Slashdot Mirror
Microsoft FAT Licensing Plan - No Big Deal?
prostoalex writes "InternetNews.com describes the reaction to Microsoft's decision on FAT licensing. It doesn't look like the company is expecting to make any significant money out of licenses (there's also a cap of $250K, so none of the big guys will have to pay millions to Microsoft). It also doesn't look like Linux companies are stressed over this decision. "We are only accessing FAT32 file systems, not using them. This licensing program is of little interest to SuSE", a Novell/SuSE spokesperson said."
The point has to be securing the patent. If they charge people to license something, they're establishing the fact that they do "own" this technology.
This isn't about making money, and it isn't about protecting a patent. This is simply an exercise for the benefit of the courts: Microsoft will be holding FAT (along with lots of other things) up as examples of how they're making their standards available to the competition.
Tarsnap: Online backups for the truly paranoid
It seems more likely that Microsoft is hoping this will bring 'complete' adoption of NTFS and WinFS, not that they were ever looking to make gobs of money from the licensing in the first place.
What are the current costs associated with licensing NTFS? Any information regarding the cost of licensing WinFS when Longhorn strikes?
If they got a patent when they created FAT, it would be near the end of its life (FAT32 might have a few years to go). But other FAT implementations have been around since before the patent was applied for OR microsoft has simply ignored them. If you don't defend your patent with due diligance, (say, wait 15 years to sue someone) you can run into legal difficulties.
And since FAT implementations are nearly ubiquitous today, trying to sue them all would be an exercise in wasting time and the patent would expire before you've even begun litigation.
It seems pretty pointless to try and enforce the patent at this point. The article addresses this to an extent in m-soft's IP policy ideals...
-
FAT is positively not good for flash drives because these systems will begin to fail have a set number of writes to a specific area of the memory.
FAT has no methodology for spreading writes around the system, in fact it will write to the logical begining of the filespace most frequently, with the number of writes dropping as you travel further into the system. This leads to a flash drive to premature failure, thousands of writes to the begining of the disk (like say, the FAT table) making the entire memory space unusable (for FAT, anyways) while the end has been largely untouched.
Other, GPLed systems are much better for flash, such as JFFS. The only reason to use FAT is if you're interfacing with a windows box and are too lazy to provide translation.
Any sufficiently advanced technology is indistinguishable from a rigged demo.
They could always centralize kernel development in another country where the FAT patents are not recognized by law.
.
If you prefer, look at it this way - perhaps MS realise how much of a mess they've made of implementing certain things in the past, and so recognise that if they can mess it up, so can anyone else. That may be part of the reason why - acknowledging that they're nothing special when it comes to writing software.
Hell, recently there was a problem with LG not implementing the "cache flush" instruction on some of their CD drives, instead using it for "update firmware". That's a monumentally stupid thing to do when implementing an accepted standard. Given that that happened, don't you think it possible that some large manufacturer could mess up their FAT implementation?
It's official. Most of you are morons.
Also, aren't there other LFN-enabled filesystems that could be used (and read/written to by various interpretive utils, if it needs to interact with a VFAT system) if one declines to pay their license fee? Which itself struck me as reasonable in terms of business expenses, especially with the cap. If you're selling a few million flash memory cards, and *everyone* making such cards has to add the same few cents apiece in extra cost for the license, well, that's not going to impact anyone's competitiveness even in today's tight-margined market.
~REZ~ #43301. Who'd fake being me anyway?
You may go the other way: use the ext2 or anything best suited to flash and supply the opensource Windows driver for it.
The Free file system best suited to write-few-read-many memory would probably be JFFS, but Google doesn't know about any efforts to implement a generic JFFS driver on Windows.