dretblog

What's the Web, Anyway?

there is a great story about how the british Advertising Standards Authority (ASA) has decided that apple is not allowed to say that the iPhone gives users access to all parts of the Internet. the reason is that the iPhone does not support Flash and Java, and thus sites using Flash or Java cannot be accessed (they were probably talking about the web), the official conclusion is:

The ad breached CAP (Broadcast) TV Advertising Standards Code rules 5.1 (Misleading advertising), 5.2.1 (Evidence) and 5.2.2 (Implications). The ad must not be broadcast again in its current form.

this is a serious misunderstanding of the web. there is a great difference between some content being on the web and content being of the web. web standards such as HTML and GIF and JPEG are part of the web, so any web access not supporting these formats would be seriously limiting. proprietary formats such as Flash, Java, PDF, MS office, Silverlight, and who knows what, can be easily made available on the web, but that does not mean they are web standards. anybody making information available through these formats does so at their own risk.

when the iPhone was released without Flash, i hoped this would be the death of Flash. that was maybe a bit too optimistic (at least in the short run). but the increasing variety of web-enabled devices will produce more and more tensions between web content, and companies just using the web for transporting proprietary content.

the iPhone story of course is just one little part of the bigger development, and there really is a bigger story behind all of these technologies that are pushed by vendors. developers are lured into using them by giving them polished development environments which conveniently build on the closed-world assumption of these technologies. developers also get prepackaged presentations and examples about how these technologies are the next big thing. all of this conveniently ignores the fact that the web thrives exactly because it is not a closed world, and that interacting in an open world will always require some trade-offs.

as a starter, it would be helpful if people would at least make an effort to differentiate between the Internet and the Web. even if apple uses sloppy terminology in its ads, the ASA ruling could have used technically correct terms.

Go Up

Go Up has been one of my favorite firefox add-ons for a long time: for any site with a well-defined URI structure, it is easier to just go up a level rather than searching for the navigation links on the web page.

the only thing i did not like was the fact that the extension did not treat fragment identifiers as navigational structures. so going up from http://dret.net/netdret/publications#wil02h took you to http://dret.net/netdret/ instead of http://dret.net/netdret/publications. for web sites using fragment identifiers, it would be more convenient to be able to go up from a fragment to the complete resource (i.e., to simply strip the fragment identifier from a URI if there is one).

after proposing this change to the author of Go Up, he responded almost immediately (thanks, stephen!) with a new version implementing the new behavior (it will be released soon). the nice thing is that this change breaks nothing, and makes navigation on sites using fragment identifiers much more convenient.

this is one more reason to install Go Up, it is just a small button in the navigation bar and you will like it and wonder why it's not there by default. it will also make badly designed sites a bit less attractive, because you cannot navigate them using URIs.

Internationalized Quotes

one of the things HTML is good at is that it supports at least some semantics, for example for using quoted text by using the <q> element, if you are using a decent browser (specifically, not IE). you can even nest quotes and they will display differently. that's how good HTML can actually be when used correctly (and when used with a non-crappy browser).

the problem is that different languages have different rules for using and nesting quotes. so while decent browsers support quotes, they don't support them in a language-sensitive way. consider the following test file for built-in support for language sensitive quotes:

<p>Some <q>language-neutral <q>quotes</q></q>.</p>
<p lang="en" xml:lang="en">Some <q>English <q>quotes</q></q>.</p>
<p lang="en-US" xml:lang="en-US">Some <q>U.S. English <q>quotes</q></q>.</p>
<p lang="en-GB" xml:lang="en-GB">Some <q>U.K. English <q>quotes</q></q>.</p>
<p lang="de" xml:lang="de">Some <q>German <q>quotes</q></q>.</p>
<p lang="de-CH" xml:lang="de-CH">Some <q>Swiss German <q>quotes</q></q>.</p>
<p lang="de-AT" xml:lang="de-AT">Some <q>Austrian German <q>quotes</q></q>.</p>
<p lang="fr" xml:lang="fr">Some <q>French <q>quotes</q></q>.</p>
<p lang="fr-CH" xml:lang="fr-CH">Some <q>Swiss French <q>quotes</q></q>.</p>

none of the browsers i tested actually display the quotes as they should be displayed; firefox and opera at least support nested quotes by using different quotation marks; safari is too stupid to properly handle nested quotes; IE is too stupid to even support quotes at all.

this means that the browers' built-in stylesheets apparently are not very sophisticated. correct language-sensitive quote formatting could be easily implemented by CSS code like this:

q             { quotes: '"' '"' "'" "'" }
q:lang(en)    { quotes: '“' '”' '‘' '’' }
q:lang(en-US) { quotes: '“' '”' '‘' '’' }
q:lang(en-GB) { quotes: '‘' '’' '“' '”' }
q:lang(de)    { quotes: '„' '“' '‚' '‘' }
q:lang(de-CH) { quotes: '«' '»' '‹' '›' }
q:lang(fr)    { quotes: '«' '»' '“' '”' }
q:lang(fr-CH) { quotes: '«' '»' '‹' '›' }

and if you look at the above HTML using this improved CSS for quotes, then firefox and opera are getting it right. both browsers even understand that de-AT is a variant of de and use the de rule as fallback, because there is no explicit de-AT rule. that's pretty neat. but safari still only supports one kind of quotes; is WebKit really that bad?

all of this simply translates to: if you want proper internationalization for quotation marks, don't trust the browsers. if you absolutely need quotation marks to be displayed (IE simply ignores <q> elements), you might want to insert them using scripting (don't insert them as text!), because IE once again does not implement standards.

Helvetica

i just watched Helvetica, a documentary about — surprise — the typeface Helvetica. it's very well done and really entertaining, my favorite screen shot is shown here, a really wonderful skateboard design! (the film is really not like that, i probably picked the most controversial image possible.)

what i particularly liked about the film is that both sides are being heard: the helvetica fundamentalists who think that typefaces should be neutral and that legibility is of utmost importance and that helvetica is the way. and on the other hand the designers who think that typeface should be specific for a job and that the ubiquity of helvetica calls for a bit more variety, but of course sometimes overdoing it.

what i missed was a bit more context. the film nicely describes helvetica's history and how designers felt about it during various trends in the designer world (and i for my part am very happy that the grunge design phase is over now). but the film only very briefly mentions Arial and does not even mention the whole issue around how to make money with typefaces. and the film also does not mention more modern fonts, such as Univers, which do have some success and simply are not as widespread because you don't get them for free on your mac...