Re: Proposal: Cookie Priorities

[Amos Jeffries <[e-mail]>]

On 4/03/2016 1:52 p.m., Matthew Kerwin wrote:
> On 04/03/2016 10:09 AM, "Mark Nottingham" wrote:
>>
>> Thanks, Mike.
>>
>> As I understand it, this is already implemented in one browser, which is
> good in that we're looking for implementation.
>>
>> What do folks -- both other browser implementers and site folks -- think
> about this?
>>
> 
> It's a practical issue with a simple (demonstrated) improvement. And hints
> like this usually fall under "handy at best, harmless at worst" so I see no
> harm in and of itself.
> 
> That said, maybe I've been listening to PHK for too long, but I wonder if
> all these cookie patches aren't just putting band-aids over a fundamentally
> flawed system. How many (and how big) cookies do you need to receive before
> this priority comes to the fore?
> 

Oh, its absolutely a battered, abused, and broken mechanism. I am
doubtful that many would even attempt to argue the contrary.

My take has been from the beginning that we should start pruning away
pieces of Cookie to prevent bad usages a much as possible.

This particular baindaid proposal looks like it might double as a nice
way to allow any recipient or relay to proactively prune away the
low-priority Cookies in traffic when bandwidth gets overloaded. Can it
be the beginning of an efficient Cookie deletion mechanism?

Amos