Forum Moderators: open
Check out for yourself at [validator.w3.org...] , and great work W3C! :)
[edited by: korkus2000 at 2:51 pm (utc) on May 7, 2004]
[edit reason] fixed URL [/edit]
If you write programs in Pascal, C, C++, etc..., then compiler warns you of the errors and makes you fix them, before it will compile the program. This is not true with browsers. Browsers just handle errors in whichever manner they think it's best (and all of the browsers have it their own way). Validator helps to find those little bugs (open tags, missed quotes, unencoded special characters, etc.).
Your code not necessarily has to validate 100%, just make sure there are no severe errors. But why not go an extra mile and make it fully valid ;)?
You can try and make a transitional doctype valid first. After, when you get a grip of that, try to validate it strict. Eventually, you will know what is incorrect in terms of coding practice and writing valid code will be a habit.
Browsers just handle errors in whichever manner they think it's best (and all of the browsers have it their own way).
That's a good point. Validated HTML doesn't guarantee the site will work in all browsers, but unvalidated HTML is taking a great risk that it won't.
Also worth remembering that spiders, eg Googlebot, have to read the HTML too. Unless you can guarantee the bugs you (or your coder) have inserted into the HTML cause no problems for (eg) Googlebot, you are taking quite a risk.
If your HTML is generate by software and it produces unvalidated HTML, get the package fixed or replaced. Why take the risk, small though it may be, otherwise? It could be your job on the line.
Jennifer
New W3C Validator [webmasterworld.com]
