Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!samsung!munnari.oz.au!labtam!graeme From: graeme@labtam.labtam.oz (Graeme Gill) Newsgroups: comp.graphics Subject: Re: Pixar's patent on stochastic image generation Summary: Xor Cadtrack etc. Message-ID: <5520@labtam.labtam.oz> Date: 1 Nov 90 01:03:20 GMT References: <67527@iuvax.cs.indiana.edu> <1990Oct29.203607.403@heitis1.uucp> Organization: Labtam Information Systems Pty. Ltd., Melbourne, Australia Lines: 60 > In article <1990Oct26.131122.2169@abcfd20.larc.nasa.gov> jcburt@ipsun.larc.nasa.gov (John Burton) writes: > > I just read the article. I don't understand what the real threat to X is. > It should be possible for MIT to continue to use backing store and the > exclusive-or cursor. Since they do not charge anyone for the sources, > the royalties should be minimal ;-). If the X system is distributed with > this capability, and a company wishes to develop and sell a server for X, > they could not enable the code for backing-store, this will make it slower, > but will only affect the systems which do not have a supplied server with > the X system. Of course, the XOR cursor is gonna be a lot tougher to get > around. > > brian It is not true that X uses an Xor cursor. The basic X supported cursor actually overlays the screen graphics, this being done using overlay planes in hardware, or software cursor code that saves the underlying images so that they can be restored when the cursor moves. Some applications (i.e. window managers etc.) make use of an "Xor" cursor, but they do so by making use of one of the 16 boolean operations X supports, i.e. Xor or more usually invert. I have had a look at the Cadtrack claims and patents (and not being a patent lawyer, I found them rather difficult to follow), and in my opinion they should not be a worry for most X vendors. As I understand it, Cadtrack is claiming patents on hardware supported Xor, split screen, panning, double buffering, textured lines and areas. Since the MIT code is based on a dumb frame buffer, all the above operations are simulated in a more laborious way by software (rather than hardware) to achieve a similar visual result, so the Cadtrack patents are probably not being infringed. As well as the above points, I don't believe the Cadtrack patents should stand up to broad interpretations (i.e. patenting the Xor operation) because of the obviousness of the techniques, prior use, and the mathematical basis of logical operations between two boolean variables. It is really like claiming a patent on addition, when applied to calculating a number for display on a numeric display. Things like double buffering are also not very novel, as it is just another form of pipelining, a long established technique in digital systems. (Of course proving this legally needs luck and LOTS of money). The double buffering patent is an example of an obvious idea that any practitioner of the art would come up with when asked to solve the problem of smooth picture updates. The point is that it wasn't worth publishing or building anything to use this technique until memory prices became more reasonable. i.e. the reason nobody used this technique in the 50's and 60's wasn't that it was such a non-obvious idea, but that the technology simply wasn't available to make it practical. Many of the patents that are causing controversy seem to fall into this class. Another point about Cadtrack are that they have been very slow in following up potential infringements. At least one manufacturer has been shipping terminals worldwide for over 10 years that have made use of the Xor cursor technique, and it is only now that Cadtrack have done anything about it. I understand that this would not look good in the eyes of a court. i.e. they haven't vigorously pursued their claims. Graeme Gill Design Engineer Labtam I.S.D. Pty Ltd