On 2005 Feb 16 , at 19.18, Starlink Software wrote:
> Module: kaprh
> Status: failed
> Build Log:
> http://dev.starlink.ac.uk/build/SPARC-SOL9/logs/kaprh.1108528828.log
[etc]
Components atools, kaprh, kappa and hdstools all failed because prohlp
wasn't available; extreme failed because prolat wasn't available.
This is part of sst, which is declared as a sourceset dependency in
atools' and hdstools' configure.ac; it wasn't in kaprh's or kappa's
configure.ac, but it is now. It was in extreme's configure.ac, as a
build dependency, but I've changed that to a sourceset dependency. You
need sourceset dependencies in order to make a distribution; you need
build dependencies, but not sourceset dependencies, in order to build
from that distribution; in this current context, building from CVS,
there's no distinction between them.
Steve, can you tell if the sourceset dependencies are being picked up
correctly? (I doubt this is the problem, since extreme was declaring a
straightforward build dependency, and they must be working). But if
that's not it, I'm a bit puzzled.
<http://dev.starlink.ac.uk/build/SPARC-SOL9/> says that sst was built
successfully, but I can't tell if it was built before or after.
Certainly the configure output for atools and kaprh shows
checking for prohlp... prohlp
which is what it shows if it can't find prohlp -- it crosses its
fingers and hopes it ends up in the path (I'm not convinced this is the
best plan).
> Module: sgmlkit
> Status: failed
> Build Log:
> http://dev.starlink.ac.uk/build/SPARC-SOL9/logs/sgmlkit.1108528828.log
This was successful, as usual.
But at least there's no evidence of failures because of *FLAGS. My
fingers remain crossed.
See you,
Norman
--
----------------------------------------------------------------------
Norman Gray : Physics & Astronomy, Glasgow University, UK
http://www.astro.gla.ac.uk/users/norman/ : www.starlink.ac.uk
|