Proto REPL, a New Clojure Development and Visualization Tool [video] | Hacker News Reader

Proto REPL, a New Clojure Development and Visualization Tool [video]

youtube.com

129 pointstosh11 comments

Proto REPL has some great features in its earliest incarnations. Kudos to Jason for such a great start. The drill-down pretty printed eval output is very cool. And I love the built in charting/graphing plugin and how easy it is to use within this IDE.

I'm an Emacs/CIDER hardcore, so it takes a lot to sway me to try something new.

I'm looking forward to this being refined to the point of fully usable for daily work. The Clojure community would really benefit from a "just works" IDE - especially for ClojureBridge where we need to get people up and running quickly and without teaching them a new editor. Nightcode, and now Proto REPL seem like leading candidates.

Current minor quibbles: * Eval of charting expressions creates 2 tabs each time. * Eval popup gets truncated if size exceeds size of frame * Several charting examples either render improperly or throw exceptions. This may just be version incompatibilities.

Some of these may be pilot error. I'll confirm and try to report the bugs soon.

I must say, I miss paredit. Parinfer (at least in this incarnation) drives me completely bonkers, especially in indent mode. It's magic, but horrible evil black magic that works "sometimes" and steals and inserts closing ])} willy nilly. I guess I'll just turn it off unless someone can testify that it can become tolerable with more experience.

Again, great work Jason.

FYI I've been using proto-repl, parinfer, and lisp-paredit in Atom and like that combination. If you like paredit more you can do without parinfer all together and get the paredit experience you're used to.

Thanks. After some more effort I found out much of my frustration was due some bugs and some bad config settings.

Lisp-paredit is buggy for me (must use a shortcut twice to get one motion, can't hold down modifier keys when changing direction e.g. from slurp to barf). Very troublesome. Also lisp-paredit strict mode fights with parinfer, so you must turn it off to get the expected parinfer paren closing behaviors. I'll give parinfer some more time to sink in now that I have it working better.

Adding some familiar paredit keybindings (to keymap.cson) helps a lot too:

    'atom-text-editor[data-grammar~="clojure"]':
      'ctrl-right': 'lisp-paredit:slurp-forwards'
      'ctrl-left': 'lisp-paredit:barf-forwards'
      'ctrl-alt-right': 'lisp-paredit:barf-backwards'
      'ctrl-alt-left': 'lisp-paredit:slurp-backwards'

Wow! This looks great.

I really like the power of emacs. It seems like Atom might be more powerful, thanks to its browser capabilities.

I've read that Atom is not great when it comes to how many resources is requires. Has it gotten better? Is there a timeline for Atom performance improvements?

Repo: https://github.com/jasongilman/proto-repl

Here is a good checklist on how to get started with Clojure in Atom.

https://gist.github.com/jasongilman/d1f70507bed021b48625

Great video. I need to do some work on a medium large Clojure web app that I wrote five years ago, with few recent changes. This talk motivates me to change my development environment to use Proto REPL.

The demos in the talk remind me of some great data visualization demos I have seen for Pharo Smalltalk.

Atom? ick. Atom spies on you: https://discuss.atom.io/t/collecting-metrics-in-atom-core/14...

HN downvotes you without providing feedback, because your negativity is correctly placed, and no one can summon the words to contradict the substance of your response.

They take chromium, and they repurpose the spyware that gets baked into the Google Chrome product, rather than removing it. Take note of this and any app based on the Electron framework.

It's opt in now:

http://blog.atom.io/2016/10/11/atom-1-11.html

Thank you for taking the time to explain why the parent post may have been downvoted. It's useful information that should be communicated. It'd be nice if more users explained whether they were downvoting incorrect information, rather than a preference for tone of voice or conventional community sentiment.

Reply on news.ycombinator.com