CodeScene's company blog, by Adam Tornhill's team, on code quality, technical debt, and mining Git history for risk areas.
HN commenters mention CodeScene's hotspot and knowledge-island analysis of Git repos, though one notes the author's company and asks for more detail.
"You seem to confuse"
"Your analogy with the automakers seems puzzlingly irrelevant"
"Your take is rather romantic."
That's pretty charged language focused on the person not the argument, so if you're surprised why I'm annoyed, start there.
Meta has one: https://arxiv.org/abs/2410.08806
Another, edited in above: https://www.linkedin.com/pulse/multi-repo-ai-assisted-refactoring-unlocks-large-scale-code-kwigc/
Another: https://codescene.com/product/ai-coding
However, I still don't recognise the names. The one I saw had no pricing but had worked with some big names.
Edit: WAIT found the one I was thinking of: https://www.youtube.com/watch?v=Ve-akpov78Q - company is https://about.grit.io
In an enterprise setting, I'm the one hitting the brakes on using LLM's. There are huge risks to attaching them to e.g. customer-facing outputs. In a startup setting, full speed ahead. Match the opinion to the needs, keep two opposing thoughts in mind, etc.
This is one of the features of https://codescene.com/
It looks for knowledge islands and relates those to frequently modified code, to identify hotspot, or areas of high risk due to low knowledge distribution in areas of high change.
Another use is if someone hands in their notice you can easily see all the code that only they know, so your handover planning is mapped out easily.
I’ve never thought of it being used maliciously, it’s for visibility. It would be a shitty manager that would use it that way and if they’re already shitty then this tool won’t change that.
Disclaimer: I work at CodeScene.
Here's a link to some docs about how code health is calculated: https://codescene.io/docs/guides/technical/code-health.html
In simple terms we try to measure cognitive complexity of source code.
A more in-depth description of one of the factors:
https://codescene.com/blog/bumpy-road-code-complexity-in-context/
That's a great visual presentation, but not really an innovation. The CodeScene tool has that built in together with a set of deep analyses on top (see https://codescene.com/).
There are several public conference talks that cover this visualization and related use cases: https://www.youtube.com/watch?v=fl4aZ2KXBsQ
That's the same visualisation used by CodeScene (https://codescene.com/how-it-works/), but there is a more elaborate one that represents affinity (number of connections) as distance to create something like a geographical map: https://homepages.ecs.vuw.ac.nz/~craig/publications/vissoft2015-hawes.pdf
This is one of the key concepts in how Codescene analyses a code base by mining the Git repo.
It is quite useful to find find problem areas and tacit knowledge such as “when you change this API these two client side adapters should probably also be updated”
See https://codescene.com/