On 2013-05-16 12:48, Sergey Beryozkin wrote:
> Hi
>
> My understanding is that "EEE, dd MMM yyyy HH:mm:ss zzz" format
> correctly represents HTTP dates, Locale: US, TimeZone: GMT.
>
> Note the above format does not capture milliseconds and I'm not seeing
> any of related RFCs recommending the use of milliseconds in Date
> representations or at
>
> http://tools.ietf.org/html/draft-ietf-httpbis-p2-semantics-22#page-63
>
> I'm seeing a test failing due to mismatch in milliseconds (between the
> original Date used by the test on the client side and the one created
> from the date constructed as per the above format, I think when Date is
> being reconstructed from the above format, the current milliseconds are
> set on Date).
>
> Question: are implementations expected to include milliseconds ? I think
> the answer No, but asking just in case...
> ...
In the "Date" header field? No.
Best regards, Julian