>> >> Ah, but that applies for *large* lists. Adding 8 bytes to >> each list >> object applies to *all* lists. I betcha that many programs >> use many >> tiny lists. >> > > Even most tiny-ish lists are wasting memory, though, according to this > sequence: > > 4, 8, 16, 25, ... >
That is only is they are being grown with list.append(). Otherwise, list sizes are exact. For example, [0,1,2] uses space for just three entries and [] for none. I get the impression that 1) you've already made up your mind and are ignoring input from the other developers, 2) that you're ignoring the python-dev discussions of long ago where this very idea was rejected and deques came in to being instead, 3) over-estimating the prevalence of use cases for pop(0), and 4) under-estimating the impact on small lists, 5) under-estimating the impact on psyco or other implementations of Python, 6) under-estimating the impact on third-party extensions and debugging tools. Deques already provide a solution to the FIFO problem and they do so without huge wastes of memory or calls to memmove(). They handle both insertion and deletion from the front and back. In comparison, the proposed changes to lists seem like a complete hack. Just because it can be done, doesn't mean that it should. ISTM that you're attacking an already solved problem and that you're playing fast and loose with a core data structure. Raymond
_______________________________________________ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
