[desktop entry spec] new FullName key
shaunm at gnome.org
Mon Aug 3 07:54:17 PDT 2009
On Mon, 2009-08-03 at 17:36 +0300, John Tapsell wrote:
> 2009/8/3 Shaun McCance <shaunm at gnome.org>:
> > On Mon, 2009-08-03 at 09:41 +0300, John Tapsell wrote:
> >> > Not sure what you mean with 'supporting GenericName', but the current
> >> > GNOME HIG recommendations are the way they are precisely because of the
> >> > translatability concerns of combining Name and GenericName
> >> > programatically.
> >> Could someone give an example where programmatically combining would
> >> fail, if the combining was done as i18n("%1 %2") or even i18n("%1 -
> >> %2") ?
> > In languages that have case declensions, "%1 %2" and "%1 - %2"
> > could involve the GenericName being written differently. So
> > you might write "Epiphany - Web Browser", but "Epiphany Webo
> > Browsero". (Completely made up example, of course.)
> I don't get what you're saying. I wrote i18n("%1 %2"), but you might
> have missed that. Is that the confusion?
No, I understood what you meant. There are languages that will
do things like put different endings on a word based on where
it's being used in a sentence. So if you asked a translator
to give you a phrase that looks like "Epiphany Web Browser",
he might use different spellings than if you ask him to give
you a phrase that looks like "Epiphany - Web Browser".
So let's say you use "%1 %2", and you pass it to gettext or
whatever you like to use so that translators can turn it into
"%2 %1" for word order issues and the like. That's fine, as
long as the GenericName that was supplied by all translators
of all projects has the appropriate declension.
Now let's say I decide to go with "%1 - %2" in my software,
and I pass it to gettext, also thinking that solves all of
my i18n problems. Now we have a problem. For some languages,
the choice of GenericName that is correct in your software is
different than the choice of GenericName that is correct in
Is that more clear?
More information about the xdg