Hi Pierre,
On Mon, Mar 16, 2015 at 10:48 AM, Pierre Joye <pierre....@gmail.com> wrote:
> > Coercive type for general developers and strict type for library
> developers
> > will
> > eliminate more type related bugs, more natural to average PHP users.
> IMHO.
>
> And change casting rules, open a pandora box. And even worst, it is a
> zero compromise RFC. Same as before, no need to discuss that any
> longer.
>
"The same way before" is going to cause problems, or hide problems at
least.
As Zeev's patch revealed, people do mistakes and the patch helps to find
them.
"The same as before" will just hide these hard to find bugs.
>
> > It's natural that we have different point of views, but we can easily
> > understand/guess
> > the consequence of the RFC. Weak mode is simply too weak to be useful.
> > Strict mode will hide type bugs by errorless casts.
>
> Show me examples when something not in strict mode behave differently
> and it will be fixed. But saying that is per se wrong and double
> standard in regard of voting. Or why did you vote in favor of other
> RFCs which obviously had or still have bugs?
>
This code is an example that I posted in other thread.
e.g.
<?php
function check_num_range(int $num) { if ($num < 0 || $num > 100)
trigger_error('Invalid range'); }
// Somewhere far from function definition.
$num = $GET['num'];
// Somewhere far from $num definition.
check_num_range($num); // Trying to check validity, int and range.
echo 'You have '.$num. ' now <br />'; // But $num could have any string.
//
"check_num_range((int)$num)" wouldn't help also.
?>
Simple cast hides bugs, not eliminates type bugs.
This is just an example and there are many cases that cast hides bugs in
real world codes.
> > I hope both you and Zeev work together to satisfy most needs because
> > hiding bugs are not good thing to do.
>
> It does but I do not claim that there are no bug, every code has bugs
> and they should be fixed.
I'm not saying the proposed patch has bugs, but I'm saying it hides "users'
code bugs".
Hiding hard to find bugs does not make much sense while there is the
proposal that finds it.
What I'm trying to say is "Why don't we collaborate?". Let users fix bugs
at first, then
introduce more strict type checks if it's needed. Coercive type and strict
type can co-exist
together well, but weak type and strict type cannot.
Regards,
--
Yasuo Ohgaki
yohg...@ohgaki.net
