What does the 'M' in Haskell's MVars stand for? Following references gets me to https://link.springer.com/chapter/10.1007/3540543961_26 … which calls them "M-structures" and never says explicitly what the 'M' is. Mutable? Mutex?
-
M-Structures were the subject of Paul Barth's PhD thesis at MIT (I was supervisor).
"M" = "mutable". Previously, Id had no mutable data. Certain array-like "accumulating" computations (e.g., histograms) seemed difficult/impossible to express efficiently with purely functional data structures.
-
-
Id already had array-like data structures called "I-structures" ("I" for "incremental"). We had array-like assignments, but each slot could only be written once (run-time error if attempted twice).
-
Semantically, an I-structure slot could only transition from 'bottom' to 'non-bottom'. As in Haskell, "synchronization" between a reader and a writer was implicit (reader just waits for non-bottom). M-structures were a natural extension to mutable arrays.
- Još 3 druga odgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.