Franz Fellner wrote: > On Thu, 16 Oct 2014 15:30:11 -0700, Ian Main <imain@redhat.com> wrote: > > Franz Fellner wrote: > > > > > > > > > > I'm starting to realize that I could default to using 'enter' to both > > > > open URI's and view attachments. Any other ideas welcome. > > > > > > - make some of the functions public so users can bind them to keys they > > > want > > > - introduce show_[prev,next]_unread_msg, probably factor out > > > "show_scroll_to_msg()" and implement prev/next msg with that? > > > > You know I was just reading a thread and hit 'a' to archive a message > > within the thread and for some reason it hung for a few seconds.. but it > > made me think we should be able to press 'a' (or whatever archive key) > > and have it archive and move to next unread. That seems to me to be a > > good behavior with a natural flow. > > Yes, sounds good. > Today I thought of a more generic approach: > pass a query that a message should match. > :notmuch-show-next-message tag:unread # go to the next unread > message in thread next unread message in thread, and if there are no more in thread, the next unread message in the next thread? > :notmuch-show-next-message from:"Ian Main" # got to the next > message in thread sent by "Ian Main" > ... > The user then can map keys to custom commands. > > Good or too generic? I'm not sure how easy it would be to invoke that.. if you think you'd use it I'd say it's fine but I don't think I would. > > Furthermore I thought of rearranging the sources: > - Put files in Vundle/*-compatible dirs, would simplify development. > (Currently I have several files symlinked from notmuch-src into .vim) > - Put the "heavy" ruby code into its own file. Today I got an error I > could not deal with because line numbers did not match. Also I read > about the possibility that the ruby interpreter might optimize only > seperate files. This could lead to better performance. > Yes, that's a good idea too. I notice that the github version has a different layout. I think it is vundle compatible? I know it makes moving patches from the official repo to the github one challenging. Ian