Re: short-time client side caching and clock skew [#212]

Tracking this as
  http://trac.tools.ietf.org/wg/httpbis/trac/ticket/212


On 01/06/2010, at 10:50 PM, Jorrit Jongma wrote:

> I thought #29 was already in -09, due to the comments of the report. It
> appears that isn't so, so re-done, incorporating #29 / -latest would be:
> 
> 
> response_delay = response_time - request_time;
> corrected_age_value = age_value + response_delay;
> if (reponse_version_1_1 && !response_via_1_0) {
> 	corrected_initial_age = corrected_age_value;
> } else {
> 	apparent_age = max(0, response_time - date_value);
> 	corrected_initial_age = max(apparent_age, corrected_age_value);
> }	
> resident_time = now - response_time;
> current_age = corrected_initial_age + resident_time;
> 
> 
> Interesting to see that this issue was raised before. Hopefully it will
> finally be resolved.
> 
> Thanks,
> Jorrit
> 
> -----Original Message-----
> From: Mark Nottingham [mailto:mnot@mnot.net] 
> Sent: dinsdag 1 juni 2010 13:57
> To: Jorrit Jongma
> Cc: 'Henrik Nordström'; ietf-http-wg@w3.org
> Subject: Re: short-time client side caching and clock skew
> 
> See also:
>  http://www.w3.org/mid/0BD82EFC-9F0D-4122-89FE-E245E32AE22C@yahoo-inc.com
> 
> If you're going to suggest replacement text, please incorporate the changes
> in -latest (if you haven't already); see 
>  http://trac.tools.ietf.org/wg/httpbis/trac/ticket/29
> 
> Cheers,
> 
> 
> On 31/05/2010, at 11:04 PM, Jorrit Jongma wrote:
> 
>> Henrik,
>> 
>> Thank you for your insights and corrections.
>> 
>> If I understand you correctly, the following pseudo-code would cover all
> the cases, taking -09 as base, and completely bypassing date_value if
> possible:
>> 
>> if (response_version_1_1 && !response_via_1_0) {
>> 	corrected_initial_age = age_value + (response_time - request_time);
>> } else {
>> 	apparent_age = max(0, response_time - date_value);
>> 	corrected_received_age = max(apparent_age, age_value);
>> 	response_delay = response_time - request_time;
>> 	corrected_initial_age = corrected_received_age + response_delay;
>> }
>> resident_time = now - response_time;
>> current_age = correct_initial_age + resident_time;
>> 
>> I have tested current operation in a few browsers, and the results are not
> encouraging; if the above is judged correct it is my intent to submit bug
> tickets to some of these browser manufacturers and see if they will update.
>> 
>> Regards,
>> Jorrit
>> 
>> 
>> 
> 
> 
> --
> Mark Nottingham     http://www.mnot.net/
> 


--
Mark Nottingham     http://www.mnot.net/

Received on Wednesday, 2 June 2010 06:56:06 UTC