On 2008.11.15., at 14:23, Phil Jordan wrote: > > Hi Attila, > > I can't comment on the other issues, but: > > Attila Szegedi wrote: >> 2. Keyword and Ref define compareTo, but don't redefine equals (and >> hashCode) to be consistent with it. It ain't necessarily a problem if >> you know what you're doing, but since they're public it's usually a >> good rule of thumb to have equals consistent with compareTo (and then >> hashCode consistent with equals). > > This complaint makes no sense to me. Keywords and Refs have identity > equality semantics. No two keywords with the same name can exist, > therefore if two keywords are equal, they are the same object. Using a > pointer comparison is therefore semantically correct, and has the nice > side effect of being faster.
Yeah, you're right -- they're canonicalized, therefore it's indeed perfectly okay. My mistake; I didn't analyze this enough to conclude it's a false alarm. > Likewise, ref compareTo is based on > comparing the refs' unique IDs. Given that they're *unique*, two refs > are equal iff they are identical. That should be okay too, except I have a question: what's the role of the UUID in the Ref at all? As far as I can tell, no code within Clojure ever calls getUUID(). If that's so, then it only seems to serve to provide an (essentially random) sorting order. If you don't mind me trying your patience with stupid questions, what purpose is there in a sorting order based on a value of a 122-bit random sequence that the objects carry internally? Is it purely for the ability to use them as keys in a sorted map? If so, wouldn't sequence numbers produced from a java.lang.concurrent.atomic.AtomicLong#incrementAndGet() fare better for the purpose? (I know it's unlikely to hit the 1/(2^122) chance of UUID collision, but still... :-) ) Thanks, Attila. > ~phil --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/clojure?hl=en -~----------~----~----~----~------~----~------~--~---
