LaTeX2HTML

But today in many fields online documents are very common. And DVI and PostScript (at least without enhancements to support hyperlinks) aren't as good as HTML or PDF for that.

LaTeX2HTML (written by Nikos Drakos and a cast of hundreds) tries to plug the gap. Impossible to guess from the imaginative name, it translates LaTeX documents to HTML ones.

Where possible, the text structure is preserved. It doesn't look as nice as the properly formatted and kerned text that LaTeX produces, but on the other hand you can copy-paste the text. Pages have no meaning in HTML; instead, an HTML page is generated for each subsection or section. Nice toplevel index.html files are produced, along with navigation buttons going next / up / previous / contents / index.

Some things you can't translate to HTML, at least not to what's commonly implemented. So formulae and other things get translated (using LaTeX, so you get the usual superb quality!) into images, which are included in the output.

LaTeX2HTML doesn't support obscure LaTeX packages, which may be a problem. XY-Pic is said to be supported, though.

Output is nowhere near as nice as printed LaTeX output. Most browsers have a problem keeping baselines straight when a line contains both text and equations; since HTML gives no way of specifying such micro-alignment, it seems no good solution exists.

LaTeX2HTML uses a slew of tools to perform its work. It's written in Perl. Apart from LaTeX, you'll need GhostScript and possibly the PNM tools,. Installation can be mildly unpleasant, depending on what you already have on your system (usually you need more rather than less than you've got; this is a Good Thing, as it's easier to add than subtract). But it's worth it.

The program works well "out of the box"; then there are hundreds of options to control the exact details of the conversion process. You probably only want to look for an option when you need it.