I have read up on this and I stand corrected -- I saw a 20,000,000 number on the top of a machine I had in pieces a week ago and must have misread it. I realise the extra space would be wasted but I just figured that MS had bought larger drives because no-one makes 8gig ones anymore, making it cheaper to waste 60% of the disk than to hold a production line open just for the Xbox.
The disk I was looking at had instruction on the sticker on the top of it that walked you through running fdisk and putting it into a Windows PC, amusingly.
The 10Gb disk that came out of my own (v1.0) Xbox is marked up with "Quality Control: Failed" stickers. Hopefully, there were bad sectors in the tail 2Gb of the disk, although you have to wonder...
The Xbox has already been through a major hardware revision (v1.0 to 1.1) which offered many of the same improvements: it eliminated a 40mm fan from the GPU, which quietened the unit considerably, and replaced the bloody awful Thomson DVD drive with a shiny Samsung that can, for the first time in an Xbox, play CD-Rs (note that some units do still come with the middle-of-the-road Philips drive though). Oh, and it has a 20gig hard disk instead of the 8 in the original Xbox, too.
Funny how Sony made a song and a dance about it and Microsoft didn't, though. Not like MS to miss a marketing trick.
In April 2002, M. J. Dunwoody produced a five-page paper that purports to prove the conjecture. However, according to the rules of the Clay Institute, the paper must survive two years of academic scrutiny before the prize can be collected.
So, why the excitment about this later Perelman paper? Has the Dunwoody paper been debunked?
Go and read Carmack's.plan file, archived here; he's already said the Xbox can run Doom3 equivalently to a PC. Doom3 only has no support for high level shaders used in DX9 class hardware.
I've seen a lot of responses in this thread from people saying "bugger CPU time, just spend a few hundred dollars on more/faster CPU/RAM. Dev time is expensive!"
Well, yes it is, but it's not always that simple.
I have a Java app here I'm performance tuning for my PhD that allocates frequency hop sets to mobile phone networks. Running on a 2.2Ghz Athlon wih 512Mb RAM for a 15-transmitter test case takes an hour, it scales exponentially with transmitter size, and I want to address a 458 transmitter case. It's about 10^500 calculations, or it was before I started improving the algorithm. Even so, it's still going to be billions of iterations through the inner loop. Even a 1% speedup is hours off my runtime and that's a big thing for me.
So when you dismiss all performance tuning with a wave of your hand, remember us poor beleagured scientists. We actually need all this stuff.
See Transgaming's WineX, which works exactly like this. Of course, the detailed howtos on Franksworld about how to compile WineX out of the (open access) CVS does make me less likely to subscribe.
But doesn't this just encourage developers to ship slightly broken source? Say, all the Makefiles deleted or something?
I looked into the legal aspects back then, when I was busy shouting at Kodak. There are a few interlocking issues.
Firstly, a contract is only formed if one party makes an offer to tender and the other party accepts this offer. For most web retailers, the most you get back is a "confirmation of order" email that has a dubious legal standing. Kodak's email, however, states in very clear terms that it is a receipt, suggesting a contract has been made.
Secondly, if one party can be shown to have acted in bad faith, the contract can be annulled. This occured in the case of people buying a £299 TV for £2.99 from Argos: the price was ruled to be clearly in error, the buyers must have known it was an error, so they were acting in bad faith. In Kodak's case it was not so clear cut; the price was low, but it was marked as being a special offer, so not utterly incredible.
In the end, Kodak decided the legal waters were muddy enough that it didn't fancy trying them out, so it gave in and sold the cameras at the advertised price. But that's not to say it would have lost, merely that the bad PR of fighting it and the chance of losing would be too costly.
In any event, if you were quick enough to get one of these PDAs ordered, don't get too excited: I'll be shocked if you get one. Amazon wouldn't lose if this went to court.
This sort of frequency hopping happens all the time in, for example, GSM and Bluetooth. It doesn't make the interference go away.
The "few errors" you refer to are still interference. With a sensible frequency hopping pattern, the interference will spread out around users and be evenly spread in time, hopefully to the point where error correcting codes can catch it and compensate. But add more users and the error rate will pile up until your network falls apart, just as with non-hopping.
This effect is called "interference diversity" and is well studied in the literature.
Additionally, your throwaway line about "ask anyone who's signal you can see to choose a different color or time division on that particular color" would be enormously, insanely complex to implement. The amount of traffic necessary to keep this sort of scheme working would dwarf the useful traffic the network would handle; plus, this whilst it would improve things for a single user, it would likely make the next user over worse. It would not lead to a better network overall.
[Disclaimer: frequency hopping is my PhD thesis topic]
How can this well designed receiver tell two signals apart? Barring directional antennae, which are impractical beyond belief, if I have a stream of 1.1Mhz photons coming from over there and a stream of 1.1Mhz photons coming from over here, how does any receiver tell the here photons from the there photons?
And Shannon's law limits bandwidth to a known amount.
Spread spectrum does not -- it cannot -- reduce interference. What it does is average interference across multiple users in the network and across time. This is known as interference diversity (my PhD is on frequency hopping).
To make sense of this, you have to remember that error correcting codes used in mobile communications can cope with a low bit error rate without data loss but will fall apart at higher rates. So, low errors all the time are OK, but errors that peak at moments in time are bad because that will lose the signal altogether.
I'm currently 18 months through a PhD revolving around the assignment of frequencies in a frequency hopping spread spectrum network (more details here) so I know a bit about this stuff. And that article is not fantastically insightful.
Interference, as it says, is not a law of nature. It's what happens when you are trying to listen to, say, a 1.1Mhz signal coming from over there and someone over here is also transmitting on 1.1Mhz. How can the radio receiver tell the difference between those signals? As the article hints, it's an engineering issue; but it's a non-trivial one. Radio engineers all over the world will not read this article and rejoice. Reclassifying the problem in some bizarre colour analogy has not magically solved it.
Now as for the politics of spectrum allocation and the potential improvements of a free spectrum policy: now that's a more interesting issue, but one the article doesn't address in any but the most superficial of ways.
Bah, I say to it.
My personal gripe with the Gnome file selector
on
Has GNOME Become LAME?
·
· Score: 2, Interesting
Now, don't get me wrong. I love Gnome, have used it for years (although I am typing this on OSX), and never got on with KDE. I think the article is a big bunch of trolling hooey. But...
Many moons ago, I ran the Ximian Gnome packages, probably Gnome 1.0 or possibly 1.2. The File Selector back then had a fourth button at the top of it, next to new folder / delete file / rename file buttons. This button took you to ~ when you clicked it. Where has this bloody button gone?
I remember reading that it was a Ximian patch to the file selector, and presumably it never made it into the upstream sources. When I left Ximian (the day I switched to Debian testing), I lost that button forever and I still miss it. It seems like such an obvious thing to me. Bah! Humbug!
He's not whining, he's just stating his position. He can pay to write closed source Qt apps or he can write closed source GTK apps for free. He doesn't prefer Qt on any technical reasons so he chooses GTK. It's a business decision. Trolltech knew that people would do that when they chose Qt's licensing and that was a business decision too. I don't see any hypocrisy here.
WIthout responding to your flame about the quality of software on Linux, I'm just going to point out that the article said it was a fileserver. Mount it over SMB or NFS to the OS of your choice and work in the app of your choice. Happy troll?
Hey, Wagoo, is that you? Given that your email address appears to have a well-known IRC network in it I assume so. It's me, g1yndwr from the hahs. Hello. *waves*
Firstly, ntl support NAT setups. 1Gb a day for me isn't bad. 1Gb a day between me and my three housemates, with a bit of streaming audio, a bit of movie trailer watching, a bit of game playing, a little bit of Xbox Live (3Mb/min, I am told, that one), keeping four installs of Windows, two of Mandrake and two of Debian up to date... now that looks rather more intrusive. Even keeping a single copy of Win2k in patches can consume gigabytes a month!
I pay for a 1Mb cable modem connection that can saturate my 1Gb limit in under three hours. That doesn't sound like the "unlimited internet" I was sold.
Secondly, this is almost certainly the thin end of the wedge, as many other people with capacity limited broadband around the world have discovered. 1Gb/day now, 750Mb tomorrow, 250Mb next week. After all, no matter how many users you kick off, 80% of bandwidth will always be used by 20% of the users because of the shape of the bell curve. And those 20% of users are always in a minority, and that 80% of bandwidth sure is expensive.
I've been staying on top of this right over the weekend (and had a/. story about it rejected 36 hours ago, grrrrr), so for those new to it, some links:
Massive thread on nthellworld.com, a offical ntl gripe site.
Basically, ntl are somewhat losing their nerve. I've exchanged emails with the MD of their home products range who claims to have only found out about this key strategic business decision on Saturday morning; he's either lying or incompetent, I suggest. The biggest gripe amongst the sane posters (barring all the "I pay for 24/7 and I'm going to damn well get it" breast-beating") is that the 128bps, 600kbps and 1024kps services all have the same download limit, making you wonder why you pay for the higher speed service.
It should also be pointed out that, unlike many other ISP's schemes, NTL offer no FTP mirror service with "free" bandwidth and recently started dropping alt.binaries groups from their newsspool, which is in any event so slow as to be unusable. So for big alt.binaries downloads or Linux ISOs, for example, customers are forced to external sites, pushing up ntl's bandwidth.
The biggest fear is that this is the thin end of the wedge. In the last two weeks, ntl have dropped a few warez newsgroups and introduced a fairly generous cap that won't inconvenience too many people. That's all well and good, but many think it won't stop there; once you get the caps in place and the groups erased, you can squeeze them down and down. ntl is desperately short of cash, newly emerged from Chapter 11 protection, and this would appear to be a beancounter-led efficiency drive that is turning into a PR nightmare.
I was part of a similar revolt over a no-servers line in the AUP a few years back (more info) and ntl backed down and clarified their position with a set of clear-cut and sensible rules. Let's hope that happens again.
Actually, I replaced my Win2k (I hadn't discovered mplayer when I built it) based DVD/DivX/mp3 box with a Xbox.
Reasons? The Xbox was a lot smaller, it had an IR remote (this was a big thing), it had good codec support out of the box (XBMP is based on mplayer so plays pretty much anything). It's noisy but not much more noisy. I recycled the PC to a family member and couldn't be happier with the result. The biggest plus is the hard drive I have chock-full with SNES, Genesis and MAME ROMs, all complete with a decent joypad to play them with.
I reckon you can now put together a chipped Xbox with a suitably big hard disk for about £300 or so, and at that price, I think it's a very compelling device.
Last time I looked, I got the impression you could only compile XBMP using a warezed copy of the Xbox SDK. This was going to be the case until the OpenSDK project reached some maturity. This is why the XBMP website only supplies source and I have to scavenge binaries from Usenet or IRC.
Assuming I was right in the first place, is this still the case? Be kind of a bummer if the only way to test your patch for this competition is via an illegally ripped MS SDK.
No they don't, the iBooks are $200 cheaper across the range. As I mentioned in my submitted Slashdot story. Which was rejected. Not that I'm bitter.
I'm convinced, anyway; the midrange iBook looks very compelling to me. All the OSS stuff I need plus Powerpoint without rebooting. I'll hopefully be ordering one later.
Slashdot Mirror
I have read up on this and I stand corrected -- I saw a 20,000,000 number on the top of a machine I had in pieces a week ago and must have misread it. I realise the extra space would be wasted but I just figured that MS had bought larger drives because no-one makes 8gig ones anymore, making it cheaper to waste 60% of the disk than to hold a production line open just for the Xbox.
The disk I was looking at had instruction on the sticker on the top of it that walked you through running fdisk and putting it into a Windows PC, amusingly.
The 10Gb disk that came out of my own (v1.0) Xbox is marked up with "Quality Control: Failed" stickers. Hopefully, there were bad sectors in the tail 2Gb of the disk, although you have to wonder...
Take it to pieces :o)
h acker.net/
Seriously, there's a way of telling from the serial numbers but I don't know it offhand. I've only ever ripped them to bits.
Some links for you:
http://www.xbox-scene.com/
http://www.xbox
Aha, here's what you're after:
http://www.xbox-scene.com/versions.php
The Xbox has already been through a major hardware revision (v1.0 to 1.1) which offered many of the same improvements: it eliminated a 40mm fan from the GPU, which quietened the unit considerably, and replaced the bloody awful Thomson DVD drive with a shiny Samsung that can, for the first time in an Xbox, play CD-Rs (note that some units do still come with the middle-of-the-road Philips drive though). Oh, and it has a 20gig hard disk instead of the 8 in the original Xbox, too.
Funny how Sony made a song and a dance about it and Microsoft didn't, though. Not like MS to miss a marketing trick.
I prefer to think of it as
public static void main (String[] args) {
doStuff();
}
The link to mathworld.wolfram.com from the post says:
So, why the excitment about this later Perelman paper? Has the Dunwoody paper been debunked?
So, Taco doesn't RTFA now.
Worst... news... post... EVARR!
Go and read Carmack's .plan file, archived here; he's already said the Xbox can run Doom3 equivalently to a PC. Doom3 only has no support for high level shaders used in DX9 class hardware.
I've seen a lot of responses in this thread from people saying "bugger CPU time, just spend a few hundred dollars on more/faster CPU/RAM. Dev time is expensive!"
Well, yes it is, but it's not always that simple.
I have a Java app here I'm performance tuning for my PhD that allocates frequency hop sets to mobile phone networks. Running on a 2.2Ghz Athlon wih 512Mb RAM for a 15-transmitter test case takes an hour, it scales exponentially with transmitter size, and I want to address a 458 transmitter case. It's about 10^500 calculations, or it was before I started improving the algorithm. Even so, it's still going to be billions of iterations through the inner loop. Even a 1% speedup is hours off my runtime and that's a big thing for me.
So when you dismiss all performance tuning with a wave of your hand, remember us poor beleagured scientists. We actually need all this stuff.
RTFA. It's a real language. It's bizarre and pointless and funny but this isn't an April Fool, it's a real, working language.
Admittedly, I had to download the binary to convince myself it wasn't an extremely elaborate hoax.
Now, the real question: who wants to write a whitespace VM in whitespace?
See Transgaming's WineX, which works exactly like this. Of course, the detailed howtos on Franksworld about how to compile WineX out of the (open access) CVS does make me less likely to subscribe.
But doesn't this just encourage developers to ship slightly broken source? Say, all the Makefiles deleted or something?
I was one of the lucky ones who got a Kodak DX3700 digital camera for £100, so these things do sometimes pan out.
I looked into the legal aspects back then, when I was busy shouting at Kodak. There are a few interlocking issues.
Firstly, a contract is only formed if one party makes an offer to tender and the other party accepts this offer. For most web retailers, the most you get back is a "confirmation of order" email that has a dubious legal standing. Kodak's email, however, states in very clear terms that it is a receipt, suggesting a contract has been made.
Secondly, if one party can be shown to have acted in bad faith, the contract can be annulled. This occured in the case of people buying a £299 TV for £2.99 from Argos: the price was ruled to be clearly in error, the buyers must have known it was an error, so they were acting in bad faith. In Kodak's case it was not so clear cut; the price was low, but it was marked as being a special offer, so not utterly incredible.
In the end, Kodak decided the legal waters were muddy enough that it didn't fancy trying them out, so it gave in and sold the cameras at the advertised price. But that's not to say it would have lost, merely that the bad PR of fighting it and the chance of losing would be too costly.
In any event, if you were quick enough to get one of these PDAs ordered, don't get too excited: I'll be shocked if you get one. Amazon wouldn't lose if this went to court.
(IANAL, of course)
This sort of frequency hopping happens all the time in, for example, GSM and Bluetooth. It doesn't make the interference go away.
The "few errors" you refer to are still interference. With a sensible frequency hopping pattern, the interference will spread out around users and be evenly spread in time, hopefully to the point where error correcting codes can catch it and compensate. But add more users and the error rate will pile up until your network falls apart, just as with non-hopping.
This effect is called "interference diversity" and is well studied in the literature.
Additionally, your throwaway line about "ask anyone who's signal you can see to choose a different color or time division on that particular color" would be enormously, insanely complex to implement. The amount of traffic necessary to keep this sort of scheme working would dwarf the useful traffic the network would handle; plus, this whilst it would improve things for a single user, it would likely make the next user over worse. It would not lead to a better network overall.
[Disclaimer: frequency hopping is my PhD thesis topic]
How can this well designed receiver tell two signals apart? Barring directional antennae, which are impractical beyond belief, if I have a stream of 1.1Mhz photons coming from over there and a stream of 1.1Mhz photons coming from over here, how does any receiver tell the here photons from the there photons?
And Shannon's law limits bandwidth to a known amount.
Spread spectrum does not -- it cannot -- reduce interference. What it does is average interference across multiple users in the network and across time. This is known as interference diversity (my PhD is on frequency hopping).
To make sense of this, you have to remember that error correcting codes used in mobile communications can cope with a low bit error rate without data loss but will fall apart at higher rates. So, low errors all the time are OK, but errors that peak at moments in time are bad because that will lose the signal altogether.
I'm currently 18 months through a PhD revolving around the assignment of frequencies in a frequency hopping spread spectrum network (more details here) so I know a bit about this stuff. And that article is not fantastically insightful.
Interference, as it says, is not a law of nature. It's what happens when you are trying to listen to, say, a 1.1Mhz signal coming from over there and someone over here is also transmitting on 1.1Mhz. How can the radio receiver tell the difference between those signals? As the article hints, it's an engineering issue; but it's a non-trivial one. Radio engineers all over the world will not read this article and rejoice. Reclassifying the problem in some bizarre colour analogy has not magically solved it.
Now as for the politics of spectrum allocation and the potential improvements of a free spectrum policy: now that's a more interesting issue, but one the article doesn't address in any but the most superficial of ways.
Bah, I say to it.
Now, don't get me wrong. I love Gnome, have used it for years (although I am typing this on OSX), and never got on with KDE. I think the article is a big bunch of trolling hooey. But... Many moons ago, I ran the Ximian Gnome packages, probably Gnome 1.0 or possibly 1.2. The File Selector back then had a fourth button at the top of it, next to new folder / delete file / rename file buttons. This button took you to ~ when you clicked it. Where has this bloody button gone? I remember reading that it was a Ximian patch to the file selector, and presumably it never made it into the upstream sources. When I left Ximian (the day I switched to Debian testing), I lost that button forever and I still miss it. It seems like such an obvious thing to me. Bah! Humbug!
He's not whining, he's just stating his position. He can pay to write closed source Qt apps or he can write closed source GTK apps for free. He doesn't prefer Qt on any technical reasons so he chooses GTK. It's a business decision. Trolltech knew that people would do that when they chose Qt's licensing and that was a business decision too. I don't see any hypocrisy here.
I love descriptive mountpoints. I have a /mnt/stuff too, but it's a measly 60 gig unfortunately.
WIthout responding to your flame about the quality of software on Linux, I'm just going to point out that the article said it was a fileserver. Mount it over SMB or NFS to the OS of your choice and work in the app of your choice. Happy troll?
Hey, Wagoo, is that you? Given that your email address appears to have a well-known IRC network in it I assume so. It's me, g1yndwr from the hahs. Hello. *waves*
Two things:
Firstly, ntl support NAT setups. 1Gb a day for me isn't bad. 1Gb a day between me and my three housemates, with a bit of streaming audio, a bit of movie trailer watching, a bit of game playing, a little bit of Xbox Live (3Mb/min, I am told, that one), keeping four installs of Windows, two of Mandrake and two of Debian up to date... now that looks rather more intrusive. Even keeping a single copy of Win2k in patches can consume gigabytes a month!
I pay for a 1Mb cable modem connection that can saturate my 1Gb limit in under three hours. That doesn't sound like the "unlimited internet" I was sold.
Secondly, this is almost certainly the thin end of the wedge, as many other people with capacity limited broadband around the world have discovered. 1Gb/day now, 750Mb tomorrow, 250Mb next week. After all, no matter how many users you kick off, 80% of bandwidth will always be used by 20% of the users because of the shape of the bell curve. And those 20% of users are always in a minority, and that 80% of bandwidth sure is expensive.
I've been staying on top of this right over the weekend (and had a /. story about it rejected 36 hours ago, grrrrr), so for those new to it, some links:
Massive thread on nthellworld.com, a offical ntl gripe site.
Complaint site
Basically, ntl are somewhat losing their nerve. I've exchanged emails with the MD of their home products range who claims to have only found out about this key strategic business decision on Saturday morning; he's either lying or incompetent, I suggest. The biggest gripe amongst the sane posters (barring all the "I pay for 24/7 and I'm going to damn well get it" breast-beating") is that the 128bps, 600kbps and 1024kps services all have the same download limit, making you wonder why you pay for the higher speed service.
It should also be pointed out that, unlike many other ISP's schemes, NTL offer no FTP mirror service with "free" bandwidth and recently started dropping alt.binaries groups from their newsspool, which is in any event so slow as to be unusable. So for big alt.binaries downloads or Linux ISOs, for example, customers are forced to external sites, pushing up ntl's bandwidth.
The biggest fear is that this is the thin end of the wedge. In the last two weeks, ntl have dropped a few warez newsgroups and introduced a fairly generous cap that won't inconvenience too many people. That's all well and good, but many think it won't stop there; once you get the caps in place and the groups erased, you can squeeze them down and down. ntl is desperately short of cash, newly emerged from Chapter 11 protection, and this would appear to be a beancounter-led efficiency drive that is turning into a PR nightmare.
I was part of a similar revolt over a no-servers line in the AUP a few years back (more info) and ntl backed down and clarified their position with a set of clear-cut and sensible rules. Let's hope that happens again.
Actually, I replaced my Win2k (I hadn't discovered mplayer when I built it) based DVD/DivX/mp3 box with a Xbox.
Reasons? The Xbox was a lot smaller, it had an IR remote (this was a big thing), it had good codec support out of the box (XBMP is based on mplayer so plays pretty much anything). It's noisy but not much more noisy. I recycled the PC to a family member and couldn't be happier with the result. The biggest plus is the hard drive I have chock-full with SNES, Genesis and MAME ROMs, all complete with a decent joypad to play them with.
I reckon you can now put together a chipped Xbox with a suitably big hard disk for about £300 or so, and at that price, I think it's a very compelling device.
Last time I looked, I got the impression you could only compile XBMP using a warezed copy of the Xbox SDK. This was going to be the case until the OpenSDK project reached some maturity. This is why the XBMP website only supplies source and I have to scavenge binaries from Usenet or IRC.
Assuming I was right in the first place, is this still the case? Be kind of a bummer if the only way to test your patch for this competition is via an illegally ripped MS SDK.
No they don't, the iBooks are $200 cheaper across the range. As I mentioned in my submitted Slashdot story. Which was rejected. Not that I'm bitter.
I'm convinced, anyway; the midrange iBook looks very compelling to me. All the OSS stuff I need plus Powerpoint without rebooting. I'll hopefully be ordering one later.