Aller au contenu | Aller au menu | Aller à la recherche

Sunday 30 April 2006

Full unicode widespread fonts : The Holy Grail of I18N

Unicode, which is a consortium founded in 1991, aims to provide a universal standards to allow each glyphs in whatever scripting to have a unique code and is close to release version 5.0.


Fundamental Web use

For the web, a full unicode fonts, implies that it will be possible to display all web pages in whatever languages to reach informations. For example, Wikipedia make a heavy use of Unicode.

Math use

For math, in the web, Unicode is just mandatory ! A lot of informations use specific characters (more than 2000 are available in Unicode), so they need to be properly displayed. Useful requirements on this could be found here.



Stixfonts is an old project (since 1995) which has tried to give to all a set of fonts that serve the scientific and engineering community. They are lacking of 152 glyphs to complete their requirements.


DejaVu is a recent project (since 2004), hosted by Sourceforge, which is now giving good results and is free as in speech and as in beer. Its coverage is, for the moment, not high but grow up fast at each version (current is 2.5).


And what about Arial, Unimath or Lucida ?

These fonts don't cover a high part of recent Unicode version and are not so widely accessible due to license requirement.

May be we have to wait for a Wikifont ?

Friday 21 April 2006

Automatic validation of Web good practices

This is french (cocorico) and it has just been released

A beta version of the opquast automatic validator !

It has been developped by a young and smart developper of OpenWeb.

Unique problem (for the moment?), it is entirely in french.

So if you catch some words of french, go and take a look Opquast.

Found via Standblog

Sunday 16 April 2006

PVL, a Path Validation Language

The author of Schematron, Rick Jelliffe, has attempted to publish another rule based validation format, whose name is PVL for "Path Validation Language".

The validation format is simple, but not simpler.

The idea behind this is to allow event-based-application (SAX or StAX) to make on the fly validation.

The other difference with Schematron is that PVL define an output where all unnecessary information could be dropped (i.e. whitespaces, comments, etc.).

Here is the core syntax with some adjustment :

    ((((prefix ":")? name)? "/")?         // parent
           ((prefix ":")? (("@")? name))  // element or attribute
            | "#WS"                       // whitespaces
            | "#DATA"                     // datas
            | "#COMMENT"                  // comments
            | "#PI" )                     // PIs
    | "#DOCTYPE"                          // doctype everywhere (at the root I presume !)

For convenience, the namespace has been fixed to "http://pvl.sf.net/pvl/2006"

As a sample here is a PVL validation instance for PVL itself :

<pvl:schema xmlns:pvl="http://pvl.sf.net/pvl/2006">
  <pvl:ns prefix="xpvl" uri="http://pvl.sf.net/pvl/2006"/>
    /xpvl:schema             +   <!-- root element must be 'schema' with the proper namespace -->
    xpvl:schema/#WS          + - <!-- allow whitespace in schema but trim them -->
    xpvl:schema/xpvl:ns      +   <!-- allow 'ns' (namespaces) element      # notice : we cannot enforce cardinality -->
    xpvl:schema/xpvl:actions +   <!-- allow 'actions' element              # notice : we cannot enforce cardinality -->
    xpvl:actions/#DATA       +   <!-- allow data content in actions -->
    xpvl:ns/@prefix          +   <!-- allow 'prefix' attribute -->
    xpvl:ns/@uri             +   <!-- allow 'uri' attribute -->
    #COMMENT                 + - <!-- allow comments in schema but trim them -->
    #PI                      + - <!-- allow PI in schema but trim them -->
    #DOCTYPE                 w - <!-- warn about doctype but remove it -->

This schema implied a lot of fallback : all what isn't precised is error of type 'X 0'

More information will be available soon on http://pvl.sf.net

Information caught via XMLfr

Wednesday 12 April 2006

CSS 2.1, a new version

Each new draft of CSS is an event here at Innovimax.

CSS 2.1 has a new draft out and this is a good news, because it's an important information about the healthiness of the work done around web standards and CSS in particular.

What are the news in this draft ?

Well, first, because it seems to be the last step as a Working Draft, so the next step should be Candidate Recommendation, and last CSS Recommendation is CSS 1 which has been released in May 1998 !

Then a lot of cleaning has been made : typos and duplications have been corrected.

Some very useful precisions have been added especially in tables section.

HTML 4 examples has been updated to HTML 4.01, but some HTML&nbsp;4.0 are still in the doc, and all the doctypes of longdesc are 4.0 Transitional.

And last but not least, because we know now that Wium, in Håkon Wium Lie in an additional-name, information that we haven't in the past !! But what can we say about Jan Roland Eriksson ?

But the real question is why the spec style sheet is not CSS conformant ?

[fr] Creation du blog d'Innovimax

Voici enfin le blog d'Innovimax. Le site web est sur le point de sortir

Innovimax est une société privée créée en fin d'année dernière.

Notre but est de fournir les dernières innovations du web à nos clients, les entreprises.

Le coeur de métier d'Innovimax est les médias : éditeurs papier et fournisseurs de contenus Web.

Nous avons un savoir faire important en XML, SGML et son écosystème (HTML, CSS, SVG, XSL).

Si vous avez des données et que vous voulez les obtenir sur différents supports (Papier, HTML, PDF, Word, WAP, etc.), mais que vous avez des doutes sur la mise en oeuvre, contactez-nous !

Nous sommes un membre actif au W3C

Plus d'informations sous peu

Innovimax's blog

This is the blog of the Innovimax Company. For the moment, our website is about to be released soon, so stay tuned.

Innovimax is a french privately helded company, which has been created at the end of last year.

Our aim is to provide web innovations to our customers, the companies.

The core of Innovimax is specialized in media companies : book publishers, web content providers.

We have a strong knowledge of XML, SGML and its ecosystem.

If you have datas, you want them in multiple different formats (Paper, HTML, PDF, Word, WAP, etc.), you're not sure how to do this, please contact us.

We are an active W3C Member.

More informations soon.