W3C home > Mailing lists > Public > public-media-fragment@w3.org > January 2009

Re: ISSUE-2: What is the mime type of a media fragment? What is its relation with its parent resource?

From: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
Date: Wed, 28 Jan 2009 00:55:57 +1100
Message-ID: <2c0e02830901270555y2a31db72rd75398fa1ce40f58@mail.gmail.com>
To: RaphaŽl Troncy <Raphael.Troncy@cwi.nl>
Cc: Michael Hausenblas <michael.hausenblas@deri.org>, Media Fragment <public-media-fragment@w3.org>

On Wed, Jan 28, 2009 at 12:49 AM, RaphaŽl Troncy <Raphael.Troncy@cwi.nl> wrote:
> Hi Michael,
>
>> Though I'm not Yves, my take on it is:
>
> All opinions are more than welcome ...

I am terribly sorry if anybody should have felt exempt. I particularly
ask for Yves opinion because he has the http background. Most
creatinly all opinions count! Sorry!

> [skip]
>
>>> From RFC2046 we learn that the MIME Type Registrations for JPEG 2000
>>
>> (ISO/IEC15444) is specified in RFC3745 [15] where no fragments are
>> defined,
>> hence the general rules from RFC3986 apply.
>>
>> My conclusion: in order to obey the rules of the game, we'd need to update
>> all registries of targeted media types.
>
> First, that is not possible, as we have already discussed it, since for
> updating IANA registries, you need to 'own' the format, which obviously we
> don't. It seems to me insane to try to register the fragment semantics for
> the 100+ multimedia formats, without counting all new ones appearing every
> year ...

The fragment spec on the jpg would require this, but I agree: it is
not the best way to do this, in particular since it requires changes
to many formats.

> Second, it is from my point of view not necessary. XPointer lives without
> for a period of time without problem. I cannot find a registry for the SVG
> fragment? Is someone aware of that?
>
> Third, that was not really the question asked in the ISSUE, although, I
> agree it is not completely clear. The server specifies using http header
> what is the type of the resource is serving. Silvia's suggestion is to
> encode in the URI this type, for example, in the case we extract a keyframe
> from the video. Is it the only way to go?

I took that out of the referenced email discussion. I cannot really
come up with a better way of specifying it, though it seems silly.

Cheers,
Silvia.
Received on Tuesday, 27 January 2009 13:56:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 21 September 2011 12:13:32 GMT