Ted Unangst
|
2518c4be4a
|
touch up redeliver loop
|
2 years ago |
|
Ted Unangst
|
7eef6ae5d0
|
flexible logging
|
2 years ago |
|
Ted Unangst
|
69fea996ad
|
don't need to retry low prio communications
|
4 years ago |
|
Ted Unangst
|
0e1ec60422
|
clear remaining retries after a final delivery failure
|
4 years ago |
|
Ted Unangst
|
b2da24a9f8
|
let's do 40 deliveries at a time
|
4 years ago |
|
Ted Unangst
|
4f386cf558
|
put common init code in honk.go
|
5 years ago |
|
Ted Unangst
|
f27a4d9ce2
|
new day, new libs. use gate library.
|
5 years ago |
|
Ted Unangst
|
3f67140c76
|
rework delivery and key loading to use userid
|
5 years ago |
|
Ted Unangst
|
ec80662d2c
|
delivery chan should buffer one
|
5 years ago |
|
Ted Unangst
|
b1e42fdae9
|
use cache interface for some more things
|
5 years ago |
|
Ted Unangst
|
2213f0e935
|
make sure every rows is closed
|
5 years ago |
|
Ted Unangst
|
fbe47a0895
|
every database exec must have error checking
|
5 years ago |
|
Ted Unangst
|
d6929a5644
|
handle two more db error conditions
|
5 years ago |
|
Ted Unangst
|
fd5d297188
|
add another retry at 4 hours
|
5 years ago |
|
Ted Unangst
|
d6e536fed6
|
increase max concurrent deliveries to 20
|
5 years ago |
|
Ted Unangst
|
0faf2b0556
|
allow replying to private messages with private replies
|
5 years ago |
|
Ted Unangst
|
aed70531ff
|
reduce the deliverator hysteresis to reduce clumping
|
5 years ago |
|
Ted Unangst
|
7d538d465f
|
put a limit on how many deliveries we actually send it in parallel
|
5 years ago |
|
Ted Unangst
|
0997e57525
|
on redelivery, try shared again too
|
5 years ago |
|
Ted Unangst
|
6ec4352d69
|
use shared inbox for delivery
|
5 years ago |
|
Ted Unangst
|
ece50727f6
|
math/rand is notrand. seed it with time.
|
5 years ago |
|
Ted Unangst
|
bf323bd0e5
|
the spice must flow. redeliverate with extreme prejudice.
|
5 years ago |
|
Ted Unangst
|
b24f484a84
|
further prep for retries
|
5 years ago |
