L18L wrote:
automation of gettext
The developer has to decide which phrases have to be internationalized.
If he has one then it is basically just (in shell scripts) to sourround the message string by $gettext('
and ')
..and defining one TEXTDOMAIN.
The rest of this process is handled by MoManager
and one or more human native language speakers.
Hi L18L,
Thank you for your explanation.
I am old enough to know that protesting a historically grown situation loudly, usually doesn't result in a change.
But today I feel I want to protest loudly, because in my humble opinion the situation is just crazy.
1. The developer does
not have to decide which phrases have to be internationalized! He marks the strings he wants his users to see ('echo', 'read', 'label', etc..). He wants those strings to be visible, not only to his English-speaking users, but to anyone, from China to Brazil.
2. The developer has
already marked the strings that need to be seen by the users! (echo, etc...)
3.
L18L wrote:it is basically just (in shell scripts) to sourround the message string by $gettext('
and ')
..and defining one TEXTDOMAIN.
With just a little bit more intelligence, this shouldn't be necessary. Again, the developer already has marked the strings needed for his users.
The sting is in the word "basically" in your reply.
If it were that simple, every developer would
eagerly gettext his/her work.
As an exercise I tried to gettext the
Uextracttool by SFR.
It was easy enough to gettext the simple AppRun (the main window), but then I started to gettext the executable uextract itself. After many hours of trying, learning how to use eval_gettext and triple backslashes... I now have only one syntax error left, but it is completely mysterious and prevents the program from running.
SFR replied:
BTW, I had attempted to gettext it, but discouragement hit me instantly.
So when I speak about "automating" gettext, I am really surprised it has not happened already, as far as I can see. Is this due to habits, conservatism, GNU-elitism,... I don't know.
But the situation is that not many Puppy developers are able to gettext as soon as the script becomes more complicated ( $'s within the string). So, their work remains English-only , unless someone likes their work and decides to gettext it. "Someone" in this case means either Sergey, Victor or L18L..
How big are the chances that the gettexting will happen??
I hope this post makes sense