UUID documentation incorrect?



In the documentation it is claimed that UUID is generated based on  "... MAC addresses ..." ; but reading through the sources it seems that currently only the time is involved. I guess that you changed the code and forgot the comment. Am i wrong?


Ah yes, sorry - that's out of date now, I'll update it. Thanks!


Speaking of UUIDs.. why not making the class compliant with RFC 4122?

Since it is completely random generated, it qualifies as a Version 4 UUID, which should adhere to these very few and simple requirements in order to meet the standard:





Well, I could make that work for newly-generated ones, but if you were to load a uuid object from a previously-stored string, it'd fail to meet the RFC 4122 requirements.


Yes, I was thinking about following RFC 4122 only upon generation: loaded UUIDs may come from other sources and should be correctly read from their string representation regardless of where they came from


Thanks - I've added that now.