#general
- f
freezing-motorcycle-14222
09/29/2015, 3:56 PMOh yeah, and there were some other things, too - f
freezing-motorcycle-14222
09/29/2015, 3:57 PMFor example,
always bootstraps the JVM tools when running./pants
, even for a non-JVM targetbench - f
freezing-motorcycle-14222
09/29/2015, 3:58 PMIf nobody objects,then I will go through and fix the JVM tasks - b
bored-art-40741
09/29/2015, 3:58 PMYeah, we are very JVM centric, and there's no particular reason except history - b
bored-art-40741
09/29/2015, 3:59 PMSince this horse will never die: the new engine helps here - f
freezing-motorcycle-14222
09/29/2015, 3:59 PMYou mean the stuff in
?src/python/pants/engine/ - b
bored-art-40741
09/29/2015, 4:00 PMI think that's the round engine? Unless someone has been doing work on the one one - b
bored-art-40741
09/29/2015, 4:00 PMNo, "new engine" refers to the rather radical refactor we've been cooking in our heads for the past 6-12 months - b
bored-art-40741
09/29/2015, 4:00 PMIt's not something you should block on - b
bored-art-40741
09/29/2015, 4:00 PMI was just mentioning it - f
freezing-motorcycle-14222
09/29/2015, 4:00 PMThen I won't block on it, although I'm still curious about what the radical refactor would look like - f
freezing-motorcycle-14222
09/29/2015, 4:01 PMIs there a write-up about it somewhere? - b
bored-art-40741
09/29/2015, 4:01 PMI think there's a doc. If there isn't, one is definitely coming up - b
bored-art-40741
09/29/2015, 4:02 PMIf you want to refactor certain tasks/goals to make them less JVM specific, you might want to consider breaking more things out into the product model, and it might help to write up a doc of the current products/tasks and how you imagine a refactor would look - b
bored-art-40741
09/29/2015, 4:03 PMThere's a good deal of not-entirely-surfaced design principle that goes into pants' structure, and it's good to make sure everyone is on the same page before committing to big changes - f
freezing-motorcycle-14222
09/29/2015, 4:06 PMOn that subject, would there be interest in me adding more detailed `README`s explaining the architecture of underlying components (i.e. a better
for example)?src/python/pants/option/README.md - f
freezing-motorcycle-14222
09/29/2015, 4:06 PMI've been learning a lot about these components while digging through internals and thought it might help to document what I learned - b
bored-art-40741
09/29/2015, 4:06 PMAbsolutely - f
freezing-motorcycle-14222
09/29/2015, 4:06 PMAlright, then I'll also take a stab at that, too - b
bored-art-40741
09/29/2015, 4:08 PMIntegrating new languages is going to present some interesting challenges - f
freezing-motorcycle-14222
09/29/2015, 4:20 PMI think an easy way to detect these sorts of issues is to add a reserved target that is not associated with any tasks and running various pants goals on that target - b
bored-art-40741
09/29/2015, 4:22 PMSorry, I don't follow - b
bored-art-40741
09/29/2015, 4:23 PMDo you mean
?target(...) - b
bored-art-40741
09/29/2015, 4:23 PMPreviously known asdependencies - w
witty-crayon-22786
09/29/2015, 4:28 PM@freezing-motorcycle-14222: to resolve the
issue, take a look at how it is solved for./pants bench./pants repl - w
witty-crayon-22786
09/29/2015, 4:29 PMand probably rename the existing
task tobench
, since we'll want to make room forbench.caliper
anywayjmh - f
freezing-motorcycle-14222
09/29/2015, 4:47 PMYeah, that was the other reason I was looking into this - f
freezing-motorcycle-14222
09/29/2015, 4:47 PMMy team wanted to add
benchmarks to our projectsjmh - f
freezing-motorcycle-14222
09/29/2015, 4:48 PM@bored-art-40741: I mean the sort of target that you put under the
directory of a backend. I might be using the wrong wordtargets/ - b
bored-art-40741
09/29/2015, 4:50 PMRight, that's any subclass of
really. How does that help detect issues with new languages though?Target