On Fri, Feb 18, 2011 at 8:35 AM, Michael Foord <fuzzyman at voidspace.org.uk> wrote: >> And the number 1 reason I consider messing with the numbering to be a bad >> idea: >> >>>>> "3.2">= "3.2.0" >> >> False >>>>> >>>>> (3, 2)>= (3, 2, 0) >> >> False >> >> If we miss anything, it could easily lead to errors like the two >> above. > > How are those errors? Surely what matters is that the following *is* True: > >>>> (3, 2, 0) >= (3, 2) > True >>>> "3.2.0" >= "3.2" > True They aren't errors per se, but they're different from the answer you get with a "3.2" or "(3, 2)" on both sides of the equation (as behavioural changes go, such a change is probably a good thing, since it makes naive version checks less likely to break when 3.2.1 hits, but it's a concrete behavioural change in the release that isn't really an option until we start working on 3.3). Cheers, Nick. -- Nick Coghlan | ncoghlan at gmail.com | Brisbane, Australia
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4