Ergonomic, hermetic, user-friendly software build system for Python, Go, Java, Kotlin, Protobuf, Scala, Shell. Pants lets you fearlessly scale up your codebase!

Pants

but how abotu in the cumulative time per function view?

i think.. theres a parallel there to showing “what a threadpool worker is doing” in terms of “time spent displayed in the UI” is the visual representation of “cumulative time per function in a profiler”

yes. and we don't have "cumulative time per function in a profiler" for async code

isnt that what the labeling and timing bit you’re working on does tho?

it would definitely look.. choppy.. in the UI - much like it would in a visual profiler

because of the smaller units of execution

sorry, i'm going to focus on this right now

sorry, just trying to get on the same page

it all boils down to: "there are thousands of stacks, which do you render"

i think i have an idea that might work, but... yea.

<@U054MQUPL>: sorry. i shouldn't have implied that this is clear cut. it's not. it's possible that sampling synthetic stacks from threads could work.

if i see `Exception message: 'ascii' codec can't encode character u'\u2018' in position 979: ordinal not in range(128)`, is there an expression with `.decode()` or `.encode()` that should likely be happening somewhere? it comes up in ci environments sometimes when there's an error when running the pex `Installer` class, and i can see where to change, but not what to change

hard to say for sure with just that one liner but I’d guess that relates to default character encoding at the sources level? <http://www.fileformat.info/info/unicode/char/2018/index.htm> is the problematic char (at char 979 of whatever source file)

<https://www.python.org/dev/peps/pep-0263/#defining-the-encoding>

&gt; Python will default to ASCII as standard encoding if no other encoding hints are given.

so if someone wanted to put utf-8 chars in their source code they’d have to declare `# -*- coding: utf-8 -*-`

thanks for the link! i asked because i've tried to answer this in a previous instantiation of the problem and spent many many hours, and got a solution which is probably stashed in my pex clone now that i think about it, but that was over a month ago

hm, we may not be doing that in our setup.py in testprojects

or maybe we want to `s/Unicode Character 'LEFT SINGLE QUOTATION MARK' /'/g`?

pretty easy to accidentall paste likewise utf-8 chars thinking you’re copying ascii

hm, we are. i figured out the error without being able to see the stderr this time