I love the idea... unfortunately something is seriously wrong with their website. They're selling fonts you can use on other websites. I'd expect that they know a lot about fonts - yet, when you look at their own source:
font-family:arial,sans-serif;
Why don't they buy at least one really high quality font from (for example) Linotype? Not even for reselling, but for use on their own website - to show the difference it can make... Why do they try to sell fonts for websites, but don't try to blow my mind by using one of ITCs, or Vialog, or something else for their own main text.
I meant the normal body text to be honest. Instead of showing off with one frame, why not typeset the whole page in one of the fonts? I couldn't find anything both toned down and "different" in there, but this might come close: http://typekit.com/fonts/392 Give it more spacing and you could use it for "normal" text (or even http://typekit.com/fonts/129 if they want to stay with web2.0-ish sans)
Probably because on Windows most of the fonts they have available would look blurry at less than 18px tall. This is due to the way font-hinting and anti-aliasing works (at least in Vista and XP).
i'm not a designer, so i'm obviously out of the loop. still curious about what's going on, though. thus: how much difference do fonts make? any a/b tests? any qualitative urls you can recommend that will blow my mind? keep in mind i use a mac...windows and buttons and fonts already look pretty.
Jeff, Greg and Ryan (3/4s of Typekit) spent a year redesigning Analytics after Google bought Measure Map.
Jeff once said that -- using the Analytics APIs they had access to -- Greg and Ryan basically prototyped a fully working version of Analytics as we know it today in a/b a month. It then took the remaining 11 months of them working with Goog engineers, PMs, etc. to get the thing out the door.
They're an amazing team and I'm looking forward to following along w/ the success of Typekit.
I really like the idea, but there are some implementation problems I haven't yet seen addressed. I'm very much in the target market (web designer / developer), and I'd like very much for this to work, but I just don't see it.
Here's my issues / questions, would be interested in hearing why I'm wrong:
- The fonts available on typekit aren't the ones I want to use. Without support for the Linotype / ITC / Adobe fonts, I'm left choosing a font from what you do offer, and at that point I might as well just choose a free one and host it myself. Any plans to get the big-name foundries on board?
FWIW, the fonts I most often need this for are the various flavors of Helvetica, Gotham, Trajan Pro, and Myriad Pro.
- The main benefit to cufon/sIFR/etc is that it works with all browsers, specifically those that don't support @font-face. This doesn't solve that problem. I know typekit falls back to standard fonts, but I can do that.
- I usually avoid fancy fonts in all but headings and other large elements. If the text is static its not a big deal to save the text in Photoshop and use text-indent to replace the plaintext. Dynamic text, like blog post titles, is a pain in the ass, but cufon / sIFR handles that case well enough.
- Its one more server and one more fee added to the equation. Furthermore, I don't have enough confidence in this business model to trust the service to remain around forever.
1) sIFR, Cufon, image-replacement, etc. each have their own problems. Hard to setup, can't select the text, tedious to maintain, requires opening Flash CS4, etc.
2) @font-face is the future, but most paid font foundries refuse to support it b/c it requires the font file to be linked to on the server. So designers can legally use free fonts with @font-face but not fonts they've purchsed.
So, Typekit is an attempt to keep the font foundries happy by hosting the font file and preventing unauthorized downloading, while keeping an eye on ease-of-use and the future of the web by doing most of the @font-face heavy-lifting.
It supports text selection, requires zero thought once it's set up, and works in every major browser. The only major downside is that you need to convert each font to typeface.js' JSON font format, which adds a bit of heft.
At the same time of course, I definitely support what Typekit is trying to do, since most font foundries won't be happy with (m)any other solutions. Too bad this discounted pricing won't last!
Forgive me if I'm missing something obvious, but how does it prevent unauthorized downloading? It looks like the font gets base64 encoded and put right into a CSS file?
It's a pity that to use the fonts I would like to on my single website, I'd have to sign up for the portfolio plan, which is a reasonable $49.99 for the first year, but jumping up to $17 a month thereafter.
It's too much for me to justify, when the font could be purchased for $40, with a license that allowed the use of sIFR.
Just dropping in to note Kernest (http://kernest.com/), which uses @font-face as well, but without the Javascript that makes Typekit somewhat undesirable.
Then I used it in my site redesign, which conveniently used a font Kernest was already using.
Kernest's weaknesses on the site are offset by its strengths in implementation and ease of use. Typekit, however, uses javascript for DRM when it isn't really necessary. Kernest uses a server-side whitelist solution which is much cleaner to load.
So there's my pitch for Kernest. Typekit seemed cool at first, but the cost struck me as too high and I'd rather use javascript for functionality, not design.
I made a point on their GetSatisfaction feedback page, and will state again, if they can somehow lay down the cash for one or two of the most-used designer fonts, say, Helvetica Neue, or Gotham, or Garamond, then designers and web designers would flock to this service. Additionally, if they were to add a few of the popular-to-designer bit type faces, such as any of the FFF fonts, then again, designers would eat that up. Designers just have their standard kit of faces they turn to.
They need to concentrate on making this work in closed environments where the developer/designer is certain that a compatible browser is being used; it will be many years before someone can use this for a practical application on the "open web."
Basically, iPhone-specific websites could make them all of their money, if they could make it work correctly with the browser (the front page looked fine, but the font sample page seemed to be a bit off). In fact, I would use and pay for this myself if it were marketed as an enhanced typography library for iPhone web apps.
There are several outfits trying to do this lately, but I just can't see these services gaining much traction.
They seem to exist to provide for the whims of font foundries who want to eat their cake and have it, in a way that no-one else gets. The stock images and icons on web sites can all be downloaded and used on other sites illegally as well, yet in practice this hasn't killed either the stock suppliers or the web in general!
Meanwhile, why would any professional web design firm recommend to their clients that they commit to writing a blank cheque, indefinitely, after the design contract is finished, (edit: and relying on a service outside of their control!), just to keep their site up and running and looking correct?
It's not as if there aren't alternatives for the cases where use of non-web-safe fonts makes a difference: we've been using custom images, even dynamic image replacement techniques, since forever. These don't really scale from occasional uses like headings and pull quotes to the full volume of body text, but for typical body text sizes using custom fonts is unlikely to gain any useful improvement anyway; even most pro-grade don't have the kind of hinting that the Microsoft and Apple standard fonts have.
So I have to wonder, why should the world bend to the font foundries' obsolete business models, when there are numerous alternative technologies available that get the same wow factor where it counts, there are many free fonts available (including a few genuinely of professional quality these days), and there just isn't much of a compelling selling point from the client's point of view to signing up for fonts-as-a-service?
Of all the copyright-holding businesses affected by the Internet, it seems to me the foundries are among the coolest. Their work is rampantly pirated. Only a tiny fraction of the people using SIFR and image replacement have actually licensed their type for embedding on the web. They could be hassling hundreds of sites over this, and they'd be well within their rights to do so.
What I get off this comment is that you think it's "obsolete" to suggest people should pay for typefaces. But typefaces, even moreso than music and movies, take years and years of painstaking work to create, and the truly useful faces are created by a tiny group of people who've dedicated their lives to the craft.
> What I get off this comment is that you think it's "obsolete" to suggest people should pay for typefaces.
Then, respectfully, you should reread it and then reply to what I actually wrote.
The point is not that I think fonts should be free. I have personally paid the equivalent of hundreds of dollars for pro-grade fonts, because I like my work to look good.
The point is that it's unrealistic for the foundries to expect that instead of paying for a font and then being able to use it, as we've managed in the civilised world for decades, we will now all move to a fonts-as-a-service model, where we will pay ongoing fees for DRM-crippled products served via third party systems out of our control.
There simply isn't any upside to that for the consumer, other than perhaps cheaper experiments with different fonts, which is of very limited value anyway if you're working with a professional agency with access to a large collection of fonts. The downsides, on the other hand, are potentially severe.
Image replacement is a huge pain in the ass. SIFR will get less compatible as time goes on, not more (already, I lose SIFR fonts because of ClickToFlash, and multiple bigcorps are moving to ban flash on corp desktops).
Convenience would be paying for a font and then using it on my web site, just the way I do with photos, icons, and other such resources.
Convenience is not jumping through hoops to integrate with an external service of unknown reliability, understanding non-trivial legal agreements with an industry infamous for trying to screw its legitimate customers, and paying more in a single year for font-as-a-service than I pay for permanent use of a font today just so I can legally use the font on a web site in plain text form instead of via images.
Incidentally, I notice that have mentioned elsewhere in this discussion that image replacement technologies are likely to violate the licensing terms of fonts. As far as I'm aware, there haven't been any significant test cases yet in this area in any major western jurisdictions, and the legalities are far from clear, because the font foundries can only protect their rights up to the extent that copyright law grants them. Copyright law incorporates concepts such as fair use (or your local equivalent). Moreover, in some jurisdictions, the design of a font is not subject to copyright at all, only the specific description (the font files being treated as software for this purpose). Even if the font foundries would like to limit your right to buy a font, put it on a server, and have the server generate renderings of specific text in that font, it's not entirely clear what their legal basis for doing so might be in unfavourable jurisdictions, including the US.
It may not be the case that the industry as a whole is locking out Flash on their corp desktops, but it definitely is the case that there are security-conscious F-500's who're getting Flash out of their standard desktop build, so that their internal machines can't be infected simply by browsing the web.
I can't give examples, for obvious reasons, but you can probably take my word for this.
It's not really the same to use images... It's not uncommon to see websites with a light feel, where main pages contain actual paragraphs of text at size 16-18pt - you can actually make a really good impression by changing the font to something elegant in that case. You don't need any sacrifices for usability, SEO or content management when you design those pages. The text is there for everyone and owner of the website can edit it without going through image exports, etc.
Of course it won't look as good as it would in a pdf for a couple more years, but it's a good start anyways.
Can you give any examples of pages where you think the use of an alternative font has improved the display of body text? I'd be happy to find cases where it makes a real difference, but it has been my experience that there just isn't enough variation at typical screen resolutions between, say, one mainstream serif font and another, that the choice of font is the deciding factor in terms of impact and legibility.
Good appearance on-screen seems to be governed much more by basic design characteristics such as x-height, and by the quality of the hinting, and frankly even the serious font foundries suck at those things in comparison to the Microsoft C-fonts that were designed for anti-aliasing, the old stand-bys of Georgia and Verdana that were designed with screen use in mind, etc.
Of pages that have improved by using alternative font - unfortunately I can't. If anyone is doing that, they certainly don't advertise it, so I can't give you any specific examples.
I based my opinion almost entirely on my experience, that even a slight change of font can make a big difference in someone's perception. I've seen people being positively surprised when I use a "different" font for normal everyday text (1-2 pages). Nothing crazy - just a bit more stylish. I also got an impression that using nice fonts (with better visible horizontal lines, etc.) got me an additional point or two at the university (no scientific test of course - I've seen very similar in content, but typographically "nicer" texts getting higher marks - just my opinion)
> I based my opinion almost entirely on my experience, that even a slight change of font can make a big difference in someone's perception.
On paper, I wouldn't necessarily disagree with you.
On screen, however, I doubt the odd pixel makes any significant difference. I'm prepared to be convinced otherwise, but I have yet to see any of these professional font foundries prove me wrong.
I do know that I have bought and paid for several pro-grade fonts, which I happily use in print, but which I wouldn't use on a web site even if I legally could. The on-screen rendering is nowhere close to as good as the fonts designed for screen use that I mentioned earlier.
It depends on how fast Typekit's server responds. I'm sure it is extra slow today. I did some unofficial benchmarks on the service while it was still in beta:
Chrome currently does not support @font-face by default so this does not work on chrome for me unless i run it with a command line option to enable it.
i love the idea, but there's no way i'm going to pay for the service until it supports IE, at least (i can accept that it doesn't work on chrome and the like, for now).
i hate IE, but its still a huge chunk of the browser segment.
It does support IE - IE has supported @font-face for the EOT format since IE5.5, it was originally a proprietary IE property that has since been standardised.
