[Openfontlibrary] FOSS and the Commercial Print World

[Raph Levien]

> I will raise it up a notch. Gustavo and Dave, what solid concrete steps
> would you both like to take on getting us there?
>
> I will try to empower you both as much as possible.
>
> What we need the most of is coding on our system, ccHost. Also, not to
> be demoted in value is good fonts...what do you guys want to contribute
> on?

This question wasn't addressed to me, but I'll answer it anyway.

I have to say that I share many of Gustavo's concerns. Basically, the
idea is good, but right now there is a pretty big gap between what's
there and what it should be. I realize it's still in its early days,
and that it will take a lot of work to build a good site, so please
take what I have to say in the spirit of constructive criticism.

Because there are obviously limited resources now (hopefully something
that will change as the site attracts more juice), I think it's
important to make a distinction between the basics and frills.
Ultimately, I'd love to see a site where people can create their own
personal instance of meta-like fonts, select their favorite
alternates, host fonts with version control, and so on. Those are all
frills, and will take a significant amount of code and other work to
make happen.

The basics are fonts. Approximately 30% of the screen area for the
OFLIB home page should be showings of the fonts available. Right now,
that figure is 0%. If you click on the "Fonts" tab, that page also has
0%. Not until you click on an actual font do you see anything
resembling a showing, and even then all you get is an unclickable
thumbnail (approximately 1.5% of page area!) that's been downrezzed
from the original.

I must confess I'm not a big fan of ccHost for this work. Our needs
differ significantly from those of the CC project in general. The
default design has a _tremendous_ amount of clutter, and the structure
does not reflect what's needed for fonts. On the programming side, I
personally can't get excited at all about a PHP codebase.

I'm in the process of developing ghilbert.org using Django, and I am
thrilled with this framework. It's very simple to get started, and
very powerful and flexible once you get going. Best of all, Django is
in Python, and Python is my most favorite language for hacking on
fonts. In fact, I've since come to the conclusion that Py-cairo is
probably the best possible platfrom in which to write the font-tester.
I'll be implemening Py-cairo on ghilbert.org within the next few weeks
to do rendering of math to PNG's.

In short, if OFLib were based on Django, I'd be willing to do
development work on it.

As I said before, we need to focus on basics before we get to the
fancier stuff. Thumbnails can be statically generated (by the
submitter, or perhaps by OFLib gnomes) until dynamic displays are
available. The default display in myfonts is a 395x55 png with the
string "AaBbCcDdEeFfGgHh" set in 48 point type. When you get to the
font, you see a 531x243 "mini-map" (approx 24pt), and clicking on that
gets you a 535x2k (approx, 46pt) showing of the full character
complement. There are, of course, plenty of other options, but for
right now this is what I'd recommend as the minimum.

I must also echo Gustavo's concerns about quality. Not to be too
elitist, but throwing up one-to-five rating scales with public voting
is not the same as ensuring quality. Ok, so I am an elitist. Sue me.

A large part of the problem is that the vast majority of free fonts
are in the "novelty" category. The number of original free fonts
suitable for text that have a complete complement in all four variants
is _very_ small (another fact which I hope will change significantly).
Unless there is more guidance or structure, a voting system will by
necessity rank most fonts on the scale of novelty fonts, not text
fonts.

I would hope that the front page of OFLib would not be dominated by
these novelty fonts. Otherwise, in the worst case, it will be little
more than "2,002 honest fonts" with a bit of free software rhetoric
added. (Must see link:
http://www.somethingawful.com/fake/fontsite/index.htm)

I can think of two approaches to the problem: first, have the OFLib
cabal select "featured" fonts which are actually good and useful,
rather than using a script to sort by user rankings. Second,
categorize fonts so that novelty fonts can exist in their own ghetto.
[I'm leaning toward this second alternative, as the categories can
also be a useful navigational aid, in addition to being more
democratic].

The main thing is to communicate the message that we really care about
quality, more so than just having blind faith in the ability of
Slashdot-style rankings to sift the wheat from the chaff (just look
how well that works for /. :). In my role as designer who's striving
for high quality, that's the most important feature. Even a site built
with static HTML and images would suit my needs better than what's up
now, as long as it showed the fonts and emphasized quality.

I really want to see something like OFLib happen. Unfortunately, I
don't have a huge amount of time or energy I can contribute right now.
With luck, enough other people do that the site will fly, and then
I'll be a very happy participant.

Take care,

Raph

P.S. For your enjoyment, I've attached a sample of Cecco, a font I'm
working on. I'm not sure exactly where I'm going with it, but with the
right encouragement, I can see it becoming a full family suitable for
book work.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cecco.png
Type: image/png
Size: 10751 bytes
Desc: not available
Url : http://lists.freedesktop.org/archives/openfontlibrary/attachments/20070127/bafe8607/cecco.png