Re: Is mutt extensible with a programming language?

2011-06-25 Thread Dave Dodge
On Fri, Jun 24, 2011 at 11:52:39PM +0800, XeCycle wrote:
> I think it'd be nice to integrate some language
> to muttrc, thus making customization easier. (I prefer perl
> for this task~)
> 
> Or --- Is there a fork of mutt that already support this?

I use a mutt+Lua combination, but I only put Lua hooks into a few
specific spots so it's pretty limited in what it can do.  For example
I use Lua functions to label the subject lines for messages from a
whitelist of senders.

  -Dave Dodge










Re: Is mutt extensible with a programming language?

2011-06-25 Thread lee
XeCycle  writes:

> On Sat, Jun 25, 2011 at 03:04:09AM +0200, lee wrote:
>> 
>> Gnus has one big disadvantage: It can be awfully slow.
>
> Yes... It's more than *awfully* slow. Now I use it as a
> newsreader, subscribing several groups at gmane and also
> some RSS feeds. Mixing them with mail will probably be even
> more slow, which is also a reason for my insisting on mutt.

So far, I have found only two things gnus is slow with:


* processing incoming email is *awfully* slow

* creating a summary buffer when there are *many* unread messages to
  show


The second one isn't really an issue. The first one is currently very
annoying because I still have a lot of messages that need to be
processed in the process of switching from maildir to nnml --- and
that's mainly my fault because I started doing it the wrong way. Still
that should be much faster. It's not a fair comparison, though, because
it's something mutt doesn't do.

>> Mutt is awesome and totally reliable; gnus is awesome and extremely
>> powerful. (I can't tell yet how reliable gnus is.)
>
> Recently I'm also trying Chromium; I use Firefox with
> Pentadactyl now. In Chromium, I didn't find anything
> comparable to Vimperator/Pentadactyl, but it's much faster
> than Firefox. Well, these comparisons are similar.

Chromium is a MUA? I tried it and found it can't even display PDF
documents and cannot be configured to do anything but display a web
page. It isn't noticeably faster than konqueror (with webkit) or
seamonkey.


Re: Revisiting: application/pgp-signature is unsupported

2011-06-25 Thread Brendan Cully
On Friday, 24 June 2011 at 00:00, Lars Hecking wrote:
>  
> >  I am facing the same problem now, but may have a bit more information.
>  
>  Original message has a multipart/signed structure with a qp text and
>  a signature part, all of which is wrapped into a multipart/mixed
>  structure, presumably done by the list server for the purpose of adding
>  a (text, 7-bit) footer. Exchange flattens the struture into a multipart/
>  mixed with a qp text part, a application/pgp-signature part in base64,
>  and another base64 text part.
> 
>  I presume mutt needs multipart/signed.

Sounds like a reasonable guess. I don't suppose you could send in a
message that has been mangled this way? It would make it easier to
test.


Re: mutt shows wrong size of message on mime messages

2011-06-25 Thread Brendan Cully
On Wednesday, 22 June 2011 at 08:35, jurri...@rivierenland.xs4all.nl wrote:
> From: David Champion 
> Date: Tue, Jun 21, 2011 at 03:25:46PM -0500
> > * On 21 Jun 2011, Jurriaan wrote: 
> > > on screen, I see things like
> > > 
> > >  50 21-06-11 Xx Xxxx (  1.2K) Xx  x
> > >  51 21-06-11 Xxx (13433K) Xxxx
> > > 
> > > The sizes of messages with mime components (mostly html messages) are
> > > much to big. Given that this is a router, without strace or gdb (well,
> > 
> > Are you certain?  That doesn't look too surprising for an HTML message
> > with inline images.
> 
> If I save the message to disk, it's only 13 kilobytes. There are no
> inline images, it a text-part and the same as html, but no images. I'll
> append the headers below. I'm not too sure I'd be happy receiving 133
> megabytes messages...
> 
> > 
> > Have you tried copying it to another machine and running mutt on it
> > there?
> 
> Yes, on a regular AMD64 machine running Debian Unstable and mutt
> 1.5.21-5 from Debian, it shows as 13K in the index.
> There is no Content-Length field in the headers at all. Still, I think
> mutt is doing something to determine the length of the message - perhaps
> something else that just looking at the file.

Might be related to this patch:
https://dev.mutt.org/trac/changeset/5d96f24efa85

If you can rebuild mutt on your router, try reverting that.