Hi Florian, "pelzflorian (Florian Pelz)" <pelzflor...@pelzflorian.de> skribis:
> On Wed, Apr 08, 2020 at 03:22:34PM +0200, Ludovic Courtès wrote: >> Go for it, thank you! > > I added some more heuristics on whether a layout is Latin or not and > after much testing pushed as 91c231a2223440081426929828a23c7baa0214fd. > There are still few false positives like non-Kana Japanese layouts > that are Latin but toggle to U.S. They can be fixed later. Awesome! > More importantly with 7bc71025ccd0fabd739803889825e10341fbabdf I also > fixed a layout’s empty variant "" not being a choice in the installer. > The "" variant is the only variant of the Azerbaijani layout providing > the Latin script Azerbaijan uses since a few decades, so it is > important. I hope the code is alright; maybe I should not have just > pushed, but I believe it is alright. Good, it’s great you’re paying close attention to this. The patch seems reasonable to me. > P.S. Don’t brag with Arabic support in Guix’ installer just yet > because Arabs still have to confusedly read from left to right. > > https://bugs.gnu.org/35319 > > But that is something to solve later. Yeah. I’ll run the “gui-installed” tests to make sure everything’s is still fine, then run “make release” and upload files for testing later today. Thanks a lot, Florian! Ludo’.
