The LaTeX2HTML Translator

Drakos/Moore

To: Internationalisation

Variation with HTML Versions

The Hypertext Mark-up Language (HTML) is an evolving standard, with different versions supporting different features. In order to make your documents viewable by the widest possible audience, you should use the most advanced HTML version with widely-accepted usage.

Currently the most advanced is HTML 4.0. However this has only recently become an officially recommended version. Not all of its features are fully implemented in popular browsers, and the level of usage is unclear. Hence the default version for L^ATEX2HTML, version 98.1 remains at HTML 3.2. Further work is required before L^ATEX2HTML can fully exploit the features available using HTML 4.0.

This provides support for alignment of headings, images and text (including text-flow around images), tables with separate captions and alignment of rows and columns, variable sizes and colors for text and color or patterns for the background as well as images, server-side image-maps, interactive forms, and the minimal typographic elements (bold, italic and teletype) that were supported already in HTML version 2.0. Furthermore, HTML version 3.2 adheres to the ISO-Latin-1 (ISO-8879) character set.

Note: Although many people still use old browsers that implement only features available with HTML 2.0, this is not a good reason to limit translation of documents to using only these effects. Most of the translation done by L^ATEX2HTML will still give acceptable results on older browsers. The deficiencies due to lack of super/subscripts, tables and some alignment effects should eventually convince such users to overcome the inertia, and update their browsers to later versions that correctly support these effects.

Sometimes it is known that the audience, for which a specific document is intended, has limited browser capabilities. Or perhaps special extended capabilities are known to be available. The L^ATEX2HTML translation may be customised to suit the available functionality.

Other HTML versions and extensions supported by L^ATEX2HTML are described below. See the description of the -html_version command-line option switch.

Version 2.0

This provides only the functionality of the HTML 2.0 standard. There is little provision for aligning headings, paragraphs or images nor for super/sub-scripts to be generated. Images are created for tables and other environments that use <TABLE> tags with HTML 3.2; e.g. eqnarray and equation with equation numbering.

i18n (internationalised fonts) Version 2.1

This extension (formerly known as HTML version 2.1) provides extensions for internationalisation. Most importantly, the default character set is no longer ISO-8859-1 but ISO-10646 (Unicode). This is a 16-bit character set and can thus display a much larger set of characters. There are also provisions for bidirectional languages (e.g. in Arabic the text is written from right to left, but numerals from left to right), and provisions in HTML to determine the character set and the language used.

Not all of the symbols are available in TEX, L^ATEX2HTML, or any browser yet available. However the `i18n' extension to L^ATEX2HTML is in preparation for when such browsers do become available, and such characters will be required in Web-accessible documents.

tables (HTML3 model)Version 2.2

Although HTML 3.2 implements tables using <TABLE> tags, the capabilities available to specify details of the table-layout are not as extensive as were originally proposed in the HTML3 Table Model. This extension (formerly referred to as HTML version 2.2) provides the full capabilities of this model.

Note that current browsers may not correctly interpret all the features of tables constructed using this extension. Tables will be constructed, perhaps with some cells mis-aligned or without the desired merging of adjacent cells, etc. This feature was already available in many HTML browsers, including Netscape Navigator V1.2, so should be still available with later versions of these browsers.

HTML 3.0

This version of HTML was never accepted to become a recognised standard; perhaps because some of its models were ``too advanced'' at the time (notably the HTML-Math and the Table Model). The proposed HTML 3.0 ``standard'' was withdrawn and re-drafted to create the HTML 3.2 standard which is in current use. Standard textual formatting features, including centering, flush-right, flush-left and underlining are among the features retained.

math (HTML3 model) Version 3.1

This extension (formerly referred to as HTML version 3.1) adds support for the HTML-Math model, originally part of the proposed HTML 3.0 standard, see above. The only available browser which can display this mark-up is Arena. Originally developed by the World Wide Web Consortium as a test-bed browser, it is no longer supported by them.

There has been a recent proposal for a Mathematical Markup Language (MathML) from the W3C Math Working Group. This would suggest that the HTML-Math model is unlikely ever to be adopted; better things being expected in the near future using MathML.

See also another page for a discussion the the mechanisms available with L^ATEX2HTML for handling mathematical equations and expressions.

To: Internationalisation

Ross Moore 1991-03-26

Processed by LaTeX2HTML-FU