[4868] Question Verification Bug On Registration

[Xarcell]

SMF 2.0RC5

In question verifications on registration, the answer cannot be "0". It can be "zero", but not "0", otherwise it returns an error saying the guest did not answer the questions properly...

If this bug has already been reported, I apologize. I don't have the time to spend looking for it.

[Arantor]

Yes, this is well known.

Cause is an empty() check in Subs-Editor.

[emanuele]

Should this be considered a bug and fixed?

If so probably:

Code (find) Select Expand

if (empty($_REQUEST[$verificationOptions['id'] . '_vv']['q'][$row['id_comment']]) || trim($smcFunc['htmlspecialchars'](strtolower($_REQUEST[$verificationOptions['id'] . '_vv']['q'][$row['id_comment']]))) != strtolower($row['answer']))

Code (replace with) Select Expand

if (!isset($_REQUEST[$verificationOptions['id'] . '_vv']['q'][$row['id_comment']]) || trim($_REQUEST[$verificationOptions['id'] . '_vv']['q'][$row['id_comment']]) == '' || trim($smcFunc['htmlspecialchars'](strtolower($_REQUEST[$verificationOptions['id'] . '_vv']['q'][$row['id_comment']]))) != strtolower($row['answer']))
in Subs-Editor.php should work ans considering only empty answers as invalid.

Or "by design" and kept as it is?

[Norv]

No, just an oversight. I remember fixing this long ago... I must have forgotten to commit the fix, oops. Or, it's gremlins!

Ema, please feel free to kill it.

[Joshua Dickerson]

Norv, is it tracked?

[emanuele]

Ema, please feel free to kill it.

Only a doubt, is it enough:

Code Select Expand

trim($_REQUEST[$verificationOptions['id'] . '_vv']['q'][$row['id_comment']]) == ''
or better:

Code Select Expand

trim($smcFunc['htmlspecialchars']($_REQUEST[$verificationOptions['id'] . '_vv']['q'][$row['id_comment']])) == ''

Norv, is it tracked?

Don't think so.

[Joshua Dickerson]

We should track these things (even if we have the fix in hand) so we can measure how much we've done.

[emanuele]

Fine, I'll use mantis a bit more then.