<?xml version="1.0" encoding="windows-1252"?>
<node id="121567" title="(Ovid) Re(2): Sticky HTML Templates" created="2001-10-25 22:51:22" updated="2005-08-02 00:31:35">
<type id="11">
note</type>
<author id="17000">
Ovid</author>
<data>
<field name="doctext">
&lt;p&gt;I read through [cpan://HTML::FillInForm] and have to admit that it would be a nifty solution.  I can see pros and cons, however.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Pros:&lt;/strong&gt;  As the HTML would be left relatively pristine, it would be much easier for our designers to work with it.  Further, as the HTML changes, there would less synchronization.  This is a huge win.  This also appears to be a relatively easy solution to implement.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Cons:&lt;/strong&gt; We're running in an ISAPI environment here.  While that's faster than straight CGI, it does not compare to the performance of mod_perl.  Until such time that we can convince the Powers That Be that mod_perl (and Apache, and Linux and ...) is the way to go, we have to be careful about the performance of our applications.  Already, using Template Toolkit is a serious issue, in terms of performance.  I'm concerned that HTML::FillInForm would be the death knell.  Any experience with how this might affect real-world performance?  It seems to me that inserting the template markup tags directly into the HTML document would be the least performance-intensive alternative.&lt;/p&gt;

&lt;p&gt;Cheers,&lt;br /&gt;
&lt;a href="/index.pl?node=Ovid&amp;lastnode_id=1072"&gt;Ovid&lt;/a&gt;&lt;/p&gt;


&lt;p&gt;&lt;small&gt;Join the &lt;a href="http://setiathome.ssl.berkeley.edu/stats/team/team_86606.html"&gt;Perlmonks Setiathome Group&lt;/a&gt; or just click on the the link and check out our stats.&lt;/small&gt;&lt;/p&gt;</field>
<field name="root_node">
121510</field>
<field name="parent_node">
121533</field>
</data>
</node>
