tag:www.post-issue.org,2010:/2009/4/21/all-forms-of-possible-prior-art-priority-date-of-april-1-1992/comments Mephisto Drax 2009-05-03T19:05:21Z Eugene tag:www.post-issue.org,2009-04-21:2563:3415 2009-05-03T12:33:29Z 2009-05-03T12:33:29Z Should this patent even apply to the case when the algorithm is being implemented in the same filesystem? If it does, this allows a company to become a monopolist by not allowing tools/drivers developed by others to work with their filesystem. Leland Wallace tag:www.post-issue.org,2009-04-21:2563:3353 2009-05-02T04:02:55Z 2009-05-02T04:02:55Z Apple had a shortname/longname scheme in their AppleShare product in 1989-1990 for their AppleShare PC product. It was also used for compatibility with the IIGS. Look at the CNode naming section of http://developer.apple.com/MacOs/opentransport/docs/dev/Inside_AppleTalk.pdf David A. Wheeler tag:www.post-issue.org,2009-04-21:2563:3292 2009-04-30T17:29:15Z 2009-04-30T17:29:15Z The "Rock Ridge" extensions almost certainly predate this; they also provide long filenames for a filesystem with short names (9660 in this case). http://en.wikipedia.org/wiki/Rock_Ridge has more. This was eventually released by IEEE as IEEE P1282 (ROCK RIDGE INTERCHANGE PROTOCOL, DRAFT STANDARD VERSION 1.12), which was Adopted 1994-07-08 - but since it was ADOPTED in 1994, its development and initial release began before and predates the priority date. For example, here's a Linux-kernel posting from January 2003 discussing implementation of Rock Ridge: http://lkml.indiana.edu/hypermail/linux/kernel/0301.1/1992.html Greg Bishop tag:www.post-issue.org,2009-04-21:2563:3282 2009-04-30T14:18:00Z 2009-04-30T14:18:00Z Replacing a name with a longer name held elsewhere is very obvious, and non-patentable. Just as links in a Unix file system may have names different from the files they represent, creating links to longer (or shorter) names and only showing the longer (or shorter) names when listing a directory is in no way an innovation. This capability has been in unix for 30 years now. Dirk tag:www.post-issue.org,2009-04-21:2563:3270 2009-04-30T09:57:25Z 2009-04-30T09:57:25Z I suppose the community had already had a close look at MiNT and how they implemented long filenames on a DOS-like partition. If not, it has become somewhat vague for me, but according to Wikipedia the first version of MiNT dates from 1990. It was adopted by Atari as an official alternative for the Falcon (late 1992). http://en.wikipedia.org/wiki/MiNT Atari used a DOS like filesystem on floppy's for there 680x0 based computers. Still there where not readable by DOS computers due to a very small detail in the header that made the DOS computers reject them. It was been set that the Atari way of formatting was consistent with Microsoft's documentation, DOS way of doing it not. A mass of utility's appeared that allowed to format on a Atari floppy's readable by DOS computers. http://www.compsoc.man.ac.uk/~xav/atari_software.html The filesystem on harddisks was, I believe, just an extension (longer fat tables) of the one on a floppy. http://linux.die.net/man/8/mkfs.vfat Apparently Minix offered long filenames on top of the Atari file system. If this was not already done, Atari's where wide-spread, Mint was somewhat more marginal, the way Eric Smith did it should perhaps be checked against both patents. Perhaps Microsoft already did before claiming them. :-) http://www2.gol.com/users/oskelton/mint.html Mike Protts tag:www.post-issue.org,2009-04-21:2563:3242 2009-04-29T22:01:27Z 2009-04-29T22:01:27Z The CA product Universe had long and short names, around 1990. This was a relational database for mainframes, and originally had 8 character limits (due to mainframe limits), but had a lookup that mapped longer names to the short ones.