Michael J Gruber

I agree that this is more natural,but how do you get the current behaviour when you need it? For that we would need a search buffer which displays matching messages...

I use alot with a couple of private additions, and `per-message-subject` is one of them. Similarly, `unset hide_thread_subject` was one of my first settings when I (partially) switched to (neo)mutt....

... and now I noticed the real content of the question... So, deferring `decode_header()` like @kbingham does is certainly a good idea unless you expect to call `get_subject()` anyways (which...

The thing is that `git send-email` and some people do the "wrong thing": they specify both `In-Reply-To` and `References` when indeed the patches are not a reply to the cover...

Interesting. I think what I meant was slightly different, but both works. The differece is in whether you replace a match for the original subject (if it matches) or for...

Sure, what I meant is: Starting at root, set `threadsubject = subject of root`. Then (pseudocode) for each child repeat: ``` if threadsubject in currentsubject: replace threadsubject by ... in...

Update: yes, #1523 greatly helps. Though one could say that it covers up the slowness of the ANSItext widget ...

Well, I literally bisected this (`git bisect`) to the commit that I mentioned (with plain pazz/alot, no own patches). So, unless the ANSIText widget uses mailcap and the old one...

> I really don't see the problem that you're solving, sorry. That means I failed to describe it ;) In a search buffer, all displayed threads originally come from matching...

> I think I see now. For instance if I have messages A