Hi Andrej, > Hi, I am working on a translation for shepherd-1.0.1rc1, and I thought I'd > give some msgid feedback on a few small issues that I noticed. Here goes:
This is very valuable feedback. > There are several msgids with strings indented with two spaces at the > beginning. Could these two spaces perhaps be placed outside of translated > strings? Or maybe there are some languages where such indenting needs to be > translated differently? Even then, a separate msgid containing just those > indenting spaces could be made instead. I agree, but… initially I didn’t want to “lose” existing translations, so I kept it this way. We should change it at some point though! > There is a msgid that is just "~a", without any explanation. Maybe there is a > reason why this is translatable, but without at least a translators' comment, > it seems unnecessary. Done. > There are two msgids with the string "~:[~a~;~*local process~]", what exactly > is the part in square brackets? should it be kept verbatim, or should "local > process" be translated in a regular manner? I added a comment. > The software name "Shepherd" is mentioned in several msgids which look like > help strings, but sometimes the starting letter is uppercase, sometimes > lowercase. Seems inconsistent. It’s usually “the Shepherd” when referring to the package as a whole and “shepherd” (lowercase) when referring to the daemon. > There are several msgids which need to take advantage of the multiple plural > support of gettext: > - "service names: ~a; heap: ~,2f MiB; file descriptors: ~a" > - "Monitoring logging period changed to ~a seconds." > - "Closing ~a system log ports." > - "Process ~a of timer '~a' terminated with status ~a after ~a seconds." > - "Terminating process ~a of timer '~a' after maximum duration of ~a > seconds." > - "warning: already ~a threads running, disabling 'signalfd' support" Fixed these, except the first and last one, which are always plural. :-) > There are two or three msgids which start with "calendar-event:". Is that > name > of some subcomponent, and as such should be kept verbatim? Fixed. > "Timer spec lacks 'at TIME'." Added a comment. I’ve just pushed these and they will make it into the 1.0.2 POT file. Thanks for your feedback! Ludo’.
