I agree... discounting a few things that HTML::Template may be useful for, I always found it hard to actually 'wrap' my code around a template. I know that if separating code from content is an issue, then wouldn't it be simpler to move on to Mason? However, I'm not particular fan of this either since many Mason scripts I looked at appeared to be quite a mess.
PS: my native language is Russian so don't mind my English! ;-)
Having said that, I am actually trying to move more towards splitting display logic from program logic. One way to do this, I think, would be to have your program generate and store all of its 'working' data mean to be displayed in some fashion in a hash or some complex enough structure which yet could be looped through/accessed via a templating script (such as <TMPL_LOOP ... > </TMPL_LOOP> implemented in HTML::Template).
Since I'm yet not a strong believer in Perl templating utilities, would anyone with some related experience share in his/her joys of using them? I'd be interested in hearing from someone who was in a position similar to I'm in now (that is 'hardly decided') and how they managed to integrate this concept in their code.
"There is no system but GNU, and Linux is one of its kernels." -- Confession of Faith