W3C home > Mailing lists > Public > www-validator@w3.org > July 2015

comments on code validator

From: Patrik Eriksson <pat.eri@hotmail.com>
Date: Mon, 20 Jul 2015 06:54:20 +0200
Message-ID: <DUB111-W1395F23A74D6684C6A66489FA850@phx.gbl>
To: "www-validator@w3.org" <www-validator@w3.org>
some comments about parsing errors

<script type="text/javascript" type="application/javascript">
is reported incorrecly as having duplicate type attributes
its actually defining two MIME types for it one for older parsers
who will fail if encuntering application/javascript 
and one for newer as text/javascript is obsolete

text/javascript is the default value - but its absolete so application/javascript
must alwas be specified

<div onmousenter="something();" onmouseover="somethinelse();" onmouseout="somemore();">
This work in all browsers since IE5 DOM1</div>

onmouseenter and such reported as not being supported on div elements at this time - they are 

<iframe width="100px" height="100px" class="mainFrame" name="main" id="mainFrame" src="Page..html">
    <p>This text should never be displayed</p>

cases an pareser error width/height in specified in pixels when parsed as html 5/5.1
-again 100 should never be used as a measurment of pixels instead 100px should be
(as specified by the standard), in html5 it must be specified in pixels, assuming 100 means 100 pixels
and not an old html 4.xx percent value is dangerous and possibly(likely) page breaking, it also is dengerous or atleast limiting for furter expansion of the language, the only legal solution at this moment is to specify 0 as in zero for both width and height and use css to alter them, as a zero value dont need any specification of whats its zero of.

...also the parser say something about negative valuse not allowed, well iframes are inline elements so negative values should be perfectly legal (even if they forces the page to reflow), it makes no sense use negative values - but it should be perfectly allright to use negaive values,
it also semms to work as intended on tested browsers IE/Firefox.

ps. fix the typo in the specs that say screen.colorDepth always should return 24 ....

Patrik Eriksson

Received on Monday, 20 July 2015 06:02:41 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:18:13 UTC