Fork me on GitHub
#vim
<
2016-09-22
>
juhoteperi10:09:00

Looks like the bencode impl has several bugs

dominicm10:09:19

@juhoteperi Are you using my watchable branch?

juhoteperi10:09:27

Nope, master branch

dominicm10:09:53

That bug doesn't look the same one anyway. sigh

juhoteperi10:09:18

Your branch still uses same nrepl-client?

juhoteperi10:09:43

Does the branch fix problems about closed connection?

dominicm10:09:48

Oh, I guess I only fixed it locally (probably because of updates)

dominicm10:09:33

closed connection = trying to connect to a dead nrepl? Not on a dry connection (should just be a try/catch, I've just never hit that because of usage I guess?). But it does if a connection dies beneath you whilst being watched.

juhoteperi10:09:30

In some cases deoplete just blows up while working, while repl is just fine still (fireplace works etc.)

juhoteperi10:09:37

And I have to restart Vim

juhoteperi10:09:09

But could be related to other completions maybe, I feel like it often happens when switching to Less files...

dominicm10:09:52

Hmm, not sure on the cause of that. I don't use deoplete for much outside clojure tbh. Get me the stacktrace if you can, I'd be interested in seeing it.

dominicm10:09:57

The watchable branch is my daily driver now, the only thing holding me off merging to master was waiting to see if cemerick would merge my patch or not. I guess not. I need to play and see if git submodule update or whatever the command is, will change the submodule url if it changes (my cursory google said no)

juhoteperi10:09:20

hmm, git submodule sync should do that

dominicm10:09:34

Yeah? Maybe it's okay then. I didn't test.

dominicm10:09:13

I can safely point it at my branch of python_nrepl_client, we might want to move that into clojure-vim if we use it though.

juhoteperi10:09:32

maybe we should have the nrepl-client at clojure-vim?

juhoteperi10:09:37

other plugins might want to use it also

dominicm10:09:55

@juhoteperi you mean cemerick's, or another?

juhoteperi10:09:23

or whatever we want to use

dominicm10:09:25

Sorry, yeah, that's what I was trying to say.

dominicm10:09:05

I'd be happy to write some kind of WTFPL or CC nrepl client.

juhoteperi10:09:57

hmm license?

juhoteperi10:09:10

is there some problem with MIT?

dominicm10:09:36

Possibly not. I just don't understand licenses at all. I want something people can do whatever they want with pretty much. I don't want future people to worry about falling foul of a license by forking our library and including it in their repo to fix things.

dominicm10:09:38

I kinda understand that building on a fork is okay though, which is why submodules are okay. So building on https://github.com/SevereOverfl0w/nrepl-python-client is probably okay.

juhoteperi10:09:44

Problem with the more permissive licenses like WTFPL and CC0 is that they are probably not valid in many jurisdictions

dominicm10:09:18

Is public license a thing in all jurisdictions?

juhoteperi10:09:15

If I recall there is some problem in declaring stuff public license somewhere

juhoteperi10:09:10

> It's not global. It doesn't make sense outside of a commonwealth ecosystem, is explicitly illegal in some places (Germany), and of unclear legality in others (Australia)

dominicm10:09:13

MIT looks pretty similar based on the summaries anyway. The only real restriction is retaining copyright notices. Which is fine by me.

juhoteperi10:09:35

> Its applicability is unpredictable The license is short, clearly expressing intent, at the cost of not carefully addressing common license, copy-right and warranty issues. It leaves a lot of leeway interpretation - meaning that, in the US, it will take a few trials before you can reliably know when the license is applicable, and how.

juhoteperi10:09:08

MIT should allow forks and including code as part of other projects

dominicm10:09:09

That sucks.

dominicm10:09:49

MIT looks pretty good now I'm comparing it to something that tries to be public domain. I don't care much for the copyright notice, I assume it only applies to the source code it's included in though.

dominicm10:09:51

nrepl-python-client is MIT too. So we can build on it. I could copy/paste it into clj-async-omni too if needed, as there's no copyright notices.

dominicm10:09:32

I wonder how MIT applies with contributors (is each intended to add their copyright?). I should probably buy a book on license laws for developers.

juhoteperi11:09:29

Probably enough that maintainer has copyright notice

dominicm12:09:20

Can I use MIT-licensed code in closed-source software?

Yes, however the copyright notice and permission notice must be included in the documentation or EULA of your Software.

For example: "This software uses <library name> - see library-license.txt".


How do I display the MIT License if my project has no documentation or “about” feature?

Contact the author. If the author agrees, a notice on your product website may suffice as documentation.

juhoteperi12:09:12

> One might argue in court that by submitting a pull request, that person has granted you an implicit license, and that's what a lot of open source projects rely on

dominicm12:09:33

But not copyright.

dominicm12:09:55

I've seen a project where a developer went back and revoked his code as it was still under his copyright. To kill a project.

juhoteperi12:09:39

Yeah, well it can happen

juhoteperi12:09:49

But I wouldn't think about licenses too much

juhoteperi12:09:10

Author probably wont care, or might even be happy that the code is being used

dominicm12:09:28

Yeah, exactly.

dominicm12:09:49

Having read that FAQ, I think using Cemerick's library via copy/paste would be easy. We just have to add a line somewhere.

dominicm12:09:58

Probably a header of some kind.

juhoteperi12:09:16

like https://github.com/adzerk-oss/boot-cljs#license, just another line "Copyright ©2013 Chas Emerick and other contributors, for original nrepl-python-client code"

dominicm12:09:55

But this isn't a problem if the submodule works as want anyway.

zamaterian13:09:24

Great initiativ with gh clojure-vim 🙂 !

karlharnois22:09:56

does anyone know of anything equivalent to vim-clojure-highlight for clojurescript?

karlharnois22:09:51

my current vim setup for clojure is near-perfect but not having highlighted local/referred vars for cljs is annoying 😛