Micro$oft to license FAT

Picture Transfer Protocol

could become very popular suddenly.

What? Please cite a reference.

Something that used encryption or compression may indeed have validly patentable components. The FAT file system however is merely a data layout with a table lookup and bitmap of free sectors, none of which is novel and all of which precedes Microsoft's existence. (The broken patent system however could patent anything, requiring mountains of cash to overturn it, so who knows...)

But there were lots of minicomputers and mainframes which used floppies, mag tapes, and hard disks, with LOTS of experience having file systems. There were also lots of homebrew systems (there must have been, because that was Microsoft's initial market for its BASIC). There isn't anything new in Microsoft's original FAT file system.

The only reason this issue has come up is because so many products have standardized on FAT because of easily obtainable public domain software and interchangeability. No one uses FAT because it's well designed. Maybe if Microsoft uses it's clout to intimidate people, then the industry will switch to somethign else instead.

The first implementation that uses the claims of the patents was published in August 1995.

Microsoft Standalone Disk BASIC.

No, the short/long filename thing can be used with FAT12 and FAT16 as well.

If you don't implement long filenames, you won't infringe the patents.

And their patents.

If you do your own implementation, using their patent claims (short & long filenames), you may be infringing. If you buy their license, it's not a problem. If you only implement short filenames, it's not a problem.

So it was. I'm very surprised.

But you can implement long filenames in a different manner if you want, except that this gets back to the issue of people wanting to use existing solutions instead of redesigning software.

was

I am pretty sure this is not the first. I graduated in 1977 as a chemist and worked under a Rusty Marr at the Bureau of Mines. He showed me a patent on a Fortran program that he had obtained and claimed that this was essentially unheard of. I can't say it was the "first", but I am sure of the date. Now I only need to verify the veracity of his claim.

The simplicity of the components have nothing to do with the novelty of the invention. Motorola patented the "flip phone", a cell phone with a hinge in the middle. Certainly the cell phone was not patentable (or already covered) and hinges have been around for hundreds of years. But the combination was new and, I guess, not obvious.

Long file names have been around in Unix a lot longer than Ms?

/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ \/\/\/\/\ Chris Hills Staffs England /\/\/\/\/\ /\/\/ snipped-for-privacy@phaedsys.org

In article , John R. Strohm writes

In many parts of the world SW patents aren't legal.... just to ad to the confusion :-)

/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ \/\/\/\/\ Chris Hills Staffs England /\/\/\/\/\ /\/\/ snipped-for-privacy@phaedsys.org

What about Captain Kirk comunicator?

to

I think the Patent Office has lost sight of the meaning "Trivial". Just because you nail two things together that have never been nailed together before doesn't make it patentable.

Jim

I did a quick search to see if I could verify the 1987 date -- I couldn't -- but did find many references to an important legal decision in 1994 involving this patent (re Alappat) which established a precedent for the validity of a software patent (that is, an algorithm for a general-purpose computer). One source estimated that

It was conventional wisdom before that case that software patents were risky and unlikely to be upheld. In fact, when I worked for DEC in the late 70's we were flatly told that algorithms were not patentable, since they were viewed as being equivalent to equations or mathematical formulae, long held to be unpatentable. Nonetheless, I seem to recall that for years DEC avoided Hoare's patented quicksort, using heapsort instead. Perhaps they didn't want to establish a precedent either by licensing quicksort or testing the patent's validity.

Jim McGinnis

Yes, I supose that although a patent may have been granted to Marr, it may not have stood up to a test. It concerned some computations related to bore hole drilling and analysis and may well have been very specific. There might have been little need to test the patent.

That makes you wonder, will Motorola charge Kirk when he gets his captains commission in a couple hundred years from now, or will Kirk get to charge Motorola by then for using 'his' technology for the past centuries? Time travel can be so confusing..

At some point a local magazine printed an article how to build a working Star Trek communicator using CB frequencies, back in 1980 or so. Would that count as prior art?

Rob