On Tue, Jul 19, 2005 at 08:59:33PM +0200, Rafael Garcia-Suarez wrote: > On 7/19/05, Michael G Schwern <[EMAIL PROTECTED]> wrote: > > On Tue, Jul 19, 2005 at 11:31:01AM -0400, Scott R. Godin wrote: > > > I've had this itch to rip Pod::Html to shreds for a while now, and > > > refactor it to do the job more cleanly. Would anyone object to my taking > > > a whack at it? > > > > It would probably be better to evaluate the existing POD -> HTML converters > > and wrap POD::Html around them, or just leave POD::Html alone and convert > > installhtml to use the better module, than to write Yet Another POD -> HTML > > Module. > > I can think about a couple of points : > * state of the art of html documentation has greatly improved since > the perl 5.000 days. Pod::Html generates obsolete html (old syntax, no > CSS support, etc.) > * Pod::Html is barely usable, making a good pod->html translator now > would mean designing a more complete and probably incompatible > interface.
But that doesn't rule out creating a minimal wrapper to provide Pod::Html's interface. > * People have hacked around the limitations of Pod::Html that were the > most annoying, and probably post-process its output. It's html, they > want eye-candy and stuff that looks like search.cpan.org. OK, that's a > weird backwards-compatibility argument. On the other hand, I'm quite happy to ignore anyone who gets too upset by their post-processing going awry. Pod::Html is documented as returning HTML. Not the specific current output it creates. > In short, my opinion would be to leave Pod::Html alone, fixing the > most obvious bugs, and slowly deprecating it in favor of the next big > thing. I prefer Schwern's suggestion. Then again, I'm not proposing to volunteer. my own labour here, so ultimately the choice is someone else's. But I don't think that refactoring the existing Pod::Html to create yet another convertor is the way to go. Nicholas Clark
