Hi Aaron, * Aaron Griffin on Monday, March 26, 2007 at 13:54:19 -0500: > Hmmm, well, I don't really know what to say w.r.t. all that. The best > I can do is push a new pythoncomplete out that isn't dependent on > cStringIO. I think it may have been a pre-optimization anyway. I'll > make a note of it and see what happens tonight.
I did a few experiments. As I compile with --disable-darwin, I build a Vim.app (that's how it's called on the Mac) for testing, and there things get even more confusing with a newer Python in /usr/local, as it uses the new Python in /usr/local as C-API but links against the old Python shipped by Apple in /System/Library/Frameworks/Python.framework/Versions/2.3 Then I made my usual build with --disable-darwin and called ./src/vim meaning vim in the build directory. And! everything worked. Still, after make install, vim in /usr/local/bin showed the same errors. Then I ran $ diff src/vim /usr/local/bin/vim and they differed. Then I *copied* src/vim to /usr/local/bin/vim and now omnicompletion, and of course :py import cStringIO, works! Very strange. Apparently "make install" strips something from the executable. c -- _B A U S T E L L E N_ lesen! --->> <http://www.blacktrash.org/baustellen.html>
