On Wed, Jan 30, 2008 at 06:23:09AM +0100, Tristan Gingold wrote: > On Tue, Jan 29, 2008 at 07:48:29PM +0100, Robert Millan wrote: > [...] > > AFAIK, there's no standard specifiing FAT, only a few implementations that > > act > > de-facto as a "reference". Because of this, it is up to us to decide what > > is > > "standard" and what is just an OS-dependant oddity. > > There are spefications from MS, eg: > Microsoft Extensible Firmware Initiative FAT32 File System Specification, > rev. 1.03, December 6, 2000, > (http://www.microsoft.com/whdc/system/platform/firmware/fatgen.mspx)
Ok, I have downloaded the document without accepting their license (which is completely legal where I live), and had a look it (or at least the part of it that can be deciphered from MS-Office internal format). I can't legally quote it, but it seems the requirement of case insensitivity is present. I really think it is completely broken to describe a filesystem structure saying, not how the data is organized, but how your implementation must present it to the upper layer. It is simply out of scope. Consider there was a document describing ext2 that said /dev/zero, if present, must output an endless stream of 0xff. > or even ECMA 107. Also, note that neither of these qualify as "standards". The first is a Microsoft internal document, the second is just a rubber-stamp organization. -- Robert Millan <GPLv2> I know my rights; I want my phone call! <DRM> What use is a phone call… if you are unable to speak? (as seen on /.) _______________________________________________ Grub-devel mailing list Grub-devel@gnu.org http://lists.gnu.org/mailman/listinfo/grub-devel
