On Fri, Sep 18, 2026 at 10:39:34PM +0200, Enrico Forestieri wrote: > I was speaking about *native* Windows. I compiled LyX using MinGW to test > your patch, so Cygwin doesn't come into play in this case. You were probably > misled by the fact that the string "cygwin" appeared in the error message, > but that is simply because I use the Cygwin `/tmp` directory as the > temporary directory for the native LyX. That's all.
I think I understood your test correctly. Let me explain less briefly, because there are more issues here. 1) The CAS problem on Windows proper. The one you tested and reported .bat problem. Fails with my patch and would need #ifdef for windows. 2) Cygwin problem which I originally wanted to report only in the next step. The issue is that cygwin runs in init runCommand(mount + " --show-cygdrive-prefix"); As with CAS this is safe, and the only reason I touch it is that without these two I can get rid of shell-based runCommand out of our API. Before I new about the .bat problem in 1 I thought that I'll just #ifdef runCommand to stay cornered only for cygwin to solve 2. Now with 1 it means I need to starting #ifdef for windows in CAS code as well and became hesitant and I asked whether is it still good idea to get rid of runCommand altogether... Pavel -- lyx-devel mailing list [email protected] https://lists.lyx.org/mailman/listinfo/lyx-devel
