Ramblings of an Idle Mind: Frolic through the playground of my mind

Reasons not to upgrade to PHP5

Some of you may not immediately catch the sarcasm intended to be portrayed by the title of this post, but anyone who knows my coding preferences or has heard me complain (in a good way, of course) about the server structure with which we are currently stuck with at work, you will have no issue with understanding the tongue in cheek nature of this post. I have actually been amazed at how many times in the last week I have run into servers that have yet to offer PHP5 in their configuration settings, even as an option.

I have had the privilege to make the aquaintence of a few new friends and work with them on a project, which I will cover in depth in another post, and through some web work I’ve done for the project, I’ve been made aware of three separate hosting companies (which shall remain nameless) that still do not have PHP5 installed on their servers. What’s more, after some checking, some of them apparently do not even offer it as an option. This has made me appreciate my current hosting company that much more, for they not only offer a choice between stable versions of PHP on my account, but they go to the extreme of allowing me to declare a PHP version for each individual domain or sub-domain that I wish to set up. As a developer, this sort of flexibility is invaluable.

Some readers may not fully appreciate the driving principles behind my aversion to using PHP4, and without an understanding of OOP and the exception handling that PHP5 offers, you may not be able to develop that appreciation. Without taking the time to delve into all those details at the moment, let me just cover what is probably the most obvious reason: the cessation of support for PHP4 by the PHP community. As of the end of life announcement posted July 13 of this year, PHP5 has been out for three years, and with PHP6 already well into development, PHP4 will not be supported beyond the end of the year. Obviously, as with any major application, any big security fixes will still be released, but why continue using something that has already been announced to be dying? I digress.

Suffice to say that it is amazing to me how something that seems so simple and obvious to me as a developer can be avoided only so long by companies that are intended to provide a web service to those of us in need.

You can follow any responses to this entry through the RSS 2.0 feed.