W3C home > Mailing lists > Public > whatwg@whatwg.org > July 2009

[whatwg] do not encourage use of small element for legal text

From: Ian Hickson <ian@hixie.ch>
Date: Sun, 19 Jul 2009 10:29:02 +0000 (UTC)
Message-ID: <Pine.LNX.4.62.0907191010520.23663@hixie.dreamhostps.com>
On Thu, 2 Jul 2009, Markus Ernst wrote:
> Ian Hickson schrieb:
> > On Thu, 4 Jun 2009, Andrew W. Hagen wrote:
> > 
> > > Encouraging use of small print for legalese also encourages this:
> > > 
> > > <h1>
> > > <a href="continue.html">
> > > Welcome to the BigCo web site. Click to continue.
> > > </a>
> > > </h1>
> > > <small>By clicking above, you agree that BigCo can charge your
> > > credit card $10 per visit to the BigCo web site per page clicked.</small>
> > 
> > Right, that's the case we do want to encourage. It's better than the
> > alternative, which would be:
> > 
> >  <style>
> >   .s { font-size: smaller; }
> >  </style>
> >  <h1>
> >  <a href="continue.html">
> >  Welcome to the BigCo web site. Click to continue.
> >  </a>
> >  </h1>
> >  <span class=s>By clicking above, you agree that BigCo can charge your
> >  credit card $10 per visit to the BigCo web site per page clicked.</span>
> > 
> > ...because if they use <small>, you can configure your client to go out of
> > its way to highlight <small> text, whereas you have no way to know to
> > highlight any text based on its font size or class.
> 
> Anyway that does not prevent the BigCos from using <span> or <p> or <div>, if
> they really want to style their fraudulent text the way it is hard to read.

Indeed, nothing will. But as far as possible, we want to encourage such 
test to be marked up in way that can be detected in this way.


> The more user agents will be set to display the <small> element big, the 
> less this element will be used by those who are actually addressed by 
> this encouragement.

Sure, I wouldn't expect this to be a common feature.


> Instead of keeping a purely presentational category such as "small" as 
> an HTML element, would it not be more efficient to use some kind of 
> <legal> element?

What would the benefit be?



> User agents then could be configured to ignore small text sizes or badly 
> visible colors on <legal> elements.

Surely this would hve the same problems as <small> that you describe 
above?


> Also, other ways to bar people from reading legal text, such as setting 
> it in uppercase characters, could be handled - which does not seem 
> appropriate for a <small> element.

I'm not sure I follow.


> And countries willing to protect their people from fraud could establish 
> a law that any text on a website is only legally binding when it is 
> marked up with the <legal> element.

That's an excellent reason to not use an element specifically for legal 
text -- I really don't want laws to be written about HTML. That way lies 
madness, and significant reductions in the possible ways we can evolve the 
language.


On Thu, 2 Jul 2009, Andrew W. Hagen wrote:
>
> The text from the current spec is, "Small print is typically legalese 
> describing disclaimers, caveats, legal restrictions, or copyrights. 
> Small print is also sometimes used for attribution."
> 
> By suggesting it is typical, that implicitly encourages people to use 
> small print for legal text.

I think the horse left that barn decades ago.


> One of HTML 5's design principles is to take the predominant
> practice of what people are doing on the web, and turn that
> into the standard that all HTML 5 authors should follow.

To some extent, yes.


> Certain sections in the spec are normative or non-normative. It does not 
> make sense for the HTML 5 spec to make a non- normative comment in a 
> normative section. Furthermore, it is fallacious to argue that something 
> is not "encouragement" but is "normative."

The sentence in question is explicitly marked as a non-normative note, so 
I don't think it's misleading in the way you describe. It's following the 
conventions used throughout the document.


> Leaving any suggestion in the HTML 5 spec that legal text typically is, 
> could be, or should be in small print could do a disservice to anyone 
> reading the spec.
>
> Take a hypothetical example. Joe is the owner-operator of SmallCo, a 
> building contractor. Joe decides to create a web site for his business. 
> Being technically proficient, he consults the HTML 5 web site. He reads 
> how the <small> element is typically used for disclaimers, etc. On his 
> one- page web site, he posts the following notice:
> 
> <p>Your 100% satisfaction in the work of SmallCo is guaranteed. 
> <small>Guarantee applies only to commercial buildings.</small></p>
> 
> Joe gets a new customer from the web site. The customer has him build a 
> residential house. Joe does the job. The work is done well, but the 
> customer is not 100% satisfied. The customer wants various changes that 
> Joe does not consider necessary. Joe and the customer have a serious 
> dispute. The customer claims that $45,000 worth of work must be 
> performed for the customer to be 100% satisfied. Joe claims that his 
> obligations are already fulfiled, and the 100% satisfaction guarantee 
> does not apply to the customer's job because his residential house is 
> not a commercial building. They take their dispute to court.
> 
> In this particular case, the disclaimer Joe posted on his web site 
> becomes a hotly disputed issue. The judge takes special note that Joe 
> had his disclaimer in small print. Due to the law of the particular 
> jurisdiction they are in, Joe loses. Joe now must perform a great deal 
> of work for no pay, or pay the customer $45,000 to hire another 
> contractor. The financial fortunes of SmallCo and of Joe have taken a 
> terrible turn for the worse. Joe feels like the HTML 5 spec gave him bad 
> advice.

This seems like a huge stretch. The text on the Web site would almost 
certainly have nothing whatsoever to do with the court case, since this 
kind of stuff is all covered by rules regarding contracts and so forth. In 
fact, it would be shocking if such a contractor were to have such a vague 
disclaimer on his site and not have an explicit contract with the 
customer. So I don't buy that this example is realistic.


> Whether legal text may properly appear in small print has been the 
> subject of numerous lawsuits of greatly significant value, at least in 
> American law. The implicit encouragement to use small print takes the 
> lexicographical mandate of "seeing what people are doing on the web and 
> then making that the standard" too far. It should be impossible for 
> anyone reading a tech specification to think that they are getting legal 
> advice. Yet that is not the case in the current HTML 5 spec.

No, it's really not. I think it is ridiculous to suggest that the current 
text comes even close to appearing anything like legal advice.


> The current situation in the spec is really bad. As it reads currently, 
> the spec encourages fraud. This should not be acceptable considering how 
> much of a benefit HTML 5 will be otherwise.

The spec does not encourage fraud, come now.


> Additionally, the word "legalese" has a strongly negative connotation. 
> Using the term "legalese" as a stand-in for any legal text implies that 
> all legal text is suspicious or useless at best. The spec should not sit 
> in judgment on the law. Espousing a vaguely anarchical political point 
> of view is OK, but that should not be part of HTML 5.

The term legalese just means hard-to-follow legal text, and that precisely 
describe "small print", exactly the kind of thing that I would expect 
<small> to be used for. We do a disservice to readers if we try to be 
politically correct instead of just being blunt and honest.


> What should be done is to remove all references to the law, legalese, 
> and such from the small element section. If people want to put something 
> legal in a small element, let them make their own choice. Let them come 
> up with that generally bad idea on their own, or perhaps in consultation 
> with their own attorneys who can take responsibility for their error if 
> the use of the small element causes some problem. What the law actually 
> says will vary according to your jurisdiction.

I respectfully disagree.


> To sum up, legal text generally should not be in small print, because 
> legal text is important and should be easy to read.

Indeed, that's why <small> is a good idea. It allows small print to be 
made legible when properly marked up. It's certainly better than just 
having authors style it small with no hook for the user.


> The exact legal consequences of legal text in small print will vary 
> depending on the facts of the case and on the jurisdiction.

Indeed.


> The HTML 5 spec should not make assertions, either normative or 
> non-normative, about legal matters.

It doesn't. It makes observations about typical practice.


> The HTML 5 spec is going to be a terrific advance, and removing this 
> thorn from its side will only improve it.

I do not believe this part of the spec will have the slightest effect 
whatsoever on HTML5's success or lack thereof.


On Fri, 3 Jul 2009, Eduard Pascual wrote:
>
> It's clear that, despite the spec would currently encourage this 
> example's markup, it is not a good choice. IMHO, either of these should 
> be used instead:
>
> <p>Your 100% satisfaction in the work of SmallCo is guaranteed. 
> (Guarantee applies only to commercial buildings.)</p>
>
> or
>
> <small>Your 100% satisfaction in the work of SmallCo is guaranteed. 
> (Guarantee applies only to commercial buildings.)</small>

In practice, if the author wants to make the parenthetical text smaller, 
he will. The question is whether we should encourage such small text to be 
marked up in a way distinguishable from other stylistic <span>s.


> I'm not sure if the word "legalese" was intended to refer to all kinds 
> of legal text, or just the "suspicios or useless" ones. In any case, a 
> more accurate wording would help.

This wording is vague intentionally, because it is a vague observation. I 
don't know how we could make it more accurate.


> First, leave the formal description "The small element represents
> small print or other side comments." as is: IMHO it is accurate and
> simple, and that's quite enough to ask from a spec.
>
> Next, replace the note that reads "Small print is typically legalese
> describing disclaimers, caveats, legal restrictions, or copyrights.
> Small print is also sometimes used for attribution." with something
> like this: "Small print is often used for *some* forms of legal text
> and for attribution. Defining which kinds of legal text should be on
> small print, however, is out of the scope of HTML."
>
> This makes clear that HTML (technically) allows using <small> to put 
> legal text (or anything else) in small print, but it doesn't encourage 
> any specific usage of small print.

I'm not convinced the suggested text is any better than the current text, 
to be honest. I'm also reluctant to start explicitly saying what is out of 
scope, because there's no end to that list.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Sunday, 19 July 2009 03:29:02 UTC

This archive was generated by hypermail 2.4.0 : Wednesday, 22 January 2020 16:59:14 UTC