Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:63610
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmail.com designates 209.85.215.42 as permitted sender)
MIME-Version: 1.0
In-Reply-To: <CAB75mSX-cMqZPTfv2vc8fakD_pr31qPR+kO5_O=zYiVoy4xH2w@mail.gmail.com>
References: <CAB75mSX-cMqZPTfv2vc8fakD_pr31qPR+kO5_O=zYiVoy4xH2w@mail.gmail.com>
Date: Thu, 25 Oct 2012 01:34:07 -0400
Message-ID: <CAL+t6cF=e_55=v9bZVy7Tv=V3M4CUQLueu=b5Efa+MD0ewG_qA@mail.gmail.com>
To: JJ <jawed@php.net>
Cc: internals@lists.php.net
Content-Type: text/plain; charset=ISO-8859-1
Subject: Re: [PHP-DEV] Changing the default value of "true" for CURLOPT_SSL_VERIFYHOST
From: theanomaly.is@gmail.com (Sherif Ramadan)

On Thu, Oct 25, 2012 at 1:03 AM, JJ <jawed@php.net> wrote:
> Hey all - I'd like start a discussion around pull request 221
> (https://github.com/php/php-src/pull/221).
>
> In short, there's a high volume of [incorrect] code out there which looks like:
>
> curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, true);
>
> Instead of what, in all likelyhood, the code meant to do:
>
> curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, 2);
>
> This is due to the convert_to_long_ex call which converts "true" to
> 1L. CURLOPT_SSL_VERIFYHOST being set to 1L bypasses common name
> validation within libcurl.
>
> My solution was to check the type for CURLOPT_SSL_VERIFYHOST: if it is
> boolean and true, the opt value for libcurl is set to 2L.
>
> I understand that engineers should have the proper option value to
> begin with but weighing the impact of this (MITM attacks) against
> doing what they probably meant anyways is worth the presumption.
>
> Please discuss and adjust the patch if necessary.
>
> - JJ
>
> --
> PHP Internals - PHP Runtime Development Mailing List
> To unsubscribe, visit: http://www.php.net/unsub.php
>

While I think it's a good idea to set the value of the option to 2, as
is recommended for production in the documentation, I think the idea
of implicitly converting a bool(true) to 2L internally might lead to
unexpected behavior since some people might actually depend on normal
PHP behavior to cast a bool(true) to 1 (and that might be what they
actually intended).

I understand there are people out there that don't read the
documentation and aren't aware of the difference between
curl_setopt($ch, CURLOPT_SSL_VERIFYHOST, 2); and curl_setopt($ch,
CURLOPT_SSL_VERIFYHOST, true); but still... I don't think this is a
good idea either.

We should probably just elaborate on this point a bit more in the
documentation. Perhaps add a note and an example to illustrate. I
notice that people tend to pay more attention to examples than
anything else in the docs.