pantsbuild #general

Channels

welcome

github-notifications

announce

witty-crayon-22786

09/30/2015, 10:07 PM

@full-kilobyte-93429: a simple log entry to dump the classpath, prefixed with how it was shortened.

full-kilobyte-93429

09/30/2015, 10:08 PM

sounds good

bored-art-40741

09/30/2015, 10:24 PM

Are you guys also working on the presentation layer for isolated?

bored-art-40741

09/30/2015, 10:24 PM

I see a lot of chatter about users really not liking to scroll up to find breaks

witty-crayon-22786

09/30/2015, 10:25 PM

we are not working on it. @fancy-queen-20734 had a doc about it

witty-crayon-22786

09/30/2015, 10:25 PM

@bored-art-40741: if they don't like scrolling up, tell them to rerun the command (as a workaround)

witty-crayon-22786

09/30/2015, 10:25 PM

(but yea)

bored-art-40741

09/30/2015, 10:25 PM

Yeah that's the workaround we suggest, but it's not a good long term solution

witty-crayon-22786

09/30/2015, 10:26 PM

@bored-art-40741: primary question is: pre or post engine refactor>?

bored-art-40741

09/30/2015, 10:26 PM

Sorry?

witty-crayon-22786

09/30/2015, 10:27 PM

does overhauling the output of a task that is currently internally concurrent, but will soon be externally concurrent make sense

bored-art-40741

09/30/2015, 10:27 PM

Oh yes

bored-art-40741

09/30/2015, 10:27 PM

The engine refactor is still long term

bored-art-40741

09/30/2015, 10:27 PM

I would consider this to be a follow-up on isolated shipping

witty-crayon-22786

09/30/2015, 10:27 PM

@bored-art-40741: sidenote: did you get to meet up with dan?

bored-art-40741

09/30/2015, 10:27 PM

Yup we did, very interesting convo

witty-crayon-22786

09/30/2015, 10:28 PM

great. yea, we're very invested in that.

witty-crayon-22786

09/30/2015, 11:18 PM

@bored-art-40741: anyway, to be crystal clear: we'd love better output and will happily review/accept it, but our focus will be the engine refactor

ancient-stone-50795

10/01/2015, 2:08 AM

I know there are a lot of differences between our environments, but we don’t have to scroll up very far to find problems with the abbreviated console output configured:

Copy code

[reporting]
# This cuts down on the verbosity of pants output
# Comment out these lines to see warnings and other output from javac
console_label_format: {
    'COMPILER' : 'SUPPRESS',
  }
console_tool_output_format: {
    'COMPILER' : 'SUPPRESS',
  }

ancient-stone-50795

10/01/2015, 2:09 AM

When something fails in our environment, there are usually just a few more modules that run after, and each only prints out one line of output.

bored-art-40741

10/01/2015, 2:10 AM

Hm, ours definitely have more than one line of output (each zinc invocation is at least 4), but we tend to have dozens or even hundreds of targets compiling after the first failed one

ancient-stone-50795

10/01/2015, 2:10 AM

Maybe I just haven’t put my bugs deep enough in the codebase yet...

bored-art-40741

10/01/2015, 3:31 AM

Thanks for the tip on those options, Eric. This might be quiet enough that we'll be okay with reporting as-is

fast-megabyte-33579

10/01/2015, 1:01 PM

Afternoon everyone - I'm in the process of upgrading our pants to 0.0.50 and we're having trouble loading the scrooge and node plugins - I can't see anyone on the mailing list having the same problem

fast-megabyte-33579

10/01/2015, 1:03 PM

I think this is because I tried out the 'setup' pants script and that doesn't install any other dependencies (like

pants.contrib.scrooge

pants.contrib.node

)

fast-megabyte-33579

10/01/2015, 1:04 PM

ah and I didn't read the docs

fast-megabyte-33579

10/01/2015, 1:10 PM

okay, now I'm getting this:

Could not find requested plugins: [u'pantsbuild.pants.contrib.scrooge==0.0.50', u'pantsbuild.pants.contrib.node==0.0.50']

fast-megabyte-33579

10/01/2015, 1:11 PM

but they do exist in

.cache/pants/plugins

fast-megabyte-33579

10/01/2015, 1:38 PM

ah it was this https://github.com/pantsbuild/pants/issues/1803

fast-megabyte-33579

10/01/2015, 1:43 PM

yay, all working smoothly