Fork me on GitHub
#untangled
<
2016-06-20
>
tony.kay19:06:11

@jasonjckn: Client 0.5.4-SNAPSHOT now supports turning off load markers

tony.kay19:06:22

just add :marker false to load-data

jasonjckn19:06:32

it works on all the fetch apis right?

tony.kay19:06:33

or load*action

jasonjckn19:06:43

will test now

tony.kay19:06:52

let me push to clojars 🙂

tony.kay19:06:18

done. Headed to lunch.

jasonjckn19:06:36

i added this little piece of code to user.cljs

jasonjckn19:06:38

(defonce install-app-state-diff-once (add-watch app-state :app-state-diff #(js/console.log "APP-STATE DIFF: " (diff %3 %4))))

jasonjckn19:06:01

[untangled-spec.reporters.impl.diff :refer [diff]]

tony.kay20:06:11

good idea! We had originally had something like that in an earlier version (pre-Om), but lost it during the transition. There are a number of dev tools I'd like to get going, including an app db browser plugin for Chrome.

jasonjckn21:06:42

:marker working beautifully

therabidbanana21:06:37

Anybody know off-hand if transit middleware wraps around other handlers attached to the untangled server, or just the main API endpoint?

therabidbanana21:06:17

(I was kind of expecting the middleware applies to all, but doesn't seem like it's applying to my sideband search endpoint I've added)

tony.kay21:06:50

you mean outside of /api?

tony.kay21:06:58

it should not

tony.kay21:06:11

well, actually, I think it is based on content-type

tony.kay21:06:16

I'd have to look

therabidbanana21:06:39

Looks like some of the middleware like gzip and not-modified get applied after the extra routes, but the transit specific stuff is higher up (/lower down) in the stack - before the extra routes get injected?

therabidbanana21:06:50

Oh well, not too hard to do myself.

tony.kay21:06:11

yeah. we could talk about moving it, since it could be content-type based

tony.kay21:06:29

some of the less used stuff hasn't had a critical eye on it

jasonjckn21:06:36

it's not a problem because I don't think any rendering is done between those steps

jasonjckn21:06:03

that's the output of adding a watch to the app-state and doing diffs

tony.kay21:06:06

which field?

jasonjckn21:06:18

i called load-field-action in a mutation on a field :hits

jasonjckn21:06:21

so from that 1 load-field-action the app-state [:search-results :global :hits] gets set to nil first, then to the new result

jasonjckn21:06:31

as oppose to just going from old data to new data

jasonjckn21:06:14

the UI output looks good though i just see the old data rendered, then the new data rendered

tony.kay21:06:51

yeah, implementation detailll.don't remember, but as long as it looks good 🙂

jasonjckn21:06:02

kk, just wanted to bring it to your awareness

tony.kay23:06:10

Untangled spec has lost all ability to do folding. Diagnosing, but it is being a pain

tony.kay23:06:31

nothing is acting as I would expect in the Om/React world 😞