dominikh1 changed the topic of #cinch to: The IRC Framework | Latest version: Cinch 2.0.6
jonahR has joined #cinch
Spami has quit [Quit: This computer has gone to sleep]
postmodern has quit [Quit: Leaving]
jonahR has quit [Quit: jonahR]
kludge` has quit [Ping timeout: 264 seconds]
kludge` has joined #cinch
Spami has joined #cinch
Spami has quit [Quit: This computer has gone to sleep]
postmodern has joined #cinch
v0n has joined #cinch
Spami has joined #cinch
Spami has quit [Remote host closed the connection]
postmodern has quit [Quit: Leaving]
kludge` has quit [Ping timeout: 248 seconds]
kludge` has joined #cinch
space_ has joined #cinch
Space has quit [Ping timeout: 254 seconds]
phony has joined #cinch
phony has quit [Client Quit]
rickmasta has quit [Quit: Leaving...]
v0n has quit [Ping timeout: 245 seconds]
v0n has joined #cinch
<dyoung522>
clearly I'm a pull-request newb
<dyoung522>
re-re-submitted the request on the proper branch, it looks okay from my end now afaik
<dominikh>
three time's the charm
<dyoung522>
let's hope.
<dominikh>
looks better than the first two :)
<dominikh>
by the way, you mentioned taking over development etc
<dyoung522>
heh, yes I agree. that '53 commits' looked, odd
<dominikh>
feel free to hack on as many features and bugs as you want; but I'd still expect PRs to review ;)
<dyoung522>
sure
<dominikh>
(especially after this incident :P)
<dominikh>
I'll merge the PR later
<dyoung522>
lol. I promise, I'm not a git newbie... just a PR newbie
<dyoung522>
but the question in my mind is, "better to create an extension or add a feature into the core". If you're busy on other things, I could understand that adding any new features may take a while to actually be implemented.
<dyoung522>
and I get that, I have the same time commitments too. I get a block of time every so often to work on 'side projects'.
<dominikh>
I'd argue for an extension first, see if the API works, then adapt it to fit into Cinch's core, then show me
<dyoung522>
that was my plan
<dominikh>
Personally I won't be implementing any new features, but I can gladly review changes
<dyoung522>
okay
<dominikh>
I do have somewhat strict requirements about APIs because I despise backwards incompatible changes, so the first revision mustn't be a fuckup ;)
<dominikh>
(that's pretty much why there's Cinch 2. Cinch 1 had some fuckups :))
<dominikh>
apart from that: bugfixes go on maint, new features go on master, and we use semver
<dyoung522>
that's where having a test suite helps tremendously
<dyoung522>
I've actually been brainstorming ways to implement tests into my cinchbot, using mocks and maybe a light-wieght local ircd server.
<dyoung522>
the cinch-storage extension I'm working on now has a full rpsec test-suite
<dominikh>
if you want to write a test suite for cinch, knock yourself out. but it will probably knock you out :) but a test suite doesn't protect you from adding bad APIs
<dominikh>
only from breaking them later
<dyoung522>
true, it's only a tool, but it does help keep them from breaking other shit.
<dominikh>
sure. still pretty tangential to this topic ;)
<dominikh>
anyway, writing tests for Cinch is hard
<dominikh>
too many bits depend on a running IRCd in the background
<dyoung522>
yes, yes it is.
<dyoung522>
that's the tricky part, which is where a local ircd server would help. But, I don't have a good solution for it yet... and may never.
<dyoung522>
mocks may help too, but you'd basically end up writing the bulk of a bogus ircd server just using mocks. :-/
<dyoung522>
anyway, sorry for the PR screwups, but I've learned how they work now.
<dominikh>
local IRCd is terrible for a test suite though
<dominikh>
(did consider that, as well as reviving my IRCd written in Ruby for that, but then you'd test one buggy piece of software with another buggy piece of software ;))
<dyoung522>
oh, I know
<dyoung522>
way too dependent on the local env
<dominikh>
right
<Netfeed>
and you'd only test it against one version of an ircd and not "all" versions
<Netfeed>
stuff that works on against one network can bug at another
<dominikh>
that as well. we have code that depends on the IRCd you're connected to, actually
<dyoung522>
still, can't help but feel some testing is better than none... but I agree that a local ircd is not the solution. I guess a good mock suite is probably the way to go.
<dyoung522>
you still wouldn't be able to test everything, of course.
<dominikh>
such as?
<dyoung522>
rspec is my current preference
<dominikh>
I meant what you couldn't test; and rspec is the devil
<dyoung522>
oh, I don't know, I haven't actually delved into it. It was an assumption.
<dyoung522>
heh, what's wrong with rspec?
<dominikh>
it pretends to be english.
<dyoung522>
ah, you don't like the DSL it uses.
<dyoung522>
doesn't really bother me
<dominikh>
I don't like its ideas
<dyoung522>
fair enough
<dyoung522>
rspec is what all the cool devs are using. =P
<dyoung522>
(that was very much tongue-in-cheek)
<dominikh>
and a lie :P
<dyoung522>
=)
<dyoung522>
being relatively new to [ABT]DD, I went with rspec because it was getting a lot of attention in the rails communities. Still is. I don't use cucumber because I don't generally have customers who are going to look at code (no matter how pretty it is)
<dominikh>
cucumber is just as pointless.
<dominikh>
and using *anything* because the Rails community does usually leads to ideas as bad as going with rspec ;)
<dyoung522>
lol
<dominikh>
rails is the epitome of bad decisions
<dyoung522>
wow, a lot of really influential people would disagree with you on that.
<dominikh>
you know who else are influential? Putin, Obama and back in the days, Hitler. And I'd argue their opinion isn't worth much to sane people.
<dyoung522>
although I do agree that there are too many rails developers that don't fully understand ruby.
<dominikh>
it's really no secret that Rails itself is a pile of magic and that the "Rails community" consists of hipsters, rockstars and people who'd rather be in politics than programming; where a library is popular because some popular guy wrote it
<dyoung522>
lol
<dominikh>
I think at some point Cinch's motto was "not like Rails.", in contrast to the Autumn bot framework.
<dyoung522>
so, what is your preferred framework, then?
<dyoung522>
web development framework, I mean
<dominikh>
test/unit (nowadays minitest) is pretty slick ;)
<dominikh>
oh, web dev
<dominikh>
I tend to avoid web development, but when I have to, I don't use Ruby for it ;) but sinatra is nice. ramaze is nice if you can find the docs.
<dominikh>
(which proved to be hard some time ago)
<oddmunds>
i like sinatra
<dominikh>
I believe that a lot of web dev doesn't need ridiculously huge frameworks; and extended functionality like sessions, validation etc can be done in a way more modular fashion than a full blown framework.
<dyoung522>
interesting
<dominikh>
but then I'm a huge fan of simple, and of composition
<dyoung522>
but you've moved to Go for the most part, haven't you?
<dominikh>
right
<dyoung522>
I guess I'm more of a 'right tool for the job' guy. I like rails a lot for web development... but I prefer cinch's framework over the others because of it's simplicity.
<dominikh>
I believe in the right tool for the job, too.
<dominikh>
strongly, at that
<dyoung522>
if I were writing simple static websites, ruby/sinatra is fine
<dyoung522>
if it's simple enough, hell, just plain 'ol HTML is fine
<dyoung522>
but with some dynamic content, sinatra is good... but for highly dynamic websites, I think rails does what it's supposed to do, and does it well.
<dominikh>
and I disagree on that part. Rails adds very little to "highly dynamic websites", since most of that dynamism is in the frontend, not the backend
<dyoung522>
I'm on the fence over rspec and T:U.
<dyoung522>
I like rspec merely because that's the path I started down.
<dyoung522>
but reading up on the new T:U stuff, it also looks good... with gems like Mocha to add mocks and other features (when needed)
<dominikh>
welcome to composition
<dominikh>
(in a proper design of Cinch, you would only need a minimal amount of mocks and mostly fixtures, anyway)
v0n has quit [Ping timeout: 276 seconds]
space_ is now known as Space
v0n has joined #cinch
v0n has quit [Ping timeout: 260 seconds]
v0n has joined #cinch
rickmasta has joined #cinch
Spami has joined #cinch
Spami has quit [Quit: This computer has gone to sleep]
jonahR has joined #cinch
Spami has joined #cinch
Spami has joined #cinch
v0n has quit [Ping timeout: 256 seconds]
jonahR has quit [Quit: jonahR]
jonahR has joined #cinch
jonahR has quit [Quit: jonahR]
space_ has joined #cinch
Space has quit [Ping timeout: 256 seconds]
postmodern has joined #cinch
space_ is now known as Space
Space has quit [Read error: Connection reset by peer]