Re: bug#5996: [PATCH] base64: always treat input in binary mode

[Glenn Morris]

Eric Blake wrote:

> The cause of the dual To: was that I had configured 'git send-email' to
> autosend to bug-coreutils, then manually added another --to option to
> keep 5996@ in the loop.  Sending to bug-coreutils is good for fresh
> patch series, but not so good when trying to send a patch in threaded
> reply to an existing bug report, so I have since changed my git
> configuration to no longer have an automatic sendemail.to entry.
>
> Unfortunately, since I sent that mail with 'git send-email' and did not
> bcc myself, and am not subscribed to bug-gnu-emacs, I didn't see the
> actual headers that git ended up using; your email was the first I saw
> of what I posted.  I know I changed the subject line as a result of
> sending from git, but did I even get the bug number in the subject line,
> or is that something that I will have to remember to do manually?

You can see your mail as it was received at:

http://debbugs.gnu.org/cgi/bugreport.cgi?msg=5;mbox=yes;bug=5996

Sorry, I know nothing about "git send-email". Things should work fine
if you just hit followup/reply-to-all in a normal mail client.

You can reply to either bug-coreutils or address@hidden and it should
work fine. Using the latter makes life easier. If you reply to
bug-coreutils, in order to work out the bug number we need either "Re:
bug####" in the subject, or a "Re:" prefix at the start of the subject
and a References: or In-Reply-To: header pointing to the original
report (this is a somewhat experimental feature).

If none of those attempts to guess a bug number work, you end up
creating a new bug report. The package is guessed from the To: header.
Currently, this expects only a single To: header, and looks at the
last if there are more than one. Your second To: header did not match
a maintainer address, so you ended up in the default package, which
was emacs (now debbugs.gnu.org). I will try to improve this part.