Allan McRae firstname.lastname@example.org on Tue, 2022/02/01 00:33:
On 31/1/22 21:42, Christian Hesse wrote:
Wondering why there's no feedback to this patch... Would be nice to receive an ACK or NACK at least on whether or not this is the way to go.
When two versions were posted within a day, I figured I'd wait a week for v3 before reviewing... And here we are!
Waiting too long with the review could result in me pushing even more patches. ;->
It is in the review queue - but so are patches from 8 months ago, so don't expect it reviewed any time soon.
Oh dear... I hope I can remember the details once review happens. :)
Just to remember for myself: Initially I though about making 'alpm_list_t *servers' point to a 'struct alpm_server_t' containing the server and cache state. This does not work well as we have cases in the code where we handle an url that is no longer associated with a server in the list. I think storing the cache state in modified 'struct server_error_count' is the only way to go as we can always extract the server hostname from url.