upmessage

up

major point of discussion around the nonce

we should hold off on this until @sashank makes forward progress with the date-time partitioning next week (tied to the scheduler). i think "Execution time" sounds like the time that the computation is occuring. I.e. if the computation were happening on oct-1 but processing the data from sept 15, I would guess that "Execution time" means oct 1, which is the opposite of this as I understand it.

would prefer to see startswith shenanigans go away

this is much less invasive than anticipated 👏🏻👏🏻👏🏻

with your new stuff I don't think this is necessary

use diff approach per in-person convo

this is pretty great. i just think we can just go for it and get rid of bool flag

opt-in bool just because it's experimental? seems like we want this or we don't.

looks good but some tests would be good

up

can you post screenshot?

thx my b.

upmessage

up

This is what happens on the runs page on master right now

up

up

Would love to see jaffle test working with this in follow on diff.

very nice

our first off-by-one error!

Can you also add that this fixes this issue in the release notes for 0.6.1?

you can be so clever with buildkite

good stuff. a couple minor edits

up

cool

this is better

seems reasonable

yeah let's not mutate requirements.txt

up

yeah i started an issue to track https://github.com/dagster-io/dagster/issues/1804