W3C home > Mailing lists > Public > public-multilingualweb-lt@w3.org > April 2012

RE: [all] suggestions for consolidating requirements

From: Moritz Hellwig <Moritz.Hellwig@cocomore.com>
Date: Mon, 30 Apr 2012 11:39:12 +0200
Message-ID: <A5F4BCC8EDECF74D97DBCEF820F7269F9475E9@cocont10.office.cocomore.com>
To: "Arle Lommel" <arle.lommel@gmail.com>
Cc: "David Lewis" <dave.lewis@cs.tcd.ie>, <public-multilingualweb-lt@w3.org>
There are indeed many reasons for stripping out certain metadata before the
end user will see the content. Many companies will not want to present
certain MultilingualWeb-LT meta-information to their competitors for example.


-----Original Message-----
From: Arle Lommel [mailto:arle.lommel@gmail.com]
Sent: Fri 27/04/2012 20:51
To: Moritz Hellwig
Cc: David Lewis; public-multilingualweb-lt@w3.org
Subject: Re: [all] suggestions for consolidating requirements
 
> We said metadata should only be added never be removed, right?


I'd disagree with this principle. There are times it is entirely appropriate
to strip a metadata item in a process. Mo's example is a good one: dropRule
would delete itself when used properly. But there are other cases as well.
For example, if a process marks up a text with term candidates and a
subsequent process rejects them, I would expect that process to delete them,
not leave them in. (One could argue it should alter them and mark that it
rejected them, but for most purposes simply stripping the tag/attribute would
make more sense since downstream processes probably have no need to know that
an upstream process incorrectly identified a term candidate that another
process cleared up.)

What I do think we should say is that metadata should not be removed without
reason. What we don't want is a process to simply drop metadata it doesn't
understand, but for a process that knows what it is doing to remove it is
another thing entirely.

-Arle
Received on Monday, 30 April 2012 09:39:52 UTC

This archive was generated by hypermail 2.3.1 : Sunday, 9 June 2013 00:24:55 UTC