W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2009

Re: Content Sniffing impact on HTTPbis - #155

From: Adam Barth <w3c@adambarth.com>
Date: Sat, 13 Jun 2009 10:16:29 -0700
Message-ID: <7789133a0906131016t7ac56a3m792f419425ec8a72@mail.gmail.com>
To: Jamie Lokier <jamie@shareable.org>
Cc: David Morris <dwm@xpasc.com>, ietf-http-wg@w3.org
On Sat, Jun 13, 2009 at 9:56 AM, Jamie Lokier<jamie@shareable.org> wrote:
> Does the sniffing document not apply to browsers looking at content on
> a local disk (therefore with no Content-Type), or does this mean it
> recommends sniffing the content without looking at the filename on the
> local disk?

I haven't investigated this question in detail, but I suspect the
answer will vary by browser.  There is very little interoperability
between browsers when interacting with the file system.

> I'm pretty sure Firefox and the like look at the file extension when
> looking at content found on local disk.  But surely it does sniffing
> at well, on local disk files?

Do you have evidence for this belief?  It should be fairly easy to
determine by looking at the source code.

> Does the sniffing document not apply at all in that case, or is there
> a different sniffing algorithm used which remains undocumented?

There is only one sniffing algorithm.  The question is only whether
its applied in this case.  More precisely, the question is whether the
"file" protocol handler assigns a media type using OS-specific
functionality before handing the response off to the next layer, where
content sniffing is performed on various media types (e.g., the empty
media type).

Adam
Received on Saturday, 13 June 2009 17:17:33 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:03 GMT