<snusnu>
dkubb: the problem with using define_method for concord's #initialize is the method signature .. there's no way i know define_method can be tricked into receiving exactly as many (named) args as specified in Concord.new(:foo, :bar) .. (in that case, :foo and :bar)
<snusnu>
dkubb: doing sth like: define_method :initialize { |*names| … } won't work because names contains the actual objects passed to .new at construction time
<snusnu>
dkubb: define_method :initialize { |*@names| … } leads to a syntax error (@names would be in scope and contain the actual names)
<snusnu>
dkubb: in any case (i.e. if you can't come up with a way to use define_method) a recent commit of mine does what you suggested in your comment on github .. it minimizes string trickery inside the eval, but moves "code generation" to regular instance methods (that actually get mutated)
<mbj>
I'll look into concord soon, recovering from party yesterday
<mbj>
snusnu: time to say a short hello via skype?
snusnu has quit [Ping timeout: 252 seconds]
<mbj>
dkubb: hola
zekefast has quit [Quit: Leaving.]
<dkubb>
mbj: good morning
<mbj>
dkubb: hi
<mbj>
dkubb: I like the concord changes and will merge them in
<dkubb>
mbj: cool. I was hoping to find something a bit less verbose, but I couldn't think of anything that would make a substantial difference
<dkubb>
mbj: at first I moved the assertion into a separate method, but I realized it wouldn't be much of a savings
kpwz has quit [Ping timeout: 245 seconds]
<mbj>
dkubb: just realized you pushed to master :D, apprechiated!
kpwz has joined #rom-rb
snusnu has joined #rom-rb
<mbj>
snusnu: hola
<snusnu>
mbj: yo
<mbj>
snusnu: Hunting down that mutation :D
<snusnu>
mbj: heh awesome
<snusnu>
mbj: since dkubb refactored #define_initializer that mutant/unparser bug isn't present anymore .. you can still find the backtrace in the pastie i linked tho
<snusnu>
mbj: so yeah, it boils down to this one last mutation
<snusnu>
mbj: and i'm thrilled to see what spec you write that makes it go away … while still being feasible
<mbj>
snusnu: yeah, I just started
<mbj>
gimme some time :D
<snusnu>
sure :)
<mbj>
snusnu: consuming a movie now, laters
<snusnu>
njoy
<mbj>
snusnu: btw I released vanguard-0.0.3
<mbj>
It needs a lot of work to mutation cover it with --rspec-dm2
<mbj>
I'll come up with hybrid strategies soon.
<snusnu>
yeah, —rspec-dm2 is not optimal
<mbj>
a mixture of --rspec-unit --rspec-per-file and --rspec-dm2 should be perfect.
<mbj>
Configurable per subject-group
<snusnu>
yeah, imo it'd be ideal to cover everything with —rspec-unit "style" .. but having it way faster, if necessary with config trickery :)
<mbj>
Yeah, exactly!
<mbj>
mutant should default to --rspec-unit
<mbj>
But it should be easy to seeup slow subjects with more narrow strategies.
<snusnu>
yeah
<mbj>
snusnu: I'm thinking off removing that strategy concept all together
<mbj>
So it should depend on your config, defaulting to --rspec-unit style.
<snusnu>
rspec-dm2 really gets annoying with bigger refactorings .. like with substation currently, i really start hating how many fail, just because they're there .. not because they were necessary in the first place
<snusnu>
yeah
<mbj>
BTW on the longterm I need an rspec replacement. Rspec is not COW friendly :D
<mbj>
Or lets say "efficient to use with fork(2)
<mbj>
"
<mbj>
but anyways, consuming movie, later. 2h.
<snusnu>
have fun
zekefast has joined #rom-rb
mbj has quit [Ping timeout: 264 seconds]
zekefast1 has joined #rom-rb
zekefast1 has quit [Client Quit]
zekefast has quit [Ping timeout: 246 seconds]
zekefast has joined #rom-rb
mbj has joined #rom-rb
solnic has joined #rom-rb
<solnic>
snusnu: "rspec-dm2 really gets annoying with bigger refactorings" <== yup it sucks, I told you we had too many specs because of this approach
<solnic>
I use --rspec-unit now
<solnic>
I don't even consider going back to rspec-dm2
postmodern has joined #rom-rb
<solnic>
snusnu: I went through hell refactoring virtus few months ago, I'm planning to rewrite the entire spec suite now and simplify it
<solnic>
anyhow, gotta run, ttyl
<mbj>
snusnu: bac
<mbj>
k
solnic has quit [Ping timeout: 240 seconds]
zekefast has quit [Ping timeout: 268 seconds]
zekefast has joined #rom-rb
<mbj>
snusnu: Killed mutation, IMHO this was a ruby bug!
<snusnu>
mbj: yeah, i was thinking about doing the same, but i thought maybe i'm missing something