We have heard a lot of anecdotal evidence about the usage of <? vs. <?php vs. other methods of PHP invocation. In order to clear things up a little, I have asked a friend of mine for more detailed numbers from a real world example.
The following tests have been done very much ad hoc, and the numbers are not representative. They have been made across the PHP files found at a very large german provider. The test has been running for a night until he killed it, and the friend estimated that a full run across several million domains would take about 10 days to complete. He did not do anything to filter out standard scripts such as phpMyAdmin, nor did he do anything else to postprocess the numbers. The following numbers have been generated by "find" and "grep" for all lines that have something that resembles a PHP opening tag. These lines have then been categorized and counted. The total result set had a size of approx. 1.3 million lines. <?php 84% <?= 3% <? 12% <% 0,3% <script... 37ppm If you can come up with a script or a program that a) performs and b) can take the output of "find" at stdin, I will ask for a more detailed statistic. Kristian -- PHP Development Mailing List <http://www.php.net/> To unsubscribe, visit: http://www.php.net/unsub.php
