This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2018-10-19
Channels
- # 100-days-of-code (5)
- # announcements (1)
- # aws (1)
- # beginners (112)
- # cider (135)
- # cljdoc (6)
- # clojure (111)
- # clojure-dev (8)
- # clojure-italy (3)
- # clojure-nl (5)
- # clojure-sweden (3)
- # clojure-uk (152)
- # clojurescript (101)
- # datascript (14)
- # datomic (61)
- # editors (1)
- # emacs (29)
- # events (7)
- # figwheel (3)
- # figwheel-main (15)
- # fulcro (18)
- # funcool (2)
- # graphql (1)
- # juxt (2)
- # off-topic (51)
- # om (1)
- # overtone (28)
- # perun (2)
- # reagent (1)
- # reitit (6)
- # ring-swagger (5)
- # shadow-cljs (112)
- # spacemacs (49)
- # tools-deps (10)
- # unrepl (11)
- # yada (10)
does ptransact
wait for a transaction’s network request to finish before going to the next, or does it wait for the transaction’s network request to simply start before going to the next?
ah, I just read the ptransact!
source.
> This function defers pessimistic writes, not reads.
Is there some other thing I should use for pessimistic reads?
@levitanong would df/load and post-mutation not work for you're use-cases ?
it’s several mutations that have load-action
s, so I don’t think so.
there is nothing like https://facebook.github.io/graphql/draft/#sec-Field-Alias in the query syntax right? neither fulcro nor pathom?
@thheller Think there is in pathon. At the bottom there is a edn->grahpql https://wilkerlucio.github.io/pathom/DevelopersGuide.html
ahh. Guess if pathom can do that for edn->graphql, should be able to do it for edn->aliaseedn and back based on those conventions.
I was thinking through a scenario where I have a parameterized attributes [(:foo/bar {:param 1}) (:foo/bar {:param 2}})]
you cannot combine the same prop in a single query twice…the return value is a map keyed by just the prop…no dupes there @thheller
@tony.kay hmm do you think it would be useful to add (:foo/bar {:param 1} :alias)
. taking the example from the graphql spec running two queries seems like overkill and very inefficient
@thheller the closest we get now with pathom is to use a placeholder node, this will make a separated segment but you can just pull from it, example:
[(:foo/bar {:param 1})
{:>/placeholder-alias [(:foo/bar {:param 2})]}]
Where do you normally create a new entity when loading a given route? Currently when there is a route mutation I’m loading a model with df/load
, I could fire off the creation of the form entities there or in a componentDidMount
hook. Is there some preference of one over the other?