These are chat archives for jdubray/sam

13th
Oct 2017
Jean-Jacques Dubray
@jdubray
Oct 13 2017 00:25
Not sure what's happening in React land: https://twitter.com/fkrautwald/status/918571189794205696
Jean-Jacques Dubray
@jdubray
Oct 13 2017 02:30
@jannesiera I didn't spend too much time with PlusCal, though I understand this is the "useable" part of TLA+. I don't believe it's practical to write a TLA+ spec for some front-end code. Dr. Lamport's work applies to algorithms. I have just tried to extract the first principles from TLA+ and see how it applies/could be applied to programming. The building blocks of TLA+:
  • initial state
  • next-state relation
  • action
  • pc (program counter / controler, aka control state)
  • state (aka model)
  • temporal aspects
    There is also some wiring aspects (boolean logic) which didn't apply
The mapping to SAM:
model -> initial state, next-state relation
action -> action
state function -> pc + computed properties
nap -> temporal aspects
again, I am happy to discuss the mapping, that is only my interpretation
Jean-Jacques Dubray
@jdubray
Oct 13 2017 02:36
As Dr. Lamport explains, the "prime" notation (aka proposal in SAM) was key to make TLA+ what it is today, and I believe that what we are fundamentally doing wrong when coding: assignments are not mutations.
I repeat, assignments are not mutations.
Daniel Neveux
@dagatsoin
Oct 13 2017 11:57
A quick question: in your experience, is the development time of a similar app/website the same with old fashion framework (bootstrap, foundation,...) than with Web Components framework, including all ceremonies around?