Font Squirrel has a page where you can submit an OTF/TTF file and get back a webfont kit. I think it's based around Unix/Linux programs, so if you have a Linux box or an OS-X Mac, you can probably assemble your own toolchain.

But Arno is right. A webfont is any font when it's used on the Web. As long as it looks good enough on the screen, no-one'll give you grief.

There are dedicated web font formats around: EOT (Internet Explorer), WOFF (Firefox 3.6 and later), Cufon, sIFR, typefaceJS... *swirly eyes*

They have a workflow that will convert and optimize your fonts for the use on websites.

Does this mean, that myfonts would split Bitstream Charter Pro into a regular and a small cap part for example? Otherwise I can access only the half of the glyphs and therefore it is not really usable as webfont, especially because the license does not allow, that I split it into parts by myself. Sorry, but if I pay for a webfont, I really expect, that I can access all glyphs. So I assume, it actually is a half-baked thing.

Yes. Maybe I don’t expect anymore, that I get, for which I pay, when I am adult. In case of Bitstream Charter Pro you could not access the small caps, if you use it as webfont. But you pay for a webfont. Correct me, if I am wrong, but Myfonts seems to simply put the regular fonts into a woff container. That’s all, isn’t it? And the answer to the silly question, why they do that, is, that it is a very easy way to make money.

'Arno Enslin' -- but if I pay for a webfont, I really expect, that I can access all glyphs

That's not a font issue but a browser issue, i.e. missing support of OT layout features. You're yelling at the wrong folks ...

Does MyFonts have a web font preview service? What's the return policy?

Cheers, Si

Thanks everyone for your helpful comments. Cheers for the holidays.

Too late. They've already broken out into the Hinters' Chorus from Der Bleifutz.

@Arno - If you want a browser supporting OT features, the only choice I know of is the Firefox 4 beta.

@arno
"But I would prefer, if I would be allowed, to split a font into parts and create CSS classes as replacement for the layout features"

FWIW - I've thought about it, and this would be my preference, too. This way, I could get some things happening across the board in browsers today, not ten years from now.

And whatever it means, I second that "die Klappe halten" thing you said. ;)

@ Karsten

Georgia would be in the second place of my font stack. And with regard to backwards compatibility your argument is not valid, because in case of the layout features there is a solution (font splitting), while there is no solution in case of font embedding. OT layout rules are even not compatible to the actual browser generation. In other words: If I would use those CSS rules, my website would be compatible to the most used browsers in a few years without backwards compatibility. If I split the fonts, my website is compatible to the actual browser generation and in five years it will be backwards compatible. By the way you seem to forget, that web pages can be printed. And I have a few texts, that are intended to become printed. So I am not interested in fake small caps. And except from that I don’t want to use Charter, because it looks nicer, but because it is more legible than Georgia in my opinion. Charter Pro really looks excellent. I simply will ask Bitstream, if they allow the splitting. And if they don’t allow it, they simply don’t get my money. I assume, that the foundries, which sell web fonts via Myfonts, want money. The “I want, I want, I want” is not on my side only. I only want a reward for my money. The reward is, that I am allowed to invest my time and my energy in splitting the font by myself. I even don’t expect, that they do that for me. If they want to have just a restrictive EULAA. Bad luck, but not only for me.

Cool Hand, baby! I was just reading your article and your conclusions and even some facts are wrong.
When will you stop with the "if only" agenda and take the world as it is?
You write like DirectWrite is a magic bullet that's going to end the need for hinting TT fonts. It don't, it won't. You make a reference to fonts with "just a little hinting" - isn't that sort of like "just a little pregnant". (And with the new Swiss laws regarding consensual incest, this can get pretty complicated. Ask Arno, he's the one who told me about it.)
And BTW - Cleartype is on by default in IE7 and 8. And take it from an old network engineer/desktop support guy: users never, ever change their fracking defaults.
Also, you write like desktop browsers are the be-all, end-all. They're not. iPad is big. iPhone is big. Mobile browsing in general is big, big, big, and growing by leaps and bounds.
The Chrome OS comes out this year. I expect that Ubuntu will make some strides, too.
You did get one thing right, though: the problem is as much about font weight as hinting.

And why are the fonts on Webtype so boring?

Yes, yes, I want, I want, I want. And guess what? I'm gettin' it.

Merry Xmas!

After the costs are amortized, it's bubkis, really.

Of course, Finagle may strike such that the size you want on screen is the very size where they look bloody awful...

Zoom, Zoom, Zoom.

Hi Ralf.
Just for the analytical exercise:

How many customers are there, that can split a modern webfont with OT features and preserve TrueType hinting, keep OT features and kerning working? 1 out of 50? And the other 49 will mess it up, ending up with bad looking or malfunctioning fonts and they or the viewers of their websites will blame the type foundry. So that's not a route the foundries will take.

Typekit (among others) - and by their own admission, really - delivers all the 'bad looking and malfunctioning fonts' you could possibly want now, today. So what?
You talk of "the foundries" as a monolithic entity. Has there been a grand merger I'm unaware of? Some foundries will do whatever they think a customer is willing to pay for.

If features such as small caps should work in a webfont, they should 1) be already in the font as OT feature (wont work now) or 2) be provided by the foundry as separate font (as FSI does it) or 3) be generated on the fly by the vendor's system (as Typotheque's webfont service does it)

Well, okay. All Arno is asking for is option number 2, so where's the disagreement?

You purchase a license to use a font. Foundries don't sell toolbox of glyphs and codes so you can build whatever you want from it.
Some will.

It’s the fear, that the user manipulates the typeface. I can partly understand that fear. But I don’t want to change the typeface, but the font. And I think, a solution could be, that the license allows to change anything in the font, if the manipulator keeps the original font names, but adds the suffix mod or modified. And the user should add the information to the copyright sections of the font, that the font was manipulated by the licensee with the consent of the foundry, but that it is still the property of the foundry. And additionally this information should be stored in the CSS and, for the case, that the font is not base64-encoded, in an info-file, that is stored in the same directory as the fonts.

There is another thing, which I would like to change in case of some fonts (not in Charter) and that are the metrics. If I enlarge the letter-spacing with CSS by one pixel the letter-spacing is already too wide in most cases. But if I enlarge it by fractions of one pixel, it becomes uneven. This problem could be solved, if the user would be allowed to enlarge the letter-spacing directly in the font.

With regard to the design of the glyphs, I would be more careful, but I could imagine, that it would be useful to provide an extra CSS especially for print, in which a slightly enbolded version of Charter is embedded. I think, the only point, that is really important is, that, if the font is pirated from the website of the licensee by a potential other licensee, this potential licensee knows, that the font is not that, what he would get from Bitstream/Myfonts. Why? Because either the modified version of the font is better than the original. In this case the new licensee would be disapointed, if he licenses the original. Or the modified version is worse. In this case he may decide, not to license the font.

For the moment I only can say, that I have not seen one website, that was improved by a webfont for body text. Example: "http://www.thomasphinney.com/". What the hell has he done there? The website does not profit from the embedded font. It is just the other way around. Except from the fact, that the font is ugly, it is not well hinted.

By the way, there is a thing, that is really absurd with regard to system fonts. It is absurd, that licensees of Windows are not allowed to embed system fonts in their websites. It is absurd, because websites are even displayed in system fonts on the screens of the big majority of all visitors, if they are in the font stack, but not embedded. And it is absurd, because a license for a system font probably costs as much or more as the whole operating system. And it is unfair, because those restrictions will let many thousands of people walk right into the trap of charged cease-and-desist orders. When you decided to become a type designer, one motivation probably was, to make the world more beautiful. In case of the internet the world is still ugly. There are so many websites overloaded with ornaments. They look like diversionary manoeuvres from technical hurdles like low screen resolutions. The ornaments are needed, because the text cannot catch the attention. And in the web empty spaces blend much more than in a book. In case of books I never have the feeling, that they look the same, although objectively they often look more the same than websites. Now I have lost my train of thought. Maybe this: If you only provide display fonts, that work with regard to the quality, in which they are displayed on screen, you just provide new diversionary tactics.

Richard Fink:

I was just reading your article and your conclusions and even some facts are wrong.

You must have read another text. There was one factual error indeed but you didn't even notice it. Mr Berlow did. A few remarks:

You make a reference to fonts with "just a little hinting"

The notion "just a little hinting" refers to the kind of hinting found in webfonts by FSI and others. It is "just a little hinting" in comparison to the hinting found in Microsoft's core fonts. No need for sophistery.

And BTW - Cleartype is on by default in IE7 and 8.

Had you read my text you would have noticed that this is core part of my argument: IE 7-9 and FF 4 are doing it already. And it would be helpful if the makers of only three other browsers followed this approach with future Windows-versions of their browsers.

Also, you write like desktop browsers are the be-all, end-all. [...] Mobile browsing in general is big, big, big, and growing by leaps and bounds.

Don't you think that you are exaggerating more than just a bit? Also the question is not: desktop or mobile? but: what operating system, rasterizer, browser?
With or without your kind permission, I prefer to address one issue after the other, not all at once. My note was about Windows and browsers running on Windows. No OSX (except for a quick comparison). No Unix/Linux.

On the website Linux Online you can download the Windows Core fonts (“official distribution of these fonts for Linux” / package contains [naturally] Georgia) for free. There are even Windows Vista fonts contained in the archive, that you can download there (Calibri, Cambria, Candara, Consolas, Constantia, Corbel).

On the website of Ascender Fonts you can license the Georgia family (for example) for 110 Dollar.

Except from that, that Microsoft has spread out the Core fonts itself (Correct English? itself?). It would be absurd, if it would not be allowed to embed the Core fonts, if they are still available for free.

And with regard to the look of Charter Pro I seem to be wrong. My test platform was XP / Firefox 3.6. There it looks very good. But on most of the other combinations of OS/browser it does not. And I cannot understand, why someone can be in the opinion, that fonts look better on the MAC. So I don’t know anymore, if Charter Pro would be really a good investment. But I dislike Georgia. On XP/Firefox it looks too bold and I never felt, that it is well legible. I think, I would go with Verdana, but I have some own poems and short stories and technical tutorials. I would like to use the same font for both, but Verdana is inappropriate for the poems and the short stories. I am in search for a neutral serif font, that harmonies as much as possible with all of my texts. And it should look good on all platforms. Seems to be, that I have to wait another ten years.

@ Simon

Do you really need me for pointing you to people, that possibly violate Microsoft’s copyrights? Just be happy, that Linux users are interested in Microsoft products. I mean, all you had to do is searching with Bin… – with Google. With a simple search I find round about 5000 unprotected directories, in which Calibri is stored. Not to talk about forums with links to file hosters. The link is in my previous post (Linux Online / paragraph 4.3 / tar archive). Do I interpret you correctly, that I would be allowed to embedd the Windows Core fonts in my website without an extra license? Otherwise you would not have explicitly asked for a pointer to the Vista fonts, right? And if yes, wouldn’t you say, that it is daylight robbery, if Ascenderfonts sells webfont licenses for the core fonts? :-)

@ Simon

But now you have found it? The name of the archive is "webcore-fonts-3.0.tar.gz.tar". It is hosted at another place, but there is a direct link to the directory under paragraph 4.3. Come on, it is easier than finding Windows bugs.

The free Google Font Directory version has a tiny Mac/Win Roman character set as most if not all GFD fonts have. The version available from Ascender Corp has a character set covering Latin Extended Additional (including Vietnamese), Greek, Greek Extended, Coptic, Cyrillic, etc as their website clearly says.* And for this pretty huge character set they charge not "hundreds of dollars" but meager $30/font.

* Just google or consult the Unicode website for the exact character coverage of "Latin Extended Additional", "Greek Extended", etc. I am pretty sure that people who go to the Ascender website know what they are looking for and what technical terms like these mean.

Mmm...
http://maemo.org/downloads/product/Maemo5/ttf-droid/

http://webfonts.info/wiki/index.php?title=Fonts_available_for_@font-face...

The Droid, which you can download there, seems to be the Pro version. At least it contains 894 glyphs. Naturally there are many people, that are going to use it without an extra license from Ascender Fonts. The reality creates a fait accompli. I doubt, that Ascender Fonts had any chance with a lawsuit in Germany. In the end the masses impose consuetudinary law. Never look a gift horse in the mouth. Droid was a gift. And it would be too much expected, that the users check, if it is still is a gift. In case of Charter it may be slightly different, but I assume, that even in Case of Charter the foundry had bad chances in a German court, if a user embeds Charter Pro. The main difference is, that Charter never was officially available as free webfont. Think about, before you make a gift to the web community.

RF>...it does make things simple and honest beyond all criticism and doubt doesn't it?

There are billions of products in the world and I've never heard someone criticizing differences in prices or features, nor have heard people doubt that different features of products have different cost/benefit profiles.

RF>...I just don't want to hear anything resembling a sermon...

But here it comes anyway?

RF>If you're Microsoft...[insert sermon on IP enforcement].

Thank you, and keep a fire extinguisher or bucket of water handy.

@ Richard

So anyway, Arno. What do you need for your webfont? A Charter that looks good onscreen? Along with a small caps version to ride along?

I did not test all webfonts for body text, but Charter is extremely well legible on my system (XP with activated and tuned Cleartype). But when I saw it on Myfonts, it does not look good on Windows Vista and 7. I wonder, why. Maybe because I am using the Cleartype value 1.5 and Myfonts a bigger value. It has not necessarily to be Charter, but it shall be more legible than Georgia. My priority is the legibility, but not the specialty of my website with regard to the font. The user may not recognize the beauty of the font, that I use. If the typeface would be a bit more beautiful (I would not call Charter beautiful, but simply neutral), it would be better, but I assume, that beauty can be a hurdle for the rasterizer (if there are more details for example).

I want to use the same font family for the whole website and I have very different kinds of own texts, poems, short stories, blog articles. (I become more and more political and there are certain themes like population growth, prenatal diagnostics, religious illusions and so on, about which I want to write in my own blog.) And I have technical tutorials like the translation of David Lemon’s PostScript hinting tutorial to German (I spend round about fifty hours with translating and creating the images). I still have to ask David Lemon, if he allows, that I publish the translation on my website. Otherwise I will keep it private or write my own PS hinting tutorial. If I find the time, I am going to study Adam’s TrueType hinting tutorial (maybe in a year), translate it to German and publish it likewise, if he allows. Why? Because nobody else seems to do it and if I want to understand this complex technical stuff, I need a clean translation for myself. So I am doing it primary for myself, but it is too good for not publishing it. Naturally in both cases with the names of the authors. But I want to publish it on my website, because my personal contribution was big. I could write my own PS-hinting-tutorial, but David has a very good writing style and although you probably see, that my English is not good, when you read comments like this one, the translation is very successful in my opinion.

But I am still in the beginning. I had a website, but I gave up a few years ago, because I felt too much pressed because of the technical limits; no CSS3, no font-embedding, no hyphenation for example. Fortunately these limits have partly gone. But meanwhile I have forgotten much. So I will need a while, probably at least a few months, until it is online. I hope, I am not getting bogged down in details. I want to try it with Wordpress first, but with an own template. I have ordered a web package a few days ago. (I can have as much MySQL databases, as six Gigabyte webspace allow. And the package is not expensive.)

On the other side I want not spread out my private data. I cannot exclude, that it ends up with a private website. I love and hate the word wide web and computers in equal measure.

Regarding the Microsoft fonts (Calibri, Cambria, etc.) these are available for license and download from many different font retailers: AscenderFonts.com, Fonts.com, MyFonts.com and FontShop.com.

Like many other commercial type foundries who participate on Typophile, when we find unauthorized postings of these font files we act promptly to have them removed.

Bill