W3C home > Mailing lists > Public > ietf-discuss@w3.org > February 2001

Re: Standards for machine translation in e-mail andnetnews

From: Martin Duerst <duerst@w3.org>
Date: Wed, 14 Feb 2001 12:20:04 +0900
Message-Id: <4.2.0.58.J.20010214115732.03cc7100@sh.w3.mag.keio.ac.jp>
To: Jacob Palme <jpalme@dsv.su.se>, WG-I18N@TERENA.NL, IETF mailing list on MIME and e-mail <IETF-822@imc.org>, <discuss@apps.ietf.org>
Cc: "Phil Scanlan" <phil@worldlingo.com>, <ashuttleworth@sdlintl.com>, Keld J (wrong string) xrn Simonsen <keld@dkuug.dk>, Brian Garr <bgarr@us.ibm.com>, Nick Shelness <Shelness@lotus.com>, "Pierre-Yves Focou. \"PY FOUCOU\"" <foucou@systran.fr>, Francisco Sparta <franspar@samerica.com>, Andrew Shuttleworth <ashuttleworth@sdlintl.com>, Portal Coordination <portal-coord@TERENA.NL>, REIS Mailing List <reis@TERENA.NL>, Thierry_Mayeur@lotus.com
Hello Jacob,

I have just had a short look at this. A few comments.

I don't understand what multipart/choices is for. There are
implementation problems for multipart/alternative, but they
don't get fixed by proposing something new. They can (and
should) just be fixed asap, which is easier than to implement
something new.

Please also note that the 'selector' parameter is very similar
to the 'differences' parameter proposed in RFC 1766
(e.g. http://www.innosoft.com/rfc/rfc1766.html), which yet
has to take on. It would at least be worthwhile to use the
same name.

For ftp://ftp.dsv.su.se/users/jpalme/draft-palme-email-translation-02.txt,
it says "All of these methods of transmitting information is based on the
assumption that all language versions are ready and available when a
message is sent.". This is clearly wrong, because it doesn't
apply to http. It may also be worth mentioning that some formats,
in particular SVG (http://www.w3.org/TR/SVG/struct.html#SwitchElement)
and SMIL (http://www.w3.org/TR/REC-smil/#switch), have a <switch>
construct that allows to display one of the languages contained
in the document.

The provisions regarding the "Content-Translator" header field:
 >>>>
   The "Content-Translator" header field indicates who made the
   translation. When a translation is submitted, the "From" header field
   should still indicate the original author, but the "Content-Translator"
   header field can indicate who made the translation.
 >>>>
are very scary, because they will produce very misleading impression
on user agents that don't understand Content-Translator (namely
that the mail came from the original sender). Things should be
turned around, so that such an impression is not possible.

In general, I think that translations, in particular incremental
ones, fit much better with the Web paradigm than with the message
paradigm. It might be better to get some more actual experience
on the Web side rather than to try to force things into messaging.


Regards,   Martin.


At 11:09 01/02/13 +0100, Jacob Palme wrote:
>Work on standards for language-translation in e-mail and netnews can
>take place in the mailing list LANGTRANS@SU.SE.
>
>To subscribe to this mailing list, send a message to LISTSERV@SU.SE
>which contains the text SUB[SCRIBE] LANGTRANS
>
>When your subscription has been accepted, please write a short message
>to the list describing yourself and your interest in languate translation 
>standards.
>
>More information and the present IETF drafts can be downloaded from
>http://dsv.su.se/jpalme/ietf/jp-ietf-home.html#translation
>
>The web page above earlier gave an incorrect description of how
>to join the mailing list. This has been corrected.
>--
>Jacob Palme <jpalme@dsv.su.se> (Stockholm University and KTH)
>for more info see URL: http://www.dsv.su.se/jpalme/
Received on Wednesday, 14 February 2001 00:35:23 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:08:11 UTC