samth
2017-7-19 14:18:35

@stamourv a new weird difference: https://plt.eecs.northwestern.edu/release-pkg-build/server/built/deps/dssl2.txt is only happening on the release build


samth
2017-7-19 14:18:50

also, css passed on HEAD, still failed on the release


stamourv
2017-7-19 16:19:54

@samth: Re dssl2: I see that @tov made changes to the deps between the pre-release pkg-build run and the HEAD pkg-build run. Let’s see what the next pre-release pkg-build says.


samth
2017-7-19 16:20:05

ah ok


stamourv
2017-7-19 16:20:35

@samth: Re css: Wargrey got back to me, looking more into it is on my to-do list.


samth
2017-7-19 16:20:59

also, I also ran the turnstile tests, they passed, I took it off the checklist


stamourv
2017-7-19 16:20:59

If you’d like to investigate too, I can forward you the info.


samth
2017-7-19 16:21:07

I won’t have time


samth
2017-7-19 16:22:22

But I ran some tests and checked them off


tov
2017-7-19 16:22:25

@stamourv indeed I added deps yesterday. Did I do something wrong?


stamourv
2017-7-19 16:24:17

@tov: I don’t think so? A pre-release pkg-build (see link from samth above) that happened before your changes had some deps errors. Your changes may have fixed them.


stamourv
2017-7-19 16:24:39

But we’ll have to wait for the next pre-release pkg-build to know for sure.


tov
2017-7-19 16:25:12

Oh, okay, I think I fixed them. Cheers!


stamourv
2017-7-19 16:25:44

:)


notjack
2017-7-20 05:40:18

Would it be bad to spawn a thread in a plumber flush callback that frees some resource and not wait on the thread before removing the flush callback registration? Could racket exit after the plumber is flushed but before the background thread finishes?