[Python-Dev] Fwd: Removal of GIL through refcounting removal.

[Adam Olsen]

On Sun, Nov 2, 2008 at 2:34 PM, Eric Smith <eric at trueblade.com> wrote:
> Giovanni Bajo wrote:
>>
>> [[ my 0.2: it would be a great loss if we lose reference-counting semantic
>> (eg: objects deallocated as soon as they exit the scope). I would bargain
>> that for a noticable speed increase of course, but my own experience with
>> standard GCs from other languages has been less than stellar. ]]
>
> And my $0.02:
>
> I'd gladly trade deterministic destruction (due to reference counting or any
> other mechanism) for improved performance. I've often thought of creating a
> mode where destruction didn't happen right away with reference counting,
> just so I could find places where I'm relying on it.  I consider it a bug to
> rely on reference counting to close files, for example. Maybe I should just
> run under Jython or IronPython everyone once in a while.

I've considered making files issue a warning if they've got buffered
writes and they're not explicitly closed.  Although my feeling is a
read-only file should produce the same warning, there's situations
where "proper" error handling is very difficult or impossible.


-- 
Adam Olsen, aka Rhamphoryncus