Re: Code2000 on SourceForge (was Re: [indic] Re: Lack of Complex script rendering support on Android)

From: Philippe Verdy <verdy_p_at_wanadoo.fr>
Date: Tue, 15 Nov 2011 01:26:03 +0100

We don't know exactly the reasons why James has closed his website or
even his support for this font. But if he really was alone he probably
did not havethe money to support the site, or the bandwidth used for
it (may be he's now in financial troubles, like loss of his job in
this hard times). Or may be a receive a DMCA warning and did not have
the money to defend his work.

That's the risk when you work alone without any profit from your work,
despite you wanted to contribute to a much larger community that
recognized his work. At some point, he should have realized that, to
defend his work, he should have deposited on a serious online
repository defended by lots of people and supporters (many choices:
GNU Foundation, Mozilla Foundation, Apache Foundation, Ubuntu, and
even Sourceforge), where the work would have been saved, while at the
same time allowing him to choose helpful contributors to maintain the
repository or track the discussions and bug reports, and propose
changes and enhancements.

We will miss Code2000, Code2001 and Code2002, because free and
open-sourced fonts still lack good coverage of lots of scripts with
quality fonts, and at least those full-plane fonts could be used as a
reasonable fallback to catch all those that were and are still
missing.

Now we are left with the excellent Everson Mono font, that
unfortunately only covers the BMP, but not completely due to technical
problems/limitatons in the font editing tools used by Michael. This is
also a demonstration that creating large fonts is still a difficult
challenge, even for experts. And that we still need a serious revamp
of font editing tools that would be maintained and usable by much more
people, and based on non-proprietary editable databases for glyph
definitions.

I just hope that one day, the SVG-font format will be extended to
include the script/language and typographic features (notably the
substition and positioning rules and lookup tables, at first, then the
optional support for hinting instructions and programs, something that
will now be less necessary with the growing screen resolution
including on small devices), in such a way that fonts will be less
monolithic, more easily extensible (thanks to the XML format which is
highly interchangeable), and even allowing fast and incremental
distribution of glyph definitions and features via very small request
to an online font server that will also be able to serve smartphones,
without requiring to install fonts that hard difficult to track in
their versions and not easily updatable (web distribution allows the
management of a font cache directly bounf to the HTTP client cache).
We could even have fonts directly manageable in the HTML5 dataset,
where they would be preparsed for fast rendering.

And it would be time now to promote the SVG font format because it is
much more easily updatable, and can forget lots of bad legacies
susbisting in the OpenType (or ISO open font) format. I just wonder
who is working on enhancing the SVG format to make it equally capable,
at least for international support, with OpenType (leaving hinting for
later, notably because SVG renderers all have support now for
excellent subpixel rendering, transparency, and in a more generic way
that is agnostic to the pixel geometries and alternate color spaces
that start appearing with more than 3 pigments and subpixels with
different relative geometries/surfaces depending on their native
pigment, as well as new refinements for those pigments or microoptical
lenses correcting the anisotropy of subpixels alignments, or
multilayered screen surfaces, and alternative non-orthogonal grids for
better isotropy that avoids very undesirable moiré artefacts: all
those things that ClearType for example does not cope with, as well as
basic OpenType hinting).

Basically, I am convinced now that font hinting has little or no
future, it just complicates things in trying to solve another more
typographical problem (i.e. the humane vision artefacts, which soon
will be also personnalisable as well for each user depending on his
effective vision and preferences). For the rest, the problem of
readable small fonts should be the same as the general problem of
rendering textures in photography, and trying to remove the artefacts
caused by artificially quantized pixel geometries and colorspaces.

2011/11/7 Andrew West <andrewcwest_at_gmail.com>:
> On 7 November 2011 08:34,  <anbu_at_peoplestring.com> wrote:
>>
>> Code2000 supports most BMP code points of Unicode 5.2. It is open sourced
>> from September:
>>
>> http://code2000.sourceforge.net/
>
> I have doubts as to whether this project was actually created by James
> Kass.  The project comprises the last public version of code2000.ttf
> and a 210MB "code2000.asm" file which turns out to be a dump of the
> ttf file in human-readable form, both of which could easily have been
> put onto SourceForge in contravention of copyright and license by
> someone pretending to be James who wants the font to be open source
> now that the official Code2000 site has disappeared.  James once told
> me that Code2000 was maintained as a "66 megabyte dBASE III database
> file" which is not what is on SourceForge.
>
> Andrew
>
>
Received on Mon Nov 14 2011 - 18:29:22 CST

This archive was generated by hypermail 2.2.0 : Mon Nov 14 2011 - 18:29:23 CST