Subject: Re: [ublas] [bindings] New traits system
From: Rutger ter Borg (rutger_at_[hidden])
Date: 2010-01-13 09:40:04
Thomas Klimpel wrote:
> The other commits where I cheated are:
Currently, 34/40 lapack regressions pass (this includes previously
commented-out cases like those for pptrf). The following 6 don't pass yet,
mainly due to wrong overload generation somewhere, and the band fill-in
stuff. Improvements in the parser/generator (mainly trait recognition) have
lead to a few dozen extra routines (e.g., pptrf), and fewer arguments for a
bulk of existing routines.
Can anyone explain the RFP and AFP formats? Do someone know of C++ libraries
using this for some of their containers?