Re: patenting compiler technology

Paul Biggar <>
Sat, 16 Jan 2010 11:03:24 -0800

          From comp.compilers

Related articles
patenting compiler technology (zeng jane) (2010-01-04)
Re: patenting compiler technology (Derek M. Jones) (2010-01-05)
Re: patenting compiler technology (rcmetzger) (2010-01-06)
Re: patenting compiler technology (George Neuner) (2010-01-11)
Re: patenting compiler technology (glen herrmannsfeldt) (2010-01-14)
Re: patenting compiler technology (Jeremy Wright) (2010-01-14)
Re: patenting compiler technology (Paul Biggar) (2010-01-16)
| List of all articles for this month |

From: Paul Biggar <>
Newsgroups: comp.compilers
Date: Sat, 16 Jan 2010 11:03:24 -0800
Organization: Compilers Central
References: 10-01-035
Keywords: legal
Posted-Date: 16 Jan 2010 20:42:01 EST

[Note, most of this stuff I'm not certain about, they're just nuggets
stored in my head somewhere.]

One effect of patenting algorithms is that they don't make it into
free compilers. I know GCC actively avoids patented algorithms, and
remember reading about one algorithm they would prefer to have used
that they couldn't (I think it was for register allocation).

I believe a problematic patented compiler algorithm is Steengaard's
unification based alias analysis. It was (originally at least) faster
than Andersen's, but patented and so avoided.

A point to consider for compiler patents is that no-one really makes
any money off compilers...


On Mon, Jan 4, 2010 at 10:19 PM, zeng jane <> wrote:
> B Perhaps a very different question than what gets posted but couldnt
> think of a better group than this. I am wondering if there is much
> value in patenting compiler algorithms.My reasons -

Paul Biggar

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.