W3C home > Mailing lists > Public > www-style@w3.org > December 1998

Re: Language (in)dependent hyphenation properties

From: James Aylett <dj-www-style@insigma.com>
Date: Thu, 3 Dec 1998 08:23:45 +0000
Message-ID: <19981203082345.B10589@insigma.com>
To: www-style@w3.org
On Thu, Dec 03, 1998 at 02:58:23AM -0100, Clive Bruton wrote:

> >As a designer and  typographer from the old school and a web designer
> >from the new school, I have one basic thought on hyphenation on the Web
> >- don't. Especially if it is a means for offering justified text.
> 
> Eeuch, what could be worse, justified text on a low res monitor :-)

Flashing justified text on a low-res monitor? :-)

> >That being said, if hypenation must be an option, some of the things to
> >consider include:
> 
> It seems it isn't, at least not in CSS, the only mention is in HTML 4.0.

There's actually a mention in CSS2 (or at least was for WD 04/11/97 - most
of our network is down at the moment, so I can't check). Only a mention,
though, that the construction of anonymous boxes depends on (among other
things) hyphenation, which isn't covered by CSS2.

> Where you effectively have a discretionary hyphen (hy&shy;phen rather 
> than hy-phen). Which isn't much use to anyone as it means that the setter 
> of the text has to enter all of them (I guess you could set a script to 
> analyse the text and do this for you based on an existing dictionary - or 
> maybe not).

I see this as an entirely valid way of doing things; as has been pointed
out, hyphenation isn't a simple area. In a growing world of thin clients, it
might be advisable to assume the densest hyphenation algorithms ("don't"),
and to do the work yourself (or using a tool). I personally publish all my
web sites using a script which does various things such as validation and
consistency checks - it wouldn't be difficult to add pre-processing of the
HTML or proto-HTML in some way.

Having said that, this is not a particularly good argument against having
hyphenation options available in CSS, in my opinion: we shouldn't play to
the lowest common denominator (and don't: we have colour). I simply don't
see (at least for the moment) a world where you can *guarantee* that
hyphenation hints are going to be used.

> >What of the four approaches is to be taken towards hyphenation? 
> 
> In addition to the methods you discuss, maybe browsers could have the 
> option to turn hyphenation on or off?  Also, maybe hyphenation
> dictionaries would come with the browsers, you could add further language 
> support by downloading those of your choice.

Better, the hyphenation system should be a drop-in module. However this
isn't really relevant to this list at all; it's a browser feature. (I'd
argue that it's an obvious browser feature to want, but I'm sure someone
somewhere has arguments against it ... .)

> Alternatively, browsers could use the text processing powers of things 
> like ATSUI or OpenType (do they have built in hyphenation routines?) thus 
> putting such tasks in the realm of the operating system, which is 
> probably preferable.

Really? Can you see hyphenation appearing in microkernel operating systems
such as might be used to build hand-held devices? It's not an operating
system feature at all - but (and I hope that this is what you meant) it
should be a shared resource.

Cheers,
James

-- 
/--------------------------------------------------------------------------\
  James Aylett, dj@insigma.com                    Insigma Technologies Ltd
  Tel: +44 (0)1285 643100                         Norcote Barn     Norcote
  Fax: +44 (0)1285 643600                         Cirencester      GL7 5RH
Received on Thursday, 3 December 1998 03:27:15 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:53:56 GMT