Re: Encoding issue in the digest (Re: Emacs-devel Digest, Vol 240, Issue 25)

[Jean-Christophe Helary]

> On Feb 28, 2024, at 21:54, Yuri Khan <yuri.v.khan@gmail.com> wrote:
> 
> On Wed, 28 Feb 2024 at 19:21, Eli Zaretskii <eliz@gnu.org> wrote:
> 
>>>>>> Subject:
>>>>>>  =?utf-8?Q?Traduction_fran=C3=A7aise_des_manuels_d=E2=80=99Emacs,
>>>>>>  _=C3=A9tat_des_lieux.?=
> 
>>> This is how it looks in the digest listing here:
>> 
>> Which is okay, I think.  That's how the author's MUA sent it; the
>> Subject is encoded using quoted-printable method, I think.
>> 
>> Maybe something is wrong here, then.  I'm not an expert on this stuff;
>> does anyone know whether this Subject encoding is OK or not?
> 
> I think the operative word here is “digest”.

Indeed.

> The above encoded string
> is quoted not from the Subject header field on an individual message
> but from the body of a message that combines multiple mailing list
> messages.

That's correct.

> In the ideal world, the server software producing digests will take
> each individual message, decode its subject according to the encoding
> declared, decode the body according to the Content-Type: text/plain;
> charset= header, build the digest in Unicode land, then encode to
> UTF-8 and mail that to digest subscribers.
> 
> I’m suspecting the digest formatter might not be MIME-aware with
> respect to subject encoding.

Is that something that we need to report to the GNU project? Other 
mailman lists that I use don't seem to have that issue.