On Fri, 30 Nov 2007, Benjamin Kirk wrote: >> On the other hand, it would be better if we could make the default >> behavior more user-friendly. Instead of testing for the presence of a >> --read_legacy command line switch, why can't we test for the absense >> of that new version line at the beginning of the file? > > Probably do-able...
No rush, but I think that before putting out a 0.7.0 release we should be backwards compatible with our old files without command line options, and even before putting out a 0.6.9 release we should at worst die with an informative error message when trying to read an old restart file without the right command line option. >> Based on the changes I had to make for ParallelMesh, I don't think >> any numbering counts as "invalid" anymore unless it's actually >> non-contiguous, and even that only breaks on some of our output code. > > This is non-contiguous, which only breaks when using e.g mesh.elem(e) > instead of iterators -- which isn't too bad of a restriction. Not for most code, no. The only things I remember breaking with non-continuous ids are some of the I/O classes, although I wouldn't really trust the error estimator classes on noncontiguous ids either. --- Roy ------------------------------------------------------------------------- SF.Net email is sponsored by: The Future of Linux Business White Paper from Novell. From the desktop to the data center, Linux is going mainstream. Let it simplify your IT future. http://altfarm.mediaplex.com/ad/ck/8857-50307-18918-4 _______________________________________________ Libmesh-devel mailing list Libmesh-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/libmesh-devel
