If you allow any random employee from within the organization - who probably hasn't had media training - to be quoted by reporters, then serious bad PR or misinformation can result. I think what's driving this in particular is a desire not to have Joe Engineer who's used to talking with other engineers give raw quotes to a clueless reporter and have his words completely misunderstood and thereby throw the public into a tizzy.
How can this 'misinformation' be the fault of the engineer and not of the 'clueless' reporter and his editor?
A fanciful but illustrative example: Joe Engineer from NHTSA may talk to the New York Times and use some intra-agency jargon like "in our latest survey, 99% of the bridges in the country got a designation of 'likely to collapse'." It turns out that 'likely to collapse' is an agency term measuring whether it is more likely to collapse or be struck by a meteor, but the reporter (who doesn't know any better, and wouldn't be expected to know better unless Joe explains it to him - and we've never heard before of a technical person who fails to explain their jargon) puts in the paper verbatim: "NHTSA says 99% of bridges likely to collapse." (Cue mass hysteria.)
Its not the fault of the engineer without media training to explain any of his jargon.
Its the fault of the reporter with media trainging to not ask for clarification before causing mass hysteria.
And its the fault of the editors to allow such alarmist headlines to exists just to push sales/ratings.
And sadly its also the fault of the public to a) "wanting" such headlines and b) believing such headlines.
Executives and other people in the organization who are given media training are at least taught how to choose their words for public consumption carefully (whether they do it properly or not is a different issue).
It should never be the subject of a report that gets to choose the wording. This is whats enabled all that spinning and newspeak. I really hate it, that I no longer only need to cut through all the lies and half truth in advertising, but also in the "news" that more and more only seem to be made from press releases that only are thinly veiled advertisings.
I thought it was the job of the reporter to do this cutting and presenting a more or less objective truth. If I have to do this myself anyways I can just start reading the press releases directly and stop paying for newspapers/reporters.
I don't see how it's in their best interest to always defend the Quran from people misquoting it because they didn't have a reliable translation.
But it is in there best interest to use the catch-all 'translation error' for any questions.
Otherwise they would have to find real, valid arguments when discussing several of the more anachronistic commandments of their religion.
(That not only fits discussions about Quran/Muslims but any religion. Like 'You should not kill' and cursades just to mention on glaring example of inconsistency in Christianity. Its very nice to have the easy way out as in 'Please learn language x and read the original'. Why do you think did the church try to stop Martin Luther translating the bible to german?)
I would get myself another HDD with another Windows installation just for that game.
I hope you will also pay M$ another $200 for that second install.
IANAL, but I don't think that kind of secondary install is allowed. At the very least I would expect problems activating the same key again and again when more games try this kind of stunt.
So until they openly say they are using linux and offer a way to upgrade the software on the system I will NOT buy one of these.
I did think about getting one of these. It has really nice features and if I could put my own linux system on one of the hard disk I could use it also as a dsl router and proxy (squid).
Anyone knows of a similar device with an upgradeble linux?
Yes.;) But the point was, that even if your architecture is supported right now, it does not mean it will stay this way. And on the other hand, if nVidia/ATI do not think your architecture is worth the work, then it will never get a working driver (ppc).
and that 2) anyone who is running an Itanium system (presumably rather expensive and proprietary) is likely going to have a team of engineers writing their own software (and probably drivers, if they need 3D graphics support).
Maybe now, but in the start of ia64 you could buy 'normal' workstations. There even was a ia64-windows before the amd64-windows.
On the other hand x86_64 is already a widely accepted commercial standard
Cynical question: And how long will it take for nVidia/ATI to drop support for the old 32bit CPUs? IA64 will still be around for some years because it is positioned in the server market. But very soon you will no longer be able to buy 32bit only x86 CPUs. And if all new CPUs support x86_64 why should offer nVidia/ATI 32bit driver for there next generation of cards? Yeah, I know the installed base of ia64 ist much smaller that x86, but nobody can say for sure how long which product will be supported by closed source drivers. Just look at Matrox. They had open source linux drivers for up to G450. You can use this card in 64bit linux. The closed source drivers for all newer cards are 32bit only.
If you are running an old server, then presumably it wont actually *need* any of your new software
Why shouldn't I upgrade even an older machine to get new features, if these features do not need much more memory / processing power? Like keeping up to date with samba. Many releases from samba do not need more resources than previous versions, but most of the time enable new features or include more fixes for special needs of some windows versions.
, and is even not likely to need many security patches if it's configured correctly
I don't think so. Not all of theses bugs can be worked around. Just think about the last announcements of firefox 1.5.0.2 or 1.0.8.
I dont even think of a server as something that needs a graphical display..
Yes, but my point was more abount recycling/ keep on using old hardware.
running the latest OS on old hardware, unless there really are significant security flaws with the old OS, doesn't really seem to be necessary.
But why using linux-2.4, if an upgrade linux-2.6 does not cost you anything?
You dont need SATA support for an ancient machine, and if your machine actually has SATA onboard, chances are that the processor will run your new OS fine,
But there are PCI cards with SATA controller. I am using my old desktop (the machine with the GF2) right now as fileserver for my home-LAN and as a backup desktop, if I hose my main desktop once again (gentoo...;)). Then I need ('want') this system with accelleradet X and a recent firefox to search online for clues on how to fix my main system.
and that it wont have a GF2 either, unless you have one lying around (like I have). Having said that, not having speedy graphics support on a server is not an issue. Th dual core Opteron HP Proliants I've just bought actually come with GF2 MXs.
You are lucky. The GF2 MX is the only GF2 that is still supported with the latest drivers... But for how long?
You should not expect 32 bit drivers to work for a 64 bit environment
The newer drivers are also 64bit. But only for x86_64 == amd64 == emt64. The 64bit ia64 == itanium architecture is no longer supported.
- also, if you have a GeForce 2, and still have games or programs that you expect to be able to run on that GeForce 2, then why not use the older drivers and an older version of linux to run it? With each new generation of computer hardware, software engineers tend to try to take advantage of all the processor power they can.
Maybe because I want to use my old Geforce2 using server to be part of my beowulf cluster and use ocfs2?
Using the old nVidia driver means also using an older kernel. And that means: a) I can't use the features of newer kernels. b) Nobody fixes any security bugs. c) I can't use software that needs a newer kernel. c) I can't get newer driver that are included in recent kernels. (For Example: SATA NCQ)
I'm not so happy with that when it comes to things like Vista needing 256MB/whatever just for the OS alone, but when it comes to things like games, I wouldnt expect or even want to be able to play Quake 4 on a GeForce 2 (I do actually still a machine with a GF2, but I'm only keeping it in case I want to run a server at home again).
Yeah, using it as backup server is why I keep it. But if I need to use it, I still want current software on it. KDE 3.5 is not so much slower than KDE 3.0. But if the xorg-server falls back to vesa THAT will be a major slowdown.
I think the fact that nVidia produces linux drivers at all is great
Except the latest drivers will no longer work with Linux on ia64 or with older GeForce2 cards. And at any time they can decide to no longer support i386 because "everybody" uses x86_64. Or drop support for Geforce4...
I will soon by a new pc and I'm still think on what graphic card to buy. NVidia has the better propritary driver, but there is an open source driver for the R3xx and R4xx chips (up to radeon X850 I think). But do I really what to give money to a ATI because of an open source driver, whose developer had zero support from ATI?
I did read about XGI open sourcing there drivers, but that was only the 2D part, the 3D acceleration was to come later. Has this happened or was it canceled when ati brought them?
One other thing about binary drivers: Has anybody tried to load both propritary kernel modules at the same time? If you own an SLI oder crossfire board it shouldn't be a problem to plug a nVidia card in one PEG slot and an a ATI card in the other. But as it is not even possible to run 2 Xserver with the ATI driver at the same time, I would suspect that this setup will not work. And neither nVidia nor ATI will probably help you to get that thing to work. I would expect them to just tell you to buy another of their cards.
If the drivers where open source, someone from x.org would probably try to fix this, as they to not want to sell more video cards, but to make better drivers.
I would imagine that, up to a certain point, these disadvantages could be dealt with through proper engineering.
The problem is that when size scales x the mass of an object scales at x^3. The cut-area of arms/legs/etc that is critical for the robustnes of a mechanism only scales at x^2. This problem cant be solved with engineering.
Yes, a steel bone is better than a real bone, but also weights more. Maybe carbon-fiber would be a better material. But if you scale a human from 2m to 10m you increase the weight by 125 times. The forces you need to move increase by 625 [1]. The area in the limbs only increases by 25. So the material would need to be 25 times as strong.
Material science gives us structural strength & armor protection which far exceeds the strength of almost any organic creature
I think spidersilk is still the stonges thread anyone can make.;) Nature is VERY resource-efficient.
Also, the mechs might not need to be 10m high - they might be only as big as the tanks that they are competing with (are tanks 10m long?).
A quick google showed a length of 9.83m for an M1A1 Abrahams. So 10m-tanks are OK.
And to tip it over your mech would need to lift nearly 70 tons. At my rough guess you probably would need a 10m mech to tip over a 10m tank.
cruching stones, dents in armor
OK, maybe not stones, but tank-barriers make from steel-grinders could impale the robot (or parts of it).
As for dents: Yes an flexibile 'myomer'-muscle would probably a solution for that particular problem. But (there are always buts.;) ) it would probably mean that you also need a relativ flexible outer armor. A fixed mounted armor of a tank would hold better agains gunfire. The main weakness of all plate armors where always the joints.
The problem with this reasoning, is that any form of external support that the tanks can use, the mechs can also use.
My reasoning included a budget. That you could get airsupport and the 4 tanks for the price of one mech without airsupport. But combined tactics should not be overlooked.
Cratering all streets is an idea, but tank-barriers would probably work better. (Tank with shields like bulldozers could push dirt into the crater to pave the way.)
But for preparing a battleground mines would be the easiest way. This would work against mech and against tanks. So I would leave this part out when comparing mechs and tanks.
Also, in such an engagement, you wouldn't try and flip the tanks over first - you'd disable (bend) the main gun on each tank -...
It would be a bad tactic if all tanks would enter the crater-zone at the same time. One would go first an the others will cover it. When it has reached a "save" position it can stop can cover the next tanks that one by one navigate the obstacle.
Bending the main gun will work well against normal tanks, but if you design anti-tank-mechs, you would have to face anti-mech-tanks or at least mech-proof-tanks. Trade some range of the main gun for a shorter thicker cannon that could not be bend to easily. Or attach huge blades one the side of the turret that could be swung under/over the main gun. So if a mech stands near your cannon to grab it, release the blades and hack it legs or head away.
Also, if the mechs can move fast enough in short spurts, I'd say that it would be easier for our air support to shoot the tanks than it would be for the enemy air support to shoot wildly-dodging, jumping & hiding humanoid figures.
Such actions would require much energy. That means heat , and that means a good IR-target.
OK, mechs might be difficult to hit precicly, but if you have bombs that leave craters that are diffult to drive throught with tanks, the precission does not need to be too good.
The main disadvantages I see with mechs are: 1) cost & complex
You can say that about any biped organism - but there are obvious examples all around you where nature has made it work.
The problem is the scale. A 10m Robot is much more prone to broken arms or legs then a 2m human. An 1cm insect can fall down over 1m without any brocken legs. Try a 200m fall with a 2m human...
Also a human has a protection in form of flesh that softens the impact on the ground. A steel robot would be permanently deformed, if it land on a pointy stone
I'm pretty sure they'd be fairly effective against tanks - either flipping the tank upside down, or jumping up and down on top of the tank would work.
It was 100 tanks against 1 robot. Even at 10:1 the robot would flip over 1 tank and the 9 other tanks could nail it with there main cannons. (Even if it was quick it needs some time to flip over a tank.)
An real advantage would be in heavy terrain, where tanks could not follow. Or maybe things like ducking behind buildings/boulders.
The tanks could try to minimize the terrain advantage by using air support. 1 robot against 4 tanks and 2 predator-drones. If the robot trys to flee into the mountains, where the tanks could not follow, the drones could still reach it.
The one hope is that Jodrell picks up something... assuming they don't get jammed.
From www.beagle2.com: An attempt to listen out for Beagle's call home by the Westerbork telescope array in the Netherlands was unfortunately interrupted by strong radio interference.
OLD MAN was half a share with a diameter of 200 km. Should be a little larger then the Death Star.
(It had over 20000 of the 2500m Galaxy-Class-Ultrabattleships on board)
But in the Perry Rhodan series there were the Sporeships with a diameter of a little more then 1000 km easily dwarfing the Death Stars.
Also there were the Eternal Servants, a robot race. They flew across the universe with some structures mesauring in lightmonth. They win even against the Magog Worldship or a Dyson Sphere. And they were real Starships with FTL-Drives.
Slashdot Mirror
At least since 2.6.12 you can disable this feature with a simple option on the kernel commandline.
With the newer firewire stack, it will default to off, so you do not need this option anymore.
For more information about this see the file Documentation/debugging-via-ohci1394.txt in the kernel source tree
Its the fault of the reporter with media trainging to not ask for clarification before causing mass hysteria.
And its the fault of the editors to allow such alarmist headlines to exists just to push sales/ratings.
And sadly its also the fault of the public to a) "wanting" such headlines and b) believing such headlines. It should never be the subject of a report that gets to choose the wording. This is whats enabled all that spinning and newspeak. I really hate it, that I no longer only need to cut through all the lies and half truth in advertising, but also in the "news" that more and more only seem to be made from press releases that only are thinly veiled advertisings.
I thought it was the job of the reporter to do this cutting and presenting a more or less objective truth. If I have to do this myself anyways I can just start reading the press releases directly and stop paying for newspapers/reporters.
Otherwise they would have to find real, valid arguments when discussing several of the more anachronistic commandments of their religion.
(That not only fits discussions about Quran/Muslims but any religion. Like 'You should not kill' and cursades just to mention on glaring example of inconsistency in Christianity. Its very nice to have the easy way out as in 'Please learn language x and read the original'. Why do you think did the church try to stop Martin Luther translating the bible to german?)
IANAL, but I don't think that kind of secondary install is allowed. At the very least I would expect problems activating the same key again and again when more games try this kind of stunt.
Yeah the N5200 does use linux. But I did not find any clear hint to this on the Thecus website.
Also I'm missing any documentation of how to upgrade the firmware to your own linux system.
If you want the source of their linux look here:
ftp://ftp.gpl-devices.org/pub/vendors/Thecus/
They tried to hide the linux, but without success:
http://gnumonks.org/~laforge/weblog/2006/02/24/
So until they openly say they are using linux and offer a way to upgrade the software on the system I will NOT buy one of these.
I did think about getting one of these. It has really nice features and if I could put my own linux system on one of the hard disk I could use it also as a dsl router and proxy (squid).
Anyone knows of a similar device with an upgradeble linux?
I thought that 1) Itanium was/is a major flop
;)
;)). Then I need ('want') this system with accelleradet X and a recent firefox to search online for clues on how to fix my main system.
Yes.
But the point was, that even if your architecture is supported right now, it does not mean it will stay this way.
And on the other hand, if nVidia/ATI do not think your architecture is worth the work, then it will never get a working driver (ppc).
and that 2) anyone who is running an Itanium system (presumably rather expensive and proprietary) is likely going to have a team of engineers writing their own software (and probably drivers, if they need 3D graphics support).
Maybe now, but in the start of ia64 you could buy 'normal' workstations. There even was a ia64-windows before the amd64-windows.
On the other hand x86_64 is already a widely accepted commercial standard
Cynical question: And how long will it take for nVidia/ATI to drop support for the old 32bit CPUs?
IA64 will still be around for some years because it is positioned in the server market. But very soon you will no longer be able to buy 32bit only x86 CPUs. And if all new CPUs support x86_64 why should offer nVidia/ATI 32bit driver for there next generation of cards?
Yeah, I know the installed base of ia64 ist much smaller that x86, but nobody can say for sure how long which product will be supported by closed source drivers.
Just look at Matrox. They had open source linux drivers for up to G450. You can use this card in 64bit linux. The closed source drivers for all newer cards are 32bit only.
If you are running an old server, then presumably it wont actually *need* any of your new software
Why shouldn't I upgrade even an older machine to get new features, if these features do not need much more memory / processing power? Like keeping up to date with samba. Many releases from samba do not need more resources than previous versions, but most of the time enable new features or include more fixes for special needs of some windows versions.
, and is even not likely to need many security patches if it's configured correctly
I don't think so. Not all of theses bugs can be worked around.
Just think about the last announcements of firefox 1.5.0.2 or 1.0.8.
I dont even think of a server as something that needs a graphical display..
Yes, but my point was more abount recycling/ keep on using old hardware.
running the latest OS on old hardware, unless there really are significant security flaws with the old OS, doesn't really seem to be necessary.
But why using linux-2.4, if an upgrade linux-2.6 does not cost you anything?
You dont need SATA support for an ancient machine, and if your machine actually has SATA onboard, chances are that the processor will run your new OS fine,
But there are PCI cards with SATA controller.
I am using my old desktop (the machine with the GF2) right now as fileserver for my home-LAN and as a backup desktop, if I hose my main desktop once again (gentoo...
and that it wont have a GF2 either, unless you have one lying around (like I have). Having said that, not having speedy graphics support on a server is not an issue. Th dual core Opteron HP Proliants I've just bought actually come with GF2 MXs.
You are lucky. The GF2 MX is the only GF2 that is still supported with the latest drivers... But for how long?
You should not expect 32 bit drivers to work for a 64 bit environment
The newer drivers are also 64bit. But only for x86_64 == amd64 == emt64. The 64bit ia64 == itanium architecture is no longer supported.
- also, if you have a GeForce 2, and still have games or programs that you expect to be able to run on that GeForce 2, then why not use the older drivers and an older version of linux to run it? With each new generation of computer hardware, software engineers tend to try to take advantage of all the processor power they can.
Maybe because I want to use my old Geforce2 using server to be part of my beowulf cluster and use ocfs2? Using the old nVidia driver means also using an older kernel. And that means: a) I can't use the features of newer kernels. b) Nobody fixes any security bugs. c) I can't use software that needs a newer kernel. c) I can't get newer driver that are included in recent kernels. (For Example: SATA NCQ)
I'm not so happy with that when it comes to things like Vista needing 256MB/whatever just for the OS alone, but when it comes to things like games, I wouldnt expect or even want to be able to play Quake 4 on a GeForce 2 (I do actually still a machine with a GF2, but I'm only keeping it in case I want to run a server at home again).
Yeah, using it as backup server is why I keep it. But if I need to use it, I still want current software on it. KDE 3.5 is not so much slower than KDE 3.0. But if the xorg-server falls back to vesa THAT will be a major slowdown.
I think the fact that nVidia produces linux drivers at all is great
Except the latest drivers will no longer work with Linux on ia64 or with older GeForce2 cards. And at any time they can decide to no longer support i386 because "everybody" uses x86_64. Or drop support for Geforce4...
I will soon by a new pc and I'm still think on what graphic card to buy.
NVidia has the better propritary driver, but there is an open source driver for the R3xx and R4xx chips (up to radeon X850 I think).
But do I really what to give money to a ATI because of an open source driver, whose developer had zero support from ATI?
I did read about XGI open sourcing there drivers, but that was only the 2D part, the 3D acceleration was to come later. Has this happened or was it canceled when ati brought them?
One other thing about binary drivers:
Has anybody tried to load both propritary kernel modules at the same time?
If you own an SLI oder crossfire board it shouldn't be a problem to plug a nVidia card in one PEG slot and an a ATI card in the other.
But as it is not even possible to run 2 Xserver with the ATI driver at the same time, I would suspect that this setup will not work.
And neither nVidia nor ATI will probably help you to get that thing to work.
I would expect them to just tell you to buy another of their cards.
If the drivers where open source, someone from x.org would probably try to fix this, as they to not want to sell more video cards, but to make better drivers.
The new Cell chip - first step towards octopus robots?
Octopus: central brain + 8 arm brains
Cell: central PowerPC chip + 8 SPU
The problem is that when size scales x the mass of an object scales at x^3. The cut-area of arms/legs/etc that is critical for the robustnes of a mechanism only scales at x^2. This problem cant be solved with engineering.
Yes, a steel bone is better than a real bone, but also weights more. Maybe carbon-fiber would be a better material. But if you scale a human from 2m to 10m you increase the weight by 125 times. The forces you need to move increase by 625 [1]. The area in the limbs only increases by 25. So the material would need to be 25 times as strong.
I think spidersilk is still the stonges thread anyone can make. ;) Nature is VERY resource-efficient.
A quick google showed a length of 9.83m for an M1A1 Abrahams. So 10m-tanks are OK.
And to tip it over your mech would need to lift nearly 70 tons. At my rough guess you probably would need a 10m mech to tip over a 10m tank.
OK, maybe not stones, but tank-barriers make from steel-grinders could impale the robot (or parts of it). ;) ) it would probably mean that you also need a relativ flexible outer armor. A fixed mounted armor of a tank would hold better agains gunfire. The main weakness of all plate armors where always the joints.
The problem with this reasoning, is that any form of external support that the tanks can use, the mechs can also use.
My reasoning included a budget. That you could get airsupport and the 4 tanks for the price of one mech without airsupport. But combined tactics should not be overlooked.
As for dents: Yes an flexibile 'myomer'-muscle would probably a solution for that particular problem. But (there are always buts.
Cratering all streets is an idea, but tank-barriers would probably work better. (Tank with shields like bulldozers could push dirt into the crater to pave the way.)
But for preparing a battleground mines would be the easiest way. This would work against mech and against tanks. So I would leave this part out when comparing mechs and tanks.
It would be a bad tactic if all tanks would enter the crater-zone at the same time. One would go first an the others will cover it. When it has reached a "save" position it can stop can cover the next tanks that one by one navigate the obstacle.
Bending the main gun will work well against normal tanks, but if you design anti-tank-mechs, you would have to face anti-mech-tanks or at least mech-proof-tanks. Trade some range of the main gun for a shorter thicker cannon that could not be bend to easily. Or attach huge blades one the side of the turret that could be swung under/over the main gun. So if a mech stands near your cannon to grab it, release the blades and hack it legs or head away.
Such actions would require much energy. That means heat , and that means a good IR-target.
OK, mechs might be difficult to hit precicly, but if you have bombs that leave craters that are diffult to drive throught with tanks, the precission does not need to be too good.
You can say that about any biped organism - but there are obvious examples all around you where nature has made it work.
The problem is the scale. A 10m Robot is much more prone to broken arms or legs then a 2m human. An 1cm insect can fall down over 1m without any brocken legs. Try a 200m fall with a 2m human...
Also a human has a protection in form of flesh that softens the impact on the ground. A steel robot would be permanently deformed, if it land on a pointy stone
I'm pretty sure they'd be fairly effective against tanks - either flipping the tank upside down, or jumping up and down on top of the tank would work.
It was 100 tanks against 1 robot. Even at 10:1 the robot would flip over 1 tank and the 9 other tanks could nail it with there main cannons. (Even if it was quick it needs some time to flip over a tank.)
An real advantage would be in heavy terrain, where tanks could not follow. Or maybe things like ducking behind buildings/boulders.
The tanks could try to minimize the terrain advantage by using air support. 1 robot against 4 tanks and 2 predator-drones. If the robot trys to flee into the mountains, where the tanks could not follow, the drones could still reach it.
The one hope is that Jodrell picks up something ... assuming they don't get jammed.
;)
From www.beagle2.com:
An attempt to listen out for Beagle's call home by the Westerbork telescope array in the Netherlands was unfortunately interrupted by strong radio interference.
Definitly a Conspiracy.
In Russia the Beowulf Cluster imagines you?
OLD MAN was half a share with a diameter of 200 km. Should be a little larger then the Death Star. (It had over 20000 of the 2500m Galaxy-Class-Ultrabattleships on board) But in the Perry Rhodan series there were the Sporeships with a diameter of a little more then 1000 km easily dwarfing the Death Stars. Also there were the Eternal Servants, a robot race. They flew across the universe with some structures mesauring in lightmonth. They win even against the Magog Worldship or a Dyson Sphere. And they were real Starships with FTL-Drives.