> -----Ursprüngliche Nachricht----- > Von: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] Auftrag > von Bruno Dumon > Gesendet: Dienstag, 4. Mai 2004 20:53 > An: [EMAIL PROTECTED] > Betreff: Re: JXTemplates - what's in a name? > > > On Tue, 2004-05-04 at 10:39, Derek Hohls wrote: > > Bruno > > > > I apologise then, if I misunderstood what you said. > > <snip/>
> > If I want I can start an HTTP server or send mail in the toString() > mehod of an object. But that's hardly the intended use of toString(). > Likewise, if someone implements the business logic and flow control of > his/her site in XSP, we can't stop them. that's true. the problem basically is (similar to JSPs etc.) that the abuse is made easy; 'gelegenheit macht diebe' (opportunity/chance? makes thieves): the paper also refer's to Murphy Law ;-) my point just is that if <xsp:logic> and the like wouldn't be there then the abuse would be made much harder and less likely to occur. > With JXTemplate the same: it is meant to use the data available from a > few supplied contexts: most importantly the viewData from flowscript, > and possibly environment data like the request and session objects. If > used in that way, is it still different from the 'real templates'? (Just > curious, I haven't found time yet to read the paper) not that different anymore, except e.g. being able to format the data (the template is supposed to be agnostic of the data) and to be able to call methods which could have side effects. for me it's now my preferred templating approach for *views*. I'm currently refactoring from views created by XSLT transforming XML models generated by generators and the like. it's all a lot cleaner and easier to maintain and less fragile. I do still use the other approach, of course, but am moving away from it for views. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
