var renaming, speed testing and no-cache

> I renamed a lot of vars (and a few files) that referred to a "file" > where it really should refer to "resource". This change should be > transparent, but let me know of any mistakes. :-) well, it should be done but it sure breaks a lot of my custom code. I guess I shouldn't be playing with the internals, but sometimes it is the only way -- which tells me that the API needs to be expanded. Eventually have a spec that defines what a Smarty engine should implement? As for testing, IMHO, it is **very** difficult to test an engine such as Smarty because of its compiler, cache and dynamic nature. (It hurts my head trying to write regression tests for Smarty). Benchmarks begin as lies and damn lies so I suggest that benchmarking Smarty leads to clever lies :) Anyhow, it is nice to report quicker load times but reporting them as a % change in this instance is misleading since Monte's tests apparently only show the difference in start-up times between the two engines and does not show relative performance characteristics between the two. I don't want to rain on the parade, but much more interesting will be some benchmarks of long running 2.5.0 pages that invoke many functions, different includes, etc. Also Monte, are you doing your fetch 1000 times? Try putting a section with a 1000 iterations so that you do 1000 includes instead of 1000 fetches -- much more closely related to real-performance, in my opinion :) I am dying to run these myself but I don't have the time to port my necessary patches to the new CVS changes before I go on holiday next week. I am sure I will have much more to say later in July!! Finally, messju, can you please give a few words regarding the no-cache changes that were added to the engine? Great work all! __________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com