Home Messages Index
[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index

Re: Time to Wipe Your Windows Partition for Good

In comp.os.linux.advocacy, Roy Schestowitz
on Thu, 12 Jan 2006 05:34:12 +0000
> __/ [The Ghost In The Machine] on Wednesday 11 January 2006 19:00 \__
>> In comp.os.linux.advocacy, rex.ballard@xxxxxxxxx
>> <rex.ballard@xxxxxxxxx>
>>  wrote
>> on 11 Jan 2006 05:10:36 -0800
>> <1136985036.869153.181720@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>:
>>> FAT is over 20 years old, not even covered by NDAs, and was based on
>>> technology used in CP/M and other operating systems such as RT11 and
>>> NOW Microsoft  wants patent rights?
>> I think they're referring to the VFAT hack.  Microsoft is perfectly
>> entitled to patent rights thereon, if there's no prior art.
>> I'm not sure there's much point, though.
> One main issue (as I personally choose to identify it) is that one's own
> files can be locked-in to a filesystem, which is apparently now a property.
> Forget about file formats for the sake of the argument. This is very
> different from most 'implementational' patents. This gives ownership to
> people's files and the ownership is the hands of a closed-source software
> vendor.

In its way, it always has been.  What, exactly, is a
"file"?  It's a bunch of bits on a disk; without something
to interpret it (e.g., a piece of software that knows
how to find those bits, given a name) it's a bit hard to
get at.

And the standard nightmare is a Windows system wiping its
disk clean.  (Personally, I think that's largely vanished
now that 2k/XP is out there.  However, Windows still likes
to play "wipe the bootblock".)

>>> Microsoft probably does deserve a patent for NTFS - but even the
>>> expiration date should be set to 19 years from the day of first
>>> prototype, and even then, only if Microsoft makes a full disclosure of
>>> the technology.
>> 19?  I thought it was 17 years.  Did they change it again?
> It will probably be superceded in Longhorn sooner or later.
> WinFS is probably delayed (deadline woes), but expect a patent
> on that. Dual-boot and file sharing which involves a Vista
> partition could become a cancer from a legal perspective.
> This may seem irrelevant to a home user, but what about
> businesses that use a hybrid of FOSS and Windows?

It'll have to be superceded for Microsoft to make money. :-)
Personally, I'm not sure *what* WinFS brings to the table.

>>> The real issue here is that the patent claim on retroactive technology
>>> is a form of fraud.  The market adopts certain technologies based on a
>>> number of factors which make it a better choice relative to other
>>> competitive technologies.  For example, FAT32 was competing with
>>> offerings from SUN (PCNFS), Slackware, Linux, SCO, AFS, and Novell.
>>> FAT32 won out - not because it was the superior technology, but partly
>>> because it was a royalty free technology.
>> And now Microsoft will want retroactive payments, methinks.
>> This sort of thing might stymie the creative process.
> This is reminiscent of the blind eye to piracy. Licensing was
> not as much of a concern at the start (Windows x.x), but think
> of it as a test drive.

And a test of our gullibility.  :-/

>>> If Microsoft insists on enforcing this patent, it's very likely that
>>> ext3 and other competing technologies will be doing to FAT and NTFS
>>> what FireFox has done to IE and what OpenOffice has done to MS-Office.
>> Almost a certainty.  The code is readily available (grab
>> a Linux kernel and go).  I don't know how much vendors
>> such as camera manufacturers care about such things as
>> long file names or permissions, admittedly, but I also
>> don't see things being a major impediment.  However, I'm
>> not sure what the limitations of hardware in the camera
>> market are at this point.  (I'd have to look to see what
>> can be shoved on an EEPROM -- and my knowledge thereof
>> dates way back to the 2764. :-) )
>>> This ruling is just making the case for Open Source technology that
>>> much more clear.
>>> This is a poor quality patent application which understates the prior
>>> art, does not acknowledge prior art - including GPL protected
>>> technology.  This is one of the problem with the whole software patent
>>> process anyway.
>> The USPO is overloaded, AFAICT.
>>> IBM and several other companies are now pushing "Patent Quality" and
>>> are now pushing to have all prior art possible included - in full text
>>> form, into the patent (as opposed to mere references which will not
>>> appear in patent searches).
>>> An even darker black-hole is technology protected by nondisclosure
>>> agreements which as then patented 5, 10, even 20 years later, and only
>>> because the technology application was kept away from patent office
>>> archives by the nondisclosure agreements.
>>> The whole point of the patent process was to make sure that there was
>>> public disclosure of emerging technology at the earliest possible
>>> stages, in part, to make sure that one applicant wasn't in fact
>>> stealing from the original creator.
>> There's also the issue that VFAT was pretty darned obsolete by
>> the time Linux ext2 came around anyway.
>>> I can understand the desire for defersive patents - for example,
>>> Microsoft filing for a patent on FAT to make sure that some whiplash
>>> lawyer doesn't file on behalf of some kitchen table software company
>>> who simply files the application for software that Microsoft has been
>>> publishing for 20 years.
>>> Unfortunately this is the biggest problem with the entire concept of
>>> software patents.  Software is fundamentally an algorythm - which is
>>> not patentable.  The patent office rejected attempts to patent software
>>> for at least 30 years, and THEN they decided that they wanted to allow
>>> companies to patent software based on the claim of a particular use for
>>> an algorythm.
>> "algorithm".  Presumably, an issue is that Big Business wants Big Bucks
>> for such things.
> "Algorithm" is a word with foreign origins, so I imagine that spelling is
> flexible. I still like the word "virii", regardless of how heavily it's
> being criticised.

Agreed here.  I use "virii" on occasion.

(Psst.  Don't tell our President, but I think the origins
of the word "algorithm" might have been from somewhere or
someone in the Middle East...  :-) )

> I believe that the catalyst for software patents was -- at
> least in part -- Open Source software reaching the hands of
> many. Suddenly, companies were frustrated to see their
> secrets formats being interpretable by rivals. Also,
> clones become a concern (GIMPShop, OpenOffice and XMMS to
> name a few). I can't see a perfect solution to this, but
> defending a piece of code with law is impossible. It's
> metaphysical. It's duplicable without a cost.

I wouldn't say "impossible".  However, neither is the
defense perfect.  It merely adds additional barriers and
complications to what should be a straightforward process
of ... erm, processing ... data.

In short, Tool A defines a format.  Tool B reads it.  Vendor A
patents the format.  (How much sense this makes, I'm not sure.)
Vendor A then goes after Vendor B.  Vendor B now gets to
hire lawyers to defend itself, and Vendor A is now expending
energy playing predator rather than improving its tools.

Of course the flip side is Vendor B siphoning off Vendor
A's success.  Piracy is running rampant in China, for
example, and proper remuneration for videos is not being
paid by many cheapie disk distributors over there, some
little more than a hawker standing on the sidewalk.

This is not good for Microsoft.  (It's not all that good
for the buyer from the hawker, either -- no support.)

>> Of course the USPO is not going to be a static agency; at one point,
>> for instance, it insisted on working models for all applicants.
>> (This was short-lived as they ran out of storage space.  Some of
>> the models presumably still exist.)
>>> As a result, we have a 30-40 year black hole of intellectual property
>>> for which the prior art is undocumented, poorly documented, or
>>> otherwize not part of the patent archive.  To make matters worse, most
>>> of this same software was never registered with the Copyright Office
>>> either.
>>> Keep in mind that these intellectual property archives include all of
>>> the articles posted to usenet and other mailing lists since it's
>>> inception back in 1980s, and the informal communication structures at
>>> MIT and other educational institutions who played a major role in the
>>> implementation of such technologies as UNIX, PCs, PC Software, and many
>>> features found it modern operating systems such as Windows.
>>> Clearly this whole patent process needs to be populated with that 30
>>> year archive history - as much as is possibly available.  If this means
>>> filing hundreds of thousands of patent applications which list millions
>>> of prior art "devices" that will be expensive - and leaves the entire
>>> system open to fraud and manipulation until the archive is fully
>>> populated.
>>> The article above is just another example of this type of fraud and
>>> manipulation of the patent system.
>> And the USPO's inability to deal with it, because of their being
>> overloaded.
> Could it be that Microsoft were the first to patent this month-to-ear,
> long-inspired and evolved 'method' due to the fact that application are
> MSIE-compatible only?
> At the end of the day, more aggressive companies will *dare* to patent
> something that brings guilt to mind and is known to be unjustifiable. As
> Bill Gates once said "Being aggressive is not unethical".

True.  The original intent of patents was to protect the innovator.
VFAT may well be one of Microsoft's few true "innovations", and
even then, someone's already mentioned some possibilities of
prior art, though PC-NFS would probably not be one of them; I would
think PC-NFS is either a server product (basically servicing port
2049 with a specialized program) or a product that allows access
to NFS files with a transparent naming scheme allowing for more than
8 chars dot 3 chars.

I'd have to look.

> Roy

#191, ewill3@xxxxxxxxxxxxx
It's still legal to go .sigless.

[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index