[maemo-developers] [maemo-developers] Identifiers instead of English strings

[From: Guillem Jover]

Hi,

On Fri, Feb 17, 2006 at 03:25:48PM +0200, ext Kalle Vahlman wrote:
> On 2/17/06, David D. Hagood <wowbagger at sktc.net> wrote:
> > Murray Cumming wrote:
> > > Is there some explanation somewhere about the use of these strange
> > > identifiers instead of English strings in the maemo source code. What is
> > > the advantage?

> > Internationalization (I18N) - the strings are kept in a set of  separate
> > files, one for each language. The code then accesses that file to get
> > the string. Thus, you create a new file, and suddenly your app speaks
> > Spanish.

> AFAIK the id's are there for few reasons, one being that any
> mistypings or phrase changes in the original version won't mean a
> change for all the translations already done. This would be bad if
> there's 10 translations, all handled by different people (as I'm sure
> open source translators know ;).

If you use gettext properly, and have a typo in an original
(non-logical) string, you can fix the string, regenerate the pot file,
merge it with all po files and un-fuzzy the ones that were not fuzzy
previously due to this change. Thus no change in translations, and if
the translation was already fuzzy it needed updating anyway.

If the actual meaning of the string changed, then obviusly you want
the translations to reflect that.

> Also the code side benefits as you can quickly add a string called
> "this_app_warns_about_this" instead of waiting for the real and
> perfect sentence to be thought up, since it can be done after the code
> is already done (without changes to it).

I don't see a benefit with that. Also it has some severe problems:

Compiler checks - a wrongly specified format string or a missing
argument may produce a build time warning in normal situations, bogus
data or even a runtime crash with logical strings.

 - Logical: Bogus warnings + Runtime crash

   printf(_("foo_bar"), 450000000, 'c');

   msgid "foo_bar"
   msgstr "value %ld with string %s and %s.\n"

 - English: Proper build time warnings

   printf(_("value %ld with string %s and %s.\n"), 450000, 'c');

   $ gcc -Wall -o foo foo.c
   foo.c:6: warning: format '%ld' expects type 'long int', but argument 2 has t
   foo.c:6: warning: format '%s' expects type 'char *', but argument 3 has type
   foo.c:6: warning: too few arguments for format

Translation consistency checks - some of those can be performed
automatically if using the po tools with normal strings.

 - Logical: No warnings nor errors

   # Master English
   msgid "foo_bar"
   msgstr "this is an integer %d.\n"

   # Translation to Catalan
   msgid "foo_bar"
   msgstr "això és un enter %f %s."

 - English: Fatal errors

   #, c-format
   msgid "this is an integer %d.\n"
   msgstr "això és un enter %f %s."

   $ msgfmt -c foo.po
   foo.po:14: number of format specifications in 'msgid' and 'msgstr' does not
   foo.po:17: `msgid' and `msgstr' entries do not both end with '\n'
   msgfmt: found 2 fatal errors

Fuzzy handling - changes to the original english string do not trigger
changes to the translations. Also translations marked "fuzzy" are not
displayed to the user as they are obviously wrong and not matching the
meaning of the current original text.

 - Logical:

   printf(_("foo_bar"), 4, "baz");

   # Master English
   msgid "foo_bar"
   msgstr "This is a string"

   - We fix a typo.

   msgid "foo_bar"
   msgstr "This is a string."

   - Translations do not get noticed. Manual review needed.

   # Translation to Catalan
   msgid "foo_bar"
   msgstr "Això és una cadena de text"

 - English:

   printf(_("This is a string"), 4, "baz");

   msgid "This is a string"
   msgstr "Això és una cadena de text"

   - We fix a typo. The translation gets "fuzzy".

   #, fuzzy
   msgid "This is a string."
   msgstr "Això és una cadena de text"

   - We fix the translations. No "fuzzy" anymore.

   msgid "This is a string."
   msgstr "Això és una cadena de text."

And some less severe ones:

Losing the ability to get repeated translations of the same strings
merged into the same string.

Losing the ability to do incremental translations. It's everything or
nothing, as logical strings are meaningless to the user and thus like
not having anything.

regards,
guillem