Bah, hit return to early. Meant to post this chunk as well:
Input: psmouse - reset mouse before doing intellimouse/explorer probes in case it got confused by earlier probes; switch to streaming code before setting scale and resolution, otherwise some KVMs get confused.
There are a whole load of other relevant-sounding fixes in the input system listed as well; damnit, if only I'd found the full changelog earlier!
Well, as the LKML will show it's an odd problem that comes and goes for alot of users under generally different conditions (many aren't using KVM's and such at all). The full changelog http://www.kernel.org/pub/linux/kernel/v2.6/Change Log-2.6.11 shows a number of fixes made to the input system, so I'm keeping my fingers crossed. I'll try vanilla out tonight and see if that helps any...
Aye, I've been perusing the LKML archives for a few days (since the system has only been up and running for a coupla weeks, and the first week was windows configuration) and I'm slowly compiling data about it and trying a few random patches - it's very much a known issue and goes all the way back to 2.5. If I can't fix it within a week I may well stump up the moolah (and I've got more USB->PS/2 adapters than you can shake a stick at), although getting it to someone outside the UK will be tricky.
All in all it's still a very small blemish on what is otherwise the most crap-reistant system I've ever used:) I could probably solve it all by doing away with the adapters, but it makes dual booting a pain (esp. as I find USB very unresponsive in windows under heavy load).
Does anyone know if any work is being done on the rather irritating problem I (and a fair few others) get with the 2.6 input system? Every time I use a mouse that goes through a KVM switch or a USB->PS2 adapter, the mouse would spazz around crazily and syslog would fill up with:
lost synchronisation, throwing [1|2|3] bytes away
Adding psmouse.proto=imps made the problem go away for most usage, but it still occurs under very heavy load, which makes mplaying UT impossible:(
Last I heard, the first dual cores out of the door won't share cache and will be more like two individual dies sharing the same packaging. But they'll be moving onto shared cache a la POWER at a later date.
It would be interesting if games were rewritten to run with the game logic on one core, the graphics on another core and the networking code on a third core of a multicore chip...
Isn't that what IBM/Sony are propsing with the Cell architecture? Lots of seperate cores running dedicated chunks of code?
Dunno about anyone else, but I'd pay good money to have everyone's ringtones replaced by Bender saying "PAAAAAAAAAAAAAAARRRRP! Pretty annoying, huh?".
I really, really don't see the point in paying an utter fortune for a terrible rendition of your tune du jour. And it's not helped by phones seeming to want to broadcast the ringtones to saturn by way of their new found super-loud speakers. Reminds me of the 80's stereotype of the "yoof" roaming around the inner cities with his ghetto blaster.
My nokia uses two ringtones. One goes ring ring for general use. The other does a quiet beepy chirrup which I use during work. Meetings and the like get vibrate only.
Didn't the 2200's come in both the Palomino and Tbred cores? IIRC the Palomino's all ran at fairly high temps (with the 2200 being the hottest and last of the Palomino line), whilst the 130nm Tbreds ran much cooler and went all the way up to the 2600's.
Same thing with any integrated GFX chipset that uses system RAM as VRAM.
It's not so much that they suck CPU cycles, it's just the way that the faux-GPU is constantly accessing your main memory in order to refresh your screen at 60 times a second (or higher), so your chip is choking for lack of memory bandwidth. This doesn't help when coupled with P4's that are extra-sensitive to memory performance.
Even when I'm building an ultra-low-spec computer for a friend, I'll still splash out on an AGP card for them - they're just make a much more responsive system, even if graphical performance is equal.
There's a big difference between shipping an OS that has every service known to humanity running by default and shipping an OS that will selectively enable services as and when you need them.
Have you ever tried to disable some of the windows services? They all seem to depend on one another, and half of them seem to have something to do with networking and/or some file sharing protocol or other.
Not so in UNIX; you can just turn them off as and when. The most complex that springs to mind is NFS, which only requires 4 or 5 daemons, all of which you can turn on and off without crippling some other part of the machine. UNIX seems to follow the KISS rule alot better than windows.
Not that I'm saying modern distros make it evident (and TBH I'm totally out of touch with modern GUI-based distros; I use Debian mysel) but it'd be a simple matter.
"The video card market is fast paced and volatile. If nVidia does produce garbage, the market will react accordingly, and drop them like a hot rock."
Too damned right, and this is exactly what happened with nVidia's GeForce FX series. Hot, noisy, underpowered - and everyone switched to ATI 9700's and the like. nVidia eventually pulled their socks up and released the GF6 series which have put nVidia back in the running again.
In terms of Linux drivers, ATI have produced nothing but garbage as far as I'm concerned - hence why I dropped them like a hot rock;) Even though they're Binary-Only And Therefore Evil, the nVidia drivers have proved remarkably solid even on cutting edge hardware and software. There's currently a problem with the latest driver release and my MX4000 in the TV box, but it's a world apart from the world of "emerge ati-drivers and pray" I experienced with a friends 9700 Pro.
Whilst I agree that at the moment AMD have a superior product to Intel, they don't have Intel's marketing or production clout. They can't fully capitalise on their innovations cos a) not enough people know how much their chips kick ass (thanks in no small part to the techs at IBM) and b) even if they did, AMD wouldn't be able to produce enough chips.
Intel have just been unlucky enough to back themselves into a corner (Itanium, Prescott) at the same time AMD invent a far superior product. Expect AMD to gain a slightly bigger slice of the pie before Intel manage to catch up; that P4-M/revamped P3 is doing pretty damned well.
"Now, if/when they come out with memory that can be reorganized on-the-fly, perform large-scale simple massively-parrallel operations, and do some content-addressable tricks, that will be a signifigant development."
Isn't that what NUMA is about? AFAICR all it needs is a board that can take advantage of it, along with an OS. I think some of the Opteron boards have NUMA and all that lot.
From Wikipedia: neutron stars (pulsars are typically spinning neutron stars that emit alot of EM radiation in regular bursts) are about 10-20km in diameter with masses of between 1.5 and 3 times the mass of Sol (approx. 2,000,000,000,000,000,000,000,000,000,000kg).
By that, the densisy is probably about eight LoC's per hogshead furlong.
"And then your software can be just as slow and tedious to use as the real items!"
Why? Did I say that?
It's quicker to flick through a photo album than it is to view each image individually by double clicking it. Want a thumbnail view? Click/speak the "thumbnail" button and the album will re-arrange itself into a 10 foot by 3 foot montage, or starburst in front of you, or a stack of categorised photos, or whatever.
My point was that as human beings, we're more used to accessing information (words, pictures) on a 2D plane (paper) and yet interacting with in in a 3D environment (the real world). All the 3D UI's I've seen so far just seem to use the 3D for extra eye candy; there's no real difference to keeping the windows in squished 3D trapezoids at the side of the screen as there is to keeping them minimised/maximised/stacked/tiled or whatever, or using a pager with virtual desktops.
...that 3D UI's will work until we have a way of interacting with them in 3D.
Pretty much all of the 3D UI's I've seen are still presented to the user on 2D monitor, and interacted with by way of a 2D mouse (or keyboard). I don't really think 3D will offer any advantage until we can look at and interact with 3D objects a la Metaverse/SnowCrash. Such as flicking through a photo album, juggling your files up and down and all over the place, pushing/pulling your work in front of you.
But maybe I just can't envisage a 3D world without an intuitive way of interacting with in one less dimension. Or maybe I've read too much sci-fi...
Jeepers. Please, no-one tell them about graphics card memory, which keeps a pixel-perfect copy of *the entire movie* in realtime...! I still want to be able to finish Doom3.
Slashdot Mirror
I've paid for Opera. Twice. Opera software have been kind enough to grant me a license that will let me use it on every computer in the house.
Why do I pay for a browser when FF is free? Because as great and capable as FF/Moz are, I prefer Opera.
Bah, hit return to early. Meant to post this chunk as well:
Input: psmouse - reset mouse before doing intellimouse/explorer probes in case it got confused by earlier probes; switch to streaming code before setting scale and resolution, otherwise some KVMs get confused.
There are a whole load of other relevant-sounding fixes in the input system listed as well; damnit, if only I'd found the full changelog earlier!
Well, as the LKML will show it's an odd problem that comes and goes for alot of users under generally different conditions (many aren't using KVM's and such at all). The full changelog http://www.kernel.org/pub/linux/kernel/v2.6/Change Log-2.6.11 shows a number of fixes made to the input system, so I'm keeping my fingers crossed. I'll try vanilla out tonight and see if that helps any...
Aye, I've been perusing the LKML archives for a few days (since the system has only been up and running for a coupla weeks, and the first week was windows configuration) and I'm slowly compiling data about it and trying a few random patches - it's very much a known issue and goes all the way back to 2.5. If I can't fix it within a week I may well stump up the moolah (and I've got more USB->PS/2 adapters than you can shake a stick at), although getting it to someone outside the UK will be tricky.
:) I could probably solve it all by doing away with the adapters, but it makes dual booting a pain (esp. as I find USB very unresponsive in windows under heavy load).
All in all it's still a very small blemish on what is otherwise the most crap-reistant system I've ever used
Does anyone know if any work is being done on the rather irritating problem I (and a fair few others) get with the 2.6 input system? Every time I use a mouse that goes through a KVM switch or a USB->PS2 adapter, the mouse would spazz around crazily and syslog would fill up with:
:(
lost synchronisation, throwing [1|2|3] bytes away
Adding psmouse.proto=imps made the problem go away for most usage, but it still occurs under very heavy load, which makes mplaying UT impossible
Last I heard, the first dual cores out of the door won't share cache and will be more like two individual dies sharing the same packaging. But they'll be moving onto shared cache a la POWER at a later date.
, 39164618,00.htm
Tried to find the Reg link I read a while back, and then found this one: http://news.zdnet.co.uk/hardware/chips/0,39020354
Isn't that what IBM/Sony are propsing with the Cell architecture? Lots of seperate cores running dedicated chunks of code?
Gold plated?! If I'm going to be paying higher prices for my tunes, I want oxygen-free bytes as well!
Dunno about anyone else, but I'd pay good money to have everyone's ringtones replaced by Bender saying "PAAAAAAAAAAAAAAARRRRP! Pretty annoying, huh?".
I really, really don't see the point in paying an utter fortune for a terrible rendition of your tune du jour. And it's not helped by phones seeming to want to broadcast the ringtones to saturn by way of their new found super-loud speakers. Reminds me of the 80's stereotype of the "yoof" roaming around the inner cities with his ghetto blaster.
My nokia uses two ringtones. One goes ring ring for general use. The other does a quiet beepy chirrup which I use during work. Meetings and the like get vibrate only.
It's official - the evidence will be released along with the new Debian Stable, along with DNF ;)
Didn't the 2200's come in both the Palomino and Tbred cores? IIRC the Palomino's all ran at fairly high temps (with the 2200 being the hottest and last of the Palomino line), whilst the 130nm Tbreds ran much cooler and went all the way up to the 2600's.
Same thing with any integrated GFX chipset that uses system RAM as VRAM.
It's not so much that they suck CPU cycles, it's just the way that the faux-GPU is constantly accessing your main memory in order to refresh your screen at 60 times a second (or higher), so your chip is choking for lack of memory bandwidth. This doesn't help when coupled with P4's that are extra-sensitive to memory performance.
Even when I'm building an ultra-low-spec computer for a friend, I'll still splash out on an AGP card for them - they're just make a much more responsive system, even if graphical performance is equal.
There's a big difference between shipping an OS that has every service known to humanity running by default and shipping an OS that will selectively enable services as and when you need them.
Have you ever tried to disable some of the windows services? They all seem to depend on one another, and half of them seem to have something to do with networking and/or some file sharing protocol or other.
Not so in UNIX; you can just turn them off as and when. The most complex that springs to mind is NFS, which only requires 4 or 5 daemons, all of which you can turn on and off without crippling some other part of the machine. UNIX seems to follow the KISS rule alot better than windows.
Not that I'm saying modern distros make it evident (and TBH I'm totally out of touch with modern GUI-based distros; I use Debian mysel) but it'd be a simple matter.
"The video card market is fast paced and volatile. If nVidia does produce garbage, the market will react accordingly, and drop them like a hot rock."
;) Even though they're Binary-Only And Therefore Evil, the nVidia drivers have proved remarkably solid even on cutting edge hardware and software. There's currently a problem with the latest driver release and my MX4000 in the TV box, but it's a world apart from the world of "emerge ati-drivers and pray" I experienced with a friends 9700 Pro.
Too damned right, and this is exactly what happened with nVidia's GeForce FX series. Hot, noisy, underpowered - and everyone switched to ATI 9700's and the like. nVidia eventually pulled their socks up and released the GF6 series which have put nVidia back in the running again.
In terms of Linux drivers, ATI have produced nothing but garbage as far as I'm concerned - hence why I dropped them like a hot rock
Soon only the criminals will have tinfoil!
Roast dinners will be a thing of the past.
(Humour-impaired mods - it's a lame joke)
Look at the back of your iPod.
/. eds these days ;)
It's quite definitely the shaving *mirror*.
Honestly.
I'll see your American capitalistic republic, and raise you a corporate plutocracy.
Yay! Not only do you get a bootable Linux distro, you'll also get a bootable kernel.org mirror ;)
Whilst I agree that at the moment AMD have a superior product to Intel, they don't have Intel's marketing or production clout. They can't fully capitalise on their innovations cos a) not enough people know how much their chips kick ass (thanks in no small part to the techs at IBM) and b) even if they did, AMD wouldn't be able to produce enough chips.
Intel have just been unlucky enough to back themselves into a corner (Itanium, Prescott) at the same time AMD invent a far superior product. Expect AMD to gain a slightly bigger slice of the pie before Intel manage to catch up; that P4-M/revamped P3 is doing pretty damned well.
"Now, if/when they come out with memory that can be reorganized on-the-fly, perform large-scale simple massively-parrallel operations, and do some content-addressable tricks, that will be a signifigant development."
;)
Isn't that what NUMA is about? AFAICR all it needs is a board that can take advantage of it, along with an OS. I think some of the Opteron boards have NUMA and all that lot.
N.B I'm not an expert, if you couldn't tell
From Wikipedia: neutron stars (pulsars are typically spinning neutron stars that emit alot of EM radiation in regular bursts) are about 10-20km in diameter with masses of between 1.5 and 3 times the mass of Sol (approx. 2,000,000,000,000,000,000,000,000,000,000kg).
By that, the densisy is probably about eight LoC's per hogshead furlong.
Sigh, replying to an AC troll. Ho hum.
"And then your software can be just as slow and tedious to use as the real items!"
Why? Did I say that?
It's quicker to flick through a photo album than it is to view each image individually by double clicking it. Want a thumbnail view? Click/speak the "thumbnail" button and the album will re-arrange itself into a 10 foot by 3 foot montage, or starburst in front of you, or a stack of categorised photos, or whatever.
My point was that as human beings, we're more used to accessing information (words, pictures) on a 2D plane (paper) and yet interacting with in in a 3D environment (the real world). All the 3D UI's I've seen so far just seem to use the 3D for extra eye candy; there's no real difference to keeping the windows in squished 3D trapezoids at the side of the screen as there is to keeping them minimised/maximised/stacked/tiled or whatever, or using a pager with virtual desktops.
...that 3D UI's will work until we have a way of interacting with them in 3D.
Pretty much all of the 3D UI's I've seen are still presented to the user on 2D monitor, and interacted with by way of a 2D mouse (or keyboard). I don't really think 3D will offer any advantage until we can look at and interact with 3D objects a la Metaverse/SnowCrash. Such as flicking through a photo album, juggling your files up and down and all over the place, pushing/pulling your work in front of you.
But maybe I just can't envisage a 3D world without an intuitive way of interacting with in one less dimension. Or maybe I've read too much sci-fi...
Nice try, AC. Save it for Queen Doppelpopolis.
Jeepers. Please, no-one tell them about graphics card memory, which keeps a pixel-perfect copy of *the entire movie* in realtime...! I still want to be able to finish Doom3.