On 10/12/01 6:20 PM, "Terry Steichen" <[EMAIL PROTECTED]> wrote:

> Joshua,
> 
>> Pithy answer: any customization feature powerful enough to be
>> useful, will be powerful enough to create the problem you see.
>> 
>> Therefore, if you want customization features that powerful,
>> you need to build in debugging and error catching features
>> to minimize the problems.  That is true with Velocity or
>> with anything else.  The important thing is to understand the
>> problem from the beginning, and design in those support features
>> from the beginning.
>> 
>> BTW: I'm assuing that you want people to customize their templates,
>> so the 'templates in Jar' stuff will not help you.  You are afraid
>> that they will make a mistake while customizing, and that your system
>> will pay the price for that mistake.
>> 
>> Joshua Levy
> 
> From the other responses I received, and after re-reading the docs on
> resource loaders, it seems that I can put the core templates (the ones for
> administering the system, ones that I do *not* want users to modify) into
> their own jar, which still allows users to freely create their own
> additional templates and put them somewhere else.  (I should explain that my
> software provides a publishing framework, so it is expected that users will
> create scads of content-related pages, which are quite separate in function
> from the pages I provide to administer and otherwise run the system.)
> 

Yep - that will work fine.

Will this be a public system or product?  We would be happy to mention it on
our 'powered by' page.

> And, BTW, I am not afraid that users *might* screw up my system if they have
> the option - my long experience (and Murphy's well known Law) says that they
> certainly *will*, regardless of debugging and error handling :-).

Indeed.  It's not 'if' but 'when'.

geir

-- 
Geir Magnusson Jr.     [EMAIL PROTECTED]
System and Software Consulting
"They that can give up essential liberty to obtain a little temporary safety
deserve neither liberty nor safety." - Benjamin Franklin


Reply via email to