I realize this is pretty late in the game but our compute costs are piling up and I want to gain a foothold so we devote more resources to our inference budget.
Our software and as a result, our CI pipeline currently relies on nodejs and npm. I've been watching it run and an outsize portion of CI involves needlessly slogging through the process of installing npm packages etc and starting node.
I say needlessly because bun is much faster.. This means CI runs faster and we save dramatically on compute costs.
This is purely a deep research task. It seems straightforward to me, but there may be something I'm missing.
So examine the issue deeply, let's assess how much work is involved. If possible, maybe run a simulation comparing the time in bun to the time in node and also seeing if anything breaks, then report your findings in a BUN_MIGRATION_REPORT.md
No need to alter code here unless it's both trivial and warranted, meaning it was clean with no issues and actually saved time rather than causing headaches.
I realize this is pretty late in the game but our compute costs are piling up and I want to gain a foothold so we devote more resources to our inference budget.
Our software and as a result, our CI pipeline currently relies on nodejs and npm. I've been watching it run and an outsize portion of CI involves needlessly slogging through the process of installing npm packages etc and starting node.
I say needlessly because bun is much faster.. This means CI runs faster and we save dramatically on compute costs.
This is purely a deep research task. It seems straightforward to me, but there may be something I'm missing.
So examine the issue deeply, let's assess how much work is involved. If possible, maybe run a simulation comparing the time in bun to the time in node and also seeing if anything breaks, then report your findings in a BUN_MIGRATION_REPORT.md
No need to alter code here unless it's both trivial and warranted, meaning it was clean with no issues and actually saved time rather than causing headaches.