Re: Request to reinstate project: m4-1.4.18d.pot

[Eric Blake]

On 5/8/21 5:53 AM, Eric Blake wrote:
> The following beta release includes a new m4.pot, where we would like to
> reinstate the m4 translation domain with disclaimers required (since FSF
> owns M4 copyright):
> 
> file m4-1.4.18b/po/m4.pot within
> https://alpha.gnu.org/gnu/m4/m4-1.4.18b.tar.xz
> announced at
> https://lists.gnu.org/archive/html/m4-discuss/2021-05/msg00001.html

m4.pot has been updated with many more strings as part of
1.4.18d.tar.xz, announced at
https://lists.gnu.org/archive/html/m4-discuss/2021-05/msg00002.html

Benno wrote:

> Before reinstating m4, I would like to see all of the translatable strings
> gettextized.  Of the main help text, only the "Limits control" fragment is
> gettextized.  Also, to make translation easier, it will be best to split
> the help text into smaller pieces: one option, one header, one paragraph
> per gettextized string, and preferably use a separate fputs() for the lone
> newlines (the blank lines), ungettextized.

I made changes along those lines;
https://git.savannah.gnu.org/gitweb/?p=m4.git;a=commitdiff;h=7a91d6598c4

Let me know if it is worth yet one more beta release with an updated
m4.pot for anything further, or if we are now good to reinstate the
project and await a few translations before releasing 1.4.19 as stable.

> 
> When I run 'grep -A1 M4ERROR src/*.c', there are nearly a hundred error
> messages that are not yet gettextized.  A bunch of them are internal errors,
> so maybe those have been left ungettextized on purpose?  Grepping for
> 'm4_failure' finds another twenty ungettextized strings.

Also done in that patch.  Internal error strings were not translated
(they should never be triggerable by user input), but the rest were marked.

> 
> When m4 gets reinstated at the TP, how long after that will it take for
> another release of m4?  I ask, because the last releases were something
> like three to four years apart.  It's not nice to make translators wait
> several years before their work finally appears in a release.

My goal is to get 1.4.19 out the door this month.  It may then be a
while longer before the next release, as m4 is a somewhat stable project
with a very low commit volume lately, so it's harder to predict that
timeline (this is still mostly on my free time, which had been scarce as
of late, rather than something that I am assigned from Red Hat).  But
I'm also amenable to the idea of a 1.4.20 in a few more months to
address any portability bugs uncovered by 1.4.19, and to pick up any
additional translations collected in the meantime.  Now that
translations are an actual driving factor where they were not
previously, that alone may warrant me paying more attention than once
every few years.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org