Top validation errors

by cientifico • 14 years ago

39 49 14 years ago

This shows how well browsers handle bad HTML.

It always surprises me when I make a mistake in my code and I don't notice it until I next few the source because the browser has understood what I was meant to do.

I guess although this is good for the end user it isn't always a great thing for developers learning HTML if they're able to write bad HTML without ever knowing they're doing something wrong.

pyre• 14 years ago

  > This shows how well browsers handle bad HTML

It's not that hard to handle something like an <img> tag with no alt attribute.

nodesocket• 14 years ago

Anybody else think its a little absurd that `alt` on img tags are still required for valid HTML?

olavk8• 14 years ago

In HTML5, the `alt`-attribute can be left out there is no way to provide an appropriate alternative text. The spec provides the example of "a blind photographer sharing an image on his blog".

In general though, the `alt`-attribute should be present. If the image is purely decorative, the value should be an empty string (although CSS rather than `img`-tags are recommended for purely decorative images anyway).

The purpose of the alt-text is some text to replace the image if you are browsing without images (e.g. using lynx or a screen-reader). Empty string means no replacement text, while a missing alt-attribute could be replaced with a text saying eg. "[unknown image]".

It makes sense.

0x03• 14 years ago

It's absolutely ridiculous, especially since you can get away with alt="".

TazeTSchnitzel6• 14 years ago

No it's not. An empty alt specifies that it's a presentational image. On the other hand, no alt is unhelpful, as you can't tell if the omission was intentional, or the developer is just lazy.

DanBC5• 14 years ago

No. Why is it absurd?

snorkel• 14 years ago

The two most popular are a crock:

* Required attribute alt missing: Most images are decorative, having alt is pointless in most cases. Should be optional.

* & didn't start a character reference: proving this was a bad choice for an escape character to begin with

olavk8• 14 years ago

You indicate than an image is purely decorative by providing the attribute value `alt=""`.

icebraining• 14 years ago

Which provides no more information than no alt attribute at all.

DanBC5• 14 years ago

> Required attribute alt missing: Most images are decorative, having alt is pointless in most cases. Should be optional.

NO. ALT attribute must not be optional. Just put an empty alt in for decorative images and learn how to do sensible alts or long_desc for other images.

And many of the supposedly decorative images are used functionally, in which case they need proper alts.

bickfordb• 14 years ago

I agree, the specification seems to set authors up for failure. I never understood why the language needed & followed by WHITESPACE to be escaped. It seems like it would be pretty easy to ignore this sequence in a parser?

voyou• 14 years ago

AFAIK, using an ampersand character to represent an ampersand, rather than starting a character reference, is no longer a parse error in HTML 5 if it occurs in an attribute. See: http://dev.w3.org/html5/spec/tokenization.html#consume-a-cha...

jaimeiniesta5• 14 years ago

yeah I too agree that the "alt" attribute should not be considered an error. Maybe just a warning.

There's a separate chart for the top 100 warnings:

http://w3clove.com/charts/warnings

jaimeiniesta5• 14 years ago

It's interesting to see the validation for Hacker News:

http://w3clove.com/sitemaps/check?url=http%3A%2F%2Fnews.ycom...

You wouldn't tell there were so many images, would you? Should they have an "alt" attribute?

olavk8• 14 years ago

It is probably the up/down arrows.

jaimeiniesta5• 14 years ago

sure, it's the arrows. Should this be done with characters, instead? Maybe CSS? This would be indeed a good case of using "alt"... alt="vote up", alt="vote down"

DanBC5• 14 years ago

They are functional. They need an alt tag. Having [grayarrow.gif] and [graydown.gif] is sub-optimal.

TazeTSchnitzel6• 14 years ago

While we're at it, we should stop using tables, too.

mistercow• 14 years ago

I think most of those are up/down arrows.

viraptor• 14 years ago

I'd be glad if they changed the format of entities to something like &non-whitespace; suddenly all uses of & on its own becomes legal and is backwards compatible

TazeTSchnitzel6• 14 years ago

This is one of those improvements that could have gone into HTML5. Maybe we'll see it in HTML6.

It's nice how they actually made writing HTML much nicer in HTML5, even without this. The simplification of the doctype and charset (<!doctype html><meta charset=utf-8>) was really nice.

jacobr• 14 years ago

Why use character entities at all when serving documents as UTF-8?

olavk8• 14 years ago

You need them for < > and &.

andrewcooke• 14 years ago

what on earth is "mibenumid" (last entry)? googling for it shows that it's a popular name for a url query param, but i'm not sure how that's relevant (perhaps the error comes from the "&" separator in a URL not being escaped?). is it used by some popular middleware? the pages google is throwing up seem to be asp...

x1• 14 years ago

http://www.iana.org/assignments/character-sets

>The MIBenum value is a unique value for use in MIBs to identify coded character sets.

>The value space for MIBenum values has been divided into three regions. The first region (3-999) consists of coded character sets that have been standardized by some standard setting organization. This region is intended for standards that do not have subset implementations. The second region (1000-1999) is for the Unicode and ISO/IEC 10646 coded character sets together with a specification of a (set of) sub-repertoires that may occur. The third region (>1999) is intended for vendor specific coded character sets.

However that specific error (reference to entity "x" for which no system identifier could be generated) has this message:

>This is usually a cascading error caused by a an undefined entity reference or use of an unencoded ampersand (&) in an URL or body text. See the previous message for further details.

I notice this on one website that is using MIBEnumID:

<span

  class="st_facebook_large"  

  displaytext="Facebook"     

  st_url='http://www.hardrock.com/locations/cafes3/events.aspx?locationid=108&MIBenumID

Could it be that it is just throwing this error because of the unencoded ampersand? And that since it is part of some MS/ASP.net setup it is everywhere?

mbostock• 14 years ago

Pie charts are intended to show how one constituent value compares to the whole. Since these errors are not mutually exclusive (a single page can fail validation for multiple reasons), and since the top 10 reasons constitute an arbitrary whole, they should really be using a bar chart.

Um, like this: http://bl.ocks.org/2941604

jaimeiniesta5• 14 years ago

Sounds good, I'll make an alternative version using bars.

jaimeiniesta5• 14 years ago

Thanks Mike, I've updated the charts. It really looks better as a bar chart:

http://w3clove.com/charts/errors/bar

ehutch79• 14 years ago

uh, forge tthe alt, should the no closing img tag be on the list? or not closing br? i've never used those, ever. they make no sense.

Top validation errors

Related Stories