Front page | perl.perl5.porters |
Postings from October 2012
On 10/24/2012 11:36 PM, Alexander Hartmaier wrote:
> > [...]
> > But if it doesn't make sense any more because most data doesn't come
> > from disk but the network today it might be a better idea to
> > deprecated it (slowly as usual).
>
> -1
>
> I'm for the compile time option that is discussed here - but I am
> strongly against a deprecation of such a powerful feature.
>
> It seems many people think that way:
> http://stackoverflow.com/questions/2228457/is-perls-taint-mode-useful
For the record. I have never and won't advocate deprecating taint mode.
I think that would be a grave mistake.
> The ideal solution would be to transform the feature from global to
> scoped. For example in a module use taint; and in some performance
> critical sub { no taint; }
> Is this possible from a technical point of view?
> Can you think of a way to keep the command line options and add the new
> keyword/module?
That takes away any performance benefit of my patch. It specifically
compiles out all "do we do taint support here?" checks.
--Steffen
Thread Previous
|
Thread Next