This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
- # admin-announcements (17)
- # beginners (78)
- # boot (162)
- # braid-chat (2)
- # cider (20)
- # cljs-dev (9)
- # cljsjs (41)
- # cljsrn (17)
- # clojure (98)
- # clojure-austin (5)
- # clojure-brasil (1)
- # clojure-dusseldorf (1)
- # clojure-greece (1)
- # clojure-ireland (2)
- # clojure-italy (1)
- # clojure-japan (5)
- # clojure-russia (128)
- # clojure-uk (2)
- # clojurescript (29)
- # core-async (1)
- # core-logic (7)
- # css (1)
- # cursive (12)
- # datomic (18)
- # devcards (1)
- # dirac (6)
- # emacs (31)
- # funcool (28)
- # hoplon (208)
- # jaunt (66)
- # jobs (1)
- # juxt (6)
- # lein-figwheel (14)
- # off-topic (9)
- # om (83)
- # om-next (6)
- # onyx (63)
- # overtone (1)
- # parinfer (2)
- # protorepl (23)
- # re-frame (27)
- # reagent (14)
- # ring-swagger (8)
- # slack-help (2)
- # spacemacs (1)
- # untangled (56)
So, I’m about to rant because I can’t help it. If slack has a mute feature, you might need it.
And my rant also can be prefixed with “oshit @bronsa is in here” but. Can’t help it.
I think that clojure-in-clojure, tools.analyzer.* are kind of doomed because of the contribution process.
They are “stop-the-world” kinds of integrations, the way it is currently managed. We can’t seem to move to something new piece-wise.
I’d love just to pick off some Java piece that doesn’t need to be Java and make it clojure, test it, and make a PR.
Fair. Jaunt is entirely a protest against the Clojure contribution process and I'll make no bones about that.
As to self-hosting as a goal... I honestly kinda agree with Rich here, self-hosting isn't really a goal for me.
Not specifically a goal for me, but less splintery and more target platforms are goals for me.
I signed the contributor agreement specifically because people were talking about LLVM targeting, which never came to be.
Yeah. People talk about native targeting, or unifying CLJ and CLJS... idk. I'm content on the JVM, my thing is that I think the language and the compiler could be meaningfully polished in some ways which won't survive the contrib process and so Jaunt is my own sandbox where I can try to polish as such.
For instance my work to date in Jaunt is pretty much entirely focused around improving warnings and improving the code climate of the implementation.
@eraserhd: FWIW I'm significantly unhappy and frustrated with the clojure contribution process myself
At this point I'm quite deliberately on
0.2.0-SNAPSHOT, because I don't even know if in the long term I want to keep Clojure's same namespace structure.
Lots of room for questioning the status quo and exploring, but compatibility is for now a goal.
Breaking changes are and will be well documented/considered. Unless they're in the Java implementation which is a free fire zone of me refactoring everything.
Yeah. I'm waaaay happier now that I can take satisfaction in incrementally improving Jaunt as an end in and of itself rather than feeling frustrated because I know that whatever change I make will at best not go anywhere for a year.
I used to maintain my own fork of clojure with tons of patches applied a while ago too - - unfortunately I don't think anymore that maintaining a fork that nobody is likely to use is worth it
Honestly I look at Jaunt as a piece of performance/protest art. I don't care if nobody uses it, I take enough satisfaction in just improving this or that to justify a little bit of effort every day.
wait till your colleagues waste two days trying to figure out a bug you fixed 3years ago
Which is why I want to keep Jaunt source compatible, so someone else can drop it in in dev, get good warnings/lintings/help and go right back to CLJ.
Also if I really break anything I loose a whole lot of value in terms of my being able to just grab a Clojure library and get back to work.
I was joking for a while about making Jaunt/West shirts and going to the con, but I can't justify the money/time while I'm still in school.
@onealexharms: we should talk about making Avi use jaunt instead of official clojure
@eraserhd: I have been thinking the same. I just now realized there’s an #C0F4MB8AY channel, and joined it.
I'd love to have someone running Jaunt snapshots in test, I would advise against it for prod tho 😛
Avi has pretty close to 100% test coverage. As long as we get non-snapshot versions (so they don’t change out from under us), it’ll be stable.
Given that we’re separating spec and implementation, we can run against both, and Avi can be a test for Jaunt
Seriously one of the first things I did when I started setting CI up was add whitespace linting. That's the reason I've been throwing all kindsa patches at cljfmt and rewrite-clj