xml-xsp-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Allan Erskine <a.ersk...@cs.ucl.ac.uk>
Subject Re: Avoiding "language impedance" in XSP (was Re: [RT] Rationalizing XSP (was: XSP and aspects))
Date Wed, 07 Mar 2001 16:47:57 GMT
> > If you're talking about side effects of an XSP page on whatever's
running in
> > your server environment, then I think this should be kept to a minimum
> > within <xsp:page> tags, since generating a page isn't about doing other
> > things.
>
> I actually meant side effects on a page. XSP's taglib design pretty much
> requires everything associated with one taglib to happen within some sort
> of enclosing tag (e.g. <esql:connection>). But it is incredibly useful to
> have side effects (despite all the CS nonsense out there about it creating
> hard to maintain code) within code, e.g. if (something) { var = one } else
> { var = two }, and then way down the page making use of var. While it's
> not much of a design pattern, it is certainly used a lot, and very useful.
>

Oh...that sort of side-effect.  I do this sort of thing so much I can barely
bring myself to look at the XSP pages I write.  However, I blame this on
there not being far cooler ways to do things.  I guess we'd be trying to
minimise this sort of thing, but obviously until there's a really good way
to do it, people are going to want to keep their programming tricks.  It's a
good point...

----- Original Message -----
From: "Matt Sergeant" <matt@sergeant.org>
To: "Allan Erskine" <a.erskine@cs.ucl.ac.uk>
Sent: Wednesday, March 07, 2001 9:33 AM
Subject: Re: Avoiding "language impedance" in XSP (was Re: [RT]
Rationalizing XSP (was: XSP and aspects))


> On Wed, 7 Mar 2001, Allan Erskine wrote:
>
> > > > On my first reading of Ricardo's email, I thought his focus was bang
on.
> > > > <xsp:if> is exactly the sort of direct programming construct that
we
> > DON'T
> > > > want to need to use.  You rightly say that programming languages can
do
> > a
> > > > lot...however so can taglibs, and they have the potential to be
exactly
> > > > aligned to multiple dimensions of the entire XML serving domain in
ways
> > no
> > > > (current) general purpose language could (not even Perl).  So IMO it
> > comes
> > > > down to much more than just reuse.
> > >
> > > What about side effects?
> >
> > If you're talking about side effects of an XSP page on whatever's
running in
> > your server environment, then I think this should be kept to a minimum
> > within <xsp:page> tags, since generating a page isn't about doing other
> > things.
>
> I actually meant side effects on a page. XSP's taglib design pretty much
> requires everything associated with one taglib to happen within some sort
> of enclosing tag (e.g. <esql:connection>). But it is incredibly useful to
> have side effects (despite all the CS nonsense out there about it creating
> hard to maintain code) within code, e.g. if (something) { var = one } else
> { var = two }, and then way down the page making use of var. While it's
> not much of a design pattern, it is certainly used a lot, and very useful.
>
> --
> <Matt/>
>
>     /||    ** Founder and CTO  **  **   http://axkit.com/     **
>    //||    **  AxKit.com Ltd   **  ** XML Application Serving **
>   // ||    ** http://axkit.org **  ** XSLT, XPathScript, XSP  **
>  // \\| // ** mod_perl news and resources: http://take23.org  **
>      \\//
>      //\\
>     //  \\
>


Mime
View raw message