Re: auto-dbm quirks

[Daniele Arena <]

Hi Steven and All the db-wg,

The feature you ask for (parsing MIME messages sent to auto-dbm@localhost)
is currently being implemented in Version 3 of the RIPE Database.

Version 3 is a complete re-implementation of the RIPE Database software;
it has been agreed not to implement new features in the current version of
the software.

Please see http://www.ripe.net/ripencc/pub-services/db/reimp/ for more
information about the re-implementation project, and particularly see the
"Transition to RIPE DB v3.0" for a description of new features and a
timeline of the implementation.

About the MIME parser itself, it will be able to parse
multipart/alternative content-types (discarding text/html and keeping
text/plain), as well as other content-types: application/pgp,
multipart/mixed, multipart/digest, message/rfc822 and multipart/signed.
These are the supported MIME content-types that have been presented at the
RIPE37 Meeting in the DB Working Group.

Concerning your query about encodings, we decided not to support any type
of encoding. But we are eager to open a discussion if the community feels
that it is desirable to support encodings.

Cheers,

Daniele.

--------------------------------------------------------------------------
Daniele Arena			RIPE NCC - Database Group
phone  : +31 20 535 4444	Singel 258
fax    : +31 20 535 4445	1016AB Amsterdam
e-mail : daniele@localhost	The Netherlands

On Thu, 21 Sep 2000, Steven Bakker wrote:

> Hi folks,
> 
>   A few people have asked me to help them out with a nasty problem they're
> having.  They are stuck behind a braindead ms-exchange server and corporate
> standards apparently prevent the use of SMTP, POP or IMAP.  Hence, these
> people have to communicate using the Exchange protocol.  Now, when the
> Exchange server has to send mail to the real^H^H^H^Houtside world, it will
> use a pesky "connector". And here's the catch: that connector has been
> configured to send mail as both plain text _and_ HTML, resulting in, you
> guessed it, one of those wonderful "multipart/alternative" hybrid freaks
> (can you tell I'm not a fan?).
> 
> So, what's the big deal, you ask?  These friends of mine are trying to send
> updates to the RIPE database, making do with the crummy mail software
> they have at their disposal.  Unfortunately, the auto-dbm robot does not
> understand multipart/alternative, nor any other MIME encoding.  In fact,
> as far as I can tell from its diagnostics, it completely ignores any
> Content-Type: header and simply treats the message as plain text.
> 
> Now, I agree, the Exchange mail setup leaves a lot to be desired, but the
> mail it sends out _is_ standards compliant.  The auto-dbm is the one that's
> non-compliant.  There are two ways that I can see to fix this:
> 
>  1. Look for a content-type:- header and INSIST on text/plain.
> 
>     This doesn't solve my friends' problems, but it would at least
>     provide more intelligent diagnostics.
> 
>  2. Do some MIME parsing, allowing both text/plain and multipart/alternative.
> 
>     In the case of multipart/alternative, INSIST on a text/plain part.
> 
>     Maybe even decode base64 or quoted-printable?  Would be handy in case 
>     there are more crummy mail systems out there...
> 
> I don't think the MIME parsing would be so difficult.  Decoding base64 or 
> quoted-printable shouldn't be that hard either.
> 
> Is this something that can be done quickly, or at least put on a priority
> list?  I foresee that in the near future more people will be forced to 
> embrace the MS-Office dogma and suddenly find they cannot update their RIPE 
> objects anymore.
> 
> Cheers,
> Steven
> 
> 
>