Skip to content

Conversation

bgamari
Copy link
Contributor

@bgamari bgamari commented Apr 30, 2020

Previously this would throw an exception in the (not uncommon) case of a
truncated eventlog. Now we simply declare the thread to stop running at
the time of the last event that we saw.

@Mikolaj
Copy link
Member

Mikolaj commented Apr 30, 2020

But aren't we lying? Perhaps warnings on stderr (via unsafePerformIO?) would make the user aware the output we provide is not accurate?

@bgamari
Copy link
Contributor Author

bgamari commented May 24, 2020

@Mikolaj I wouldn't say we are lying per se. However, we could indeed print a warning. I'll do so.

@bgamari bgamari force-pushed the wip/truncation-robustness branch from 1ca6509 to 952dcbc Compare May 24, 2020 17:42
Previously this would throw an exception in the (not uncommon) case of a
truncated eventlog. Now we simply declare the thread to stop running at
the time of the last event that we saw.
@bgamari bgamari force-pushed the wip/truncation-robustness branch from 952dcbc to 630a4f7 Compare May 24, 2020 17:45
@Mikolaj
Copy link
Member

Mikolaj commented May 24, 2020

Haven't tested, but LGTM.

@maoe
Copy link
Member

maoe commented May 25, 2020

This PR slipped under my rader somehow. Sorry. I’ll take a look tonight.

@maoe
Copy link
Member

maoe commented May 25, 2020

Ah I just remembered the context. I did recognize this and found AppVeyor was failing for some reason. It seemed this PR was innocent though. Anyway I’ll take a look again tonight.

@maoe
Copy link
Member

maoe commented May 25, 2020

LGTM. The AppVeyor are failing when installing library dependencies with pacman so this PR is not responsible.

@maoe maoe merged commit e808c2d into haskell:master May 25, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants