[Users] Deleting one long thread

[blind Pete]

On Wed, 30 Sep 2015 07:53:28 +0200
wwp <subscript at free.fr> wrote:

> Hello Francis,
> 
> 
> On Wed, 30 Sep 2015 07:38:02 +0200 Francis Moreau
> <francis.moro at gmail.com> wrote:
> 
> > On Tue, Sep 29, 2015 at 8:46 PM, Paul <claws at thewildbeast.co.uk>
> > wrote:
> > > On Tue, 29 Sep 2015 20:30:34 +0200
> > > Ralf Hutter <rhutter at posteo.de> wrote:
> > >  
> > >> To be fair: Out of the 3 people giving their opinion before your
> > >> last email, paul (in your first email you didn't seem to give an
> > >> opinion but rather a hint on a kind of workaround), 2 were saying
> > >> that they dislike the current way it works.  
> > >
> > > The way it currently works is not an accident, but represents the
> > > dev team's opinion on how it should work. Since this is implicit,
> > > I didn't think it was worth a mention, but since you seem to be
> > > not taking that into consideration, I've mentioned it now.
> > >  
> > 
> > Ok, the dev team thinks it's how it should work but haven't
> > explained why they think so:
> > 
> > Why does the dev team think that  collapsing the entire thread and
> > press SUPPR on the root email of the thread should delete the root
> > email only and let a partially broken thread (ie a thread with
> > missing emails) ?
> > 
> > Seriously I really don't see any reason this is superior than delete
> > the whole thread.
> 
> I tend to agree, things could be better there, like del key to delete
> a message, shift+del to delete message and children, with no regards
> to what's collapsed or not, root node or whatever. Usually, GUI's may
> behave differently when handling collapsed things, but doing things in
> blind mode without a warning about deleting a tree or a specific
> action (shift+.., ctrl+..) scares me a bit.

I like that, but how about; 
del -> message to trash
shift+del -> really delete message
ctl+del -> message and children to trash
shift+ctl+del -> really delete message and children. 
Maybe ask for confirmation the first time.  

-- 
testing
bP