If I opened a PR that moved map to sequtils (using the {.deprecated.} pragma), then imported sequtils as necessary throughout the standard library, would it be accepted?
Rationale: I've recently been using an algorithms module I wrote to be lazy, composable, a bit more generic, and multidimensional-array-aware. Currently, my map definitions are shadowed by the definitions in system. If I moved system.map to sequtils, (1) I'd free the name up for library developers and (2) map could live with its friends (filter, zip, and the fold twins).
TL;DR: map is a generic concept that's useful even to people who don't want to allocate memory or operate on seq values. Is anyone opposed to moving it out of the system namespace?
Sorry; that was not my intention.
I figured since the upcoming release involves many other aesthetic changes (the language name change, case-sensitivity, T/P prefix removal, Babel -> Nimble), the community was open to small aesthetic/polishing tweaks before 1.0.
I also remember reading somewhere that you (Araq) thought the system module was too cluttered, so I thought I was offering to help (slightly) in that regard.
Anyway, I'm a researcher, not a software engineer, so I'll read (i.e. lurk :) ) more to get a sense of the kind of ideas/conversations people here find more helpful.
So review it. I do not see much that's there that shouldn't be there. On the contrary, things like 'mitems' are still missing. And everybody wants his favorite shortcut in system.nim like
template `===`(a, b): expr = system.`==`(a, b)