Fork me on GitHub
#cursive
<
2020-02-04
>
kenny03:02:11

It seems my new projects are now all called "Imported From Deps". Can I change this to match the directory name?

kenny03:02:41

I see a .name file that has "Imported From Deps"

kenny03:02:38

Changing that and then reopening the project updated the name. I think Cursive used to set the parent name to the directory name though.

cfleming04:02:04

Yeah, that seems to have changed in 2019.3. Are you creating the projects in IntelliJ, or creating using the CLI and then importing?

kenny04:02:34

I usually duplicate an existing project and then import.

cfleming04:02:00

I see, interesting. I can’t always reproduce this, I’ll try that.

kenny04:02:22

I try to always delete the .idea, any .iml and .git before importing.

cfleming04:02:30

Ok, thanks.

d-t-w09:02:49

I'm finding Intellij 2019.3.1 and 2019.3.2 painfully slow, anyone else the same? To be clear I don't think it's a Cursive issue (though I guess it's possible). Super laggy scrolling through cljs files in the editor pane, super unresponsive with shortcut keys, seems to take forever to gain focus when switching into app. I'm fairly certain it's since I bumped up to 2019.3.x in early january.

jumpnbrownweasel15:02:51

That's my experience as well.

jvtrigueros15:02:08

I was noticing that yesterday too, I've disabled almost all plugins too 😞

cfleming22:02:40

I’ve noticed IntelliJ being laggy too, but it’s always difficult to tell on my dev machine since I run multiple instance of IntelliJ and debug a lot.

d-t-w06:02:12

I can imagine that your Intellij setup might be a bit bonkers Colin! I'm putting it down to suuuuper poor performance regression since the 2019.3.1 update.

d-t-w06:02:39

Worst bug report ever, like when I went to the doctor at Uni and said "I'm just like, suuuuuper tired".

imre10:02:00

Is there a bindable command to show expected/actual difference after a failed test run? I would love not having to hover-scroll-click

cfleming23:02:35

No, but it’s on my list, you’re right that that is very annoying.

thanks 4
onetom09:02:35

i started using F2 for that purpose actually. discovered it accidentally 🙂 it's jumping to other locations too, but includes this use-case too

imre09:02:20

That would be the "Next nighlighted error" command, wouldn't it?

imre09:02:29

I do have that bound, and it does bring up the hover window, however inside the hover window there's a link at the bottom saying "Show difference". When you click that, it will show the diff in Idea's diff viewer which is quite handy for large diffs. I'm looking for a way to invoke "Show difference" with the keyboard

onetom21:02:12

Yeah, that show difference action was not exposed as a bindable function the last time I asked about it a few months ago.

ghadi15:02:46

Does cursive have infra to support tap> ?

cfleming23:02:56

Not at the moment, no. What would you like to see?

ghadi00:02:35

One of our customers was asking about debugging a long ->> pipeline, and I suggested to comment out all the steps he didn’t want to run, but a step that called tap> would be more useful

ghadi00:02:13

Dunno, the easy thing would be to print the data to the repl but you can already do that today

ghadi00:02:20

Maybe a viz of some sort

cfleming00:02:20

Yeah, I’m planning on adding a data inspector so that might be a good candidate. My socket REPL support is sort of prepl-y but also has structured input, so I could potentially do things with tap values on the Cursive side there too

onetom09:02:27

i just use the dbg macro from this library usually: https://github.com/philoskim/debux#debugging-thread-macro-or

David17:02:29

If I (:require [clojure.test :as t]) in a -test.cljs file, Cursive marks 't' as unused--even when t is clearly used. If, instead, I (:require [cljs.test :as t]), then Cursive doesn't complain.

cfleming23:02:23

Interesting, I haven’t seen that. I’ll check that I can reproduce that and fix it.

David00:02:46

If you have any problems reproducing, let me know.

David17:02:42

Is there something I can do make Cursive recognize the clojure.test usage?

sam21:02:05

I’m using tools.deps with Cursive. I have the following alias that I’m trying to use with the cursive repl.

{:repl {:main-opts ["-i" "aot.clj" "-r"]}

cfleming00:02:49

Someone mentioned this to me a while back. I thought I had investigated but I can’t find my notes now. I’ll take another look.

sam21:02:56

Thanks @U0567Q30W, let me know if I should file an issue.

cfleming01:02:57

Actually, if you could that would be great, then it won’t get lost again.

sam21:02:58

Even though I added :repl to Aliases, it seems that aot.clj is not getting run like I expect it to when starting the repl.

sam21:02:50

When I use clojure.main instead of nrepl, aot.clj runs as expected. Is main-opts not supported for nrepl? I need to do some AOT compilation every time the repl starts and I’d rather not make devs have to remember to do a separate step. Is there a better way to do this?