chicken soup for the soul? It's a country song waiting to be written, if not already recorded.
Tell me, if Al'Quaeda's day of judgment had happened to be the 4th of October, would we be making much of the fact that it was "10-4 day, get it? The police use that call on their CB radios. So appropriate!"?
And I find it exceedingly cheesy. There. Glad to have got that off my chest. Please continue, pay me no mind.[*]
So that's why he kept misspelling Colombia as "Columbia". He's still thinking "Columbine".
(Actually, I read a local review of Collateral whose reviewer did the same thing - I guess it's just a common mistake by the semi-literate.)
There's also the small matter of write-back caching - any modern OS should cache writes aggressively (or at least should have the option), such that short-lived temp files (you know, the ones whose speed matters most) usually never reach the platter before being deleted anyway.
It's by analogy to vaxen, itself of course an analog of oxen, which is legit.
I think boxen was a massively parallel coinage. I'm pretty sure I started using it before seeing anyone else's usage, but I hardly think I'm influential enough to have been the originator. (Same story with Pentia meaning more than one Intel 586.)
The majority, almost certainly, but not me. I really do use cdparanoia and oggenc purely for my own convenience. I've got 10 GB of ogg files on my hard disk here at work, every last octet of which was ripped from CDs I have at home [well a few are still at work, from being ripped]. Nobody has access to this drive except me.
I appreciate what you're saying - I bet there are more people on /. who use p2p to infringe copyrights than there are people like me - but we do exist. Not that I've given you any evidence to back up my claim of personal audio copyright integrity. (:
Hey - all of us agree, CDS and similar measures suck kidney stones through Fallopian tubes, but if Big Media were to make it clear that they would only use technological measures "against" their customers and never call down the lawyers for piracy/DRM issues, I figure that's almost a fair deal. I just hate it when they call down the lawyers.
Unfortunately we know they would never agree to tie their legal hands like that....
Assuming what you get in the mail is on a paper medium. What if it was a floppy disk, or a flash memory stick?
I suppose this is (-1, Redundant), but of course the thing the music industry either hasn't noticed yet or keeps ignoring is that not every player has to circumvent the protection; it's enough for one electrical engineer to get something through a S/PDIF cable and an mp3 encoder ... and then game over. People already illegally download music, all the time, so what is to stop people from downloading something they already own on CD (which in my book is absolutely permissible - morally, if not legally)?
Your faith in the harmlessness of DRM in the face of open honest competition is touching. (:
DVDs are not required to be encrypted or region-coded. CDs are not required to be CRC-crippled. The Big Money is not forcing DRM down every content producer's throat.
Back on topic, I guess we're talking about people being forced to use MPEG-4 because that's what the hardware will support. Well, I suppose a VM-based player architecture would help there, but honestly, that's what flash rom is for.
And flash rom has the advantage that if you need to crack obnoxious features like DRM, you only have to do it once, not once per disc.
Good point, but I still believe the support will come. Various reasons not to use vorbis are falling apart, and when Monty finishes with his integer-only decoder with performance comparable to mp3 (two reasons not to do embedded vorbis, currently), most of the pieces will be in place. There is still brand recognition, to be sure....
Perhaps, but that only matters if the MPEG-4 Forum needs to recoup their costs, which is another way of saying if the MPEG-4 Forum is not funded in some other way. What I'm saying is that there are other ways to fund R&D than the promise of patent royalties.
The X Consortium is a well-known example of an entity which made no money at all from any sort of license fees, but only existed to develop technology to benefit its member companies. It is undeniable than having everyone implement the X Window System (15-year-old tech and still very much alive, but that's a different thread) was a better solution than everyone having his own private windowing system like NeWS or NeXTstep, as cool as those might have been (studlycaps and all).
Of course, many parties benefitted from X11 without paying for X Consortium memberships. However, contrary to popular belief, IP economics is not a zero-sum game. The existence of freeloaders does not detract from the value to the paying sponsors - at least not the primary value, which is having a robust, cross-platform, widely supported standard for workstation graphics. In fact the freeloaders actually helped achieve this!
I see no reason the various parties who benefit from good media codecs can't form a similar consortium, sponsored the same way. (Whether they would or not is another question, but they could.)
Egomaniac. (:
Indeed, ogg is just a container format - but think of how confusing that would be! If you show anyone a "*.ogg" file today, they assume it's Vorbis-coded audio. (Assuming they have any clue what it is at all, of course.) With your proposal, it could be any sort of multimedia file.
Not that there's anything wrong with a little confusion, but people would have to start learning to say "ogg vorbis file" instead of just "ogg file". (:
That is absolutely the best way to introduce all sorts of exciting digital rights management schemes. Remember, this is what some "copy-protected CDs" do today - they autoload a player binary under Windows.
Imagine - the player bytecode would checksum your hardware player's ROM to make sure it is an "authorized AOLTW playback device", then try to connect to the web. If your media station isn't plugged into your phone line it'll refuse to play. Then it will make the necessary pay-per-play arrangements between your credit card company and AOLTW, and finally you get your movie playback. (Oh, and btw, no freeze-frame - since you chose not to pay the extra 15% for that privilege.)
I prefer smart (and openly-implementable) players and dumb media, thanks very much.
The problem with your argument is that you assume that the only people who profit from MPEG-4 are the people who license the patent.
Fact is, anyone who wants to provide streaming video - or for that matter any digital video - profits from said R&D. Digital camera companies, DVD publishers, digital cable providers, pr0n sites ... lots of people stand to be able to provide their customers with an improved service when someone develops an improved codec.
It is in the digital video industry's best interest to sponsor R&D of new compression technology - by way of working groups, consortiums, research grants, whatever. Also, for many or most of these parties, having one's own proprietary codec is of no use at all - they need open standards to drive consumer acceptance, which means they need to cooperate, on some level. (Again, consortiums are the usual vehicle for this.)
Side note: this is similar to the Microsoft fallacy of "GPL software is bad for business". Their key assumption is that the harm GPL software does to the software-producing industry somehow outweighs the benefit to the whole rest of the software-consuming industry. That's a pretty arrogant assumption, if you think about it. (Now don't get me wrong, I'm not accusing you of being arrogant!)
Yeah you did (: - you and I both know it's because they're ill-informed and they just assume stuff. Or perhaps they were talking about mixing hardware architectures - something which if you're careful with htonl() and friends you can still do just fine on Linux. (For obvious reasons you can't do it on OSX!)
Shouldn't be necessary - you should already be running your batch jobs at a nice value of say 5 or 10. After all, if the machine is otherwise idle, nice processes still get the same amount of CPU.
Perhaps now ... but it only came out for the mac like a year ago.
As for the number crunching - Maya has two modes of operation, interactive use and rendering. The interactive part can make good use of a beefy system, but I'm not sure it's even multithreaded, much less clusterable.
The render part is what you use a cluster for. And for that, you don't need Macs - Alias|Wavefront supports their render engine on IRIX, NT and Linux (and perhaps other platforms by now). Sure, they don't support distributed rendering out of the box except for IRIX, but any company serious enough about doing large renders will develop or purchase a system to handle the cluster glue.
Huh? What I signed was a contract between me, the merchant, and the Greenwood Trust Company. What on Earth does that have to do with any supposed contract between me and the software house?
I know, I know, never underestimate the power of the legal eagles, but what you suggest would be ridiculous.
Incidentally, I imagine click-through would be upheld in court. Part of that is my innate cynicism, of course. Part is the fact that if a court were shown the consequences of following common sense (collapse of current off-the-shelf software sales method; requirement for PKI / certs for every e-commerce consumer), they would be afraid to rock this particular boat.
But the clincher: I recently noticed that even the US government is now using click-through contracts. My taxes are still simple enough that I can file via phone, and I noticed that the process includes a step where you "sign" a statement by punching in a 5-or-so-digit code.
No, it's a bug. 4Mb pages are a supported feature on Athlons - but we now know that the CPU has a small problem invalidating such pages when AGP activity is going on, resulting in memory corruption, which in turn can cause crashes or lockups.
Bad example - you can't work around that class of bug in the OS. Best thing you can do is get Intel to ship you a replacement - which I guess they did, on request.
You should have picked the Pentium F00F bug from a few years back. That's one you can work around in software. Ingo Molnar produced a working Linux patch over the span of a weekend, and it was shipped with a new kernel rev (2.0.35 or so) a couple days later. Microsoft didn't fix it at all in NT4 - presumably they did in Win2k, if they felt the Pentium-MMX 233 and below was still worth supporting by then.
None of which had much to do with Intel, except that they were very cooperative distributing intel about the bug and suggested workaround to the Linux people.
They are only used for kernel memory, which is never paged out. Memory for user processes and kernel modules always uses 4k pages. (I say always, but of course I mean always on the i386 architecture. The Alpha for example uses 8k pages natively.)
You underestimate kernel hackers. They'll definitely check steppings. If Linux ends up downgrading all Athlons to 4k pages, just email the kernel list and demonstrate / assert / prove that certain Athlons are OK. Alan or Dave Jones will patch the kernel correctly in no time, I'd put money on it.
GIF is lossless? Only if your source image uses 256 or fewer colors. Since GIF only supports an 8-bit colormap, it's horribly lossy for any sort of photo-quality image. JPEG may blur your edges but at least it can handle 24-bit color.
Using people as your pawns, you mean?
Just used up my mod points. Figures. That was +1 Interesting.
David Parsons created a distro a few years ago with pretty much exactly this in mind. Parsons on the GNU utilities: "Unfortunately, these tools come with a few albatrosses around their neck: FSF bloat, FSF philosophy, and Richard Stallman." One of his goals was to minimise the amount of GPL code in his distro.
It's called Mastodon Linux.