[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#109171: Use Maildir format by default

[I know this thread is old.]

On Sun, Aug 19, 2001 at 08:31:58AM -0700, Aaron Lehmann wrote:
> I'm interested in performance differences. A big flaw of the mbox
> format is that every byte of the file must be read to extract headers
> of the messages .. i.e. display a list of the messages without
> necessarily showing any data. Does maildir employ a summarization
> technique or must an MUA open every single file to extract the mailbox
> headers?


Also note that mbox format need not be lossy -- MDA should quote with >
any line matching regexp "^>*From ", MUA should always remove one level
of quoting at display time.  Any software not following this is losing
information (and therefore buggy).

Also, for reference:

Date: 8 Jan 1997 01:15:39 -0000
Message-ID: <19970108011539.27537.qmail@koobera.math.uic.edu>
From: "D. J. Bernstein" <djb@koobera.math.uic.edu>
To: djb-qmail@koobera.math.uic.edu
Subject: Re: wishlist -- shared Maildir/

> Therefore, why not expand the Maildir definition to support the issues that
> an MUA has as well?

It's already adequate for most MUAs, with info serving the same role for
maildir as Status serves for mbox. (Yes, agreement between MUAs on the
info format would be nice. Yes, I will coordinate.)

It's not adequate for MH-style sequences. Handling sequences reliably
over NFS is exceedingly difficult; I certainly wouldn't want to try.

As a practical matter, I don't see one-file-per-message as a good format
for long-term message archiving. It's bitten by all of the filesystem's
traditional performance problems. If I were writing an MUA, I'd want an
archive format designed for low space, very fast lookup, and fast
append: e.g., each message gzipped individually, concatenated into a
single file, with a separate random-access index listing positions and



I think the koobera.math.uic.edu address is still good, but his
cannonical address is djb@cr.yp.to


Reply to: