00:55
postmodern has joined #rom-rb
01:00
cored has quit [Ping timeout: 264 seconds]
01:34
mbj_ has joined #rom-rb
01:37
mbj has quit [Ping timeout: 260 seconds]
03:49
misfo has joined #rom-rb
04:04
misfo has quit [Quit: misfo]
04:06
misfo has joined #rom-rb
04:36
misfo has quit [Quit: misfo]
04:49
postmodern has quit [Quit: Leaving]
05:49
postmodern has joined #rom-rb
09:55
zekefast has joined #rom-rb
10:14
misfo has joined #rom-rb
10:33
basex has joined #rom-rb
11:09
misfo has quit [Quit: misfo]
11:12
cored has joined #rom-rb
11:12
cored has quit [Changing host]
11:12
cored has joined #rom-rb
11:41
solnic has left #rom-rb [#rom-rb]
11:47
mbj_ is now known as mbj
12:16
cored has quit [Ping timeout: 264 seconds]
12:42
basex has quit [Quit: basex]
12:44
basex has joined #rom-rb
12:45
misfo has joined #rom-rb
12:56
solnic has joined #rom-rb
12:56
<
solnic >
mbj: hey, can you release new version of descendants_tracker? I prepared everything, just pull, build the gem and push it
13:31
misfo has quit [Quit: misfo]
13:39
misfo has joined #rom-rb
13:51
misfo has quit [Quit: misfo]
14:12
misfo has joined #rom-rb
14:16
skade has joined #rom-rb
14:24
misfo has quit [Quit: misfo]
14:28
<
mbj >
solnic: I merged PR #5 with licence metadata and just released 0.0.3 with no additional changes.
14:30
basex has quit [Ping timeout: 256 seconds]
14:33
mbj has quit [Ping timeout: 246 seconds]
15:48
_whitelogger has joined #rom-rb
15:59
<
dbussink >
solnic: around?
16:03
<
dbussink >
mbj: g'day
16:04
<
mbj >
dbussink: You have a rom question? Or a virtus one?
16:04
<
mbj >
dbussink: Better: Can I help you?
16:06
<
mbj >
dbussink: I also ran into this once. But I dont have a more narrow reproduction, surry.
16:07
<
mbj >
dbussink: BTW congratz on rbx 2.0 ;)
16:07
<
dbussink >
mbj: thnx
16:11
<
dbussink >
mbj: the thing is, i'm pretty puzzled still why it even works in mri :)
16:12
<
mbj >
dbussink: heh
16:12
<
mbj >
dbussink: I had no problem to track down behavior in rbx, and jruby.
16:12
<
mbj >
dbussink: But I cannot easily follow MRI code.
16:12
<
dbussink >
well, it's more a general thing as in why does this ruby code work
16:13
<
dbussink >
not about mri internals
16:13
<
mbj >
dbussink: Ahh, you are talking about the issue above.
16:13
<
mbj >
dbussink: Not about 2.0 compat
16:14
<
mbj >
dbussink: Dkubb and solnic worked on axiom-types, I for myself did not had the time to participate.
16:14
dkubb has joined #rom-rb
16:14
<
mbj >
dbussink: Dunno any internals.
16:14
<
dkubb >
good morning
16:14
<
dkubb >
dbussink: congrats on rubinius 2.0 being released
16:16
<
mbj >
dkubb: Question, did you ever understood what ducktrap is?
16:16
<
mbj >
dkubb: In case the answer is "No", I plan to test the upcoming readme on your brain ;)
16:17
<
dkubb >
mbj: I only have a vague understanding
16:17
<
mbj >
dkubb: perfekt.
16:17
<
mbj >
dkubb: I'm writing the short and long version of the reasons behind this lib. My wife proof reads it during the next week.
16:18
<
mbj >
dkubb: So I could need some reviews ;)
16:19
<
mbj >
But now, back to clients work ;)
16:25
travis-ci has joined #rom-rb
16:25
<
travis-ci >
[travis-ci] dkubb/abstract_type#59 (master - 59d1640 : Dan Kubb): The build was broken.
16:25
travis-ci has left #rom-rb [#rom-rb]
16:33
misfo has joined #rom-rb
16:44
misfo has quit [Quit: misfo]
16:48
<
dkubb >
mbj: wdyt about making it so rubocop warnings fail the build?
16:49
<
mbj >
mom, my coffe had fallen on the desk
16:52
<
dbussink >
mbj: ah, found it
16:52
<
dbussink >
we use #coerce as a method in rubinius internally
16:52
travis-ci has joined #rom-rb
16:52
<
travis-ci >
[travis-ci] dkubb/abstract_type#61 (master - 6b1cc80 : Dan Kubb): The build was broken.
16:52
travis-ci has left #rom-rb [#rom-rb]
16:52
<
dbussink >
so here axiom-types overwrites that coerce method
16:52
<
dbussink >
and it spirals into that infinite loop
16:53
postmodern has quit [Quit: Leaving]
16:56
<
mbj >
dbussink: interesting
16:56
<
mbj >
dbussink: how to solve it? From my initial feelings I think rbx should move.
16:57
<
mbj >
dbussink: rename to something like __rbx_coerce ?
16:57
<
dkubb >
mbj: I think #coerce is actually a ruby convention/interface
16:57
<
dbussink >
mbj: well, there are default coerce methods
16:57
<
dkubb >
it returns an Array of two objects
16:57
<
dbussink >
so we can't do that
16:57
<
dbussink >
since then that wouldn't work
16:57
<
dbussink >
basically here axiom-types overwrites it and breaks it
16:57
<
dbussink >
it mri you can away with it
16:58
<
dbussink >
it probably calls the original c method under the hood
16:58
<
dbussink >
the coerce method also uses a case statement
16:58
<
dbussink >
that ends up using === and this ==
16:58
<
dbussink >
and <=> for comparable
16:58
<
dbussink >
dkubb: yeah
16:59
<
dbussink >
dkubb: yeah
17:00
<
dkubb >
I could change that to an if/else/elsif statement and use kind_of? instead
17:00
<
dbussink >
dkubb: yeah, you probably have to
17:00
<
dkubb >
I think I just used a case statement because it resulted in nice compact code
17:01
<
dbussink >
dkubb: since those are more primitive in a sense
17:01
<
dbussink >
dkubb: you got lucky ;)
17:01
<
dbussink >
all coerce code in rbx itself uses instance_of? / kind_of
17:01
<
dkubb >
I suppose I could also change it to ->(object) { object.kind_of?(self) }
17:01
<
dbussink >
exactly because of reasons like this
17:01
<
mbj >
Gonna start coding for client again. And to focus I have to quit IRC.
17:01
mbj has quit [Quit: leaving]
17:01
<
dbussink >
since coercion protocols for <=> etc. depend on this lower level behavior
17:06
<
dbussink >
dkubb: well, question is whether it's worth the effort
17:06
<
dbussink >
we wouldn't hold anyone back on it
17:06
<
dbussink >
but probably won't do it outselves
17:06
misfo has joined #rom-rb
17:08
misfo has quit [Client Quit]
17:09
misfo has joined #rom-rb
17:15
<
dkubb >
dbussink: I don't know what the rbx parser looks like atm, but it's been really nice for some of the tools we use to have something stand-alone like this
17:15
<
dkubb >
dbussink: I just wondered if anyone has brought it up with you before
17:16
<
dbussink >
dkubb: yeah, it's come up
17:16
<
dbussink >
dkubb: basically our view has been do it and show us ;)
17:17
<
dbussink >
but i gotta run
17:17
<
dbussink >
back later
17:20
misfo has quit [Quit: misfo]
18:07
zekefast has quit [Quit: Leaving.]
18:10
misfo has joined #rom-rb
18:28
mbj has joined #rom-rb
19:11
zekefast has joined #rom-rb
19:37
dkubb has joined #rom-rb
20:44
mbj has quit [Ping timeout: 245 seconds]
21:03
zekefast has quit [Quit: Leaving.]
21:17
zekefast has joined #rom-rb
21:30
skade has joined #rom-rb
21:57
snusnu has quit [Ping timeout: 264 seconds]
22:19
<
dbussink >
dkubb: wouldn't this be an example of using methods in a method that are of the same level of abstraction :)
22:19
<
dbussink >
dkubb: although that can be pretty ill defined ;)
22:38
zekefast has quit [Quit: Leaving.]
22:38
zekefast has joined #rom-rb
23:23
<
dkubb >
dbussink: do you mean the coerce stuff in axiom-types?