cursive

cfleming 2025-05-22T03:55:45.804609Z

Cursive 2025.2-eap1 is out. Support for the new IntelliJ 2025.2 EAP, some bug fixes and dependency completion in deps.edn and project.clj files (again!) https://cursive-ide.com/blog/cursive-2025.2-eap1.html

➕ 2
🎉 3
cfleming 2025-05-23T08:24:54.691449Z

@kolstae I don't think so, no, I could add one if this were problematic. Can you check your notifications view to see if you have a notification pending there?

kolstae 2025-05-23T08:28:57.066089Z

I don't have any notifications pending

kolstae 2025-05-23T08:31:31.680269Z

My bad. Somehow it had ended up in "Don't ask again" in the settings. Removed it, restarted and it's all there

👍 1
Pedro Boschi 2025-05-23T14:47:23.085499Z

Is it possible for us to trigger local/private repos indexing to work with this feature as well?

imre 2025-05-22T10:04:01.186079Z

Thanks Colin!

imre 2025-05-22T12:53:29.826889Z

Is there any way to disable the deps.edn version warning inline, per item? There are some dependencies that I do want to keep at their current version. (I don't think there's a community-wide established way for this, checked the antq docs but nothing there either)

cfleming 2025-05-22T22:54:15.790909Z

Unfortunately no, I'll look to add general inspection suppression to the next EAP.

kolstae 2025-05-23T06:56:35.458859Z

I'm not getting asked about the index download on startup. Is there a way to trigger it somehow?

cfleming 2025-05-24T22:17:10.589279Z

Not at the moment, but I'll think about how that might be done, it should be possible. Do you know if those repos are generating the Maven indexes?

imre 2025-05-26T12:29:51.535529Z

> I'll look to add general inspection suppression to the next EAP TBF I wouldn't want to turn this feature off, it's useful. What I'd like to be able to do is what depot used to have: https://github.com/Olical/depot?tab=readme-ov-file#controlling-which-part-of-the-file-are-checked (I normally use antq these days and that doesn't have an ignore functionality either. Would be lovely to have something like this work across tools)

imre 2025-05-26T12:30:32.119189Z

Love the dep bump feature btw

Pedro Boschi 2025-05-26T13:08:10.209489Z

@cfleming How can I check if these indexes are being generated? The local repos are added in deps.edn with the line:

:mvn/repos {"company-mvn" {:url ""}}

Pedro Boschi 2025-05-26T14:14:17.690709Z

@imre antq does have an "ignore" functionality: if you use DEPS, just add the meta ^:antq/exclude before the {:mvn/version "..."} you want ignored in the deps.edn (https://github.com/liquidz/antq/blob/main/doc/exclusions.adoc)

1
imre 2025-05-26T14:36:41.181209Z

Thanks Pedro!

1
cfleming 2025-05-27T02:41:30.438479Z

> TBF I wouldn't want to turn this feature off, it's useful Yes, definitely won't require turning the feature off, just localised ignores like the antq one.

Pedro Boschi 2025-05-22T13:22:26.331129Z

In REPL, when an exception is printed, would it be possible to skip from the collapsed part the exception type + message, and also the "Caused by:" line, with the cause exception's type and message? Currently those lines are often hidden by the "internal lines collapse logic". I like the collapse behavior, but I also find the info I mentioned very important to be always visible

kenny 2025-05-22T13:59:05.367489Z

I am noticing when large strings are rendered with the inline viewer, the tab with the large string becomes quite laggy. Clearing the inline result goes back to normal, snappy performance.

p-himik 2025-05-22T17:28:42.695409Z

It seems that Cursive gets confused by aliases that are the same as existing namespaces. E.g. I have (:require [chime.core :as chime]) and the Chime library has a single-segment namespace, chime. Both chime.core and chime-the-namespace have a chime-at function. And it's deprecated within chime, but isn't deprecated within chime.core. Even though [chime.core :as chime] is unambiguous, Cursive still think that chime/chime-at can refer to the function in either chime/core.clj or in chime.clj, and offers you to select which one you want to go to upon navigation. It also highlights chime/chime-at as deprecated, even though it isn't.

cfleming 2025-05-22T22:54:56.675639Z

I'll check that, thanks.

hlship 2025-05-22T23:08:22.436619Z

I often get hit by our linter for inconsistent use of aliases for the same namespace across different namespaces. It would be awesome if there was a way to change the alias of a namespace and have that take effect throughout all source that references the aliased namespace. Like an alternate refactor of alias that applies globally.

💡 2