Re: thunderbird needs graphical pinentry to use external gnupg for encrypting emails

"Xi Ruoyao" ([email protected] via blfs-dev Mailing List) <[email protected]>
Newsgroups gmane.linux.lfs.beyond.devel
Message-ID <[email protected]>
On Tue, 2024-11-19 at 13:56 +0100, Rainer Fiebig wrote:
> Am 19.11.24 um 13:42 schrieb Xi Ruoyao ([email protected] via blfs-dev
> Mailing List):
> > On Tue, 2024-11-19 at 12:36 +0100, Rainer Fiebig wrote:
> > > Am 19.11.24 um 06:00 schrieb Xi Ruoyao ([email protected] via blfs-dev
> > > Mailing List):
> > > > On Mon, 2024-11-18 at 21:58 +0800, Zhang Wen wrote:
> > > > > On 11/18/24 21:42, Xi Ruoyao ([email protected] via blfs-dev Mailing 
> > > > > List) wrote:
> > > > > > I just cannot see any way we can ever "fix" this.  Raising gcr to
> > > > > > recommended is not an option because GNOME is not used widely by BLFS
> > > > > > users.
> > > > > > 
> > > > > > 
> > > > > Maybe we can add some words on pinentry page, saying that the pinentry
> > > > > built here is not full, and can be rebuilt to be more functionally after 
> > > > > installing some graphical environments, or on the thunderbird and git 
> > > > > page? it's really borning without this information.
> > > > 
> > > > How's it different from all the optional dependencies in the book?  Or
> > > > are we going to add some text everywhere for every optional dependency?
> > > The problem here is that Thunderbird (and perhaps other) needs a
> > > graphical pinentry-box.  Else one cannot enter the passphrase.  To the
> > > user this is rather confusing because this happens even though he has
> > > followed the book to the point.
> > > 
> > > I found this out the hard way in 2019 and ever since build pinentry
> > > after GTK/QT.  A note at least in ch. Thunderbird would have helped a
> > > lot at the time.
> > 
> > ch. Thunderbird is definitely a wrong location because it does not only
> No, it's definitely _not_ a wrong location.  It's not a coincidence that
> the OP and I encountered the same problem with Thunderbird and not with
> "git commit -S".

Why it's not a coincidence?

I do really sign my BLFS commits with GPG but your (and my) posts to
blfs-dev are not signed.  So I can logically consider the git use case
more important here (while IMO both Git page and Thunderbird page are a
wrong location).

> > happens with Thunderbird.  At least I know it also happens to Git.  And
> > technically it'll happen with every GUI program invoking gpg.
> Then make it a recommended dep. for those programs, including Thunderbird.

It does not make sense.  It'd be just like making alsa-utils a
recommended dep. for all programs (potentially) using audio just because
w/o alsa-utils some audio cards cannot be configured properly.

Basically our options are:

- Making KDE Framework recommended for pinentry.  The resulted pinentry-
qt6 works for all DEs (not only KDE) but KDE Framework is a very bloat
dependency (for anyone not using KDE).
- Making gcr recommended for pinentry.  But the resulted pinentry-gnome3
does not work at all for DEs other than GNOME.

Also if both gcr and KF6 are installed, pinentry-qt6 will *always* be
used despite pinentry-gnome3 is installed.  But in GNOME the latter
looks better.  Arch uses a custom script as /usr/bin/pinentry to "fix"
the issue.

Now I'm feeling we need a lightweight pinentry implementation working on
all DEs (and just delegate to pinentry-gnome3 or pinentry-qt6 on GNOME
or KDE for "aesthetics").  But if we really pursue this we'll end up
writing our own...


-- 
Xi Ruoyao <[email protected]>
School of Aerospace Science and Technology, Xidian University

-- 
http://lists.linuxfromscratch.org/sympa/info/blfs-dev
Unsubscribe: See the above information page
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.