#general
- a
ancient-stone-50795
07/24/2015, 1:37 AMWe unpack the jars and use them as source inputs to codegen. - a
ancient-stone-50795
07/24/2015, 1:37 AMThere is an example checked in to the pants repo. - w
witty-crayon-22786
07/24/2015, 1:37 AMmy second statement is the one i was actually asking - a
ancient-stone-50795
07/24/2015, 1:38 AMIts moot, because we aren’t using binary data in the jars we import and they don’t intersect with the set of jars that we resolve for libraries - w
witty-crayon-22786
07/24/2015, 1:38 AMok, that's what i was hoping. - w
witty-crayon-22786
07/24/2015, 1:38 AMoh, actually - w
witty-crayon-22786
07/24/2015, 1:38 AMnew question: - w
witty-crayon-22786
07/24/2015, 1:39 AMdoes it matter that you get independent ivy resolves per target? - w
witty-crayon-22786
07/24/2015, 1:39 AMin the imports phase - a
ancient-stone-50795
07/24/2015, 1:40 AMNo. It would be great if there were just one ivy resolve as far as I’m concerned. It would be faster and I added a TODO saying as much. We don’t specify different versions of the same jar in different targets. - w
witty-crayon-22786
07/24/2015, 1:40 AMok, cool. - w
witty-crayon-22786
07/24/2015, 1:40 AMyea, because i would like mapjars to just use the output of ivy_resolve directyly - w
witty-crayon-22786
07/24/2015, 1:40 AMie, the same type of classpath that ends up in compile_classpath - a
ancient-stone-50795
07/24/2015, 1:41 AMI’m fine with that if you mean ivy_resolve() the method, but there still needs to be a task that runs before the ivy_resolve task (and before the codegen group) - w
witty-crayon-22786
07/24/2015, 1:41 AM(ivy_resolve the method, not resolve.ivy the task) - w
witty-crayon-22786
07/24/2015, 1:41 AMCoke! - a
ancient-stone-50795
07/24/2015, 1:41 AMsimple smile - w
witty-crayon-22786
07/24/2015, 3:22 AM@happy-kitchen-89482, @bored-art-40741 : are you guys using name hashing or scala 2.11 yet? - b
bored-art-40741
07/24/2015, 3:22 AMNeither, unfortunately - w
witty-crayon-22786
07/24/2015, 3:23 AMk - b
bored-art-40741
07/24/2015, 3:23 AMWe're still having issues with analysis/missing dep detection on name hashing - w
witty-crayon-22786
07/24/2015, 3:23 AMmm - w
witty-crayon-22786
07/24/2015, 3:23 AMyea, i just saw an issue with the parser encountering non-ascii data in the analysis too. - b
bored-art-40741
07/24/2015, 3:23 AMOh, I thought we fixed that one - w
witty-crayon-22786
07/24/2015, 3:23 AMwith name hashing - w
witty-crayon-22786
07/24/2015, 3:24 AMoh... maybe? this release is about a month old - w
witty-crayon-22786
07/24/2015, 3:24 AMlink? - b
bored-art-40741
07/24/2015, 3:24 AMMuch earlier than that - w
witty-crayon-22786
07/24/2015, 3:24 AMhm - b
bored-art-40741
07/24/2015, 3:24 AMBut the tests for analysis now specifically include injection of non-ascii characters into the anonymized analysis