On Tue, Oct 14, 2003 at 04:55:07PM +0200, Corinna Vinschen wrote: >On Tue, Oct 14, 2003 at 10:24:47AM -0400, Christopher Faylor wrote: >> On Tue, Oct 14, 2003 at 01:29:37PM +0200, Micha Nelissen wrote: >> >Corinna Vinschen wrote: >> >>This patch is a nice idea but it's not quite correct. You can't >> >>rely on "current_codepage" being ansi_cp. Since the user can set >> >>it to oem_cp in the CYGWIN environment variable, you have to memorize >> >>the old value on \E[11m and to restore to the old value on \E[10m. >> > >> >Ok, that's true. Attached is a patch with the suggested changes. >> >> I guess this is ok although it's not thread safe. It looks like >> fhandler_console isn't thread safe in general, though. >> >> Corinna do you have time to check this in? > >Erm... I was still mulling over this code since I thought there's >something wrong. It took some time until it occured to me that this >implementation overrides the original value of current_codepage, if >the application accidentally happens to send \E[11m twice. That >shouldn't be possible.
Good point. Ok, I'll just shut up and let you handle this. :-) cgf
