"John W. Holmes" <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED] > Wow... where'd you pull that from?? Any facts to back that up? :)
:) no, I was think in terms of how easy is to write template oriented software, and how flexible te results are, which are all pretty subjective things, that is why I used "I think" fomulation :). Anyhow, it was not my intent to refer to the execution efficiency. My appologoizes for beeing misleading. > FYI: I've got a template benchmark site running at > http://sepodati.realxl.net/tpl_bench/ > > Any interest in looking over the code and adding a TemplateTamer > benchmark to the list? I'll take a look, but I am not sure I could fit into the structure of the benchmark. Problem is that this benchmark oriented towards the simple templating class (or functions) that user has to initialize and make some calls in order to has his variables substituted. This benchmark then wraps this substitution calls in the loop to see how efficient is template engine at simple variable substitutions. TemplateTamer works other way around, it basically provides you a light framework, and code generator. You as a progaramer have a duty to provide the data for the variables (by returning them in the array as result of the getData() function), and there is no more you are doing. You do not initialize templates, you do not load various template files into the template engine, you do not assign variable one by one, you do not call substitute function, you do not echo the youtput in order for your page to appear. All this is handled by code that was generated by TemplateTamer or by framework. Actually first (internal) versions of TemplateTamer did not have its own template engine, they were just using FastTemplate for the task. I believe such approach can be a real time saver and more easy way to work, but it seems difficult to fit into your benchmark, but I will give it a try if nothing else to get one more link :) . Another problem is that thing that is benchmarked is plain page without structure. No subtemplates, not conditional parts, no repeated subtemplates. And I believe this is where the real action is for the templates, and this is all you have in the real world situations. Also I believe that in real world situations, time that is spent in the any template engine is not that important, usually the bottleneck is somewhere else, database most usually. But I am driffting. Thanks for the response! rush -- http://www.templatetamer.com/ -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
