#general
- b
bored-art-40741
08/31/2015, 8:33 PMthanks - e
enough-analyst-54434
08/31/2015, 8:33 PMNone of any of those - b
bright-dinner-81701
08/31/2015, 8:33 PMin the design doc i added another target type - b
bright-dinner-81701
08/31/2015, 8:34 PMweb_bundle … but all it was was a wrapper for “npm run [script]” so you could depend on the product of an npm script run - w
witty-crayon-22786
08/31/2015, 8:34 PM@bright-dinner-81701: i expect one of us should focus on filling in the
directoryexamples - e
enough-analyst-54434
08/31/2015, 8:34 PMThat would be awesome - b
bright-dinner-81701
08/31/2015, 8:34 PMthat sounds like it requires the least knowledge of Pants, so maybe i should do that 😛 - b
bright-dinner-81701
08/31/2015, 8:34 PMpants’s codebase i mean - w
witty-crayon-22786
08/31/2015, 8:34 PMaye - e
enough-analyst-54434
08/31/2015, 8:34 PMI have plenty of server-side node examples from Medium both private and public - e
enough-analyst-54434
08/31/2015, 8:35 PMbut client-side webpack stuff would be nice to see examples of - b
bright-dinner-81701
08/31/2015, 8:35 PMsure - b
bright-dinner-81701
08/31/2015, 8:35 PMalso somebody wrote a comment on the design doc - b
bright-dinner-81701
08/31/2015, 8:35 PMtoday - b
bright-dinner-81701
08/31/2015, 8:35 PMabout my suggestion that we offer an option to symlink node_modules from .pants.d into the source tree - b
bright-dinner-81701
08/31/2015, 8:35 PMas a kind of hack for local dev - e
enough-analyst-54434
08/31/2015, 8:35 PMThey appear to use make + grunt for client bundling / prep - e
enough-analyst-54434
08/31/2015, 8:36 PMI'd love to avoid that hack and instead offer a changed driver - e
enough-analyst-54434
08/31/2015, 8:36 PMbut we'll see where we are when we get there - hack may be needed to bridge - b
bright-dinner-81701
08/31/2015, 8:37 PMyeah, its just attractive bc it would make all the existing node watch tools work locally with such little effort - b
bright-dinner-81701
08/31/2015, 8:37 PMbut yeah in an ideal situation pants would somehow let all that work without exposing anything linking into .pants.d - b
bored-art-40741
08/31/2015, 8:37 PMHey John or Eric, could I get another quick ack on https://rbcommons.com/s/twitter/r/2741/ ? - b
bored-art-40741
08/31/2015, 8:37 PMIt's a very quick CR - e
enough-analyst-54434
08/31/2015, 8:38 PMif we had a changed loop driver the tools would work just as seamlessly - just out of .pants.d , pants would ship changed files in there, the existing tool --watch functionality would work above that - e
enough-analyst-54434
08/31/2015, 8:38 PMBut the driver is clearly some work - w
witty-crayon-22786
08/31/2015, 8:39 PMthat was my original suggestion here... i'd still like to polish one of those. - w
witty-crayon-22786
08/31/2015, 8:40 PMi'm still worried that providing transparency into
for external tools is risky.pants.d - w
witty-crayon-22786
08/31/2015, 8:40 PM@bright-dinner-81701: where did we settle on that? that if pants supported re-running a command on change (ie, re-launching the webpack server) it would be "good enough"? - b
bright-dinner-81701
08/31/2015, 8:41 PMmaybe, although i think JS developers would see it as a degradation of their workflow - e
enough-analyst-54434
08/31/2015, 8:42 PMIt would be easy enough to provide a ReadOnlyNodePath product