I am running PHP 5.1.4 on a Win2K Server w/SP4. When processing scripts which produce a significant amount of HTML, the HTML is corrupt when it reaches the browser. Clicking refresh repeatedly produces a different result each time with the corrupted characters in a different place. What I mean by corrupted characters is that the characters are being replaced. For example, 'script=' came out 's2">S=' -- the right number of characters, just the wrong ones.
I have tried turning off output_buffering and turning on implicit_flush in php.ini without success.
I have also tried the same script on a Mac OSX 10.4.8 Server accessing the MySQL database on the Win2K server remotely which does NOT corrupt the HTML. I believe that rules out MySQL as a contributor to the problem.
I am convinced that PHP for Windows is the root of the problem. Can anyone suggest a configuration change that might fix it?
I should also point out that an ASP/MS Access script which produces the same HTML runs well on this same server -- without corruption -- as do PHP/MySQL scripts which produce less HTML.
Start Free Trial