Mark,
> Not sure I understand what the astAt/astINVOKE changes are about
> (something to do with reporting an error at line XXX of file YYY.c?),
> but won't these require a change in JNIAST?
I can't immediately think of any. If astNOERR is left undefined, then the
behaviour is unchanged from earlier versions - which is presumably what
JNIAST wants. This new macro is to allow Tim to insert his own calls to
astAt in his perl wrapper functions so that error messages issued by
AST refer to the line number in the perl code rather than the C
wrapper function.
> For the others, to take advantage of the changes Peter and I would
> need to rebuild the native libs on the supported platforms (no longer
> part of the normal build-from-CVS process). Not sure how significant
> the new features are - since Peter is main client of AST and the
> person who it's most effort for to effect the rebuild, I'll let him decide
> whether we want to rebuild for V3.2-5.
Things are a bit fluid regarding AST at the moment. The work being done to
produce perl wrapers is throwing up several (mostly minor) issues which
will need fixing. So it may be best to wait until the rate at which bugs
are discovered slows down a bit.
David
----------------------------------------------------------------------
Dr David S. Berry ([log in to unmask])
STARLINK project | Centre for Astrophysics
(http://www.starlink.ac.uk/) | University of Central Lancashire
Rutherford Appleton Laboratory | PRESTON
DIDCOT | United Kingdom
United Kingdom | PR1 2HE
OX11 0QX Tel. 01772 893733
01257 273192
|