[Issue 13936] groupBy must be redone

via Digitalmars-d-bugs digitalmars-d-bugs at puremagic.com
Wed Jan 14 13:30:06 PST 2015


https://issues.dlang.org/show_bug.cgi?id=13936

--- Comment #38 from Andrei Alexandrescu <andrei at erdani.com> ---
(In reply to hsteoh from comment #37)
> As far as groupByAdjacent is concerned, I am still of the opinion that it
> belongs as part of groupBy. But since you apparently regard groupBy as
> release-critical, perhaps it may be wiser to defer this issue to later -- we
> can always extend groupBy later to encompass the additional functionality,
> even if it involves unhappy users filing bugs about why groupBy can't handle
> certain kinds of predicates. Whereas if we release with support for both
> now, it will be hard to retract one later (at the very least it will involve
> the dreaded deprecation cycle and consequent unhappy users filing
> regressions about their predicates breaking). So perhaps for this release we
> should only include groupBy with support only for equivalence relations, and
> then we'll have more time to fight it out over whether groupBy should be
> extended or groupByAdjacent should be introduced. :-P
> 
> But, that aside, since you have already written most of the code, I may just
> copy-n-paste it and fix some of the most egregious flaws in it before
> submitting it. :-P  No predicting how the reviewers will react, though.

Sounds like a plan. Thanks much! I'll be looking for your PR. -- Andrei

--


More information about the Digitalmars-d-bugs mailing list