Re: What are advantages and disadvantages of RMAIL and mh-rmail ?

[wgreenhouse@riseup.net (W. Greenhouse)]

Don Saklad <dsaklad@gnu.org> writes:

> What are advantages and disadvantages of RMAIL and mh-rmail ?

RMAIL and MH-E (of which the function `mh-rmail' is a part) are similar
in that they assume the presence of a traditional UNIX mail system (Mail
Transfer Agents such as sendmail, postfix, exim, etc., and possibly Mail
Delivery Agents such as procmail, maildrop, etc.).  That is because they
normally only handle mail arriving at the system mailbox for the user,
leaving it to the user or the user's site administrator to arrange for
that to happen.  There are exceptions to this; RMAIL offers some basic
ability to deal with POP3 on its own, and some MH implementations
(notably the GNU Mailutils version of MH) can handle POP and IMAP in a
simple fashion.

They differ in the handling and formatting of the mail once it arrives.
RMAIL stores the mail in one big file, similar to UNIX mbox format and
usable with external tools designed for that.  MH-E deals with the mail
in a directory structure, one message per file and one directory per
mail folder.  Unlike RMAIL, which manages its own mailbox on its own,
MH-E is a frontend and set of Lisp functions to exploit the RAND Mail
Handler (mh)[1], one of the first mail systems to use a "message-as-file"
concept.  Modern mh implementations suitable for use with MH-E include
nmh and GNU Mailutils.  With MH-E you have the option of using the MH
implementation's command line tools to manipulate your mail as well as
using Emacs.

So, in short, you need an MH to use MH-E, which uses the legacy mh mail
format.  Or you can use RMAIL which also uses a legacy format.  Or, you
can use Gnus as `read-mail-command' instead of either of these, which is
my own preference.

Footnotes: 
[1]  http://rand-mh.sourceforge.net/

--
Regards,
WGG