On Tue, Feb 16, 2010 at 12:50 PM, Thijs Kinkhorst <kink@...> wrote:
> On Tue, February 16, 2010 19:15, Paul Lesniewski wrote:
>> On Tue, Feb 16, 2010 at 5:48 AM, Thijs Kinkhorst <kink@...>
>>> Hi Alexandros,
>>> On Tue, February 16, 2010 12:14, Alexandros Vellis wrote:
>>>> ...due to use of function stripos().
>>>> Something like this fixes it:
>>> Thanks! I have applied the patch to the stable branch.
>>> However, Michelle is right. PHP4 is out of support. The Debian release
>>> that still has PHP4, Debian Etch, has also just ran out of security
>>> support. That means that the base for supporting PHP4 gets very small at
>>> this point.
>>> I believe the following is a useful strategy:
>>> 1) Keep PHP4 support working for stable when reasonably possible. I
>>> think it's wise to keep running PHP4, but I also see no need to break
>>> the stable branch for it.
>>> 2) On the other hand, don't invest the same effort for devel. It will be
>>> a while before devel becomes a new stable release, and by that time
>>> will be even longer out of support.
>>> I propose to hence change the documented minimum version in devel to
>> The reason the project goals include supporting things like PHP4 and
>> possible. Statistics still show a lot of PHP4 out there. I believe
>> we should wait until that changes until considering any big changes.
>> To your last question, if the code still supports PHP4, then why would
>> we make a misleading statement in the documentation to the contrary?
> One reason for making that statement is that testing whether changes work
> with PHP4 is as far as I know done rarely or not at all.
Then that's a problem with the developers wherein they should pay
close attention to their work. If the code supports PHP4, that's what
we should say. If we want to *encourage* PHP5, that's a different
Please support Open Source Software by donating to SquirrelMail!