Python's garbage collection was Re: Python reliability

Diez B. Roggisch deets at nospam.web.de
Thu Oct 13 11:49:01 CEST 2005


Paul Rubin wrote:
> "Diez B. Roggisch" <deets at nospam.web.de> writes:
> 
>>AFAIK some LISPs do a similar trick to carry int values on
>>cons-cells. And by this tehy reduce integer precision to 28 bit or
>>something. Surely _not_ going to pass a regression test suite :)
> 
> 
> Lisps often use just one tag bit, to distinguish between an immediate
> object and a heap object.  With int/long unification, Python shouldn't
> be able to tell the difference between an immediate int and a heap int.

That particular implementation used 3 or 4 tag-bits. Of course you are 
right that nowadays python won't notice the difference, as larger nums 
get implicitely converted to a suitable representation. But then the 
efficiency goes away... Basically I think that trying to come up with 
all sorts of optimizations for rather marginal problems (number 
crunching should be - if a python domain at all - done using Numarray) 
simply distracts and complicates the code-base. Speeding up dictionary 
lookups OTOH would have a tremendous impact (and if I'm n ot mistaken 
was one of the reasons for the 30% speed increase  between 2.2 and 2.3)

DIEZ



More information about the Python-list mailing list