Fork me on GitHub
#clojure-dev
<
2016-06-15
>
bronsa08:06:42

@cfleming: type hints on the argvec have been fixed indeed

bronsa08:06:21

var type hints aren't really deprecated, type hinting the argvec has been the preferrable way of type hinting for a while

bronsa15:06:15

@alexmiller: any reason I'm missing why vectors are Comparable but sequences aren't or is it just an oversight?

bronsa15:06:59

thought I could write (sort-by (juxt val key) the-map) as (sort-by rseq the-map) or (sort-by reverse the-map) but apparently not

Alex Miller (Clojure team)15:06:29

There's a ticket out there re making lists comparable

bronsa15:06:45

just lists or all seqs?

Alex Miller (Clojure team)15:06:48

I guess sequences would be similar

Alex Miller (Clojure team)15:06:16

I'm not sure though if there is some reason why they aren't already

bronsa15:06:53

I'd expect any IPC should be Comparable (maybe sorted colls would be weird though?)

Alex Miller (Clojure team)15:06:36

The vector behavior is specifically doc'ed on the data structures ref page so it seems like it was a conscious choice

Alex Miller (Clojure team)15:06:57

But can't say I know why

bronsa15:06:20

ok, thanks. I'll leave a comment on CLJ-1467

andrewhr17:06:48

@alexmiller: @bronsa maybe they're not comparable because of lazyness? Aka, compare would force to evaluate the whole sequence. Not saying they should not be, but that could point why they aren't yet

bronsa17:06:33

it wouldn't necessarily have to evaluate the whole sequence, just enough elements to find a difference

bronsa17:06:56

obviously comparing two identical infinite sequences wouldn't terminate, but I'd say that would be expected behaviour