|
|
Log in / Subscribe / Register

Gentium: An award-winning font joins the free software world (NewsForge)

[Posted January 26, 2006 by cook]

NewsForge looks at the Gentium font, which is available under the SIL Open Font Licence. "Gentium is something new in fonts. Its design is a mixture of the practical and aesthetically pleasing. It supports the diacritical marks needed to render a wide range of Latin and Greek characters, yet it is also designed for readability, compactness, and visual appeal. What is really unusual is that its designer, Victor Gaultney, has released it under a free licence and is developing it as a free and open source project."
to post comments

yet another license

Posted Jan 26, 2006 23:46 UTC (Thu) by stevenj (guest, #421) [Link] (7 responses)

Just what we need, complete with poorly thought-out provisions like:
  1. Neither the Font Software nor any of its individual components, in Standard or Modified Versions, may be sold by itself.
  2. Standard or Modified Versions of the Font Software may be bundled, redistributed and sold with any software, provided that each copy contains the above copyright notice and this license. These can be included either as stand-alone text files, human-readable headers or in the appropriate machine-readable metadata fields within text or binary files as long as those fields can be easily viewed by the user.
  3. ...

So you can't sell the font by itself, but it's fine as long as you bundle a two-line "hello world" script. Compounding the confusion, their FAQ claims, "The OFL is compatible with most FLOSS (Free/Libre and Open Source Software) licenses."

Sigh...will people ever learn?

yet another license

Posted Jan 27, 2006 1:50 UTC (Fri) by joedrew (guest, #828) [Link]

It *is* compatible with, e.g., the DFSG, though. This sounds like a font license very similar to the
Bitstream Vera font license, which has requirements so easy to fulfill that it's considered totally
Free.

yet another license

Posted Jan 27, 2006 9:56 UTC (Fri) by yosch (guest, #4675) [Link] (5 responses)

Well... ever wondered why there are so few high-quality modifiable and redistributable fonts available? One of the big reasons is that most font designers really hate finding their work sold through cheap font cd collections.

Your friendly comment focuses on one small restriction designed to alleviate that problem while still allowing wide packaging and distribution, but you seem to be ignoring all that the license will now enable for worldwide collaborative typography. This area is not as simple as you think.

Please read the FAQ a bit more closely: http://scripts.sil.org/OFL-FAQ_web.
Especially entries 1.4 to 1.8.

The issues of embedding and derivative works are much more important and many think that the OFL is providing a much nicer legal framework for that.

This is not yet another license but a license that was much needed to bring better non-restricted font support to the free desktop. The Bitstream Vera licensing model certainly helped but more needed to be done in that crucial area, there were still *many* grey areas.

The OFL is a well-researched licensing model which has been reviewed, commented and refined with the FLOSS community. See http://scripts.sil.org/OFL_feedback

We think it caters to the needs of font designers while complying with the free software values and methodologies. See this for more details: http://scripts.sil.org/OFL#6c6c80bc

Note also how the license is neutral and not project-specific like most other font licenses out there so that it can be easily reused by any designer or foundry. How's that for fighting license proliferation?

The OFL has received good support from various key members of the commmunity:

The FSF actually thinks the license is a free software license for fonts:
http://www.fsf.org/licensing/licenses/

George Williams, the author and maintainer of Fontforge recommends the license:
http://fontforge.sourceforge.net/faq.html#license

Raph Levien contemplates using the OFL for his Inconsolata project:
http://www.levien.com/type/myfonts/inconsolata.html

Stepan Roh from the Dejavu project (the main Vera derivative) likes the license:
http://openlists.sil.org/archives/ofl-discuss/2005-Octobe...

Various people from GNOME, OpenOffice.org and Debian also see this license as useful.

Check out the OFL-discuss archives for more:
http://openlists.sil.org/archives/ofl-discuss/

So, it is a fairly solid license, various designers are going to use it for upcoming font releases and we are in contact with the maintainers of the major distros for inclusion in main but, if you have any good suggestions for future refinements, free free to post them on the ofl-discuss list.

--
Nicolas Spalinger, SIL Non-Roman Script Initiative volunteer
http://scripts.sil.org

yet another license

Posted Jan 27, 2006 17:31 UTC (Fri) by stevenj (guest, #421) [Link] (4 responses)

One of the big reasons is that most font designers really hate finding their work sold through cheap font cd collections.

And the SIL license does absolutely nothing to prevent that. As I pointed out, you can obey the letter of the license by bundling any trivial program with your font cd. (Indeed, if the license succeeded in prohibiting cheap font cds, it wouldn't qualify as FLOSS.)

I have no doubt that the intentions of the license authors were good, but it was obviously written without the advice of people experienced in FLOSS licensing, because it falls into errors that have been made time and time again by well-meaning people who thought they could dash off a new license. (Yes, I know the FAQ claims otherwise, but the proof is in the pudding.) As another example, it clearly tries to be a copyleft (by requiring that derived works be distributed under the same license), but copyleft licenses are notoriously subtle to get right (to avoid loopholes and unintended consequences) and I don't think this one succeeds.

There are lots of places that one can go for advice on this — the FSF, OSI, and even the debian-legal mailing list. I strongly urge that the SIL authors to take advantage of the available expertise in order to revise or replace the license.

did you actually *read* the answer?

Posted Jan 27, 2006 20:29 UTC (Fri) by gvy (guest, #11981) [Link] (3 responses)

especially bottom half

yes, I read it, and the "support" for OFL is not nearly as strong as is claimed

Posted Jan 27, 2006 21:19 UTC (Fri) by stevenj (guest, #421) [Link] (2 responses)

Just because the FSF says it is a free-software license (something I do not contest) doesn't mean it is a well-crafted license. The FSF considers lots of licenses "free" that they don't particularly recommend.

I looked on the OFL-discuss list, and I don't see any representative of a group experienced with FLOSS licensing saying that the OFL is particularly a good idea. I do see Jim Gettys posting expressing concern about "license proliferation" and suggesting that they should get input from Eben Moglen of the FSF. I see Gervase Markham of Mozilla.org (who is also not a lawyer) also making some substantial criticisms of the license and also expressing concern about license proliferation.

The FontForge link is someone pointing it out as an example of a license designed for fonts, with no particular analysis of whether it is a well-crafted license or not. Besides, random developers are notoriously bad at evaluating licenses...that's one of the reasons why there are so many poorly-crafted ones floating around (cough, Artistic license, cough).

You really need to get competent legal counsel, experienced in free software, involved in such a thing. The FSF, OSI, and CreativeCommons are good resources; Groklaw might also be helpful. I can see the need for a good free font license, but this is not a task for amateurs.

yes, I read it, and the "support" for OFL is not nearly as strong as is claimed

Posted Jan 28, 2006 16:50 UTC (Sat) by yosch (guest, #4675) [Link] (1 responses)

Well, thanks again for your respectful insights.

We actually submitted the license to many experts in the community including Lawrence Lessig, Eben Moglen and Bruce Perens. We are in touch with Debian and Ubuntu people and are discussing the license with debian-legal. The discussion with the community is by no means stopping.

No license is ever perfect. The Open Font License 1.0 is designed to enable font designers to work together and improve the state of non-restricted typography and I think we're getting there.

Of course, there will always be those who prefer criticizing...
If everything is so obvious to you, then we'd love to see your suggestions and advice in helping improve future versions of the license.

--
Nicolas Spalinger, SIL Non-Roman Script Initiative volunteer
http://scripts.sil.org

yes, I read it, and the "support" for OFL is not nearly as strong as is claimed

Posted Jan 29, 2006 13:36 UTC (Sun) by rehdon (guest, #29998) [Link]

I, for one, am grateful for your efforts in providing a high quality font to open source user.

rehdon

Gentium: An award-winning font joins the free software world (NewsForge)

Posted Jan 27, 2006 2:14 UTC (Fri) by Richard_J_Neill (subscriber, #23093) [Link] (10 responses)

Is it properly hinted? So far, the only fonts which are ideally readable under Linux are the MS corefonts. The correct solution to displaying fonts is *not* antialiasing: see here.

Gentium: An award-winning font joins the free software world (NewsForge)

Posted Jan 27, 2006 7:02 UTC (Fri) by Interneci (guest, #34645) [Link] (1 responses)

> So far, the only fonts which are ideally readable under Linux
> are the MS corefonts.

That's not true. The free Bitstream Vera fonts are at least just as good,
in my eyes even better.

Regarding anti-aliasing, my impression is that AA especially helps with
well-hinted fonts. I use antialiased Vera Sans as my default font on a
TFT, and it's *really* crisp and sharp. Horizontal und vertical lines
exactly match the pixel lines, so AA is only used on diagonal and curved
parts of the types and desn't make the letters "blurry".

Even *very* tiny font sizes can be read well with AA, while they become
completely unreadable without. (The same holds for the fonts distributed
by MS.)

(slightly off-topic) DIsappointment with Bitstream Vera

Posted Jan 29, 2006 5:00 UTC (Sun) by pr1268 (guest, #24648) [Link]

I notice how people and distributions seem to praise Bitstream Vera fonts - Every distro I've used from Fedora to SuSE to RH to my current Slackware all have the BSV fonts. Even OpenOffice's default serif font is the BSV font.

What I don't get is why BSV (serif) doesn't have an italics-capable font, at least not in OpenOffice. Either (1) BSV came to all the distros without an italicized font or (2) I'm living under a rock when it comes to configuring and setting up my X11 Fonts configuration to make BSV fonts italicized.

I will be writing a graduate thesis pretty soon and foresee the need to use italics quite often in my paper. I'll be reconfiguing OOo to use Utopia as the default serif font because it actually goes italic when I hit <Ctrl-i> on selected text. If this can be fixed for BSV serif and someone knows how, I'd be forever grateful for the knowledge....

Hopefully, license debate notwithstanding, the Gentium font won't have this disability. Just my $0.02.

Gentium: An award-winning font joins the free software world (NewsForge)

Posted Jan 27, 2006 8:36 UTC (Fri) by cga2001 (guest, #35510) [Link] (4 responses)

Thanks for stating the obvious - and providing a perfect example.

At best, font antialising is a small a quick'n dirty fix to disguise poorly-hinted fonts and try to make low quality displays tolerable. And at worst I suspect it is a clever ploy designed by the optician lobby to broaden their market and sell their wares to individuals who would otherwise have maintained perfect vision.

In other words it is a rip-off..

What's next..?? anti-aliased books..??

Gentium: An award-winning font joins the free software world (NewsForge)

Posted Jan 27, 2006 17:58 UTC (Fri) by Ross (guest, #4065) [Link] (3 responses)

I really don't understand the statement that anti-aliasing is bad for fonts.

Yes, using it makes bad fonts look blurry, because they weren't designed for pixel-based display (at least not at the used resolution), but those fonts would look bad anyway, with parts being two thick, too thin, and gaps being filled in. Anti-aliasing is not inherently bad, and it is required to maximize the utilization of the screen. If all output were binary, even the best font would still be a little blocky except for straight lines which were completely aligned with the pixel grid or at 45 degrees.

I also don't understand the suggestions of specific point sizes to match up the output pixel-perfect. Screens don't all have the same dpi... so 10pt on one screen may have a different number of pixels than on another. A good font will look nice at a number of different pixel sizes and render at different resolutions while preserving its distinctive look.

Gentium: An award-winning font joins the free software world (NewsForge)

Posted Jan 28, 2006 6:50 UTC (Sat) by cga2001 (guest, #35510) [Link] (2 responses)

All I'm saying is that w/o AA we might stand a better chance of seeing affordable 300dpi+ displays and quality free fonts in the near future.. and as a result, preserve our eye vision. But due to the existence of this cheap - in every sense - AA hack.. neither hardware manufacturers nor font designers feel much of an urge to do much to improve the quality of text rendering on computer screens. With adequate AA.. so-so fonts like the bitstream vera.. nimbus etc. series globally look on a par with truly well-designed screen fonts (verdana.. tahoma..). So why bother..?

I am obvioulsy no font-rendering expert.. and I won't mind if someone more knowledgeable proves me wrong. In that painfully underdocumented area it would be nice to learn a little something..!

But I do believe that w/ the perverse effects I'm hinting at above, AA has done us more harm than good.. The mind boggles at its popularity within the OSS community.

As to the point/pixel ratio aspects I guess you were replying to somebody else's post..?

Gentium: An award-winning font joins the free software world (NewsForge)

Posted Jan 28, 2006 7:04 UTC (Sat) by Ross (guest, #4065) [Link]

The problem is that display resolutions haven't been going up at the same rate they used to. I'm still stuck using 1024x768 at work. Given that restriction AA is a very nice thing. I totally agree more resolution would be better... but it's not here today, at least for most people. Now correctly done, AA is not harmful and actually improves the display even on high-resolution screens -- in fact it works much better when the pixel widths are less than half of the width of the smallest lines in the font.

My point about font sizes was in reference to the article linked by Richard_J_Neill which talked about why AA was not the right solution.

Gentium: An award-winning font joins the free software world (NewsForge)

Posted Jan 29, 2006 1:07 UTC (Sun) by bk (guest, #25617) [Link]

With adequate AA.. so-so fonts like the bitstream vera.. nimbus etc. series globally look on a par with truly well-designed screen fonts (verdana.. tahoma..).

Then AA seems like a very efficient, useful "hack". Not only does it lower the hardware requirements, but saves significant human effort at the expense of a slight increase in software complexity.

If what you're saying is true, the "harm" is entirely psychological. If you are unable to visually distinguish "well designed" fonts from anti-aliased "so-so" fonts then your preference for the former is merely ideological bias.

Gentium: An award-winning font joins the free software world (NewsForge)

Posted Jan 27, 2006 8:43 UTC (Fri) by cga2001 (guest, #35510) [Link] (2 responses)

Naturally my initial post was in reply to Richard_J_Neill. I did not realize the posts were not threaded.

Threaded discussion

Posted Jan 27, 2006 13:14 UTC (Fri) by man_ls (guest, #15091) [Link] (1 responses)

Posts are threaded. Just choose a "nested" view.

Threaded discussion

Posted Jan 28, 2006 6:56 UTC (Sat) by cga2001 (guest, #35510) [Link]

Can't find it.. but I did notice that the boxes around each post are somewhat indented. Thank you.


Copyright © 2006, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds