Fork me on GitHub
#re-frame
<
2017-08-19
>
lumpy14:08:30

@mikethompson from your answer above what is a level 1 subscription? Would that be a static value. Coulndt find it in the graphic

sandbags22:08:23

Anyone remember the handle of the gentleman who wrote re-frisk? I think he changed it and I can't think to what

mikethompson22:08:31

@lumpy See my notes above for "Level 1"

mikethompson22:08:44

And see the infographic. Described in both places. But I do mix the terms "Level 1" and "Layer 1"

sandbags22:08:25

also if anyone has any ideas about tracking down "re-frame: expected a vector, but got: null" followed by "re-frame: no subscription handler registered for: '', Returning a nil subscription". Neither appear to touch any of my app code.

sandbags22:08:06

seems to be going on somewhere in the re-frisk setup code but i've no idea why it's started to happen

lumpy22:08:10

ah, ok I see it now, thanks

mikethompson22:08:39

@sandbags I'd guess that means there's a (subscribe) somewhere?

mikethompson22:08:55

Ie. no vector provided as an arg

sandbags22:08:10

@mikethompson right, just not sure why that should suddenly be true for re-frisk

lumpy22:08:27

Layer 1 entire app-db, Layer 2 portions of app-db, Layer 3 multiple portions of app-db combined together. is that the gist of it?

mikethompson22:08:00

Layer 1 - app-db Layer 2 - extractors Layer 3 - derived values (from extracted values)

mikethompson22:08:23

Layer 3 sometimes involves time-consuming computation

lumpy22:08:48

got it, thanks for the clarification

sandbags22:08:13

the only change i think i made was adding re-frame-datatable perhaps there is some incompatibility

sandbags22:08:37

although it seems unlikely

mikethompson22:08:08

I'd put a breakpoint in the re-frame itself, at the line of code which produces the error message and have a look at the call stack. Or perhaps you already have ... and that's how you narrowed down to re-frisk?

sandbags22:08:47

re-frisk is implicated in the call-stack, it's happening inside of enable-re-frisk!

mikethompson22:08:45

That's the handle of Mr re-frisk .... or at least it used to be ... I think

sandbags22:08:02

yeah but he changed it

sandbags22:08:18

hrmm... okay... so removing re-frame-datatable appears to have fixed the problem

sandbags22:08:24

or it's some coincidence

sandbags22:08:12

hrmmm... put the dependency back in and the problem does not recur

sandbags22:08:18

stupid non-determinism

sandbags22:08:36

Ah, but requiring the namespace does trigger it

sandbags22:08:59

so, something about the way re-frame-datatable works is b0rking re-frisk

sandbags22:08:51

ok i see the errant subscribe but i'm not sure why it's broken. Presumably re-frisk is fishing around in kind->id->handler before its fully formed

sandbags22:08:59

the simple thing to do would just be patch in a nil test into the re-frisk code i guess

sandbags22:08:16

i'm not sure what the right thing to do is

sandbags23:08:15

Okay, logged an error for Andre... best I can do this evening