Description of problem: Currently anaconda lists all languages in language code order, after the default language based on the user's location. Anaconda supports a surprisingly large number of languages, which is great, but we would like to suggest listing the most important major languages first making them super-easy to find and choose with one click, and then listing the rest of the other languages below. This will make anaconda more accessible to the vast majority of international users. The list of major languages can be agree on and defined in langtable (maybe on the order of ~10 languages). Additionally, we think any other existing locale languages for the user's country should also be displayed at the top of the language list, again using the existing data from langtable. Steps to Reproduce: 1. choosing the language on the first page of anaconda Actual results: The large number of languages spread out makes it hard to find common major languages (and alternative national languages). Expected results: List the major common international languages first (and also any other national languages for the location) and then other languages. Additional info: This would expand on the recent work Sundeep contributed to improving the keyboard layout ordering.
WIP PR https://github.com/rhinstaller/anaconda/pull/2935
I'm responsible for the upstream Region and Language settings. I recently came to the conclusion that the language shortlist we have there isn't desirable, and removed it from the mockups [1]. There were a couple of related reasons for this. The first is that the language shortlist is normative - there's a perceived value judgement which prefers some languages over others. If you have the bad luck not to use a short-listed language, it feels like you're somehow less valued. This ties in with a second related factor, which is that the logic behind a shortlist isn't always apparent and can easily feel arbitrary. So, while the long language list is clearly not a great user experience, it is at least easy to understand and treats users equally. Ideally we'd make suggestions based on personal aspects of the user - where they are located, and so on. [1] https://gitlab.gnome.org/Teams/Design/settings-mockups/-/blob/master/region-and-language/region-and-language-no-input.png
(In reply to Allan Day from comment #2) > Ideally we'd make suggestions based on personal aspects of the user - where > they are located Anaconda already does that (at least when the geo-location works).
(Just to link) https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/L5G3KAHKG5BLLP7NDUWK24RO3UX25TLQ/
This bug appears to have been reported against 'rawhide' during the Fedora 34 development cycle. Changing version to 34.
this has already been fixed/done, and available in anaconda f35.