Re: Patentability of HTML tags.

Benjamin Franz (snowhare@netimages.com)
Sun, 22 Feb 1998 00:43:31 -0800 (PST)


Date: Sun, 22 Feb 1998 00:43:31 -0800 (PST)
From: Benjamin Franz <snowhare@netimages.com>
To: www-html@w3.org
In-Reply-To: <34EFCC81.5745@ozemail.com.au>
Message-ID: <Pine.LNX.3.96.980222001111.20176A-100000@ns.viet.net>
Subject: Re: Patentability of HTML tags.

On Sun, 22 Feb 1998, Dudley Mills wrote:

> There has been some discussion in this forum about the patentability and
> desirability of patenting inventions based on specific new HTML (or
> SGML) tags such as in:
> 
> At the very least the granting of these patents demonstrates that
> inventions having a special new HTML or SGML tag or element as an
> essential component are considered patentable by their US Patent
> Examiners.

[snip]

The Patent Office *often* grants patents that fail to hold up when
challenged on grounds of being unpatentable due to obviousness,
impossibility or prior art.  For example, people slip 'perpetual motion
machines' past them every year despite the fact the Patent Office
*specifically* forbids the patenting of Perpetual Motion Machines (at
least without an actual *working* model). 

I seems to recall someone patenting 'unrolled loops' for software. Never
mind people have been using them since at least the sixties and it is
probably been 'invented' independantly many many thousands of times by
programmers needing to speed up a piece of software with deep loops. They
are *OBVIOUS*. So obvious that probably almost every large piece of
software has the trick embedded into them by optimizing compilers
somewhere. Obvious enough that computer scientists working with
'Artificial Life' programs have witnessed it being 'invented' by programs
themselves. 

And a couple of years ago some idiots over at AT&T patented the idea of
'thumbnail images' for a web browser 'page history' (A web browser (not
from AT&T) was doing that years before their patent application). And yet
*another* idiot tried to patent the idea of 'downloading and executing
programs' in web browsers (only 3-5 years after they were first proposed
and implemented and a decade or so after the idea first circulated - the
guy actually had the balls to try and *collect* royalties - whatever
happened to him and his 'R' something or another Mosaic-license browser
anyway?  Didn't we come to the conclusion that Postscript would have come
under his patent if it was valid?)

The US Patent Office is well known for its lack of cluefulness in the
computer field. It is one of the reasons many people argue that the Patent
Office should get out of the business of patenting software algorithms
entirely (which is a questionable practice under the law anyway).

-- 
Benjamin Franz