On Sat, 2003-01-04 at 22:18, Stuart Donaldson wrote:
> I lean towards fixing the example in the docs. Some of the code looks
> like XML, in particular the <psp:method> and <psp:include> commands.
> I would argue that if they look that much like XML then they should
> behave like it too.
We shouldn't even pretend PSP has any relation to XML -- lots of
templating languages do that, and it's dumb, because they either are
wordy (ZPT) or are a long way from proper XML (Albatross) or they are
horribly arcane (XSL). PSP is a text processing layer, built to be
vaguely similar to ASP/JSP -- the chosen markup isn't any deeper than
that. It's simple because it deals with text, not markup.
The only problem with not accepting a space is that PSP won't signal any
error (I believe), and the developer might be mystified. But I don't
think it happens that much, so whatever. But there's nothing *wrong*
with allowing a space. It's not proper XML either way.
> We could make it slightly more friendly and allow for the space, but
> does that mean we should allow for "< %" as well? The parser is
> fairly simple and works on simple tokens, trying to fix it for
> whitespace would make it more complicated too.
I wouldn't want < % to be allowed. That just looks weird.
> Just out of curiousity, why do we have both the <%...> and <psp:...>
> syntaxes? It seems inconsistent to me at first glance.
I don't use PSP, so I haven't thought much about its design, but I
assume it's modeled after either ASP or JSP. <psp:...>'s are really
compiler directives, and can't be folded into <%...>. I thought ASP
used <@...> for such directives, but maybe JSP uses <jsp:...>.
Ian Bicking Colorstudy Web Development
PGP: gpg --keyserver pgp.mit.edu --recv-keys 0x9B9E28B7
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241