Clojurians Log v2
Clojure programming
Channels
Apps
missionary
timur 2024-12-10T14:45:50.772479Z
Is there a specific reason for deprecation of m/?=? Is it just because a more general version (m/?> ##NaN ...) exists, or unlimited parallelization is explicitly discouraged for some reason?
For example, when I want to watch for all values from thousands of atoms from a vector atoms, seems like the cleanest option is to do (m/ap (m/?< (m/watch (m/?= (m/seed atoms))))) — it works fine. Should I be wary of something?
leonoel 2024-12-10T15:03:39.204769Z
> Is there a specific reason for deprecation of m/?=? Is it just because a more general version (m/?> ##NaN ...) exists, or unlimited parallelization is explicitly discouraged for some reason?
Both. Unlimited parallelization is discouraged because there's no such thing, as no machine has unlimited memory. Explicit parallelism forces you to think about the space boundaries of your problem.
> Should I be wary of something?
You should be wary of atoms being not too large for your available memory, and definitely not an infinite sequence.
✅ 1