And then you get into needing lots of buttons for game play. Todays controller do have tons of buttons, it really wouldn't be all that hard to map everything WoW has to offer on a controller, which after all only has 12 action buttons in its default GUI, while game controller have 14 action buttons. And that is not even considering that a console MMORPG could be build up from the ground for a controller and thus be made to work with that perfectly instead of trying to retrofit a PC game to the console controller, after all there are tons of console RPGs that work just fine. Chat with a controller would indeed be a problem, but then there are things like the Microsoft Chatpad and of course all consoles these days support USB keyboards. Maybe one could even do speech recognition to do transparent voice to text translation.
The technical side of things seems to be well covered. If anything the comfort could be a problem, typing on a USB keyboard on a couch doesn't work as well as on a desk.
olpc was using a e-paper screen iirc. The OLPC isn't using e-paper, but a special display based around normal LCD screen, it is able to do 200dpi in black&white and has a sunlight readable screen, so its quite similar to e-paper in that regard. The OLPC screen is however, with backlight off, a lot darker then normal paper, so in-door it has some limitations in comparison to e-paper. The area where the OLPC however just flat out kills e-paper is refresh time, the OLPC works just like a normal monitor, so you can watch video, browse webpages and all that stuff. All e-paper devices on the other side that I have seen so far are completly unusable for that, since they have refresh times around one second and not even a clean refresh, they fill the screen all black before they move on to the next page. Which might be usable for books, but rather useless for classic GUI or scrolling, which is why I have some serious doubt if e-paper will ever catch on.
I know of at least a few software programmers who have used this sort of technique to produce fairly profitable pieces of shareware. Well, that kind of proves my point, doesn't it?
While true in theory, in practice its kind of wrong, since the GPL pretty much ensures that the normal business models used for software no longer work.
It had a story line Doom and Quake had one too, which of course doesn't mean that either of those storylines was good. And when it comes to real storylines DarkForces still beats HalfLife any day and there is of course also stuff like SystemShock. The new thing that HalfLife did wasn't really the story, since it really didn't have much of that either, the new thing was the presentation, you no longer had levels or missions screens, but instead everything happened in the game engine itself, the whole game, including intro, was one seamless whole and not just a series of levels.
SSD will never reach parity with hard disks It doesn't have to. Most uses don't need 1TB, even less 4TB or 16TB, storage is quickly reaching the point where you simply don't need more. Simple example would be MP3 players, 1TB gives you a full year of non-stop audio in good quality, who is ever going to listen to all of that or even collecting that much music in the first place? Most people just don't have a use for that much storage in an MP3 player, they are more then happy with 10GB and if size is becoming a non-issue other things get much more important like robustness, power usage and stuff like that. Another thing to keep in mind is that spinning disks seem to have a upper pricecap. I can get 1TB for $200, but I can't get 1GB for $0.20 or 10GB for $2. The cheapest spinning disk always seems to be around $50, size might increase, but the price will not go down. For some applications however $50 is just to much and SSD seems to have much less issue with scaling down into much lower price regions, which is why you don't see a spinning disk in the OLPC, the Eee or Apples cheaper iPod variants.
A few years down the road I fully expect to have a SSD in my computer, for larger storage and archival I might still have an additional spinning disk attached, but for daily use (OS,/home) even something small like 100GB would be plenty.
Jordan Mechner was involved with Prince of Persia: Sands of Time as game designer and writer, in the second and third he wasn't involved at all and well, it shows. Sands of Time is simply a game that wouldn't have needed a sequel and what they did in Warrior Within is just disgusting.
There is quite a difference between a series of small glitches and an army of robots trashing down a police station, which happens even before the police knows what is going on, i.e. the robots *attack*, they don't just defend. And anyway, nothing of this chances that the red-glowing robots are in total violation to the first law.
The Three Laws *did* work. The first law starts with "A robot may not injure a human being". In I, Robot the movie however they *directly* attacked humans, threw them around, snap their necks, throw cars at them, killed them, and all kind of other shit that you would expect from robots going on a killing spree, which is a total violation of that first law.
Yes, Asimov's work was about unintended consequences of the laws, but it was never about "Lets just ignore the laws and do some shit".
His argument focuses on novices and teaching and there I fully agree, forcing people to do indention teaches good practice and produces code with a minimum amount of readability from everybody. However languages like Python where whitespace is the only way to indicate a block still drive me nuts. They make it way to easy to break code when you copy&paste stuff around and sometimes even impossible when you want to copy code from a webpage or forum. Whitespace is just way to easily lost or damaged. I guess I would mind mandatory whitespace much less if it was used in combination with standard block start/end tokens instead of being the token itself.
I agree that I, Robot wasn't complete garbage, but it could have been a hell of a lot better. One of the central points of Asimovs books is that the robotic laws actually do work, not always as intended of course, but they got created exactly to avoid robots going on a killing spree, yet in I, Robot the movie they do exactly that and of course in the most stupid looking way possible (look, it's red, it must be evil...). The resolution at the end was of course not much better, switch the mainframe of and everything is back to normal. Would they have replaced that whole killing spree thing with something more Asimovian or went the route of Matrix's The Second Renaissance (which wouldn't have been Asimovian, but at least it would have made more sense) it could have ended up really good, but so its just a few good ideas and a lot of garbage in between.
The only real problem Sugar has is that applications take 10 seconds to start, even trivial "hello world" ones, which start pretty much instantly when you start them directly. Starting of Sugar itself is really a non-issue and much of the booting time is taking up by other stuff. So I really doubt EDE would be a good choice. On a small screen you really want something that doesn't force you to mess around with tiny little windows, but instead something that takes up the whole screen and lays out everything well right from the start and Sugar does exactly that.
Sugar simply needs to a decent amount of optimization and some cleanup with the Journal stuff, other then that it works great for the OLPC.
Do those bootstrap server step over any law? They don't tell you where to download files, they simply tell you where other clients are, kind of like DNS.
For a text based format one thing that can get out of date is charset, but then that is relatively easily converted and we also have Unicode now which should be good enough for a while to come.
Then there is of course the whole markup, that can easily get lost or forgotten in a 100 years. You still might be able to make some sense of the document, but getting an exact rendering of the document might become tricky, unless you also dig out a copy of OpenOffice and have an emulator that can run it.
In practical terms I think there are mainly two dangers of ODF. First there is the Internet and ePaper, both of which will sooner or later make paper obsolete. Why print something, when you can just email it around? And why print it, when the screen you are having already reads as easy as paper? So you might simply end up using other formats then ODF far more often and no longer end up having a copy of OpenOffice around in a few years. The second and bigger danger I think are however subtle improvements in the format. When ODF gets expanded and improved they might try to get backward compatibility going, but you likely never reach 100%. So you will have tiny little differences and when ODF1.1 is no longer used, since everybody is using ODF5.0, software might no longer support ODF1.1 properly either. It might still render, but the output might be wrong in the details. This would be basically what we have today already, you won't have much of a problem getting an old Word document to open, but you will have a very big issue if you want to get the exact rendering that the Word version it was created with produced.
Yep, and one also has to keep in mind that the voice actor is only a small part of what makes the character. You also have the artists that designed the character, the artists that created the 3d model, those that animated it, then maybe a stunt guy that did the motion capture for it and then sometimes even a special guy for the lip syncing and facial animation.
$100k sounds like a pretty good pay for just recording the dialog.
That said, I wouldn't mind if all those people involved got a share of the money that gets made, but its not like $100k is a rip off.
Best definition: "It'll be ready when it's ready." Which in the case of a distribution would be never. With thousands of software projects packaged into a distribution you reach never a point where every single one of them is ready and if you don't have at least a little bit of synchronization you end up with a distribution that is bleeding edge in some areas and outdated in others. With six month release cycles this is much less of a problem of course then with Debian's endless spans of time between new releases, but it can still be really annoying if you are stuck with a buggy piece of software for six month that is already fixed upstream just because the new upstream released missed the distributions release by a few days.
In the end however I don't think synchronization is the solution, it is the workaround. I think we simple should start to abandon the idea to get a whole distribution "stable" at some random point in time, instead isolate some core components and make sure that they are stable, but leave everything else to be updated as soon as upstream makes a new release. There is no reason why a new SuperTux release for example should have to wait for six month to be shipped to the user, it should get shipped as soon as its ready and with the current way distributions work that just doesn't really work.
PS: It would of course be nice if distributions would adopt the idea to have different versions of the same package installed at the same time, since the forced upgrade to a new version that you get today, is really one of the main reasons why we have that mess in the first place.
i thought that Debian had some kind of policies about testing each package before committing changes in testing/stable branches. This is a common myth, but Debian doesn't do any explicit testing, they pack things into unstable and then just let them go into testing/stable if no bug reports come in. There is no standard code audit or anything going on.
As a distro maintainer I can tell you that upstream providers generally do not care about distro-specific patches. This was not a distro specific patch. It was messing around with code because it triggered a warning in valgrind and this code happens to be deep down in OpenSSL. If it would be some random path in a Makefile I would agree, upstream doesn't care about that, but messing around with the code itself, especially in a security relevant package, should never happen without upstream.
Hm, could you explain a little more what is different between Blender under Windows and Blender under Linux? I was under the impression that its pretty much the same thing, since it has all its own GUI components and doesn't really make much if any use of OS specific features.
USENET did not have to evolve, Did not have to evolve? Because it didn't evolve we are now stuck with dozens of web forums with proprietary data storage and no way to retrieve posts other then the HTML interface. Yeah, I know USENET still exist, but pretty much everything these days happens on either mailing lists or web forums which both lack a lot of features that USENET had back then 20 years ago.
When you don't evolve stuff you have a very good chance to end up with a whole bunch of ugly ad hoc fixes.
What if there's child porn on my computer because of freenet? 1) Encrypt your hardrive.
2) Freenet doesn't transmit complete files, it chops them up into tiny blocks and encrypts those with themselves, so you won't ever find a JPG on your drive, you instead might find a little bit of random data, that is meaningless unless you have the key and also have all the blocks that are needed to make up the file.
3) Freenet isn't a fix for unjust laws, i.e. when they outlaw running a Freenet node due to childporn, terrorist or whatever, then you can't really fix that by running a Freenet node.
In the end Freenet really isn't so much a fix for lack of Freedom of Speech, but a test of it. If they let you run Freenet, good, you might still have some freedoms left, if they don't allow you to run Freenet, well, better start the revolution now, since Freedom of Speech is a thing of the past.
Freedom of speech does not mean - nor has it ever meant - that I have to open my home to provide services for the pornographer. On the other side it also doesn't mean that you can just slap him in the face when he happens to drive by your house. With Freenet you simply cache what is requested by the users of the net, it is kind of like a democracy, if you don't want it, don't request it. That doesn't stop content from being cached on your note, but it makes it less likely.
Isn't "Freedom of Speech" exactly *about* allowing those messages to spread that you disagree with? With Freenet it is really rather simple, when people don't request the content, then it will fall out of the network and disappear, if they do request it, then it gets cached, maybe even on your node. You might not like that, but that is the price to pay for Freedom of Speech.
Slashdot Mirror
The technical side of things seems to be well covered. If anything the comfort could be a problem, typing on a USB keyboard on a couch doesn't work as well as on a desk.
While true in theory, in practice its kind of wrong, since the GPL pretty much ensures that the normal business models used for software no longer work.
A few years down the road I fully expect to have a SSD in my computer, for larger storage and archival I might still have an additional spinning disk attached, but for daily use (OS,
Jordan Mechner was involved with Prince of Persia: Sands of Time as game designer and writer, in the second and third he wasn't involved at all and well, it shows. Sands of Time is simply a game that wouldn't have needed a sequel and what they did in Warrior Within is just disgusting.
There is quite a difference between a series of small glitches and an army of robots trashing down a police station, which happens even before the police knows what is going on, i.e. the robots *attack*, they don't just defend. And anyway, nothing of this chances that the red-glowing robots are in total violation to the first law.
Yes, Asimov's work was about unintended consequences of the laws, but it was never about "Lets just ignore the laws and do some shit".
His argument focuses on novices and teaching and there I fully agree, forcing people to do indention teaches good practice and produces code with a minimum amount of readability from everybody. However languages like Python where whitespace is the only way to indicate a block still drive me nuts. They make it way to easy to break code when you copy&paste stuff around and sometimes even impossible when you want to copy code from a webpage or forum. Whitespace is just way to easily lost or damaged.
I guess I would mind mandatory whitespace much less if it was used in combination with standard block start/end tokens instead of being the token itself.
I agree that I, Robot wasn't complete garbage, but it could have been a hell of a lot better. One of the central points of Asimovs books is that the robotic laws actually do work, not always as intended of course, but they got created exactly to avoid robots going on a killing spree, yet in I, Robot the movie they do exactly that and of course in the most stupid looking way possible (look, it's red, it must be evil...). The resolution at the end was of course not much better, switch the mainframe of and everything is back to normal. Would they have replaced that whole killing spree thing with something more Asimovian or went the route of Matrix's The Second Renaissance (which wouldn't have been Asimovian, but at least it would have made more sense) it could have ended up really good, but so its just a few good ideas and a lot of garbage in between.
The only real problem Sugar has is that applications take 10 seconds to start, even trivial "hello world" ones, which start pretty much instantly when you start them directly. Starting of Sugar itself is really a non-issue and much of the booting time is taking up by other stuff. So I really doubt EDE would be a good choice. On a small screen you really want something that doesn't force you to mess around with tiny little windows, but instead something that takes up the whole screen and lays out everything well right from the start and Sugar does exactly that.
Sugar simply needs to a decent amount of optimization and some cleanup with the Journal stuff, other then that it works great for the OLPC.
Do those bootstrap server step over any law? They don't tell you where to download files, they simply tell you where other clients are, kind of like DNS.
For a text based format one thing that can get out of date is charset, but then that is relatively easily converted and we also have Unicode now which should be good enough for a while to come.
Then there is of course the whole markup, that can easily get lost or forgotten in a 100 years. You still might be able to make some sense of the document, but getting an exact rendering of the document might become tricky, unless you also dig out a copy of OpenOffice and have an emulator that can run it.
In practical terms I think there are mainly two dangers of ODF. First there is the Internet and ePaper, both of which will sooner or later make paper obsolete. Why print something, when you can just email it around? And why print it, when the screen you are having already reads as easy as paper? So you might simply end up using other formats then ODF far more often and no longer end up having a copy of OpenOffice around in a few years. The second and bigger danger I think are however subtle improvements in the format. When ODF gets expanded and improved they might try to get backward compatibility going, but you likely never reach 100%. So you will have tiny little differences and when ODF1.1 is no longer used, since everybody is using ODF5.0, software might no longer support ODF1.1 properly either. It might still render, but the output might be wrong in the details. This would be basically what we have today already, you won't have much of a problem getting an old Word document to open, but you will have a very big issue if you want to get the exact rendering that the Word version it was created with produced.
Yep, and one also has to keep in mind that the voice actor is only a small part of what makes the character. You also have the artists that designed the character, the artists that created the 3d model, those that animated it, then maybe a stunt guy that did the motion capture for it and then sometimes even a special guy for the lip syncing and facial animation.
$100k sounds like a pretty good pay for just recording the dialog.
That said, I wouldn't mind if all those people involved got a share of the money that gets made, but its not like $100k is a rip off.
In the end however I don't think synchronization is the solution, it is the workaround. I think we simple should start to abandon the idea to get a whole distribution "stable" at some random point in time, instead isolate some core components and make sure that they are stable, but leave everything else to be updated as soon as upstream makes a new release. There is no reason why a new SuperTux release for example should have to wait for six month to be shipped to the user, it should get shipped as soon as its ready and with the current way distributions work that just doesn't really work.
PS: It would of course be nice if distributions would adopt the idea to have different versions of the same package installed at the same time, since the forced upgrade to a new version that you get today, is really one of the main reasons why we have that mess in the first place.
The SNES shoulder buttons don't click, the GCN shoulder buttons do.
Hm, could you explain a little more what is different between Blender under Windows and Blender under Linux? I was under the impression that its pretty much the same thing, since it has all its own GUI components and doesn't really make much if any use of OS specific features.
When you don't evolve stuff you have a very good chance to end up with a whole bunch of ugly ad hoc fixes.
When everything fails you could always just install a webcam, TrackIR or Wiimote and do some headtracking.
2) Freenet doesn't transmit complete files, it chops them up into tiny blocks and encrypts those with themselves, so you won't ever find a JPG on your drive, you instead might find a little bit of random data, that is meaningless unless you have the key and also have all the blocks that are needed to make up the file.
3) Freenet isn't a fix for unjust laws, i.e. when they outlaw running a Freenet node due to childporn, terrorist or whatever, then you can't really fix that by running a Freenet node.
In the end Freenet really isn't so much a fix for lack of Freedom of Speech, but a test of it. If they let you run Freenet, good, you might still have some freedoms left, if they don't allow you to run Freenet, well, better start the revolution now, since Freedom of Speech is a thing of the past.
Isn't "Freedom of Speech" exactly *about* allowing those messages to spread that you disagree with? With Freenet it is really rather simple, when people don't request the content, then it will fall out of the network and disappear, if they do request it, then it gets cached, maybe even on your node. You might not like that, but that is the price to pay for Freedom of Speech.