Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:31463
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: error (pb1.pair.com: domain haxx.se from 91.191.140.26 cause and error)
Date: Sun, 5 Aug 2007 23:06:34 +0200 (CEST)
To: internals@lists.php.net
In-Reply-To: <48557.209.254.223.2.1186285643.squirrel@www.l-i-e.com>
Message-ID: <Pine.LNX.4.64.0708052243430.2754@yvahk3.pbagnpgbe.fr>
References: <Pine.LNX.4.64.0707301316440.22586@yvahk3.pbagnpgbe.fr>   
 <059601c7d353$b17cf050$49fc1f3e@foxbox> <48557.209.254.223.2.1186285643.squirrel@www.l-i-e.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
Subject: Re: [PHP-DEV] Regarding the ext/curl extension
From: daniel@haxx.se (Daniel Stenberg)

On Sat, 4 Aug 2007, Richard Lynch wrote:

> That said, I personally never had a problem searching the haxx.se site and 
> finding out what CURLOPT_XYZ meant, except for the php-specific 
> CURLOPT_BINARYTRANSFER... :-)

The binding has a 1:1 mapping on all its option names (except 
CURLOPT_BINARYTRANSFER) to libcurl's as far as I know, but libcurl offers a 
lot more options than what the binding offers. There's also the not so minor 
detail that the PHP binding has its own way of interpreting and acting on some 
options that isn't done by libcurl. (I'm mainly thinking about 
CURLOPT_HTTPPOST.)

The funtion names of the binding are also all different than the ones used by 
libcurl.

> I'd be hesitant to try and duplicate the libcurl docs onto the PHP site, as 
> the whole point of PHP being a "glue" language and maintaining as much as 
> possible of external lib idioms is to get documentation re-use out of the 
> original docs, presumably maintained better by the original authors -- and 
> they're pretty darn nice docs over on haxx.se, at least when I went looking 
> for stuff.

Thanks.

I'm all for trying to get things as good as possible, but the libcurl docs 
have the minor drawbacks to PHP users that they speak of the C API so it 
refers to data types and operations etc that PHP user won't care about or in 
worst cases will get confused by.

So, while I wouldn't mind adapt somewhat or even do additional hacks in order 
to help others to host or extract partical docs from the curl site or the curl 
docs (since most of our docs are in fact the man pages converted to HTML), I 
just don't know how or what I could do.