Re: Call for adoption: draft-nottingham-http-cache-groups

I'm also in support of adoption.

On Mon, 23 Oct 2023 at 22:52, Patrick Meenan <patmeenan@gmail.com> wrote:

> I'm in support of adoption to shake out the details.
>
> I don't know that it makes sense for client caches to do anything with it
> but I can see where a standard way to flush portions of intermediary caches
> would simplify building apps that work across CDNs. At least on the Chrome
> side, our caches are keyed by URL and purging by something like a group
> would either be slow (with a full scan) or require a rewrite of the
> indexing.  All good things to discuss in the wg though.
>

We have a similar architecture in Firefox, but I believe we can get away
without rewriting the index by introducing an additional hashtable (which
may or not be persisted across restarts) that holds the revalidation status
of a cache group. Subsequent loads would consult the hashtable to determine
if a cache entry needs revalidation depending on its cache group.


> On Fri, Oct 13, 2023 at 10:42 PM Tommy Pauly <tpauly@apple.com> wrote:
>
>> Hello HTTP WG,
>>
>> This email starts a call for adoption
>> for draft-nottingham-http-cache-groups, which was presented and discussed
>> at the IETF 117 meeting.
>>
>> The current version of the document is here:
>> https://www.ietf.org/archive/id/draft-nottingham-http-cache-groups-00.html
>>
>> Please review the document and reply with your reviews and thoughts on if
>> this is something this WG should take on.
>>
>> This call will last until *Monday, October 30*.
>>
>> Best,
>> Tommy
>>
>