Fork me on GitHub
#hoplon
<
2017-04-05
>
alandipert20:04:52

^^ pls try this out! [hoplon "7.0.0-SNAPSHOT"] we made a change that dramatically increases performance

alandipert20:04:59

cc @jumblerg ⏫

alandipert20:04:16

it should help a lot with initial loads

jumblerg20:04:24

@alandipert: already on it!

jumblerg20:04:30

micha tipped me off.

jumblerg20:04:39

pretty psyched about this one.

alandipert20:04:26

in our little test app, adding 1000 lis to a ul went from 5000ms to ~30ms

jumblerg20:04:39

i also love that we just skipped hoplon 6. seems appropriate, it was getting kinda like perl.

alandipert20:04:48

lol yeah the -alpha business

alandipert20:04:52

onward and upward

jumblerg20:04:05

i'm with ya.

flyboarder20:04:59

^ stoked for this!!!

flyboarder20:04:48

@ejemba in that example get-state is run and then again every second

flyboarder20:04:25

@alandipert any other breaking changes you can think of that may want to go into 7?

alandipert20:04:23

i don't think so

alandipert20:04:51

also this latest change shouldn't be breaking

flyboarder20:04:04

So I have been working on a project, it’s still WIP and there is little code available but I will be launching this at an upcoming conference; https://github.com/degree9/meta

flyboarder20:04:06

Opinions welcome πŸ™‚

alandipert21:04:04

hm - did we never fix nil children stopping appends?

alandipert21:04:10

i forget where that landed

jjttjj21:04:43

hoplon7! πŸ‘

alandipert21:04:48

^^ removed because it made hoplon task 6x slower for our app

alandipert21:04:15

hm i think we can actually keep it, it just shouldn't be used by hoplon task by default maybe. anyone use it/like it?

jumblerg21:04:35

i'm for ditching it as well.

jumblerg21:04:01

i think newer versions of cljs make it less necessary anyway.

jumblerg21:04:34

i think less is more here; it will probably only frustrate new users.

jjttjj22:04:39

Agreed. I used it for a bit but now I never do. Just use ns with hoplon/page metadata when necessary

jumblerg22:04:06

i don't use the page macro anymore either.

jumblerg22:04:23

or the cljs.hl files by extension.

alandipert22:04:44

i use the page macro and will for the foreseeable future, because our massive app uses it

alandipert22:04:02

and i use it for experiments, because the template does lol

alandipert22:04:11

but not using it is a pretty good idea too i think

jumblerg22:04:20

for anyone listening in this looks like:

(ns ^{:hoplon/page "index.html"} mybiz.application ...)

jumblerg22:04:36

i doubt that's documented anywhere.

jumblerg22:04:17

i'm thinking about documenting it though. and it's the thought that counts, they say.

alandipert22:04:30

i'm down to just switching over to it wherever we have examples

alandipert22:04:53

at least, in the immediate docs and templates

jumblerg22:04:43

makes sense

alandipert22:04:56

a cool thing about that is people could use the hoplon task together with reagent or whatever

alandipert22:04:14

to be rid of .html stubs

jumblerg22:04:08

maybe it should just be html now.

jumblerg22:04:55

i'm about to listen to dan bricklin give a talk.

alandipert22:04:54

πŸ™‡ πŸ™‡ not worthy

chromalchemy23:04:46

For a for-tpl structure to edit element values in the UI:

(let [data (cell [1 2 3])]
    (for-tpl [x data]
      (elem :click #(swap! x inc) @x)))
As I understand it, x is formula cell that carries the value of part of the data sequence. So by default it does nothing to the data cell. And x is not reset! directly. Instead you update the underlying value in data. What is the idiomatic way to get the new values for x to use in other components? Like perhaps a formula cell that contains all the updated x values for data?

flyboarder23:04:01

@alandipert Can I suggest that we abstract the boot-hoplon dsl stuff out? I think it is both convenient to be able to build your own dsl and confusing for Hoplon users how much magic is being done with the .hl files