@micha have you read the https://rfc.zeromq.org/spec:22/C4/ yet? i just came across it again because the author (Pieter Hintjens) just died a few days ago. it's a little abstract but worth internalizing. i was just thinking why am i not asking for permissions for pushing releases to https://clojars.org/groups/hoplon and i realized that i wont trust myself yet to release anything without your review first.

thats when i realized there are no explicit rules setup for governing boot and hoplon

i think it would be a good idea to establish rules

like what are the expectations for compatibility

from one version to the next

@micha, were you talking to me? If so, what did you think you saw?

@vigilancetech yes, the 2.5.0 release doesn't have some of the commits from master

@micha so I should go back to 2.4.0?

What is the substantial difference between boot and boot-hoplon? Why maintain two different code bases?

boot-hoplon is the task that handles building hoplon stuff

boot is a more general type of program

boot-hoplon ships a task that runs in boot

and composes with other tasks in boot

ah, ok, so boot-hoplon is just a task UNDER boot, ok, thanks for clearing that up.

for example you combine it with the cljs task

which isn't specific to hoplon

@vigilancetech i just pushed boot-hoplon 0.3.0-SNAPSHOT

@micha ok, thanks

I just built a new task from the seancorfield/boot-new template and this is what it gave me: `(deftask dev "Build nt2 for local development." [] (comp (watch) (speak) (hoplon) (reload) (cljs) (serve :port 8000)))`

looks good, no?

but it has cljs in there

sure, you will need to compile cljs -> js

which is what the cljs task does

the hoplon task actually generates cljs source files and html files

the cljs needs to be compiled into javascript by the cljs task

the comp there sets up a pipeline of tasks

ok then I'm not understanding what you meant by >for example you combine it with the cljs task >[5:28]  >which isn't specific to hoplon

yeah that's exactly what i was referring to

(comp (task1) (task2) (task3))

that creates a pipeline of 3 steps

oh, ok, sorry, yes I get it now. cljs is another of boot's tasks. I thought you were telling me I was doing something wrong by including it.

oh, no sorry

I'll get it eventually 🙂

it's a ridiculous amount of machinery to make a webpage 🙂

yes, but it's very cool. Thanks for your and Alan's work on this. It's an exciting project!

hopefully someday soon I'll know enough to actually contribute

the whole js/html thing is like some kind of bad nightmare that's almost over thanks to you guys 🙂

and for that whole reason, I've been kind of staying away from it.

thanks 🙂

yeah i feel like we're getting close to freedom

very much! I'm ramping up on clojure, cljs, html, etc... to do a project all at the same time and my head is kind of spinning. Sorry if I appear confused at times. Most of my lisp experience is with elisp and probably less than half of that transfers over to clj/cljs

welcome! hope the process is more fun than hassle 🙂

Thanks, the only hassle is my personal deadlines 😕

@micha I'm still getting java.lang.IllegalArgumentException: No matching clause: ns+ after pulling 0.3.0-SNAPSHOT. I've cleaned up things a lot (I think). I decided to go from a new (known working) template and only add/change what I think I need: http://pastebin.ca/3727408

@vigilancetech can you do a quick git grep ns+?

in my clone boot-hoplon repo? Or in my app?

http://pastebin.ca/3727411

maybe my build.boot's require should have this instead? '[hoplon.boot-hoplon :refer [hoplon prerender ns+]] (although ui's test doesn't, and it seems to work)

no, you don't need to :refer ns+

hm i should check something

ok, the prelude is the only place in my code that has the ns+ in it then

the prelude?

yeah, the top of my *.hl file

last part of the last past

*paste

well, not the grep paste, the one before

is the project repo accessible?

like am i able to clone it to try?

I can do that, b4b

*brb

https://github.com/vigilancetech-com/nt2.git

is there a quick workaround for ns+?

@vigilancetech ns+ is unnecessary. It's only sugar. What do you want to do?

I'm trying to build an app based on hoplon/ui. So far, the only one's I've been able to get to compile were the ui/test and computable-law examples. Somewhere it ran off the rails and I started getting this error. Right now I'm going back and putting my code into computable-law and adapting it. Hopefully I can get that to compile again

Having code compiling all the time is really useful for warnings, I recommend that you look at that. I would ignore ns+ and use regular ns for the time being.

@mynomoto well, its back compiling again, whew! Going to stick with getting the elems up on the page for a while 🙂

@micha I might have found out something important about that ns+ issue. It only does it if the file is named .hl , it's fine if the file is .cljs. Will that mean that hoplon isn't processing the file?

@micha as part of our upgrade to clojure 1.9, i made this ^ minor change. can we publish a 3.0.0-alpha5 pls? or if you can add me to the hoplon clojars org i can do that. i finally have a gpg setup and i just successfully published a release to clojars: https://clojars.org/onetom/boot-lein-generate

hmm... now that im saying "successfully" i mean clojars didn't complain, but if i look at my jar it seems unsigned:

pair@tomplus ~/g/h/castra> jarsigner -verify -verbose /Users/pair/.m2/repository/onetom/boot-lein-generate/0.1.2/boot-lein-generate-0.1.2.jar

           0 Tue Oct 11 12:17:42 HKT 2016 META-INF/
           0 Tue Oct 11 12:17:42 HKT 2016 META-INF/maven/
           0 Tue Oct 11 12:17:42 HKT 2016 META-INF/maven/onetom/
           0 Tue Oct 11 12:17:42 HKT 2016 META-INF/maven/onetom/boot-lein-generate/
        1111 Tue Oct 11 12:17:42 HKT 2016 META-INF/maven/onetom/boot-lein-generate/pom.xml
         136 Tue Oct 11 12:17:42 HKT 2016 META-INF/maven/onetom/boot-lein-generate/pom.properties
           0 Tue Oct 11 12:17:42 HKT 2016 boot/
        1746 Tue Oct 11 12:15:51 HKT 2016 boot/lein.clj
          25 Tue Oct 11 12:17:42 HKT 2016 META-INF/MANIFEST.MF

  s = signature was verified
  m = entry is listed in manifest
  k = at least one certificate was found in keystore
  i = at least one certificate was found in identity scope

jar is unsigned. (signatures missing or not parsable)

maybe it's better if u publish it for now...

@onetom i think gpg is a different signature mechanism

you can try boot show -v i tink

think

indeed, thanks!

pair@tomplus ~/g/o/boot-lein-generate> boot show -v
:signed   └── [adzerk/bootlaces "0.1.13" :scope "test"]

or like boot -BP -d onetom/boot-lein-generate show -v to check that specific jar

thx. i've also tried:

> cd ~/.m2/repository/onetom/boot-lein-generate/0.1.2/
> curl -O 
> gpg --verify ~/.m2/repository/onetom/boot-lein-generate/0.1.2/boot-lein-generate-0.1.2.jar.asc

gpg: assuming signed data in '/Users/pair/.m2/repository/onetom/boot-lein-generate/0.1.2/boot-lein-generate-0.1.2.jar'
gpg: Signature made Tue Oct 11 12:17:54 2016 HKT using RSA key ID 054F2F74
gpg: Good signature from "Tamas Herman <[email protected]>" [ultimate]

so would a 3.0.0-alpha5 release be possible?

@vigilancetech boot-hoplon only does things to .hl files, so ns+ only works there.

@onetom 3.0.0-alpha5 version of which library?

castra

ah sure

wow that's a weird one

i have no idea why i would have done (import java.io.File) like that

@onetom done

is there a way in a cell= to get a reference to a named cell, not it’s deref’d value? Like if I wanted to update some cell as a side-effect in a cell=?

@adamfrey direct formula is a way, also add-watch

ok, thanks Alan

(page "index.html")

(defc x 123)
(defc y 321)

((formula (fn [v] (swap! y + v))) x)

(js/setInterval #(swap! x inc) 1000)

(html
  (body
    (h1 (cell= (str "x=" x ", y=" y)))))

cool, that seems easy enough.

i would probly do (add-watch x ::swap-y (fn [_ _ _ v] (swap! y + v))) instead of formula

Btw did anyone get a chance to look at my javelin-bug minimal repo? https://github.com/AdamFrey/javelin-bug I’m curious if it can be reproduced by anyone else

hm i did not, i will look right now

thanks so much

@adamfrey do i understand right that the warning appears, but the code works as expected?

let me double check

yes it does look like the code works

I think the warnings scared me off a bit.

@adamfrey for updating a cell inside cell= there is also ~(cell my-cell) to get the my-cell cell.

@mynomoto oh interesting

thanks

@mynomoto that's a good one... i guess that creates a new anonymous cell containing my-cell that is instantly deref'd?

@alandipert I'm not sure of how it works, @micha taught me that one. There is code on the demos using it castra-chat and tictactoe.

hehe yeah, i think so. since cell= boxes all free variables in a call to cell-deref*. so you can sneak cells in by making them the value of a cell

@adamfrey in any case, confirmed. exploring a fix now

@mynomoto you can also do this to update a cell inside a cell=:

(cell= (if foo (~(partial swap! bar) inc)))

the ~ evaluates the form when the cell= is created, instead of evaluating it every time the cell= updates

We need a javelin emoticon 😉

@adamfrey i'm not sure how the way we expand macros doesn't always spew lots of warnings

well, the warning occurs in the normal version without the mx macro

ok so micha wheeled over and we looked at the warning thing

it's coming from the call to analyzer/macroexpand-1

really weird

i wonder about macroexpanding as we go instead of in 1 shot

i don't see how that would fix the problem. maybe we could look at core.async and see what they do

@micha so #141 #142 and #143 are good to go and fix all the attribute stuff we previously discussed.

defaults are in core and jquery overrides default browser behaviour with hoplon.jquery

I think im going to roll them into a separate PR tho to simplify all the changes

much better :thumbsup:

@flyboarder awesome, i fixed boot-hoplon last night

ready for the changes

it should still be compatible with older versions of hoplon also (ones that don't have hoplon.jquery namespace)

cause it checks to make sure the namespace is on the classpath when you do :refers

ignores if it isn't

oh thats great!

so I'm trying to do ui and I laid out an image with a 3 subimages (icons) on it and it's wanting to group them all together but I want them to spread out. Does anyone know off the top of their head what attrib I should be using?