Am 09/08/2014 11:28 PM, schrieb YoheY - OpenOffice:
thanks for your help, now I have a more clear view in it.
Further questions:
And what if a new language appears in the list and the translators of
other languages are on holiday?
All 'holiday' site will present the download links incorrectly, until it
is presented by somebody who is visiting the page or tested/supervised
by the translator back form holiday.
then it will be broken until it gets fixed. Either from the volunteer
when she/he is back or from someone else. Sorry, but we cannot cover
every case of problems. ;-)
And what about sequence?
This download list is ordered alphabetically, which is a very good point
for making selection easier.
For English visitors.
But not for the others, the native site users.
If we translate the language names, we will need a different
alphabetical order.
Which generates a faulty download link again.
Or, leaving only translated, but keeping the original order, we will
have a pack of translated, but messy dropdown list for native languages...
Right, it was simply not taken into account that also the language names
could be translated. Now we have indeed the problem that the order is
not correct in every language.
What about using one global csv file containing all native translations,
iso codes, preferred language selection, translation of 'other versions'.
Like this:
iso,native name,Language 1,Language 2,Language 3,Language 4,Language 5
en,English,English,German,Italian,Hungarian,Spanish
de,Deutsch,Englisch,Deutsch,Italienisch,Ungarisch,Spanier
it,Italiano,Inglese,Germano,Italiano,Ungarico,Spagnuolo
hu,Magyar,Angol,Német,Olasz,Magyar,Spanyol
es,Espanol,…,…,…,…,…
…,…,…,…,…,…,…
Getting the country from the UA string or IP, js selects the correct
language - or in case of non-recognised, not-translated language selects
English as default -, sort them alphabetically, chooses the correct iso
and preferred language, and anything else which is needed.
Sounds good. But I'm unsure about the sorting. How should this work? Do
you have some examples?
One file to handle, no duplicates to cause conflicts, correct
alphabetical ordering, correct iso selection for the main site and for
all native subsite.
We have one file to handle - this is per language. At least this already
OK: ;-)
Marcus
On 2014-09-08 21:02, Marcus wrote:
Am 09/08/2014 04:52 PM, schrieb YoheY - OpenOffice:
Hi!
I found another issue, why xx variables in xx package should be revised.
In test-hu at the moment ( http://www.openoffice.org/test-hu/download/ )
Khmer is selected automatically, although DL.NL_LANG = "hu"; change has
been made (And it worked perfectly previously).
The reason - I gues only, since I can't change the files on svn - that
the 3 new available languages are
left out from the newly uploaded msg_prop_l10n_test_hu.js file. Since 3
left out (Catalan 1,2,3), which were inserted before Hungarian, it
counts 3 more after Hungarian, which is Khmer. It seems, it does not
care about DL.NL_LANG always, it seems it stores a value somewhere
behind, telling to js which number should be used in the sequence of
languages.
when filling the drop-down-box, the strings from the array in
"msg_prop_l10n_<xy>.js" will be used when existing. Otherwise from the
global "msg_prop_l10n.js" file.
But when the language from DL.NL_LANG should be shown as pre-selected
in the drop-down-box, only the strings from the array in the global
"msg_prop_l10n.js" file are used as only here the ISO codes are listed.
The array in "msg_prop_l10n_<xy>.js" contains only the localized
strings. Therefore these are ignored because here are no ISO codes
listed that could be compared with DL.NL_LANG.
Changing this would result in changing all arrays in all
"msg_prop_l10n_<xy>.js" files which is IMHO not worth the effort.
Rather I would use this miss-match as a good indicator that the
languages in the global and localized JS files do not match.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
