Clojurians Log v2
fulcro
hey there, is there another way to add no-value attributes (like disabled) to HTML tags aside from (dom/div {:disabled ""} ...)? I combed through the book and this seems to be the only way to do so, but honestly it looks kinda… weird so is there a more elegant way to do it?
Use: {:disabled true}
This might be more of a pathom question, but is the typical way of setting up a resolver where you might not know all the outputs from the api response something like:
(pc/defresolver some-resolver
[__ {:keys [resource/id]}]
{::pc/input #{:resource/id}
::pc/output [:id]}
(let [resource (get-resource-by-id id)]
resource))where I could still query other parts of the response like: [{[:resource/id 1] [:id :name]}]
but I wouldn’t get the auto-complete in the fulcro inspect
@njustice definitely a pathom question, they have a separate channel :simple_smile:
anyway, no, the output needs to include :name
if you want to be able to query for :name
hmm.. odd with what I have above I query any keyword in the response of resource
my example query works, and returns {[:resource/id 1] {:name "foo", :id 1}}
(at least it works in fc inspect query)
hmm, might be a feature of a newer version? or i'm misremembering
it would be tricky for me to know all the properties the api will respond with, I.e. a new field is added to the response
I recalled Tony saying something (in the videos) about how ::pc/output helps w/ auto-complete but you can query other fields - but I could also be misremembering
totally could be correct 😅
haha let me see if I can find the part in the videos 😄
Our resolver returns :stuff (not the actual name) which is a hash-map of all the stuff we're not sure will be returned and which we have little control over (just enough control to coerce everything under the stuff keyword)
Hmm, did something change/break with Fulcro and/or the Inspect tool? The tool loads an initial state, but then is "stuck" on it. Even a tab refresh doesn't un-stuck it. Meanwhile, I get about two dozen copies of this error in the console:
Do you get the warnings logged also in a browser without Inspect, eg Firefox?
Good point. I just checked, and yes I do. Although only four times.
When developing locally for a full stack application, how do we tell the client the location of the server? POST <http://localhost:8000/api> 404 (Not Found)
Are we not to develop from 8000 and have it hit 3000/api?
The /api is under the same host as the page itself, so it is also on port 3000. The other ports are for shadow-cljs and the nrepl.
@zilti versions?
Latest Fulcro requires manual update to inspect at the moment
I’ve been messaging about it all week
Ok, so deving from 3000 make sense. But why do I get this error when trying to go to a page that fetches data from the API?
core.cljs:159 WARN [com.fulcrologic.fulcro.networking.http-remote:184] - Transit decode failed!core.cljs:159 ERROR [com.fulcrologic.fulcro.networking.http-remote:344] -
#error {:message "Remote Error"output is used to build indexes in Pathom, so autocomplete won’t work without that info. You can return stuff from a resolver that isn’t advertised, but treat that as an optimization since pathom itself would not know to ask for it from that source. For example, say you have a resolver for fact A that can be retrieved via some ID, but another route can also give A if it is hit. If that route does not advertise it can return A then it will never be asked for A; however, if it is hit first and provides A anyway, then Pathom will detect that is already has A and won’t go looking to resolve it if it then finds that it needs it.
strangely enough, I can go to the server localhost:3000/index.html and navigate to that page via a link and the page loads, data renders..
3.4.1 currently. Ah, I'll try that then!
yes, 3.4.1 will NOT work with the Chrome store version of Inspect
I’ll restate it in case anyone missed it: The current Chrome Store release of Inspect will not work with Fulcro 3.4+. There are pre-release versions of the new Inspect that you can manually install here: https://github.com/fulcrologic/fulcro-inspect/releases
both Chrome and Electron
Hmm it still doesn't work though. Same errors in console, same behaviour
I would recommend using the latest Inspect (which requires 3.4.3-SNAPSHOT of Fulcro). They depend on each other.
it definitely works for me, so you’ll have to provide much more context if you want further help: OS Versions involved Errors from your console Errors from Inspect’s console (right click on Fulcro Inspect and choose Inspect)
screen shot of UI of Inspect
Well, since I also get the same errors in Firefox without Inspect, I am not even sure if Inspect is the cause
Okay, latest Inspect, Fulcro 3.4.3-SNAPSHOT. Console errors: see screenshot above. Errors from Inspect's console: none.
the console errors you’re showing me have to do with your routing and application startup mistakes (they are harmless, though), nothing to do with inspect.
all that’s happening there is you’re rendering routers before you’ve started them.
(which is harmless, but deserves a warning because that could be a problem in non-startup scenarios)
I’d recommend quitting/restarting Chrome. You must at least start a new tab to get a newly installed inspect.
and make sure you uninstalled the old one
Hmm, quitting/restarting Chrome helped. Also, good to know where the warnings are from (a bit odd though that they get repeated so many times). As always, thank you a lot!
sure. In terms of repeating: yeah, they will repeat on every render until the SM starts, so that really depends on how many animation frames you cause before a route gets triggered to start the SM. You can use app/set-root! with initialize state, then dr/initialize! or dr/change-route!, then app/mount! with NO initialize state to get rid of most or all of those. Basically: Make sure you’ve explicitly routed to a leaf (target) before mounting.
(initialize just makes sure all routers have started their state machiens)