#general
- h
happy-kitchen-89482
07/13/2015, 9:51 PM - h
happy-kitchen-89482
07/13/2015, 9:51 PMOK, what you actually need falls out of those previous changes relatively easily - e
enough-analyst-54434
07/13/2015, 10:27 PM@mammoth-autumn-2551: unfortunately the guidance is provided by deprecation warnings on the console, but you probably skipped past these - e
enough-analyst-54434
07/13/2015, 10:28 PMie, if you had upgraded to 0.0.33 or 0.0.34, .
would have spewed warnings that described corrective measures you could have taken in advance of 0.0.35/pants list - e
enough-analyst-54434
07/13/2015, 10:29 PMIt would be good to get deprecation help text in the CHANGELOG where the deprecation is introduced - e
enough-analyst-54434
07/13/2015, 10:29 PMPresumably if we had that, you'd have found it in a ~natural way - e
enough-analyst-54434
07/13/2015, 10:32 PM@mammoth-autumn-2551: this has 1 more example in tests/java/com/twitter/common/metrics/BUILD, killing with_description in favor of the description kwarg: https://github.com/twitter/commons/commit/66fc3a07ff77094cb7f09d340cf6511877539786 - m
mammoth-autumn-2551
07/13/2015, 10:33 PMThanks @enough-analyst-54434 I think I’ve migrated all our existing BUILD files - e
enough-analyst-54434
07/13/2015, 10:34 PMYW. For the type of advice we spew to console, see, for example, this change - LHS of src/python/pants/backend/core/wrapped_globs.py and src/python/pants/backend/python/tasks/pytest_run.py - the `hint_message`: https://github.com/pantsbuild/pants/commit/69a95616bc5ab498166adf939e5634e6fa2ff0ae - w
witty-crayon-22786
07/13/2015, 10:51 PMsergey will be following up to continue to specialize grouptask in https://rbcommons.com/s/twitter/r/2462/ and others... does anyone have objections to getting https://rbcommons.com/s/twitter/r/2452/ in now? - e
enough-analyst-54434
07/13/2015, 10:55 PMWell, I do but they're not super strong - I just don't understand the desire to special GroupTask when it has no natural mapping to an isolated pure zinc compile at all - w
witty-crayon-22786
07/13/2015, 11:00 PMwell, it probably wouldn't be called GroupTask anymore. - w
witty-crayon-22786
07/13/2015, 11:00 PMbut members are still needed - e
enough-analyst-54434
07/13/2015, 11:01 PMThe latter I do not understand - e
enough-analyst-54434
07/13/2015, 11:01 PMzinc gets handed a target - e
enough-analyst-54434
07/13/2015, 11:01 PMits java or its scala - e
enough-analyst-54434
07/13/2015, 11:02 PMwhy is anything special needed? - w
witty-crayon-22786
07/13/2015, 11:02 PMor apt - e
enough-analyst-54434
07/13/2015, 11:02 PM== java - w
witty-crayon-22786
07/13/2015, 11:02 PMwe don't have support for using zinc for apt though. - e
enough-analyst-54434
07/13/2015, 11:02 PMfalse - e
enough-analyst-54434
07/13/2015, 11:02 PMso - zinc calls javac - e
enough-analyst-54434
07/13/2015, 11:03 PMjavac handles both 1.) compiling apt-processors (like twitter/common/args) 2.) running apt-processors that have been compiled (lik t.c.args) - e
enough-analyst-54434
07/13/2015, 11:03 PMor am I having that wrong? - w
witty-crayon-22786
07/13/2015, 11:04 PMhaven't we done something in jmake to capture annotation processor file outputs? - e
enough-analyst-54434
07/13/2015, 11:04 PMno - w
witty-crayon-22786
07/13/2015, 11:04 PMhm - e
enough-analyst-54434
07/13/2015, 11:04 PM2 things - w
witty-crayon-22786
07/13/2015, 11:04 PMok. that might be an option then. i don't think you'd be able to avoid having a parent class for the isolated and global JvmCompiles though - e
enough-analyst-54434
07/13/2015, 11:05 PM1. you added handling for generated sources + classes just days ago in python pants code