It was incorrect metric. Such metric is totally useless.
The original problem was the fact that in new node 18.20.1 prints output gathered from react-native-builder-bob:
> react-native-keyboard-controller@1.11.6 prepare
> bob build
ā¹ Building target commonjs
ā¹ Cleaning up previous build at lib/commonjs
ā¹ Compiling 32 files in src with babel
ā Wrote files to lib/commonjs
ā¹ Building target module
ā¹ Cleaning up previous build at lib/module
ā¹ Compiling 32 files in src with babel
ā Wrote files to lib/module
ā¹ Building target typescript
ā¹ Cleaning up previous build at lib/typescript
ā¹ Generating type definitions with tsc
And such output can not be parsed.
The interesting thing is that 18.19.0 version produced expected result. So in this PR I locked node version for this particular job.
š¢ Changelog
CI
lock node to 18.19.0 for size-diff job;
š¤ How Has This Been Tested?
Tested on CI.
šø Screenshots (if appropriate):
š Checklist
[x] CI successfully passed
[x] I added new mocks and corresponding unit-tests if library API was changed
š Description
Fixed "0 bytes" diff from size-diff job report.
š” Motivation and Context
It was incorrect metric. Such metric is totally useless.
The original problem was the fact that in new node
18.20.1
prints output gathered fromreact-native-builder-bob
:And such output can not be parsed.
The interesting thing is that
18.19.0
version produced expected result. So in this PR I locked node version for this particular job.š¢ Changelog
CI
18.19.0
forsize-diff
job;š¤ How Has This Been Tested?
Tested on CI.
šø Screenshots (if appropriate):
š Checklist