- From: Lou Montulli <montulli@mozilla.com>
- Date: Mon, 13 Nov 1995 18:38:46 -0800
- To: Gavin Nicol <gtn@ebt.com>
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Gavin Nicol wrote:
>
> >> incomplete. This is great for small pages, but if you try fetching
> >> small peices of a 5MB document, it makes no sense.
> >
> >What makes you think it makes no sense.
>
> Give a 5MB HTML document to Netscape Navigator. For certain media
> types, and HTML is one, you can request peices (not byte ranges), and
> do not need to reparse the data. For example, it makes more sense to
> ask for single elements from HTML, than a byte range.
I'm not disagreeing with your application, but I am saying
that byteranges can make sense for any object, as demonstrated
by partial caching.
>
> >> Byte ranges are a lazy replacement for a general naming mechanism.
> >
> >What's your naming scheme for JPEG files? How about AVI
> >video streams? Bytes are already a general purpose naming scheme,
> >and they have been used for a number of years. There is no
> >need to invent another one.
>
> Byte ranges are not sufficient for many data types, and indeed, are
> potentially harmful. A general naming mechanism would allow byte range
> naming, as well as things like asking for elements m to n in an HTML
> document.
Ari's proposal allows for generalizable requests. Bytes is simply
one of the range request methods.
:lou
--
Lou Montulli http://www.netscape.com/people/montulli/
Netscape Communications Corp.
Received on Monday, 13 November 1995 18:44:59 UTC