From: Andrej N. G. <an...@re...> - 2014-09-13 16:49:37
|
Hello! Jerome Leclanche has written on Saturday, 13 September, at 17:32: >2014-09-07 16:02 GMT+02:00 Andrej N. Gritsenko <an...@re...>: >> >> The same example http://pootle.lxde.org/uk/lxlauncher/ as above - the >> translation has 17 strings, not 72 as Pootle shows, and yes, there are 72 >> words in those 17 strings. >I missed that part last time, sorry. Yes, Pootle shows word count >instead of string count; this gives a more accurate representation of >progress for translators. AFAIK it has always been that way. No, it never was before. Words count is completely invalid meter, as if just one dot was missing and that fails a check on string, it will not show 1 word but (may be) 20. And also what translation is more complete below? ========================== #: t.c:1 #,fuzzy msgid "Mama" msgstr "Papa" #: t.c:2 #,fuzzy msgid "Papa" msgstr "Caca" #: t.c:3 msgid "Love" msgstr "" #: t.c:4 msgid "Choose right person from the list:" msgstr "Оберіть потрібну персону:" ========================== ========================== #: t.c:1 msgid "Mama" msgstr "Мама" #: t.c:2 msgid "Papa" msgstr "Папа" #: t.c:3 msgid "Love" msgstr "Кохання" #: t.c:4 msgid "Choose right person from the list:" msgstr "" ========================== Counting by strings shows first is 1/4 complete and second is 3/4 complete. Counting by words shows first is 6/9 complete and second is 3/9 complete. I wouldn't consider counting by words correct in this example as second is better translated for people, they can suggest context but not short choices. >J. Leclanche Andriy. |