I think many comments like this are missing the point. You don't get a "Haskell/C/Java/Lisp validator". You have a compiler, which will flag obvious crap as crap. The possibly crap stuff are flagged as warnings. And then you have a linter when you want your crap to smell slightly better. But the point is that obviously wrong stuff on the web should give you errors, not a vaguely working page, sometimes, in some browser version.

It is you who has missed the point. I mention in my comment specifically that W3 validation errors are not comparable to compiler errors or run time errors in programming.

As far as things working in one browser and not another, if you're working with standard features they tend to work across all modern browsers. If you are using features which are still only candidates, then yes they may not work but you've chosen to implement non-standard features so you've done this to yourself.

Not sure that they realized that the Internets motto is "good enough for jazz" as opposed to proper OSI internet standards :-)

Even then OSI had It's "kids on the variety sunshine coach" members who couldn't implement to standard I am looking at you Sprint and ICL.

> "kids on the variety sunshine coach"

:( There are better ways to express this than shitting on special needs kids.