https://kotlinlang.org logo
Channels
Powered by
#kobalt
  • c

    christophsturm

    12/09/2015, 8:06 PM
    the contribute page says the main class is MainKt, but i find only Main. or does kotlin add the kt?
  • c

    cedric

    12/09/2015, 8:07 PM
    It does
  • e

    evanchooly

    12/09/2015, 8:08 PM
    it used to be BlahPackage but that was confusing so they changed it some other form of confusion
  • c

    christophsturm

    12/09/2015, 8:19 PM
  • c

    cedric

    12/09/2015, 8:26 PM
    Added a comment
  • c

    cedric

    12/10/2015, 4:21 PM
    @christophsturm: Merged, thanks!
  • c

    christophsturm

    12/10/2015, 4:21 PM
    yay!
  • c

    cedric

    12/10/2015, 10:47 PM
    @evanchooly: I just pushed a change that impacts 
    MavenId
    . The main modification is that the constructor is private and creation is done through the factory 
    create()
    . Read the comments if you’re curious about the justification for this change.
  • e

    evanchooly

    12/11/2015, 1:35 AM
    yeah, i saw that change coming. hopefully it doesn’t fuck me up too much. simple smile
  • c

    cedric

    12/11/2015, 1:43 AM
    It shouldn’t, take a look at the current source, it’s pretty straightforward
  • p

    pabl0rg

    12/11/2015, 3:08 AM
    Hi Cédric, i deleted ~/.kobalt and things worked. Thanks!
  • c

    cedric

    12/11/2015, 3:09 AM
    Really? That's a bit odd
  • c

    cedric

    12/11/2015, 3:09 AM
    @pabl0rg: So your PR is no longer necessary?
  • e

    evanchooly

    12/11/2015, 3:13 AM
    @cedric: which PR was it?
  • c

    cedric

    12/11/2015, 3:14 AM
  • e

    evanchooly

    12/11/2015, 3:15 AM
    interesting. i was just about to make that exact same change for different reasons simple smile
  • c

    cedric

    12/11/2015, 3:15 AM
    I've been trying to reproduce the error @pabl0rg saw but to no avail so far
  • c

    cedric

    12/11/2015, 3:15 AM
    Oh interesting, do you have a failing test case for me?
  • e

    evanchooly

    12/11/2015, 3:15 AM
    dependency resolution seems overly convoluted to me as it stands.
  • c

    cedric

    12/11/2015, 3:16 AM
    Kobalt's implementation you mean?
  • e

    evanchooly

    12/11/2015, 3:16 AM
    no there are just vars where there don’t need to be and making them val would be cleaner in places.
  • e

    evanchooly

    12/11/2015, 3:16 AM
    yeah. in kobalt.
  • c

    cedric

    12/11/2015, 3:16 AM
    Looking forward to your fixes simple smile
  • e

    evanchooly

    12/11/2015, 3:16 AM
    we’re making multiple calls to findCorrectRepo() e.g.
  • e

    evanchooly

    12/11/2015, 3:17 AM
    ultimate because we’re passing MavenId around everywhere instead of passing around RepoResult once we find it
  • e

    evanchooly

    12/11/2015, 3:17 AM
    i think...
  • c

    cedric

    12/11/2015, 3:17 AM
    Multiple calls should be cheap because ultimately, there will only be one network call and the others will wait for it, but go ahead, improvements are improvements
  • e

    evanchooly

    12/11/2015, 3:17 AM
    yeah, i need to dig in to it. you could very well be right.
  • c

    cedric

    12/11/2015, 3:18 AM
    I'm heavily relying on 
    LoadingCache
    to be smart where I'm not, here
  • e

    evanchooly

    12/11/2015, 3:18 AM
    the first call is probably the failed check on the local repo and the second inside MavenDependency where we do the actual fetch
1...139140141...826Latest