W3C home > Mailing lists > Public > www-validator@w3.org > June 2012

textaera validator rows, cols and the standard

From: bill kelley <bill@windhome.com>
Date: Thu, 07 Jun 2012 12:44:10 -0500
Message-ID: <4FD0E86A.2030203@windhome.com>
To: www-validator@w3.org
Preface, The validator is "working correctly" in that it calls out as 
error the rows and cols attributes
of TextArea, which are both new "required".

So this is a comment on the standard, as opposed to a bug report of the 

TextAreas were not broken pre "rows and cols". Not only are rows and 
cols not "required"
they are not even useful as hints.

Like all of html, TextArea is constrained by style and page size to fit 
a particular area.
And like all other input tags, It is required to handle the data given it.

Neither the data nor the area will conform to any fixed rows and cols, 
or even a "maxrows or maxcols",
Nor is it useful that it should.

Brings us back to the days editors had fixed widths, which in turn went 
back to IBM punch cards.
Even then they recognized the need for variable rows.

So, given rows=20 and space for only 2 rows, does the browser scale the 
font down to unreadable dots to
make 20 rows appear? Probably not. If data is 100 rows, is it ok for a 
scroll bar? I assume so.
So space determined the display, and data determines the, um, data. What 
does rows determine?

So in fact, the determining factor of TextArea behavior is the space 
available and the data it must be able to display.
Same for columns. Rows and Cols add new, completely  unusable 
constraints to TextArea

The one usable hint, maxlength, is now gone. Maxlength determines the 
amount of storage required
to save TextArea, to a database for example. Very useful for say a web 

Rows times cols is not the same, as most rows are less than the maximum.

I once served on the ANSI SQL committee, and saw what happens when the 
Standards group thinks
it will change the worlds requirements, and html standards are there.

I only point out the "validator", although following the new "standard" 
is calling an error
an omission of UNUSABLE attributes.

Making the validator less useful.


bill kelley
512 266 1896 *
Received on Friday, 8 June 2012 15:38:53 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:18:06 UTC