input type=file

[RSteinwand]

Hi Albert.

I have a file upload page that I'd like to increase the width. Setting the width with css doesn't seem to work, but when I used this (which does work) I get a warning:

<input type="file" name="filename" id="filename" onkeydown="return false" size="50" />

"The "size" attribute does not apply when type="file". Consider removing this attribute."

I can't find anything online that would indicate this isn't valid for xhtml strict. http://w3schools.com/tags/tag_input.asp

[Albert Wiersch]

Hello,

According to HTML5, it is not a permitted attribute for type="file":
http://dev.w3.org/html5/markup/spec.html#input.file

The permitted attributes for "input" when type="file" are:
common attributes & name & disabled & form & type★ & accept & autofocus & required & multiple

But I see that it does seem to work in IE and Firefox so I will remove that message. Thanks!

[RSteinwand]

I didn't think to check html5 specs.

Couldn't you just label it as invalid for html5? I think you already have similar warnings.

That type="file" is a bugger, you can't style it w/o a script and the default width doesn't show much.

[MikeGale]

On a matter of principle.

If a browser accepts some markup that trumps a written specification, in my book.

I want something that works in the real world. (I have a low opinion of parts of some specifications which simply look nutty to me. Implementers often seem to agree and just do what they believe is right, so I'm not alone!)

I know it's harder work for Albert, but messages which reflect the real world are valuable to me.

[Albert Wiersch]

Couldn't you just label it as invalid for html5? I think you already have similar warnings.

I suppose I could, but I don't really see the point in terms of usefulness. If it works, then I'll just leave it as is without generating another message that probably won't be useful. However, if I find out that "size" doesn't work in a particular browser then I could generate a message about that.

CSE HTML Validator has always been more about "what works in the real world" than adhering strictly to the specs.

[Lou]

On a matter of principle.

If a browser accepts some markup that trumps a written specification, in my book.

I want something that works in the real world. (I have a low opinion of parts of some specifications which simply look nutty to me. Implementers often seem to agree and just do what they believe is right, so I'm not alone!)

I know it's harder work for Albert, but messages which reflect the real world are valuable to me.

On a matter of principle -- WHAT?:????

Which is why there are so may patches/fixes/hacks to make the standard work in different browsers, because they "just do what they believe is right"

At least the standards are documented. JMHO