> Add casts (one needed, one for consistency).
FWIW, I feel that these casts are misguided. Having function pointer
casts means that type errors in the signature get suppressed. E.g.
people using such casts for tp_hash will be unable to detect the change
in the tp_hash return type that we have imp
2010/10/24 Bj Raz :
> I was looking for a way to set a function being equal to another function:
> q(m+1) = q(m)+ ((-1)^m) * exp(lnanswer);
> I was hoping to use something like this:
> (q).__setcall__.(m+1) = (q).__setcall__.(m)+ ((-1)^m) * exp(lnanswer);
> As a work around.
> But I have not found
I was looking for a way to set a function being equal to another function:
q(m+1) = q(m)+ ((-1)^m) * exp(lnanswer);
I was hoping to use something like this:
(q).__setcall__.(m+1) = (q).__setcall__.(m)+ ((-1)^m) * exp(lnanswer);
As a work around.
But I have not found the `__setcall__' built in being
> If you have anything you want me to look at please forward it,
> and if you already did then don't be afraid to nag me.
Please keep the release schedule in mind. After 3.2b1, no new
features can be accepted (until after the 3.2 release). So
there might be some urgency for some of the patches.
A
On Oct 23, 2010, at 8:10 PM, Martin v. Löwis wrote:
>> Who is doing multiprocessing maintenance these days? I thought Ask
>> Solem had been given commit privs for that, but I haven't seen any
>> activity from him; and Jesse is, mostly, absent. Is anyone working on
>> the multiprocessing issues?
>
