[thelist] HTML validators

[Erika Meyer]

It would help, to answer this better, if you gave the validator URLs. 
I don't know the WDG validator or even who WDG is, though it has a 
familiar ring....

By the way, I'm not sure why this message had an (off topic).  Seems 
like validators are on-topic.  Anyway.

Here's my validator experience:

I tried a few.  It started with one or two, but then I was hooked. 
I'd sneak over to RxHTML for a quickie, then go to the W3C for the 
hard stuff.

Sure, I got different results.  In the early days, I'd run my code 
thru multiple validators, mainly because I couldn't get the W3C 
validator to give me happy results.

I wanted validation, not complaints.

Sometime over the course of events I discovered the DTD.  The W3C 
validator will validate according to your DTD.  This can be helpful. 
Do the browsers care about your DTD?  The browsers do not.  (I'm told 
IE 5 Mac does, but I'm not really savvy about that).  Anyways, a DTD 
helps with your W3C code validation, which is reason enough.

If you don't have a <doctype> the W3C validator will validate to HTML 
4 strict, and this can be painful.  Try transitional.

Which is most reliable/authoritative/up-to-date?
The W3C validator.

Some list members also use Bobby.

How much stock to place in a validator?  If you use CSS, I'd use the 
W3C HTML validator and the W3C CSS validator and put a great deal of 
stock in it.  You still end up making educated decisions about when 
to allow invalid code because you have to have a certain effect that 
doesn't really work any other way.

Eventually the issues become: browsers vs. code.  future vs. past. 
portable vs. perfect.

Stuff like that.

Erika



>Hey all,
>
>I've been experimenting with some HTML validators online,
>notably the w3c.org validator, and the WDG validator. These things
>seem to be giving me conflicting and/or questionable results, as well
>as neurosis and a head-ache, and I wondered;
>
>1) how much stock to place on them
>2) whether list members use them
>3) which validators you'd consider to be the most
>    reliable/authoritative/up-to-date
>4) whether and when DTD declarations are really needed
>
>thanks in advance,
>
>Duncan O'Neill

erika at seastorm.com
http://www.seastorm.com