Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:21272
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Message-ID: <43A5B7E4.3050105@php.net>
Date: Sun, 18 Dec 2005 20:26:28 +0100
User-Agent: Mozilla Thunderbird 1.0.7 (Windows/20050923)
MIME-Version: 1.0
To: Ilia Alshanetsky <ilia@prohost.org>
Cc: PHP Developers Mailing List <internals@lists.php.net>
References: <43A5B487.6070405@php.net> <43A5B727.1030107@prohost.org>
In-Reply-To: <43A5B727.1030107@prohost.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Subject: Re: [PHP-DEV] supported OS/platforms
From: lsmith@php.net (Lukas Smith)

Ilia Alshanetsky wrote:
> Lukas Smith wrote:
> 
>>As per the release check list that Derick and Ilia wrote up [1] we now
>>have something called "primary testers". This is a list of projects per
>>php branch that will get directly notified of php RC's. We have one for
>>PHP 4.x [2], 5.x [3] and 6.x [4].
> 
> 
> I disagree with you in that respect, I received a fair amount of input
> from our test projects. One of the responses lead to a resolution of a
> nasty bug affecting smarty templating system and through it the project
> being tested. If we get A LOT of feedback it means our release is badly
> broken ;-).

Ah great! Last time I checked back with Derick the feedback was minimal.

>>Anyways I remember some discussions about this on internals (or maybe it
>>was a PEAR meeting) about what platforms to support. Obviously this is a
>>question on what the developers focus and also on what we test on.
> 
> 
> I see no reason to add limits on the supported platforms, the bottom
> line if you can compile PHP on your platform of choice it is technically
> supported.

Ok, I probably did not word things optimally. I guess the focus was more 
on getting more platforms in the test setup.

> gcov shows code coverage it does not tell if you the code is working or
> not. If anything it is a tool to help devise tests that would examine as
> much code behavior as possible.

But the way I read the reports we also get the test results as a "side 
benefit" including valgrind.

regards,
Lukas