Lars Petter Hauge

Perhaps add a warning to segyio-crop when no traces have been copied Considering that normal-small.sgy have xlines between 20 and 24, this will not copy any traces: `../build/applications/segyio-crop normal-small.sgy test.sgy...

I think it makes sense to first send the structure of entities and then use the flat structure to send events/logs on a per entity level. The client will have...

> > Either the `experiment server` needs to hold all chunks of logs sent to clients > > Yes, I think this is probably best. There are multiple ways to...

We've looked a bit into how the current setup is made in order to better see how the next Tracker will be made. The `sequencediagram` presented here is a draft...

Thanks for picking up on the continuation of https://github.com/equinor/ert/pull/3438. With this implementation though we are adding a fair bit of duplicated code for the `run` function, from what I can...

Hi @ming-origen, I thought there were some examples within libecl, but didn't see them right away. We do however create dummy `UNRST` files in other projects for unit tests: https://github.com/equinor/semeio/blob/9be403dcdb2ef5803bc1a1a9d1e89a07fcb745ab/tests/jobs/overburden_timeshift/ots_util.py#L19...

Hi @bartoli Unfortunately I don't believe there is an option to do anything other than to populate the entire grid model on instantiation. I agree that it would be useful...

What do you think?: `def _publish_event(self) -> def _event_publisher(self)` As this is a function that is designed to continuously publish events. Not a biggie, just a small suggestion..

Thanks for the PR, just had some small comments/questions :)

Could you please rephrase the commit message a bit? ("but" -> "by"?, should also be less than 80 chars or something, otherwise it's wrapped..)