Re: Proposed changes to ro_RO locale (Debian bug #119528)

[Petter Reinholdtsen <pere at hungry dot com>]

[Ionel Mugurel]
> Which kind of references do you need? I am Romanian, it is that good
> enough?

First of all, it isn't me you need to convince, it is one of the
people with write access to the glibc CVS.  I haven't quite understood
where the limits for accepted references are, but knowing the language
is in general not accepted.

Some official standard body or some official body demonstrating how
the romaniam locale output should look might work.  I'm still trying
to understand what will be accepted by the glibc maintainers, but I do
know that the chances improve a lot if the original author agree to
the proposed changes.

> I am a member of the Romanian Translators for Free Software
> (http://rtfs.sourceforge.net/,
> http://lists.sourceforge.net/lists/listinfo/rtfs-project) were we
> try to make a common standard for the Romanian versions of the free
> (and not only free) software.

Sounds good, but I am not sure if that is enough to convince the glibc
maintainers.

> I did not contacted him, since the e-mail address seams wrong.

Yes, I forgot a 'k' at the end of '.dk'.  Sorry about that.  I tried
to tell you about it, but must have failed.  Keld reads the libc-alpha
mailing list, so I dropped him from the CC list.

> I guess that in the previous version it was a dot. I just wanted to
> say it should be a space. But you are right, U00A0 is better than
> U0020. In fact even better would be a shorter space...

OK.  Good.

> Well, I proposed those changes in 2001... What was the significance
> of 2000 over there?

I believe it is a reference to the version of a standard.  I've since
been told that a proper value would be 'posix:1993' 'i18n:1999'.  Am
an not sure about all available values.  But I am sure it isn't the
year of the change itself.

> Yes, I can. The Romanian alphabet is: <A>, <A(>, <A>>, <B>, <C>, <D>,
> <E>, <F>, <G>, <H>, <I>, <I>>, <J>, <K>, <L>, <M>, <N>, <O>, <P>, <R>,
> <S>, <S,>, <T>, <T,>, <U>, <V>, <X>, <Z>. The correct sorting order is
> the same like in the alphabet.

Can you send me this as an UTF-8/ISO-8859-16 formatted file with one
entry/word/character on each line?  The lines must be sorted.

> The number format is like that:
> a million: "1 000 000"
> pi: "3,14 159 26..."
> monetary: "25 367,50 lei", "1,00 leu", "-5 000 lei"
> (Some times the two digits after the comma are smaller and a bit higher.)

I've created some test cases.  The current locale uses capital L in
'lei'.  Is this correct or wrong?  How is the monetary values supposed
to look if you use the international currency symbol?

I ended up with this content of tst-fmon.data.  Is it correct:

  #
  # Check the Romaniam locale (ro_RO)
  #
  ro_RO.ISO-8859-16	%n	123.45		123,45 Lei
  ro_RO.ISO-8859-16	%n	-123.45		-123,45 Lei
  ro_RO.ISO-8859-16	%n	3456.781	3 456,78 Lei
  ro_RO.ISO-8859-16	%n	-3456.781	-3 456,78 Lei

  ro_RO.ISO-8859-16	%i	1.23		1,23 ROL
  ro_RO.ISO-8859-16	%i	-1.23		-1,23 ROL

The test should use non-break space, but doesn't at the moment.

> I can look at the new regulations to see if there is anything else I
> could provide. This modification I made two years ago.

What do you mean by "new regulations"?