Re: Where do you want to go tomorrow?

[Simon Josefsson <jas@pdc.kth.se>]

Vadim Gutnik <gutnik@mtl.mit.edu> writes:

> > * Faster look for new mail (asynchronous?)
> 
> I think this is very hard. Leave this for last.

Yup. I don't think this is hard though, just select all subscribed
groups and fetch uid's of lowest/highest article in a function that is
run with the idle timer (<50 lines of code?). One problem is we can't
tell gnus about this because backend can't initiate anything. But we
can cheat..

> > * Cooperation with Gnus Agent, ie disconnected mode
> 
> This could be useful, but I don't use Agent...

Same here. But it's part of gnus and adding a new component to gnus
shouldn't break others. Well, in theory...

> > Feel free to speak your mind.
> 
> The main thing I'd like is for subscribing/unsubscribing to be a lot
> cleaner.

*Browse server*. Hint, hint, nudge, nudge. :-)

> So, ideally, in the *Group* buffer, I could "j nnimap:newimapgroup",
> then "u" to subscribe it and create it.

Hm. No. Theese commands don't create a group, they modify some Gnus
structures to make it look like a group exist, but if you try to enter
the group you'll get an error because the group doesn't exist in the
backend. Correct me if I'm wrong, didn't browse any code but did some
simple testing (`j nnml:saiufhad `u' `RET' => error).

Create groups with `G m' (which is buggy, but that's another story).

> Just like it works in other back ends. There's no reason nnimap
> should be different, is there?

Nope.