Re: [SMARTY-DEV] smarty recompiling with op-code accelerators

Robert Amos wrote:

> At 02:18 AM 9/07/2004, Monte Ohrt wrote:
>
>> Robert Amos wrote:
>>
>>>
>>>> > IMHO, the compiled Smarty files should get a new timestamp, not 
>>>> the same
>>>> > one as the originating template file. If you dump template files 
>>>> from
>>>> > another server, it should be operational procedure to touch them 
>>>> to be
>>>> > sure Smarty picks up the changes.
>>>>
>>>> this would mean we should remove the touch()-call in
>>>> core.wrtie_file.php?  i have no problem with that.
>>>
>>>
>>>
>>> I'll agree with that, if the template is changed its mtime will 
>>> still be more recent than the compiled version, so it will recompile.
>>>
>>> That should fix the accelerator problem also right? since the 
>>> compiled template assumes the compiled time, it will be different 
>>> than the templates you just cleared..
>>>
>>> -bok
>>
>>
>>
>> Yes that would clear up the problem. The only issue it introduces is 
>> if you are someone that dumps your template files (via RPM , tarball, 
>> etc.) from a developer server and expect the files to recompile if 
>> the timestamp has changed whatsoever (newer or older)
>
>
>
> That wouldn't be expected behavoir in my opinion. I'd expect the 
> templates to recompile if and only if they've been changed, so if I 
> were to copy the templates_c dir through tarballs, the compiled 
> templates would still be that of the original templates, IE the 
> templates havent changed so why would I expect the templates to 
> recompile?
>
> I'd file that as an operational issue anyway.
>
> -bok


I was referring to dumping the template files, not compiled files. The
whole reason we began touching the compiled files was when someone
brought up the issue of dumping *.tpl files onto the production server
and Smarty didn't recompile them. (timestamps weren't necessarily new.)