RE: [Exim] Maildirsize files: opinions sought

Góra strony
Delete this message
Reply to this message
Autor: Philip Hazel
Data:  
Dla: Eli
CC: 'Exim-users'
Temat: RE: [Exim] Maildirsize files: opinions sought
On Fri, 13 Feb 2004, Eli wrote:

> > Personally, I like it to count everything. I assume the default
> > of the option is an empty expression which means no regex processing
> > is being done and everything is counted?
>
> Count my vote for that as well. I'd actually think rather than a regex only
> variable, could it not be as was initially mentioned, an expanded string
> variable that can contain regexes, but could just be other variables or
> normal strings?


I am not trying to add new features. I am just trying to implement
maildirsize, as it is currently used in Maildir++ folders by existing
MUAs and other software.

> It would also mean if you don't want to count trash, you could just simply
> put "Trash" (or would it be ".Trash") as an item in the list rather than
> having to make it a regular expression.


Scanning lists takes time. If you include all the options in a regular
expression, I think things will be faster. After all, logically, a:b:c
in a list is equivalent to ^(a|b|c)$ in a regex - not *very* different.

OK, a list could be much more extensive and do all kinds of complicated
lookups. But are these *really* needed? What I understood was that
people wanted compatible support for maildirsize. So that's what I'm
currently doing. I'm short of time at the moment; pushing this design
further isn't important enough just now, I'm afraid.

> Adding some flexibility to this now
> might take a bit longer, but would probably save you from wishlist items
> down the road :)


The whole maildirsize thing is a huge hack. What should happen is that a
much better scheme comes along rather than pushing this hack further.
Designing it isn't something I have time for, I'm afraid.

--
Philip Hazel            University of Cambridge Computing Service,
ph10@???      Cambridge, England. Phone: +44 1223 334714.
Get the Exim 4 book:    http://www.uit.co.uk/exim-book