On Thu, 18 Sep 2003, Giaretta, DL (David) wrote:
> Take a look at http://aips2.nrao.edu/docs/aips++.html
> <http://aips2.nrao.edu/docs/aips++.html>
>
Old news. We knew about this early on considering that we have adopted
AIPS++ as the real-time processing engine for ACSIS (which shoulod have
been here already but is now coming next spring).
It sounds really bad in that web page. What's really happened is that the
AIPS++ work has been absorbed into the individual telescope projects to be
more focussed on the customer rather than writing an "ideal" data
reduction engine. VLA and ALMA are now pushing things. There was a worry
that there would be a fork in development but it seems that everyone has
seen sense and have agreed to share the source code repository (and they
are switching it CVS). There is a risk of a fork from GBT.
The key outcome of the review was the dropping of the glish messaging
system and replacing it with CORBA. GLISH itself should be able to stay
(they have too much code). Extending support for python as a scripting
interface, replacement of Tk with something more modern and faster. As I
understand it, the real fight is over the ALMA software since ESO are
insisting that AIPS++ is too slow and needs to be re-eritten if it is to
be adopted for ALMA.
--
Tim Jenness
JAC software
http://www.jach.hawaii.edu/~timj
|