bug fix release 0.22.1

[Tomi Ollila]

On Wed, Jun 29 2016, David Bremner <david at tethera.net> wrote:

> [ Unknown signature status ]
>
> Hi Team;
>
> A long time ago, David E. fixed a couple of annoying bugs with the emacs
> interface, and I added two commits to the release branch with the intent
> of making a bugfix point release. Then I got distracted, and never made
> that release.  Are there any other small, but important fixes in master
> that I should cherrypick? A few candidates are listed below.

>From this list:

>
> David Bremner:
>       lib: fix definition of LIBNOTMUCH_CHECK_VERSION
>       - this is a pretty embarrassing bug

#t

>       
>       doc: forbid further operations on a closed database
>       - this is not so important, but would let me close a debian
>       bug ;)

early doc fixes ftw!

>       
>       test: don't use dump and restore in a pipeline
>       - these seems like a test failing in the waiting.

potential mystical failures due to timing... #t

>       
> Mark Walters:
>       emacs: show: improve handling of mark read tagging errors
>       - not sure if this really fits the small bugfix category.

I'd say nil -- I have something to check on this ... (*)

> Stefano Zacchiroli:
>       notmuch-mutt: use env to locate perl for increased portability
>       - I don't use notmuch mutt, but it's a small fix

#t

> Tomi Ollila:
>       test: set LD_LIBRARY_PATH early and keep its old contents
>       - this seems like it could be annoying
>
>       test: test_python: set PYTHONPATH to the python execution environment
>       - I'm less sure about this one

nil

Tomi

(*) My "plan" is to:

- craft file to be added to fs and to database
- get the message to search results
- remove the file
- try to "open" the message into notmuch-show buffer

IIRC there was some peculiarity on behaviour -- but the same
peculiarity might have happened before the change was applied...