The only advantage (to Apple) would be that it would require QuickTime to be installed on every computer that used Fairplay. This may or may not be advantageous to Apple, depending on just how little money they're making on the iTunes Music Store.
The obvious disadvantages would be potentially fewer iPod sales and losing more money if indeed ITMS is a loss-leader.
It's probably a good option in the future if iPod sales start cracking. I wouldn't be surprised if they're looking into it.
Previously, the nVidia flash ROM updates were only available as OS 9 installers.
I would surmise that these are the same ROM updates as the old OS 9 ones, except they figured out how to port nvflash to OS X (by loading a KEXT hook, apparently).
They haven't increased the intelligence of the installer, as I had problems with my GF2MX working with my Sawtooth, 10.3, and >256MB of RAM. The new OS X installer said I didn't need it, as did the old OS 9 one. But, this time I extracted the appropriate ROM from the archive (thanks to the magic of packages), flashed my card on a PC and fixed the problems.
Works in Windows, Linux, Solaris, Mac OS X, IRIX, HP-UX, and AIX.
Doesn't work in Mac OS 9. Use netatalk for a while and then consider upgrading your machines.:)
Re:OK, so who's got a GOOD book on this topic?
on
Linux Clustering
·
· Score: 5, Informative
I wrote up a paper for my employer a while ago about most of those topics. The sad truth is that a comprehensive guide is not available, and most of the solutions are proprietary. However, there are a few bright lights.
Load-balancing and failover are tough nuts. You can do some stupid things like Round Robin DNS or Rotary NAT, but to be actual balancing, you need a balancer box. You can either make your own (using proprietary software or the stuff above) or buy a piece of hardware to do the job for you. I've heard Cisco makes some good ones.
NAS units usually operate using CIFS, AFP, or NFS, all of which are pretty lame options for a modern cluster. SANs are pretty cool, but you need some big-ass hardware to support them. Personally, I'm working on an OpenAFS cluster, which is pretty easy if you look into the capabilites of the software. Coda is another option of which I'm not using because it doesn't play as well with Windows.
As for clustering MySQL: If you read the Slashdot interview log they had a couple days ago, you'd see that the setup here is a master writer that replicates to a couple of reader databases. This is about as effective as it gets with MySQL. If you need higher power, I've read that commercial versions of Postgres support clustering/synchronization. More powerful than that and you're into Oracle territory.
Maybe rolling OSS out cheaply works in a sandbox where you can trust everything. Somehow I doubt that's good enough for the CIA.
While development costs could be curtailed with OSS, it is unlikely that they will be eliminated. Who writes software with the CIA in mind? Modifications to make OSS software work within the CIA model will be required, not to mention the granular security features that would have to be added. IIRC, it's either A or B grade trusted systems where they have to prevent communication via morse code using on/off file locking. Those are the kind of things I don't think anyone except the CIA/NSA/whoever thinks about when designing software.
This doesn't eliminate the testing cost either. While thorough testing doesn't seem like a big deal with many OSS projects, many corporations and (obviously) the CIA require that an application be tested in a standard fashion before being deployed. Find a bug? That's more man-hours to fix it right there.
Anyway, my point is that even if the software is free and open source, it's still probably going to cost a significant amount of money for the CIA to use it. Perhaps more, since sometimes it's harder to beat an application into doing what you want it to than to simply start it from scratch. Many OSS projects have realized this.
The GPL sections they quote only cover code that is distributed by persons other than the copyright owner. Since SCO "owns" the code, the sections do not apply if they distribute it.
This trial just needs to get started so IBM can thrash SCO into bankruptcy.
According to Apple, you can put the music on an unlimited number of iPods. I tried this. I dumped the music on my friend's iPod. It played just fine.
Now, I can only draw one conclusion from this: The iPod decrypts the AAC file without using your Apple ID. So, if somebody hacks the new iPod firmware update & gets the key out of it, anybody can play AAC files. How do I know this? Well, we tried to play the AAC file off of my friend's iBook in iTunes. Didn't work. The file was *still* encrypted, yet only iTunes cared about it.
To me, this has to be some ridiculous fast one Apple pulled on the RIAA Labels. I mean, you can still share music. You just have to use an iPod to do it.
* Twenty of your friends come over * You copy all your purchased music from your computer to their iPod * They leave and enjoy your music on the iPod without your presence
I think you just nailed why Apple doesn't support Vorbis.
There are no patents anybody _knows_ of. Who wants to be a big corporate target to begin with and then paint a sign on their back saying "we're making millions on stuff that you might own a patent to."
With AAC, they at least know a legal team spent countless man-hours checking it out. Well, I'm guessing on that part, but it *is* Sony...
Since the AMD guy went out of his way to *NOT* claim the Opteron's method was, in fact, "Dual DDR", what is your problem?
Considering that you can only grab one hunk of memory to use at one time in an operation, Dual DDR and Opteron 128 whatever do the exact same thing: grab twice as much data as needed.
Difference is thus:
Dual DDR grabs the *next requested memory address* AFTER the current operation completes. (the way the nForce chipset works)
Opteron grabs the next contiguous memory address *at the same time.* I would guess the rest gets shoved into the L1. (based on the fact that AMD guy is asserting that the extra bits on the bus actually transfer data from memory)
So, in instances where the memory you want comes in monster contiguous hunks (paging, multimedia, whatever) Opteron guarantees a 100% hit rate on the L1 cache. Even if there are only a couple bunched together, this helps.
In instances where single pieces of data are accessed in a scattered fashon in memory, Dual DDR eliminates a couple wait states. This doesn't seem as fast as the former.
Does anybody have any statistics about memory accesses that indicate a majority of them request the next contiguous piece of memory?
Evidence:
The iPod is available for Windows.
MP3s play on any computer.
OS X promotes open standards.
OS intentionally provides Windows and Linux interoperability.
Apple seems to has no interest in things that *lock* a user in. Sure, they have things that entice users to stay, like iTunes, but there's no lock-in there. I see the mentality being "well, you _could_ use something else, but why? Our stuff is SWEET!"
Will it even stand up with Mono around? Even if it is Microsoft's idea to begin with, they didn't apply for the patent until after Mono showed up.
I'm not sure it will fly. US Code title 35 Sec. 102 says something like
"A person shall be entitled to a patent unless the invention was patented or described in a printed publication in this or a foreign country or in public use or on sale in this country, more than one year prior to the date of the application for patent in the United States"
Mono has been around since July 2001, but since it's half-done, does that count?
IIRC, sometimes a company of sorts is formed for movies and its stock traded on the exchanges in order to raise money for a movie. Not sure how common the practice is, but I know they did this for Spider-Man.
Yeah, I know that no direct bearing on what you were talking about. Just a tidbit.
What X are you using? I can get X up (on my k6-500) in 3 seconds after typing xinit. Granted, I only have a xterm open with no window manager, but that is X. Now, perhaps you're talking about KDE or GNOME startup. That's where bloatage is.
Now, obscure I'll give you for X. 4.x improves upon that, but it's got a ways to go.
Right. IBM is a patent company. They just sold the unit that manufactures the newfangled stuff that comes out of the labs. They kept the labs. If you can't sell drives, at least you can still sell technology to other manufacturers.
Actually, I own a IIIe and have used a m105 for a couple months to test it out. Personally, I love Palm devices over CE, but I am pissed off that they're shooting themselves in the foot.
Battery life is important for the AAA users and people who don't like to have to recharge all the time. My IIIe lasts for a MONTH under good use. The m105 about the same.
Palm is a hardware company too. I expected them to roll out more IR->adapters->cell phone devices. Something that would integrate with the existing network other than building stupid limited-coverage limited-service networks.
Obviously you don't have cradle drain problems because you're using a rechargeable battery, which charges faster than it drains. While this problem may not exist in the pure USB versions of Palm devices, every serial device, including the rechargeables, experience this problem which has been known about since the original Palm Pilot, I belive. There are plenty sites out on the internet that tell you how to hack your cradle into not draining. It's a simple thing, *if* you're any good at electronics. The fact that Palm didn't fix this problem in the 'm' series just displays a lack of respect to their consumers
SD is just like Magic Gate Memory Sticks and Palladium. Sure they have open versions, but the potential exists to do great evil. Palm users are tech savvy. They can spot a trap a mile away. Palm should have gone with a more open format.
I made the comment about the 130 because of its cost. If they call $250 "affordable", then what's a couple more bucks to throw in flash.
I like Palm, I want them to do better, but they're not listening to the people who buy their products.
What exactly can you do with USB? Other than improve your HotSync time, what feature does USB on the Zire offer? Plug-in ethernet or 802.11b? I don't see those products anywhere. Personally, I wouldn't put USB support as a major plus to this device. Rechargeable battery yes.
The introduction of the M series for Palm was the beginning of the end for them.
* No compelling new features, minus the to-do/calendar view (which doesn't display alarms or notes)
* Switchable faceplates? Surely a sign that they've run out of *really* good ideas.
* Color? What about battery life? I thought Palm had proven color wasn't a must-have feature.
* Still-broken cradle drain problem.
* Wireless solution for the masses still not solved.
* SD/MMC Card that the technophiles already despised ("Secure" digital? We all know how much we like hardware that says that)
* m130: $250 un-upgradeable hardware. What a plan!
Palm just lost it. They had a dreat idea, ran with it, and then fumbled. Now CE is slowly replacing them. Zire just reaffirms that they still don't get it anymore.
Hey! Do you think you could tell me what you had to do with Samba and CUPS? I just finished compiling/beating Samba 2.2.5 (because the provided 2.2.Suck wouldn't read my 2.2.5 smb.conf) and I had to turn off CUPS support in order to get it to compile.
Excuse me, I should have said *Seagate* SCSI drives.:) I'm totally biased towards these guys, especially after they went private. In terms of mechanical reliablilty, using better but more expensive materials/processes over cheaper/not-as-good materials/processes. Hence a better drive. The way I see it, is the Seagate SCSI drives are the BMWs, and the IDE drives are the Toyota/Hondas. I have quite a few of both and I can tell with the tradeoffs in speed, noise, heat, etc.
That's pretty much exactly how it works with Seagate. Actually, in the case of the technology in Seagate's X15, there was a mechanism C. The orgininal X15 was a heatmonger and mostly unuseable without a dedicated drive fan (I have one) or other direct cooling system. So they release the X15 LP to fix those problems (and add some disk-to-controller speed). Now we have the Barracuda ATA V.
You're also totally right about how to make money off of this idea. Margins in the SCSI market recoup any lost market in the IDE arena. I didn't track it carefully, but I think the desire to keep the existing business model versus a "New Economy" pump-profits-till-you-die model is why Seagate went private.
Slashdot Mirror
Friend had the same problem with his 20GB 2G. Tried to fix it but the pads broke off the board.
I've got a 3G, no problems with it yet...
The only advantage (to Apple) would be that it would require QuickTime to be installed on every computer that used Fairplay. This may or may not be advantageous to Apple, depending on just how little money they're making on the iTunes Music Store.
The obvious disadvantages would be potentially fewer iPod sales and losing more money if indeed ITMS is a loss-leader.
It's probably a good option in the future if iPod sales start cracking. I wouldn't be surprised if they're looking into it.
Previously, the nVidia flash ROM updates were only available as OS 9 installers.
I would surmise that these are the same ROM updates as the old OS 9 ones, except they figured out how to port nvflash to OS X (by loading a KEXT hook, apparently).
They haven't increased the intelligence of the installer, as I had problems with my GF2MX working with my Sawtooth, 10.3, and >256MB of RAM. The new OS X installer said I didn't need it, as did the old OS 9 one. But, this time I extracted the appropriate ROM from the archive (thanks to the magic of packages), flashed my card on a PC and fixed the problems.
- Sync-on-green
- horizontal refresh of 75KHz+ (which typically means your vertical is going to be 75Hz+ too)
Had to go out and buy a new monitor cause my 1998 PGS couldn't handle it.Listen to Russ and try again
previous trolltech/canopy group discussions
Ah, so is the method of the flylady/a?
Works in Windows, Linux, Solaris, Mac OS X, IRIX, HP-UX, and AIX.
:)
Doesn't work in Mac OS 9. Use netatalk for a while and then consider upgrading your machines.
I wrote up a paper for my employer a while ago about most of those topics. The sad truth is that a comprehensive guide is not available, and most of the solutions are proprietary. However, there are a few bright lights.
Eddie: Load Balancing Software
http://eddie.sourceforge.net/
Linux Virtual Server Project: Clustering Tools
http://www.linuxvirtualserver.org/
OpenAFS: Efficient Distributed Storage
http://www.openafs.org/
Load-balancing and failover are tough nuts. You can do some stupid things like Round Robin DNS or Rotary NAT, but to be actual balancing, you need a balancer box. You can either make your own (using proprietary software or the stuff above) or buy a piece of hardware to do the job for you. I've heard Cisco makes some good ones.
NAS units usually operate using CIFS, AFP, or NFS, all of which are pretty lame options for a modern cluster. SANs are pretty cool, but you need some big-ass hardware to support them. Personally, I'm working on an OpenAFS cluster, which is pretty easy if you look into the capabilites of the software. Coda is another option of which I'm not using because it doesn't play as well with Windows.
As for clustering MySQL: If you read the Slashdot interview log they had a couple days ago, you'd see that the setup here is a master writer that replicates to a couple of reader databases. This is about as effective as it gets with MySQL. If you need higher power, I've read that commercial versions of Postgres support clustering/synchronization. More powerful than that and you're into Oracle territory.
Maybe rolling OSS out cheaply works in a sandbox where you can trust everything. Somehow I doubt that's good enough for the CIA.
While development costs could be curtailed with OSS, it is unlikely that they will be eliminated. Who writes software with the CIA in mind? Modifications to make OSS software work within the CIA model will be required, not to mention the granular security features that would have to be added. IIRC, it's either A or B grade trusted systems where they have to prevent communication via morse code using on/off file locking. Those are the kind of things I don't think anyone except the CIA/NSA/whoever thinks about when designing software.
This doesn't eliminate the testing cost either. While thorough testing doesn't seem like a big deal with many OSS projects, many corporations and (obviously) the CIA require that an application be tested in a standard fashion before being deployed. Find a bug? That's more man-hours to fix it right there.
Anyway, my point is that even if the software is free and open source, it's still probably going to cost a significant amount of money for the CIA to use it. Perhaps more, since sometimes it's harder to beat an application into doing what you want it to than to simply start it from scratch. Many OSS projects have realized this.
The GPL sections they quote only cover code that is distributed by persons other than the copyright owner. Since SCO "owns" the code, the sections do not apply if they distribute it.
This trial just needs to get started so IBM can thrash SCO into bankruptcy.
According to Apple, you can put the music on an unlimited number of iPods. I tried this. I dumped the music on my friend's iPod. It played just fine.
Now, I can only draw one conclusion from this: The iPod decrypts the AAC file without using your Apple ID. So, if somebody hacks the new iPod firmware update & gets the key out of it, anybody can play AAC files. How do I know this? Well, we tried to play the AAC file off of my friend's iBook in iTunes. Didn't work. The file was *still* encrypted, yet only iTunes cared about it.
To me, this has to be some ridiculous fast one Apple pulled on the RIAA Labels. I mean, you can still share music. You just have to use an iPod to do it.
* Twenty of your friends come over
* You copy all your purchased music from your computer to their iPod
* They leave and enjoy your music on the iPod without your presence
I think you just nailed why Apple doesn't support Vorbis.
...
There are no patents anybody _knows_ of. Who wants to be a big corporate target to begin with and then paint a sign on their back saying "we're making millions on stuff that you might own a patent to."
With AAC, they at least know a legal team spent countless man-hours checking it out. Well, I'm guessing on that part, but it *is* Sony
Since the AMD guy went out of his way to *NOT* claim the Opteron's method was, in fact, "Dual DDR", what is your problem?
Considering that you can only grab one hunk of memory to use at one time in an operation, Dual DDR and Opteron 128 whatever do the exact same thing: grab twice as much data as needed.
Difference is thus:
Dual DDR grabs the *next requested memory address* AFTER the current operation completes. (the way the nForce chipset works)
Opteron grabs the next contiguous memory address *at the same time.* I would guess the rest gets shoved into the L1. (based on the fact that AMD guy is asserting that the extra bits on the bus actually transfer data from memory)
So, in instances where the memory you want comes in monster contiguous hunks (paging, multimedia, whatever) Opteron guarantees a 100% hit rate on the L1 cache. Even if there are only a couple bunched together, this helps.
In instances where single pieces of data are accessed in a scattered fashon in memory, Dual DDR eliminates a couple wait states. This doesn't seem as fast as the former.
Does anybody have any statistics about memory accesses that indicate a majority of them request the next contiguous piece of memory?
Unlikeley.
;-)
Evidence:
The iPod is available for Windows.
MP3s play on any computer.
OS X promotes open standards.
OS intentionally provides Windows and Linux interoperability.
Apple seems to has no interest in things that *lock* a user in. Sure, they have things that entice users to stay, like iTunes, but there's no lock-in there. I see the mentality being "well, you _could_ use something else, but why? Our stuff is SWEET!"
Will it even stand up with Mono around? Even if it is Microsoft's idea to begin with, they didn't apply for the patent until after Mono showed up.
I'm not sure it will fly. US Code title 35 Sec. 102 says something like
"A person shall be entitled to a patent unless the invention was patented or described in a printed publication in this or a foreign country or in public use or on sale in this country, more than one year prior to the date of the application for patent in the United States"
Mono has been around since July 2001, but since it's half-done, does that count?
IIRC, sometimes a company of sorts is formed for movies and its stock traded on the exchanges in order to raise money for a movie. Not sure how common the practice is, but I know they did this for Spider-Man.
Yeah, I know that no direct bearing on what you were talking about. Just a tidbit.
Glad somebody modded you down. That newsforge article doesn't deny anything except Mandrake's death.
They implicitly included bankruptcy as an option in their statement. There's no honesty problems here; they just didn't know what they wanted to do.
Bankruptcy protection doesn't mean the whole operation shuts down.
What X are you using? I can get X up (on my k6-500) in 3 seconds after typing xinit. Granted, I only have a xterm open with no window manager, but that is X. Now, perhaps you're talking about KDE or GNOME startup. That's where bloatage is.
Now, obscure I'll give you for X. 4.x improves upon that, but it's got a ways to go.
Right. IBM is a patent company. They just sold the unit that manufactures the newfangled stuff that comes out of the labs. They kept the labs. If you can't sell drives, at least you can still sell technology to other manufacturers.
Actually, I own a IIIe and have used a m105 for a couple months to test it out. Personally, I love Palm devices over CE, but I am pissed off that they're shooting themselves in the foot.
Battery life is important for the AAA users and people who don't like to have to recharge all the time. My IIIe lasts for a MONTH under good use. The m105 about the same.
Palm is a hardware company too. I expected them to roll out more IR->adapters->cell phone devices. Something that would integrate with the existing network other than building stupid limited-coverage limited-service networks.
Obviously you don't have cradle drain problems because you're using a rechargeable battery, which charges faster than it drains. While this problem may not exist in the pure USB versions of Palm devices, every serial device, including the rechargeables, experience this problem which has been known about since the original Palm Pilot, I belive. There are plenty sites out on the internet that tell you how to hack your cradle into not draining. It's a simple thing, *if* you're any good at electronics. The fact that Palm didn't fix this problem in the 'm' series just displays a lack of respect to their consumers
SD is just like Magic Gate Memory Sticks and Palladium. Sure they have open versions, but the potential exists to do great evil. Palm users are tech savvy. They can spot a trap a mile away. Palm should have gone with a more open format.
I made the comment about the 130 because of its cost. If they call $250 "affordable", then what's a couple more bucks to throw in flash.
I like Palm, I want them to do better, but they're not listening to the people who buy their products.
What exactly can you do with USB? Other than improve your HotSync time, what feature does USB on the Zire offer? Plug-in ethernet or 802.11b? I don't see those products anywhere. Personally, I wouldn't put USB support as a major plus to this device. Rechargeable battery yes.
The introduction of the M series for Palm was the beginning of the end for them.
* No compelling new features, minus the to-do/calendar view (which doesn't display alarms or notes)
* Switchable faceplates? Surely a sign that they've run out of *really* good ideas.
* Color? What about battery life? I thought Palm had proven color wasn't a must-have feature.
* Still-broken cradle drain problem.
* Wireless solution for the masses still not solved.
* SD/MMC Card that the technophiles already despised ("Secure" digital? We all know how much we like hardware that says that)
* m130: $250 un-upgradeable hardware. What a plan!
Palm just lost it. They had a dreat idea, ran with it, and then fumbled. Now CE is slowly replacing them. Zire just reaffirms that they still don't get it anymore.
Hey! Do you think you could tell me what you had to do with Samba and CUPS? I just finished compiling/beating Samba 2.2.5 (because the provided 2.2.Suck wouldn't read my 2.2.5 smb.conf) and I had to turn off CUPS support in order to get it to compile.
Excuse me, I should have said *Seagate* SCSI drives. :) I'm totally biased towards these guys, especially after they went private. In terms of mechanical reliablilty, using better but more expensive materials/processes over cheaper/not-as-good materials/processes. Hence a better drive. The way I see it, is the Seagate SCSI drives are the BMWs, and the IDE drives are the Toyota/Hondas. I have quite a few of both and I can tell with the tradeoffs in speed, noise, heat, etc.
That's pretty much exactly how it works with Seagate. Actually, in the case of the technology in Seagate's X15, there was a mechanism C. The orgininal X15 was a heatmonger and mostly unuseable without a dedicated drive fan (I have one) or other direct cooling system. So they release the X15 LP to fix those problems (and add some disk-to-controller speed). Now we have the Barracuda ATA V.
You're also totally right about how to make money off of this idea. Margins in the SCSI market recoup any lost market in the IDE arena. I didn't track it carefully, but I think the desire to keep the existing business model versus a "New Economy" pump-profits-till-you-die model is why Seagate went private.