Hash inconsistency between calling db$finish() and knitting the html

[falexwolf]

Upon knitting the html and no changes to the source code, we don't want throw an error and require bumping the version.

This is likely an issue within lamindb where it does this because it assumes re-execution will change the source code hash.

I'll look into this.

[falexwolf]

Probably not. Probably @rcannood made a small change to the source code between knitting and calling db$finish()

The .Rmd is currently categorized as a "script" because Robrecht used lamindb 0.76.16 and I only fixed this in lamindb 0.77

[falexwolf]

Never calling db$finish() interactively would solve the problem.

• Ensure that auto-knitting works with db$track() in RStudio #110

[lazappi]

I think this is fixed by #120

[falexwolf]

I also think so!