Open schneems opened 12 months ago
Thanks for the report. It seems like Flowistry is generating 1.2GB of data for this one function, which is ... too much. I will investigate and try to improve the performance.
Thanks for the plugin, your work, and reviewing my issue.
Other files take a long time but eventually successfully complete. I tried it on user_errors successfully (in the log I think too). Maybe there is something helpful in the delta between those two files that might point to the deeper underlying issue.
Also to add I have commit to libcnb.rs if there’s a small modification there that you find that could significantly improve performance, it might be possible to patch there while we wait on the longer term investigation and fix.
Problem
I'm trying to use flowistry with the https://github.com/heroku/buildpacks-ruby repo on this commit:
When I try to trigger flowestry in the main.rs I get an error:
This might be because the
src/main.rs
contains code for what will eventually become two binaries. The code infn detect
andfn build
will be built into two separate binaries. This behavior comes from libcnb.rs. Docs: https://github.com/heroku/libcnb.rs/blob/6cd51ce91b58cebf72ed8e8513d52871b6106f0f/README.md#L134-L142.Running the command:
Takes quite a long time and it's then followed by a really long string.
Logs
OS: darwin (23.1.0) VSCode: 1.84.2 Error message
Full log: https://paste.rs/5gPs1