It turns out, I was wrong

In the past, I have not been shy about sharing my opinion that the impending death of PHP 4 would have wreaked all sorts of havoc over the PHP world. I am glad to say that I've been wrong—dead wrong, in fact—and that I have never been as happy to be so far off the mark before.

The fact that I was wrong, however, me no likey. After all, I did my research, and wasn't just basing my opinion on divination—I had some hard data and plenty of research statistics to back me up.

It turns out, I was simply using the wrong data. Most of the statistics on PHP usage are based on mostly-automated research done on PHP hosts. This gives you a great general view of the makeup of PHP usage on the Internet, but it really tells you nothing of the people who use it.

Interestingly enough, we just ran our yearly subscriber survey about a year or so ago—and, had I had access to the results back in the summer when I was busy lamenting the status of things, I would have been able to tell that the statistics I was using were very misleading.

If you look at the image above, you'll see that the readers who replied to the survey have switched to PHP 5. In fact, only a meager 8% sticks with PHP 4, while the vast majority uses some version of PHP 5.

Under these circumstances, therefore, the decision to EOL PHP 4 was a very good one—but the fact remains that it was made without any hard data to back it up.