#general
- f
fast-megabyte-33579
10/01/2015, 1:43 PMthanks guys simple smile - f
fast-megabyte-33579
10/01/2015, 1:52 PMokay, almost all working smoothly - now it's complaining about not finding the 'scala_jar' definition - f
fast-megabyte-33579
10/01/2015, 1:53 PMI guess that went in after 0.0.50 was released - e
enough-analyst-54434
10/01/2015, 2:03 PMYes, exactly. The
will be part of the 0.0.51 release tomorrow.scala_jar - f
fast-megabyte-33579
10/01/2015, 2:29 PMso excited simple smile - f
fast-megabyte-33579
10/01/2015, 2:29 PMDoes anyone know what
might mean?'WikiArtifact' object has no attribute 'fingerprint' - f
fast-megabyte-33579
10/01/2015, 2:29 PMit looks like it hasn't had that attribute since 2014 - in this commit https://github.com/pantsbuild/pants/commit/7bcd00e8c7c98719055411bce758d04e838a8b45 - f
fast-megabyte-33579
10/01/2015, 2:37 PMoh, I wonder if it's because I'm trying to compile .../:: - b
bored-art-40741
10/01/2015, 3:22 PM@fancy-queen-20734: I dropped some comments on https://rbcommons.com/s/twitter/r/2853/ - b
bored-art-40741
10/01/2015, 3:23 PMI'm concerned that this has broken artifact caching for zinc compilation - w
witty-crayon-22786
10/01/2015, 4:45 PM@bored-art-40741: we've cherry picked that one back and released it - b
bored-art-40741
10/01/2015, 4:45 PMEffectively meaning you needed to revert? - w
witty-crayon-22786
10/01/2015, 4:45 PMno: we wanted it in 0.0.47 - b
bored-art-40741
10/01/2015, 4:45 PMAh - w
witty-crayon-22786
10/01/2015, 4:46 PMso we cherry picked it back from master - b
bored-art-40741
10/01/2015, 4:46 PMI patched it out of our release and just pushed the release - b
bored-art-40741
10/01/2015, 4:46 PMGetting rid of it seems to have fixed the problem - w
witty-crayon-22786
10/01/2015, 4:46 PMit seems to be working quite well, but we don't have as precise a handle on hitrates as you might - b
bored-art-40741
10/01/2015, 4:46 PMI strongly believe that code is broken, and in a subtle and dangerous way - w
witty-crayon-22786
10/01/2015, 4:46 PMhm - w
witty-crayon-22786
10/01/2015, 4:46 PMhow so? - b
bored-art-40741
10/01/2015, 4:47 PMThat is, it only breaks when the order of certain upstream values change, and those probably only change due to goofy things like different versions/platforms of python - b
bored-art-40741
10/01/2015, 4:47 PMSee my comment on the RB - b
bored-art-40741
10/01/2015, 4:47 PMBasically I don't think that code should be considered correct until there's a strong line of reasoning proving the order is stable end to end - w
witty-crayon-22786
10/01/2015, 4:47 PMonly if there is a portion of the code where we're not using OrderedSet for deps, right? - w
witty-crayon-22786
10/01/2015, 4:47 PMivy is ordered - b
bored-art-40741
10/01/2015, 4:47 PMThat doesn't save you from whatever is populating the product - w
witty-crayon-22786
10/01/2015, 4:47 PMand ClasspathProducts is definitely ordered - b
bored-art-40741
10/01/2015, 4:48 PMOur resolver isn't necessarily ordered in its output - b
bored-art-40741
10/01/2015, 4:48 PMNor is that part of the contract AFAIK