Fork me on GitHub
#clojure-dev
<
2015-07-13
>
Alex Miller (Clojure team)17:07:15

I'd love to have someone address the comments raised in http://dev.clojure.org/jira/browse/CLJ-1453 so that there was a good patch on it

andrewhr18:07:28

@alexmiller: the old patches doesn't apply anymore. There are some git-fu I could use (just to maintain the old commits) or is better to start off new ones?

Alex Miller (Clojure team)18:07:07

there have been significant changes in some relevant parts of the code

Alex Miller (Clojure team)18:07:24

you could try "git apply --reject" to apply the hunks that will apply

andrewhr18:07:45

whoa git dark magic... now I'm in the middle of a rebase, just fix and continue, right?

andrewhr18:07:20

nice! thanks alex!

Alex Miller (Clojure team)18:07:22

additionally, the original evaluation likely needs to be re-done to catch anything new that was introduced (I tried not to introduce new instances of this in patches during 1.7 but I might have missed something)

andrewhr18:07:58

yep, I first will try to reapply (with related fixes) the original patches

Alex Miller (Clojure team)18:07:23

please assign to yourself too to indicate you're working on it

andrewhr18:07:30

then I will add this behavior to new implementations and cover that with tests

andrewhr18:07:44

already did it simple_smile

bronsa18:07:23

@alexmiller: the patch for CLJ-1208 breaks one of @ztellman's potemkin macros, it assumes that deftype expands to (deftype foo ..) but the patch changes it to expand to (deftype* ns/foo ..), I have a fix for potemkin, would you rather have me extend the patch for CLJ-1527 to relax deftype so that (deftype ns/foo) just ignores the ns part rather than failing at compile time?

andrewhr18:07:36

anything left to do to track progress? (need to "start progress on Jira"?)

andrewhr18:07:29

alexmiller: thanks!

Alex Miller (Clojure team)18:07:31

bronsa: I think I'd rather fix potemkin?

bronsa18:07:41

@alexmiller: yeah, just wanted to check in case you considered it a breaking change (I think potemkin is bound to break from time to time as it relies on specific macroexpansions)

ztellman18:07:45

always happy to make potemkin less questionable simple_smile