This message is in response to the previous messages in this thread.
There *IS* a viable solution to the whole problem of "adding a few extra
characters" to a font without having to wade into the potential legal
morass of individual font vendor's intellectual property rights:
The solution is to use collaborative Open Source development methodologies
to produce one or more high-quality, operating system and vendor-neutral
TTF and OpenType unicode fonts. Resulting fonts would be copyrighted and
released under a well-known Open Source license like the General Public
License (GPL, http://www.gnu.org/copyleft/gpl.html) or a BSD-style
license.
IMHO, there is a great need for a funded collaborative project of this
nature which would benefit the entire Unicoding community, the Open Source
community, the commercial software vendors who are already benefitting
from, and contributing to, Open Source technologies (Apple and Sun, just
to name two), and the computing world at large.
As many of you know, Roman Cyzborra (czyborra.com) produced a bitmap
unicode font called Unifont using Open Source methodologies. Many people
contributed, and continue to contribute, bitmap glyphs for additional
scripts. The font is now distributed in major Linux distributions and is
quite useful, for example as a console unicode font for "simple"
languages that don't have special rules for conjoining characters or
glyphs that go above/below/before/over/under other gylphs (in other words,
it's great for European languages, Chinese, Japanese, etc., but less
useful for Thai, Indic scripts, Arabic). Just the other day on this
mailing list there was an email about somebody packaging Unifont into a
TTF so that it could be used on Windows or something like that ...
Unifont proves that a collaborative Open Source font creation project can
be very successful. Of course, an Open Source TTF or OpenType unicode
font creation project is perhaps an order of magnitude more complex than
the Unifont project. Just thinking about maintaining consistency in style
and quality with multiple contributors will certainly lead some readers,
of this list to conclude that it's not possible!
Disregarding the potential difficulty, let me just note that such a
project already EXISTS!: Primoz Peterlin has such a project at
http://savannah.gnu.org/projects/freefont/. The project appears to be
making good progress and might be the answer for those of you who want to
see quick turn-around times on the addition of specific glyphs. For
fastest turn-around times, join the freefont project and add the needed
glyphs yourself ;-). Note that the project uses an Open Source font
editor, pfaedit (http://pfaedit.sourceforge.net/), which looks pretty good
at first glance. Pfaedit has at least two very important features. The
first is the ability to place a scanned image on the background of the
workspace. This is great for people like myself who are much better, or
at least faster, at drawing and designing on paper first. The second is
the display of glyphs in unicode order. Check it out!
On Fri, 30 Aug 2002, William Overington wrote:
> Peter Constable wrote as follows.
>
> >On 08/27/2002 12:08:09 AM "James Kass" wrote:
> >
> >>William Overington has mentioned the Softy editor. Please keep
> >>in mind that fonts are copyrighted material, and, mostly users
> >>are forbidden to modify them, even for internal use purposes.
> >>
> >>The best way to get characters added to a font is to ask the
> >>font's developer.
> >
> >I agree completely. Also, it's worth noting that font engineering involves
> >rather more than just "adding a few extra characters", especially when
> >"smart fonts" are involved. Note, for instance, that some tools may trash
> >the hints in a font. The overarching issue, though, as James pointed out,
> >is that very often it is simply not legal to make such changes.
> >
> >
> >
> >- Peter
> >
>
> James raises the important matter of intellectual property rights in fonts
> and suggests that the best way to get characters added to a font is to ask
> the font's developer.
>
> Peter agrees with James and adds some good computing reasons as to why, even
> if permission were available, simply adding a few extra characters without
> highly expert skills would not be an effective solution.
>
> The matter which concerns me is as to whether James' suggestion that the
> best way to get characters added to a font is to ask the font's developer,
> while probably quite true, is nevertheless, in effect, what the theory of
> procedural rules would, if that course of action were a formal motion for a
> meeting, term a pious motion. What I mean by this is that, for example, if
> someone does want a particular character added to a font, how effective, in
> practice, is such a request likely to be, qualitatively in terms of whether
> such a request would be accepted at all, and quantitatively, in terms of
> time scale and financial charge, as to how accessible such an addition would
> be to someone making such a request of a font's developer.
>
> Now, let me say at once that James has already shown in his posting that, in
> the particular case of the situation in the thread from which this
> discussion has spun, that he has reacted proactively in setting about adding
> U+FE20 and U+FE21 to his own font, and hopefully the results of that
> addition will be available to all at the next release of that font.
>
> Yet is that a response which could be anticipated as typical of font
> designers? James produces mainly one huge font which covers many Unicode
> characters and is continually adding items to produce a better version for a
> later release.
>
> What is the situation with other font developers? Is it perhaps the case
> that some font designers, or a team, produce a particular font and then
> wrap up the project, so that adding a few extra characters at a later date
> would mean a substantial restarting up of the project? I do not know the
> answer to this and I wonder if some font designers could perhaps comment
> upon the possibilities and the modalities of someone getting a font
> developer to add a few characters to an existing font please.
>
> The whole situation has led to me trying to think out the problem of how
> someone could get a few extra characters added to a font and, recognizing
> the issues and problems that James and Peter mention, I wonder if I may
> perhaps put forward a few thoughts on the matter, which might perhaps lead
> to a new infrastructural facility for end users of the Unicode system.
>
> Firstly, I mention that I know, at present, very little about font
> authorship. I have only used the Softy program and not all of the
> facilities in Softy yet. I am aware that there are various sophisticated
> font authoring packages available, which are expensive and not widely
> accessible by many end users of Unicode.
>
> When using Softy, one method of designing a glyph is to load a template,
> which is a .bmp file of a large monochrome image of the desired end result
> in a .bmp file, say about 200 pixels by 200 pixels or thereabouts, and then
> to use Softy to automatically outline the template so as to produce the
> Bézier curves for the glyph. The template file can be produced using a
> widely available package such as Microsoft Paint. I have had a lot of
> learning fun producing experimental glyphs for a few characters using Paint
> in this way, including using the line, ellipse and curve tools of Paint to
> produce two tengwar-inspired fantasy characters, namely a double thorn and a
> double thorn with tilde, in a manuscript style, by drawing upon a background
> grid of a different colour.
>
> I wonder whether it would be possible for some interested people to devise
> some basic grids in .bmp format with green and cyan lines upon them so that
> the containing boxes for letters x, h, p, a circumflex, A circumflex and so
> on were indicated, so that any interested end user could draw a desired
> character, using Paint, upon a copy of such a grid using black, then erase
> the green and cyan lines. This could have the effect that if, say, twenty
> to fifty end users each produced designs for five or more characters, that
> the artwork for an easily extendible font could become available. Clearly
> such an idea would need someone with the necessary software and skills to
> gather the glyphs into a font. I am thinking that if such a font designing
> facility had a distinctive name, then anyone producing such .bmp files
> could, if he or she so chose, place them upon his or her own website using
> that distinctive name as a key word, so that, in time, someone going to a
> search engine such as http://www.yahoo.com could search for the word and
> perhaps find all of the available glyph designs from web sites all around
> the world, together with fonts which included different subsets of the
> available glyphs. Producing such glyphs and fonts could become an
> intellectual sport, in the same manner as playing chess is an intellectual
> sport.
>
> I appreciate that there would naturally be intellectual property rights
> issues in relation to such fonts, yet it does occur to me that trying to
> produce several such designs, such as a sanserif face, an old style face and
> a manuscript style face using tightly drawn rules for how to prepare the
> template designs, could be lots of fun and could lead to some useful fonts
> which are extensible by end users for various specialised applications.
>
> I realise that there is a lot of skill in producing fonts and I wonder
> whether what I am suggesting would be feasible or whether there would be
> some factor which would make it unworkable in practice.
>
> I realize that such a discussion of font generation might be off-topic, so
> maybe the discussion would need to move to email, yet perhaps for a while a
> discussion in this forum might be permitted please?
>
> William Overington
>
> 29 August 2002
>
>
>
>
>
>
>
>
>
>
>
This archive was generated by hypermail 2.1.2 : Fri Aug 30 2002 - 11:53:51 EDT