Re: [HASMAT] Parsing Content-Type

[Adam Barth <[e-mail]>]

On Tue, Jul 13, 2010 at 4:04 AM, Julian Reschke <[e-mail]> wrote:
> On 13.07.2010 12:21, Anne van Kesteren wrote:
>> I was wondering if http://tools.ietf.org/html/draft-abarth-mime-sniff
>> should not contain the error handling rules for parsing the Content-Type
>> as well. It seems potentially dangerous if one user agent were to treat
>> the value "text/html;" as bogus and fallback to "text/plain" whereas
>> another would treat it as "text/html".

The draft used to contain the parsing rules for extracting the
charset, etc.  My guess is that this draft will be controversial
enough as is.

> If the *intent* was to send "text/html" then the only problem I see is that
> some recipients would see plain text.

Indeed.  That sort of mis-match is likely to be a bad user experience
if it occurs.

> Speaking of which: why would the fallback be "text/plain"?

It depends on how the UA treats text/potato.  It would not be
unreasonable to treat text/potato similarly to text/plain, depending
on the UA's purpose.

> If the intent was different, why send the header in the first place?

That's a question for the server operator.  I can look up this
particular sequence of bytes, but server do end up sending a bunch of
crazy things for whatever reason.  The UA is then faced with the
question of how to interpret the response.

Adam