On 10 Dec 1999, Jan Vroonhof <vroonhof(a)math.ethz.ch&gt; wrote: As per usual with development, doco comes last[1] in the development cycle. *grin* Well, I need to actually grab the source and (more importantly) the headers for the Xaw variants and make sure that I can reliably detect them. Then I would be happy to implement a thing of `look for sexy widgets, use them if found'. At the moment it prefers -lXaw3d to -lXaw, even if -lXaw is a 3D Athena library. It will refuse to use a non-3d -lXaw if you asked for something 3d. If you told it to use a 3d library, it can't find the long name version, and -lXaw is a 3d library, I assume that you know what you are doing and that you have installed /that/ 3d Xaw over the top of the normal Xaw library. I don't know. At the moment I have backward compatible configuration in there, including the blind assumptions that it made about what and where the Athena libraries were. This is actually much easier than I had though to achieve. I also have an instinct that says not breaking peoples assumptions is nice, at least for a while. I can drop the backward-compatibility stuff if that's desirable though. That will break `./configure --with-widgets=athena3d' though. Um, it's actually working for `configure' now. I don't think that `lwlib' hacking will take that long. What's the standard for posting patches to xemacs-patches? You folk prefer unified diffs, as plain text attachments, yes? Daniel Footnotes: [1] Actually, I wanted to make sure that the configure stuff passed muster here before writing helpful documentation about it since that's what I find hardest :) -- Increasing numbers of private-sector companies are discovering the efficiency of hiring bright undergraduates or recent graduates and paying them next to nothing. -- Washingtonian magazine