On dberlow's PERM table for web fonts

abattis's picture

Many foundries have expressed support for Tal Leming's web font proposal on Twitter in the last few days, and so have some influential web pundits like Gruber. Its an impressive list, to be sure, and may become the web font format for all browsers in the future.

Meanwhile, Typotheque and TypeKit (EDIT: and Kernest) are doing a good job of bringing web fonts to 99% of browsers today.

But little has been said about David Berlow's PERM table proposal, published Wednesday - www.fontbureau.com/otpermtable/ - which
he introduces on the FB blog.

So I thought I'd start a thread here.

Who supports the PERM table? :-)

blank's picture

Meanwhile, Typotheque and TypeKit are doing a good job of bringing web fonts to 99% of browsers today.

TypeKit is doing a good job blowing smoke up the ass of 99% of browsers. Let me know when they have a working demo viewable by the public.

Jens Kutilek's picture

Meanwhile, Typotheque and TypeKit are doing a good job of bringing web fonts to 99% of browsers today.

Unless I missed some launch, you seem to have a strange concept of "today". ;)

Besides, the Typotheque solution is nothing new, a similar proof of concept implementation has been presented by Ralf Herrmann nearly two years ago.

aluminum's picture

"TypeKit is doing a good job blowing smoke up the ass of 99% of browsers. Let me know when they have a working demo viewable by the public."

Not TypeKit, but Kernest launched an early version of the site yesterday:


I'm assuming TypeKit is using a similar (if not identical) technical process for their eventual product.

abattis's picture

you seem to have a strange concept of “today” Today isn't literal, but Bilak and Veen are a few weeks from launch, and anything like Tal's concrete proposal has got at least a year before its implemented.

dberlow's picture

Well, I obviously support perm. It is the first 21st-century meta-data subfile for any media type and whether generally accepted or not, it is what users and developers really want — teaching not busting, updatable, upgradeable, and likably linkable, machine/human-readable, user-interface-able at every level from founder to end user. Please support it, or Harry Parker will shoot my dog Mutton.

>Many foundries have expressed support for Tal Leming’s web font proposal on Twitter

I wish the list my best. Developing web fonts is even more informative then reading, say, 1,000 entries such as this, maybe a million. Or... what would founder X on that list be charging for such a Tal font order? Tal what are you going to charge for fonts in your format to; a previous licensee for print, a new licensee for print, and new licensee for both? We have business and development plans all the way out to 'delivery', with PERM, or no PERM, or PERM Sooner, or PERM later. In case you are interested though, I am only permitted to make legitimate offers.

Furthermore, if these listed foundries really want potential protections for all their font licensing activities, the perm table is retroactively listing fields for licensing and other activities we've needed bits for, for years. What does Tal's proposal have to say about OEM bundling? about modification rights? about transferability? Isn't it time for not-entirely web-related permissions and recommendations to be stated inside fonts? Isn't it time to move backwards and forwards a little, get a complete grip, and then take off in whatever format they decide on... already perm-ed?

Then there is Tal's list of fields which repeats much of what is already in the font, adding a little to the education of users, but e.g reinforcing the idea that answering the question of 'who designed the font?' is more important than say, how well it is suited to blinking, setting at 9 ppm or on a curved baseline? Is this really how the font designers want our view of 'what's important' to be viewed?

Is "text describing the font" supposed to handle all this? Where it the machine-readability in that? In the end, about compression? In the short term, (i.e. before broadened bandwidth makes font file sizes irrelevant in seeking instant movie downloads), compression is needed to be solved in the presence of the font, the use and the document, and not before, (if you're the protective kinda typophile).

I'd also be sickened and disgusted by the divisive processes of uncooperative proposal and counter proposal ;) and tired by 350 hours of community service since spring (without being convicted of anything, much less pleading guilty), but in about 30 seconds, I'm on the first day of a two week vacation.

The last thing that was added to the OT spec in this area, was the DSIG table... is it our turn yet?

Cheers and Laters!

k.l.'s picture

I had considered them as both-and rather than either-or.

aluminum's picture

"how well it is suited to blinking"

Oooh! Blink-optimized typefaces! ;)

jeffveen's picture


My apologies if our attempt at being transparent with Typekit came off as "blowing smoke" -- that wasn't our intention at all. Regardless, we do have the product up and running for a limited set of beta testers. Andy Clarke of For A Beautiful Web has posted a good overview of how everything is working here:


blank's picture

Jeff: Now that’s what I like to see. As someone who remembers the dotcom era, working demos mean infinitely more than announcements about investors.

Richard Fink's picture

In defense of TypeKit's vaporware approach over the past couple of months - as it was explained to me by Jeff's partner Bryan Mason at TypeCon - they were keenly aware that web fonts were a big topic of conversation now, today, and insofar as they absolutely were going to have a working product shortly it was necessary for them to find a way to get in on the conversation and get a seat at the table.
Credit to them for certainly having done so. I wish I could drum up that much support for an idea of mine!
The have a working beta preview going on right now.
Quite real.

Stephen Coles's picture

For those who care about letters: it's Typekit with a lowercase 'k'.

Syndicate content Syndicate content