11 apr 2013 kl. 14.03 skrev Stefan Sperling:
The keys people type at the conflict prompt will eventually become part of muscle memory. What is mnemonic to one person might not be mnemonic to someone else.
[...]
In which case having to type different keys at a menu prompt depending on the language can be counterproductive, especially for fast typists.
Thanks for explaining your line of reasoning. I wonder how common it really is to jump between different localisations like that. Please permit me to expand on my argument:
With English-mnemonic codes for all translations, the gain for the LANG-switching crowd, to which you belong, is a loss for the monolingual non-English user. At best, with reduced usability (arbitrary codes that are difficult to remember in the way that you mention); at worst, with increased risk of errors (when the English mnemonics are falsely suggesting a different choice).
Other than translating the codes, the obvious solution would be to use a neutral sequence -- 1, 2, 3 or a, b, c. While I wouldn't be opposed to such a scheme, it may appear a tad inelegant since different conflict prompts expose different subsets of the options, which would make them appear non-consecutive. Perhaps this would not be a serious concern in practice. However, the codes would also cease to be mnemonics for anyone. Translating them still appears to me as the best solution, only slightly inconveniencing a small group.
Comparing localisation of the mnemonics to that of GUI acceleration keys (which indeed usually is a mistake) is stretching it, I'd say. Keystrokes like Apple-Q or C-M-\ are always muscle-memory-inducing -- my hands know a lot more about Emacs than I do -- and the actions take effect immediately; they do not generate letters that turn up at the cursor for inspection before pressing Enter.
This isn't to say that I'd be a strong opponent of English interactive mnemonics; I'm just trying to find the best overall design.
