NopMap, > In my python render script, I am using render_to_file() to write the render > result to a ram disk and then read it again to do some post processing in > PIL. This works thousands of times in a row, but occasionally it fails with > the following error message. > > render_to_file( m, tmp_file ) > RuntimeError: Could not write file to r:/hill_tmp.png
It should be possible to convice it to give more details, no? Is it possible that someone else is reading the file while you are trying to write - I heard that Windows is picky about that? Either your own process, or maybe something completely unrelated, some program that indexes your RAM disk for some stupid reason or so? > Is there any chance that mapnik does not close a file properly after > render_to_file under rare circumstances? It's the only plausible idea I have > so far. If you include an auto-increment counter in the file name then that would at least reduce the chances of someone still reading the old file while you are creating the new one. Bye Frederik _______________________________________________ Mapnik-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/mapnik-users
