Keep your friends close, your fears closer, your well-wishing d[a]emons closest.

The last few weeks have been taxing with things that wrap up, and preparing for what’s ahead in every sense. Part of the former has been running long with cold storage and consolidating things.

A major concern with most of the operations has been wherever we introduced Node into the mix. Or, to be precise, the node modules black holes; the obvious joke of most tech-meme circuits.

Where it became particularly problematic was the amount of n-th degree dependencies’ modules. A certain repo that we were working with added up to about 600M. Most of it coming from the nested, never ending chain of node_modules, this is a certain pain in the neck.

The major concern looking back is the sheer number of things that can in every sense, stay out of storage.

Somehow, from my shoes, finding a Py-Ruby-NodeJS packages is a clear no. At least imo, it adds pointless indirection chains and is a running away from learning things.

When chalk uses nothing but a wrapper over a shell and the \e choices for colours, why use it?

Quite probable a contrarian opinion that has been in the past, but I stick to in a dogged fashion. So, as always decided to get back to the trusted platform for a fix.

Had been using scripts built using find for monitoring et al, but didn’t think it’d come handy here too. Stumbled right away on this wonderful SO answer. The beauty is how it uses find with -prune and -exec switches to remove the nod modules in an elegant manner.

As the anonymous maxim goes - Remember, when in doubt trust the basics.