Wisdom at #cfgMgmtCamp . If you can't reproduce it, you don't really run it. Without verifications, claims about systems are performative.
Devdas Bhagat
@f3ew
- 325 Followers
- 574 Following
- 311 Posts
Just another recovering sysadmin.
The takeaway from this talk appears to be that all computer developers and users now need to understand quantum physics.
LLMs are probabilistic, so we have to test for results within defined bounds. "Where is the electron?"
Listening to @stahnma at #cfgmgmtcamp .
Desired state is a myth.*
* Save for small values of desired and state..
git history should be linear, because it makes things easier to understand.
Work must be organized in patchsets, as a linear set of atomic changes. Distinguish between preparatory/refactoring commits and functional commits
Two examples of code commits, one good and one bad each with specific comments on why they are good/bad
Atomic Flow at #fosdem . Commit Atomicity is important. Commit messages document the why. Small scope for the commit to easily grasp what is being done.
SellaTheChemistIt is with great sadness that I am doing what I would once have thought unthinkable. Following Salim Abdool Karim's superb Faraday Lecture 10 days ago, i have decided to return my own Faraday prize, awarded in 2014.
The silence of the Royal Society in the face of Elon Musk's rampage (a member of the RS since 2018) across US science is incomprehensible to me.
We must all draw the line somewhere in the face of authoritarian attacks and this is mine.
https://www.ft.com/content/877853df-1687-4d1f-940b-9eefc0cf69c0?accessToken=zwAAAZdYtYr2kdOHeFPfFodNH9OUC57vwM9pwA.MEUCIQDNG6kcQyHa-g6Kx4G9NHE2ofD36kR8OX0yEu-KfXlV0QIgcmMFKTaT90y42QfwQTWb05khk6thpQPe-JvIGkaZzQY&segmentId=e95a9ae7-622c-6235-5f87-51e412b47e97&shareType=enterprise&shareId=5fbecded-0dc9-419a-b064-cda0dee0d73a
JPif your "beautiful, elegant, futuristic" huge ring-shaped HQ requires a truly colossal fuck-ugly parking garage that dwarfs it, it's not actually a very beautiful elegant futuristic HQ now is it
