Motivation for Vesta-lite
There is at least conjecture that vesta has too much of a ramp for new users. The worst case tradeoff between
how much i get out of vesta -------------------------------- how long it takes to learn vesta
would be for a single developer working on his project. (The best is for a large group (100+) of developers)
This page attempts to list the things which could be made simpler so that a single developer working on a single project can get benefit out of vesta with minimal investment.
Simplifications
Not learning SDL
This is a tough one. Much of the power, but especially flexibility, comes with the ability to script the build.
SDL, being a pure functional language, is part of the vesta guarantee of reproducibility.
Speculating idealisitcally, if one would write in perl, or python, or tcl or some other language a that a large majority of prospective vesta users _already know_, it would be a huge help. Not sure how/if we could do this.
Yet, at intel, few users write SDL, so maybe there's hope.
Name Cache, or hierarchy simplification
At intel we use a "name cache" which lets the average user ignore the relatively deep hierarchy of the repository.
Why? I assume it makes it easier.
If so, can we make this public, and/or integrated, somehow?