 |
 |
Peter and all, On 2004 Oct 8 , at 10.18, Peter W. Draper wrote: > The remaining big problems, are that I needed to use --without-stardocs > and AC_FC_CHECK_INTRINSICS doesn't work for all the functions we test > for, > but I think that's it. I dimly remember the latter problem coming up, but I discover it isn't on my own list of buildsystem niggles. I'm trying _not_ to spend all my time working through those, and rather knocking things off when I get a gap (which isn't often), but this means I'll inevitably forget which ones are important, and which ones are merely desirable. Is it now time we started using bugzilla consistently? I raise the prospect without much enthusiasm -- the prospect of having a queue like that, which other folk add things to, is a potentially rather depressing one -- but I feel we're in danger of going round in circles sometimes, as we one after another run into problems which it turns out were raised some while ago. I acknowledge that because it's me looking after the build system as a whole (this is not meant to be exclusive, by the way: do feel free to dive in!), it's quite often me that's holding things up. However if there were a list of outstanding buildsystem problems somewhere other than on my machine, it would mean that others could knock off ones they knew how to fix, in a fairly non-confusing fashion. That would mean I could work on TimeFrame.... > This should at least get us in the position of > thinking about a nightly build. This would be very good. Al mentioned that the GKS build was failing in an odd way, but I've no idea when it first failed, and so what it is that might have caused it. That wouldn't have solved the problem, but it might have helped. Norman -- ---------------------------------------------------------------------- Norman Gray : Physics & Astronomy, Glasgow University, UK http://www.astro.gla.ac.uk/users/norman/ : www.starlink.ac.uk