Re: [PATCH 1/3] localedata: use same comment_char/escape_char in these files

[Chris Leonard <cjlhomeaddress@gmail.com>]

On Mon, Apr 11, 2016 at 10:11 PM, Mike Frysinger <vapier@gentoo.org> wrote:

>
> my only clarification/semi-counter point is that getting languages into
> CLDR probably has a wider impact than only getting into glibc.  that db
> is pulled by a significant number of projects/companies (and i know for
> a fact it's used internally at Google in conjunction with ICU).  if it's
> only glibc though, that impact is entirely limited to people using glibc
> as i don't think there's that many people downstream from glibc's locale
> db.  and we want to try to automate/smooth the process for adding any of
> the locales that exist in CLDR today but not glibc.
>
> tl;dr: locales in glibc might get short term returns, but CLDR is the
>        only long term answer.

I totally get the utility of CLDR, but I know of 2-3 million users
that are directly downstream of glibc locales on Sugar/GNOME dual-boot
XO laptops.  Those users span an incredible array of languages that
are not commonly used on computers, yet.  Those laptops have shown
remarkable longevity in the field, I've just loaded a newly minted
build on a ten year old XO-1 and it runs like a charm.  My two main
reasons for swimming upstream to this project are to help those users
and to "show some love" to the upstream.

Nonetheless, I am onboard with the dual-tracking of locale development
to maximize utility and minimize re-work.

cjl