The fonts are dis­trib­uted under the SIL Open Font License, recog­nised by the Open Source Ini­tia­tive as an ‘open source’ license.

In my per­sonal opin­ion, I and the Web Fonts team at Google hasn’t encour­aged pub­lic par­tic­i­pa­tion in the fonts it has com­mis­sioned, but the license allows any­one to par­tic­i­pate if they wish.

In my oppin­ion you have good points, but you are los­ing some ones:

- Many fonts in GWF were a propi­etary work by pres­ti­gious design­ers and were released, some of them prize-winners. If all the fonts are «ugly, ama­teur­ish, etc» it means that also propi­etary fonts are ugly, ama­teur­ish and the recog­ni­tion of their design­ers is absolute crap.
I think this is a gen­er­al­iza­tion, there is great work there. And the high qual­ity foundries, famous design­ers and dis­trib­u­tors have lot of crap in their cat­a­log.

- If all the work is ama­teur­ish, incom­plete and awful maybe the $1750 aver­age is a great price for an ama­teur­ish designer. But many well known and suc­cess­full typog­ra­phers are there and I don’t think they are just destroy­ing their busi­ness for coins. I’m pretty sure they are get­ting a rea­son­able value and some­thing that works for their busi­ness model. Or they are stu­pid, which is a pos­si­bil­ity.

- «who can’t drop every­thing and move to Berlin or New York to learn the trade». Well, there is great type­face design­ers with great work and they didn’t move to Europe or USA. There is great schools in Asia, South Amer­ica and Aus­tralia and actu­ally many peo­ple doesn’t need to move to the Cen­ter of the Uni­verse.

Andres sets a petty low bar for “great type­face design­ers.” I think there are fewer than 20 work­ing today, and none of them is rep­re­sented in GWF.

MB, I think you’ve summed it up in a nut­shell. Impres­sive analy­sis!

@David
You can find fonts by Type-Together and AlePaul on the GWF among other great type design­ers as Pablo Cos­gaya, Eduardo Tunni, Dario Muha­fara, Ale­jan­dro Lo Celso, Juan Pablo del Peral (recent Letter2 win­ner) and many oth­ers.

@MB
Com­par­ing it to 1001 free font is totally unfair.
Here you have a list of great qual­ity type­faces on GWF that deserves much more than that:
– Pas­sion One
– Crete Round
– Almen­dra
– Over­lock
– Bub­blegum
– Gen­tium
– Abril Fat­Face
– Mate
– Bree Serif
– Galdeano
– Bit­ter
– Petrona
– Andada
– Rosario
– Chivo
Just to name a few…
Maybe you missed the lat­est addi­tions, I rec­om­mend you to check again.

Also, if you check the repos­i­tory, you will find a nice sur­prise there, soon to be released.

Regard­ing the col­lab­o­ra­tive aspect, you have to keep in mind that col­lab­o­rat­ing on “Design” stuff it’s not as easy as col­lab­o­rat­ing on “Pro­gram­ing” code. That’s why you don’t see as many col­lab­o­ra­tive design as you see col­lab­o­ra­tive pro­gram­ming.

There is no easy way to merge design ideas as there are tools to merge code, unit test­ing, etc.
So, col­lab­o­ra­tion is car­ried on as much as pos­si­ble. For exam­ple, on my own fonts, I’ve col­lab­o­rated with peo­ple who have made cor­rec­tions, extended the char-set, added Cyrillics, hand-hinting, etc.

Raph Levien is work­ing on a col­lab­o­ra­tive type design tool, witch he showed at the lat­est ATypI con­fer­ence. I hope this new tools will improve the col­lab­o­ra­tive aspect of type design when released.

Any­way.. any sug­ges­tions on how to improve col­lab­o­ra­tions is always wel­come.

Matthew raises some good issues here. I haven’t looked at GWF in a few months, but the last time I did, my con­clu­sion was the same as his — the qual­ity just is not there.

I have some expe­ri­ence in this area. I headed a team at Microsoft which pro­duced some of the best onscreen fonts in the world, includ­ing Matthew Carter’s Ver­dana and Geor­gia. We also pro­duced the excel­lent ClearType fonts, and Meiryo (intended to cre­ate tech­nol­ogy for screen-readable com­plex Asian glyph­sets, Meiryo began life with the inter­nal code-name of Verdana-J. Matthew Carter was again brought in to col­lab­o­rate, espe­cially on the har­mo­niza­tion of Romanji and Kanji).

The recipe for pro­duc­ing great fonts was the same in every case. Define the projects well, up front. Work with great exter­nal design­ers. Pay them fairly for the work they do.

Have your own great inter­nal typo­graphic experts who can over­see and man­age the projects from start to fin­ish.

Define stan­dards for fonts: char­ac­ter sets, etc.

Rec­og­nize that cre­ation of great fonts does not end with either the design or the cre­ation of the out­lines, but involves many other fac­tors includ­ing test­ing, qual­ity con­trol, etc.

There’s no rea­son why some of this work could not be done as Open Source. But, as MB says, there’s a need for benev­o­lent dic­ta­tor­ship in at least some areas. Design is not demo­c­ra­tic; art is not a com­mit­tee pur­suit.

View­ing its fonts so far, GWF strikes me as just too casual and hap­haz­ard, and seems to rely on what fonts peo­ple wish to con­tribute, rather than any sys­tem­atic approach.

Pro­duc­ing great fonts is a very expen­sive busi­ness. Microsoft, for all its faults, rec­og­nized this, and spent huge sums of money over more than 20 years on fonts for Win­dows, Office and the Web. It had per­haps 20 inter­nal employ­ees work­ing full-time on font projects, in addi­tion to out­side design­ers and con­sul­tants.

If Google is really seri­ous about Web Fonts, it needs do the same; bring a new pro­fes­sion­al­ism to the project, and hire its own experts to run it. And it needs to rec­og­nize that doing it prop­erly will take time and money. Despite the protests of some of the com­menters, there really are not that many great type design­ers out there who can pro­duce world-class fonts.

There are great type design­ers.
And there are good type design­ers.

In a nut­shell, the biggest issue is a lack of strong type direc­tion. Even a group of the best type design­ers can floun­der with­out a good type direc­tor to iden­tify what kinds of fonts are needed (or not) and how to go about mak­ing them.

Think of peo­ple like Chauncey Grif­fith, Mor­ris Fuller Ben­ton, Robert Hunter Mid­dle­ton, Mike Parker, etc … They over­saw entire depart­ments of peo­ple work­ing on col­lab­o­ra­tive type devel­op­ment projects before dig­i­tal type even existed.

With mod­ern font for­mats (UFO) and code repos­i­to­ries, there’s no rea­son type design can’t eas­ily be a col­lab­o­ra­tive effort today too.

I should be clear that my com­ments here reflect my own per­sonal opin­ion and not the views of any of my con­sul­tancy clients, includ­ing Google Inc. So when I ear­lier said that ‘Google hasn’t encour­aged pub­lic par­tic­i­pa­tion’ then I mis-spoke, and really I mean that I per­son­ally haven’t focused on encour­ag­ing the kind of pub­lic par­tic­i­pa­tion that is com­mon to libre cul­ture projects.

In work­ing to cul­ti­vate a libre type cul­ture over the last few years, I have, gen­tly and in pass­ing, encour­aged design­ers com­mis­sioned by Google Web Fonts have made some small steps to work in this way — plac­ing the font files github for exam­ple, or run­ning a font project blog actively encour­ag­ing pub­lic par­tic­i­pa­tion — but these efforts by design­ers haven’t led to much inter­est.

Ver­non Adams is one of the most active type design­ers work­ing on web font com­mis­sions for GWF, and he is pub­lish­ing all work in progress in a blog and also pub­lishes all his work­ing files in a Google Code project. I hope Vern can com­ment on how many peo­ple have con­tacted him in the last 2 years about con­tribut­ing or col­lab­o­rat­ing.

The Kick­starter projects that Google Web Fonts sup­ported have also invited par­tic­i­pa­tion from back­ers, which is some­thing per­haps Pablo can com­ment on. My under­stand­ing is that par­tic­i­pa­tion has been muted there too, despite that peo­ple have paid from their own pocket to fund the designer.

There were some lengthy dis­cus­sions on Typophile when libre fonts started grow­ing, about 6–7 years ago, and the idea of pub­lic par­tic­i­pa­tion in type design has been ridiculed; how­ever, the vision of par­tic­i­pa­tion built into the Open Font License’s Reserved Font Name fea­ture has been shared and turned into a very pop­u­lar work­flow by GitHub.

While Nick men­tions the UFO (or SFDir) for­mats and project host­ing that solve part of the prob­lem, the lack of equiv­a­lent tools for man­ag­ing revi­sions in text for fonts don’t exist — and as Pablo says, Raph Levien has been mak­ing some explo­rations in that area.

So there appears to now be a grad­ual accep­tance that pub­lic par­tic­i­pa­tion in type is real­is­tic, and its nice for me to see peo­ple call­ing for it to be done bet­ter. I will do what I can

Regard­ing the ‘sniff test’ you men­tion, there is a Google Code project ‘google­font­di­rec­tory’, where the source files for the fonts are pub­lished; this is linked to quite promi­nently from the down­load dia­log of the main GWF web­site; and the pub­lic do reg­u­larly report issues in the issue tracker there. So I think the GWF project clearly passes the sniff test.

Also, last month I started a small project to do peer review of fonts — http://typereview.wordpress.com/ — and if any­one is inter­ested in doing sim­i­lar review con­sul­tancy, I hope they’ll speak to me directly.

I was look­ing for­ward to read­ing this. How­ever at the risk of expos­ing myself as stu­pid on the super high­way can you explain the mean­ing of the “Dilu­tion” and “Real­ity” sec­tions in each “Essen­tial Qual­ity”? I am lost.

[I added a note that the dilution–reality com­par­i­son comes from my ear­lier arti­cle about the seven essen­tial qual­i­ties of open source. — MB]

I just checked the review blog Dave men­tioned. There is a type­face in progress there called “News Cycle”. The designer claims that it is “based” on News Gothic. From what I can see in the spec­i­men, “based on” is a euphemism for a straight-forward copy, aka Rip-off. Does any­body chez Google look at the sources of these “free” fonts? If you use exist­ing out­lines, mak­ing a free font doesn’t need much work. I won­der whether the pub­lish­ers of the orig­i­nal agreed to their intel­lec­tual prop­erty being made Open Source with­out their con­sent.

News Gothic was first pub­lished in 1908 and is there­fore in the pub­lic domain. News Cycle is a straight for­ward copy, and entirely legit­i­mate — just the same as Lino­type and Ben­ton Sans

The dig­i­tal out­lines are not in the pub­lic domain. And I would bet that “based on” in this case means that the designer used some­one else’s out­line.

I can­not believe that you would men­tion Ben­ton Sans here. That is a dif­fer­ent type­face alto­gether. Clearly inspired by News Gothic and named for its designer, but com­pletely redrawn. As is the new Lino­type ver­sion, albeit much closer to the orig­i­nal. Lino­type based this on their own ver­sion which they are enti­tled to use.

So there is many great type­face design­ers as great astro­nauts in the world: twenty. Wow!

Between TDC, Let­ter 2 and Tipos Lati­nos we had about 100 rec­og­nized great designs.
Some of these designs and his design­ers are in GWF or will do soon.
I agree maybe most of the fonts there are not high qual­ity. But it is not true to say:

- All in GWF is low-quality.
– Comis­sioned fonts for GWF are under paid.
– A good typog­ra­pher has to study in Europe.

«There really are not that many great type design­ers out there who can pro­duce world-class fonts». There are hun­dreds! But only a few of them will have the bud­get to do world-class fonts.

Typog­ra­phy is not quan­tum physics.

PS: The 100 design­ers were rec­og­nized only in the last year.

@Erik,
I think that as the audio is in Span­ish, you may have mis­un­der­stood the video.

What you see there, is not a type in-progress using pro­pri­etary dig­i­tal out­lines as a base font.

It’s a review ses­sion by Angel­ica Diaz, fix­ing errors and improv­ing News Cycle out­lines by Nathan Willis.

If News Cycle where based on pro­pri­etary out­lines, then there would be noth­ing to cor­rect.
The fact that cor­rec­tions where needed, proves to some extend that it was a new dig­i­ti­za­tion.

If News Cycle where based on pro­pri­etary out­lines, then there would be noth­ing to cor­rect.
The fact that cor­rec­tions where needed, proves to some extend that it was a new dig­i­ti­za­tion

Oh no, it does not. It could sim­ply mean that some­one alters some­one else’s out­lines, but not well enough. Or not far enough away from the orig­i­nal.
BTW: my Span­ish is good enough to get that.

News Cycle is absolutely not made using any other out­lines. The scans from the ATF cat­a­log are pub­lished along with the source files by the designer at https://launchpad.net/newscycle/with-samples — another exam­ple of the designer com­mis­sioned by GWF who took up my sug­ges­tion to make the font in a way that invites the pub­lic to par­tic­i­pate.

Glad to hear that, Dave. That is a good way to get a new feel out of an old pro­to­type.

Just looked at News thingy: it needs work. Apart from some stray points and a few wob­bly curves, char­ac­ters like lc e are egg-shaped, which is not what Ben­ton had in mind when he designed News Gothic.

@MB:
1. Typog­ra­phy as an indus­try wouldn’t be able to pass OSI’s ‘patch test.’ That exam­ple you link to on the OSI blog is so depen­dant on the can­di­date project being some sort of code based appli­ca­tion.

2. Not all the fonts are cre­ated at Google. They’re licensed. Or comis­sioned. So it’s up to the type designer to seek dis­trib­uted peer review.

3. A rec­om­men­da­tion to load files from a CDN vs locally down­loaded makes them anti open source? It’s well doc­u­mented that load­ing assets from a CDN is ben­e­fi­cial to the end users. If a copy of a Google Web Font is already loaded into a user’s browser cache then it won’t need to down­load that font again if another site uses it. This is why type­kit, font­deck, et al host font files with­out offer­ing down­loads. Google offers down­loads because some sites (like inter­nal intranets) can’t use hosted files. Type­kit and the like don’t have any sup­port for enter­prise.

Agree that gen­er­ally “Google has a great engi­neer­ing cul­ture, a weak design cul­ture, and no dis­cernible taste.”

How­ever, two recent projects lead me to believe things are chang­ing for the bet­ter:

http://www.thinkwithgoogle.com/quarterly/speed/note.html

http://www.googlezeitgeist.com/en/

I really enjoyed this Matthew. I’m glad to see you in the mix. One small note. I think you mean “delu­sion”, rather than “dilu­tion”.

On reflec­tion, I’ve done Alexei Vanyashin (Cyreal) and Eben Sorkin a dis­ser­vice by not men­tion­ing them as exam­ples of some the stuff that Matt is propos­ing that just isn’t very pub­li­cised. Each has a team of design­ers dis­trib­uted around the world who they work with closely with to increase the qual­ity of the work deliv­ered to Google — doing exactly this kind of ‘benev­o­lent dic­ta­tor’ coach­ing. The evi­dence for this is in the FONTLOG files in the font source direc­to­ries.

Thanks to this post, I’ve heard from some of the pro­fes­sional type-design indus­try folks who have offered to step into this ‘benev­o­lent dic­ta­tor’ role for any less expe­ri­enced design­ers who want such help and are work­ing inde­pen­dently

Google isn’t curat­ing these fonts; it’s act­ing as a repos­i­tory to make them avail­able to web­sites. In other words, it’s act­ing basi­cally like GitHub or Bit­bucket. Both of those sites host plenty of crap open-source projects. Say­ing GWF needs a “benev­o­lent dic­ta­tor” is mis­un­der­stand­ing what it is.

I wanted to thank you for talk­ing about “grift­ing.” As a musi­cian, this hap­pens all the time in my busi­ness, and the result is that pay goes down for every­body, and a lot of peo­ple get taken advan­tage of. When many peo­ple will work for free, it causes those want­ing the ser­vice to think it isn’t worth much…

What thrills me more is, the fact that most of the type design­ers doesn’t accept the fact that, you should work and get paid once, and not work­ing as a king get­ting life­time trib­utes.

Nobody pays me for every design I do. If Google pays you to design an open source font I do believe it’s ok. Why shouldn’t? Is the price the main prob­lem? Solu­tion: ask for more money.

Google isn’t curat­ing these fonts; it’s act­ing as a repos­i­tory to make them avail­able to web­sites. In other words, it’s act­ing basi­cally like GitHub or Bit­bucket.

But it surely looks like Google is offer­ing those fonts, as opposed to e.g. Google Image Search, where it’s obvi­ous that you get results from third par­ties with no involve­ment by Google.

Well, are the Google web fonts curated? Has any sub­mit­ted font so far been declined for what­ever rea­sons?

Per­haps Google doesn’t need to care if the fonts are of any par­tic­u­lar qual­ity as long as peo­ple use them. The tricky thing is, that even if peo­ple put these fonts in their web­site and feel good about hav­ing been given cool fonts from Google for free, they may be doing a dis­ser­vice to their read­ers and not even notice it – when fonts fail on a par­tic­u­lar sys­tem for tech­ni­cal or design rea­sons, deter­ring the reader. Which is the oppo­site of tak­ing respon­si­bil­ity for a bet­ter web, some­thing that I’d say would fit Google (»With great power comes great respon­si­bil­ity« – to quote Spider-man’s uncle), but prob­a­bly isn’t nec­es­sary for their busi­ness inter­ests.

“Open source rede­fines what is valu­able.”

Actu­ally, the pur­pose of all of Google’s “open source” projects is to, “destroy the mar­ket­place”. In a mar­ket­place where there is no value, it’s harder for a com­peti­tor to thrive, and eas­ier for one with deep pock­ets to cap­ture eye­balls for ads. This is the basic strat­egy in all Google’s projects over the last 10 years.

There’s a deep con­fu­sion at the heart of this piece. It’s summed up by this quote:

“And these fonts are “free” only in the triv­ial sense that Google does not charge us to use them.”

In real­ity these fonts are “free” in the rather impor­tant sense that the peo­ple who cre­ated them (which in almost every case was not Google) have released them under a licence which allows (amongst other things) for Google to then dis­trib­ute them, and you to use them in com­mer­cial works, bun­dle them with Open Source soft­ware and so on. What most peo­ple would con­sider gen­er­ally con­sider to be “open source” in fact. I realise that you have your own idio­syn­cratic and some­what sub­jec­tive def­i­n­i­tion of that term but you don’t really make clear whether you believe that it is gen­er­ally accepted by the wider com­mu­nity (in my opin­ion it’s clearly not, though whether any one def­i­n­i­tion would be acc­cepted is another ques­tion).

The rest about type appears to be mostly snob­bery, no dif­fer­ent from early crit­i­cisms of Wikipedia or Open Source soft­ware or dig­i­tal cam­eras or dig­i­tal music by snobs of a dif­fer­ent per­sua­sion.

Why you are blam­ing a sin­gle com­pany for these early teething issues is not clear though. You men­tion Red Hat, who’ve sim­i­larly done good work in get­ting fonts to dis­play bet­ter on the desk­top and funded font cre­ation. Why are they not being held solely respon­si­ble for the state of cur­rently avail­able Open Fonts instead of Google? Both have con­tributed great work to build the foun­da­tions for the future, surely there­fore they should both be crit­i­cized if we have not yet reached per­fec­tion?

Can some­body tell Andy that no one who reads Matthew But­t­er­ick thinks exper­tise is mere “snob­bery”?

Some peo­ple really do know more than you and have bet­ter taste.

@Joe,

I’m not dis­put­ing exper­tise or even taste. I’m say­ing that things can and do improve from hum­ble begin­nings. My first dig­i­tal cam­era was frankly ter­ri­ble, the pic­tures were barely rec­og­niz­able, it didn’t stop me from see­ing the poten­tial all those years ago and even then it had util­ity for cer­tain use cases. Sim­i­larly my old phone had a ter­ri­ble cam­era, that still worked as an opti­cal sen­sor that could read bar­codes, sense light lev­els and take my heart­beat.

But some peo­ple are too close to the issue and harp on about qual­ity to such an extent that you think they’d be able to rec­og­nize the obvi­ous qual­ity increases over time if they’re pay­ing atten­tion and are so attuned to sub­tlety, even if that level remains, for the time being, below some per­sonal limit that mat­ters for them. But then you realise that they’re snobs and will always believe the new way to inher­ently infe­rior.

Ale­greya, Letter2 win­ner.. just added.

Ale­greya is ugly and ama­teur­ish, just as every­thing there. Only 20 guys in the world can make great things.

For clar­ity, Google’s “agent” hand­ing out cash for fonts (“TOO MANY FONTS? NOT ENOUGH CASH?”) is whom, exactly?

I’m say­ing that things can and do improve from hum­ble begin­nings.

I can­not believe you actu­ally mean this. We are not talk­ing about a small sweat­shop that is just learn­ing about dig­i­tal type. This is Google who could buy the whole type indus­try out of petty cash! And, as Matthew has pointed out, the begin­nings of dig­i­tal type are 20 years and more behind us. For Google of all peo­ple to claim inno­cence is absurd, even cyn­i­cal.

Not sure how this fits into the larger debate, but there is at least one project that aims to do open-source typog­ra­phy in a curated fash­ion: http://www.theleagueofmoveabletype.com/ (I’m prob­a­bly not qual­i­fied to say whether they’ve suc­ceeded, but their one page of sam­ples did make me down­load the whole thing. GWF inspires no such urge.)

If the hum­ble begin­nings of dig­i­tal type are behind us why is it still a nov­elty for a web site to use any font not one of a hand­ful pro­vided by default by Microsoft? Or use the font-weight: 100–900 prop­erty to its full poten­tial? Or use lig­a­tures? Or real small caps? Or lin­ing or tab­u­lar num­bers? Or type­set maths? Or use decent jus­ti­fi­ca­tion and hyphen­ation? How many peo­ple are using devices or soft­ware that require elab­o­rate and expen­sive hand-hinting for fonts to dis­play prop­erly? We’re clearly still in the wild west of web fonts.

As I said, it is get­ting bet­ter. I can see the improve­ments that have hap­pened recently and are com­ing down the track. And open source soft­ware and/or Google is lead­ing the charge on many of these items since they both want the web to be bet­ter.

Are peo­ple approach­ing this as “Google should be pro­vid­ing amaz­ing, free, open source dig­i­tal type­faces for my use but I don’t par­tic­u­larly care about using type well on the web” or am I the one miss­ing some­thing?

Matt, curi­ous to hear what your exam­ples of qual­ity web fonts are

As a typog­ra­phy enthu­si­ast, I’d almost agree. But as a real­ist, I can’t.

Google Web Fonts doesn’t aim to make high-quality typog­ra­phy free, it sim­ply gives the ordi­nary per­son the abil­ity to exper­i­ment with type online. The same kinds of peo­ple who would pre­vi­ously use and abuse Comic Sans. There is no bet­ter way to raise the pro­file of typog­ra­phy than to give peo­ple stuff to play with.

Does it suck that the fonts are often of lower qual­ity, that the kern­ing is off, that the italic is gen­er­ated? Sure. But it’ll be a tiny bit more inspired than yet another Times New Roman or Ver­dana. And after a while, they’ll prob­a­bly fig­ure out them­selves that some things look weird and oth­ers look good.

Text doesn’t become beau­ti­ful just by using the right font any­how, it takes a lot more than that. If you care about that, you’ll either look for the few free GWT gems, invest in com­mer­cial fonts, or pay a designer to do it for you.

[…] the way, this post was inspired by Matthew Butterick’s own crit­i­cal look at Google Web Fonts. Matthew makes some good points, but over­all I still think Google Web­fonts is a great […]

I’m some­what wary of con­tribut­ing to this essay, as i’m aware there is a wide gulf between my own prac­tices & inter­ests in type & type design and the prac­tices and inter­ests of oth­ers here; a gulf too wide to make for dia­logue. For exam­ple, I don’t believe in ‘great design’, or great ‘design­ers’, and i don’t believe in pro­pri­etry pro­duc­tion meth­ods, instead i see most value in the ways in which the ‘fas­cis’ of ‘design’ can be re-shaped, sub­verted, freed, and re-used. In other words, we are not going to agree on design mat­ters How­ever, on the prac­ti­cal side, i think it’s worth not­ing a few things.
1. In my mem­ory, peo­ple have talked a lot, for a long time, about serv­ing type over the net, but not really done that much. The sit­u­a­tion is bet­ter now that, in the last few years, some peo­ple have, at least, started doing ‘some­thing’.
2. A lot of the google fonts get used a lot. Aka more than fonts from other web­font direc­to­ries. They clearly serve pur­pose & func­tion.
3. A lot of the google fonts (and the server-side itself) are ongo­ing, being improved, still being shaped, but in the open. This is appar­ently not a prob­lem for users (see point 2).
4. Users actively and help­fully feed back. This can be put, in my expe­ri­ence, into three cat­e­gories; a. “thanks, give me more”, b. “i notice some­thing wrong, please fix it”, c. “some­thing was wrong, i fixed it, file attached”.

As far as i’m con­cerned that’s a func­tion­ing, pos­i­tively evolv­ing sys­tem. How is your sys­tem doing?

As for the gen­eral idea of the font’s not being ‘open source’; it’s an inter­est­ing intel­lec­tual exer­cise (i guess), but behond that i don’t see where the magic is in claim­ing these ‘open’ fonts are ‘closed’. Where’s the trick meant to be?

Mathew, i won­der if your are approach­ing this too much from the view that ‘Google Web­fonts’ is some homoge­nous, sta­tic, whole. I sug­gest it’s not good ori­en­teer­ing. Think ‘crowd’ rather than ‘law court’. Some of the larger type foundries have cre­ated web­font direc­to­ries more on the model you seem to pre­fer, so just use those ser­vices if you are approach­ing this as a user. If your inter­est in google fonts is more int­elec­tual, then we are talk­ing of such dif­fer­ent mod­els that they are just not com­pa­ra­ble, hence you hav­ing to resort to gym­nas­tics to re-imagine font files freely and widely avail­able under open­source licenses as not really open source. And i don’t under­stand this need some have to so assertively com­pare, cri­tique and pros­e­ly­tize on who uses what font, and from where Why would type design­ers who have their own fonts for sale via com­mer­cial web­font direc­to­ries go out of their way to crit­i­cise (or even accuse of piracy) the design­ers of fonts on free web­font direc­to­ries? I can’t see why they would do that.

Ver­non, all you’re basi­cally say­ing with the “crowd” vs. “court” metaphor is that Google Web Font being a pile of fonts with qual­ity all over the place is inten­tional rather than acci­den­tal. If I already believe that “pile of fonts with qual­ity all over the place” is a bad model for ped­dling fonts, that just makes Google delib­er­ately taste­less as opposed to acci­den­tally taste­less.

Mathew, every­thing i’ve said is my own opin­ion, it in no way should be read as the view of any other designer cre­at­ing web­fonts for the font direc­tory. It espe­cially shouldn’t be read as google’s view.

It’s heart­en­ing to see the para­me­ters of your stance soft­en­ing, but are you really still insist­ing that soft­ware freely and openly released under well estab­lished open­source licenses (GPL, OFL, etc), and used freely by mil­lions of users under that license, isn’t ‘open source’? As a lawyer, are you say­ing that either the Open Font License or the Gnu Pub­lic licenses are not legally valid for licens­ing fonts? Have you con­tacted SIL or Gnu about your ideas? Curi­ous about their posi­tion?
cheers –ver­non

@anonymous
Why do you equate “crowd” with ‘lower qual­ity’, ‘acci­den­tal’, and ‘pile’ ? Are you demo­pho­bic?

Mathew, thanks for clar­i­fy­ing your ideas.

I’d really like to pin this down prop­erly with you, if pos­si­ble, because, i think you are point­ing at some­thing very impor­tant. If you, as a lawyer, with some exper­tise in type design, is mak­ing seri­ous state­ments claim­ing that a sig­nif­i­cant amount of fonts released under open­source licenses, such as the OFL, “aren’t really open source”, then i think we need to sit up and lis­ten to you. I’m cer­tainly lis­ten­ing closely, because i see that if open­source licenses are being used by design­ers to pub­lish ‘not really open­source’ fonts more than to pub­lish open­source fonts, then surely the licenses become some­what under­mined as valid open­source font licenses? That is what you are say­ing, isn’t it? That these licenses are being used to add false cre­dence to a posi­tion that adversely “shrinks down the idea of open source to noth­ing more than file shar­ing”, and that despite the license under which they are pub­lished, the fonts are ulti­mately “not really open­source”. Is that really your claim, Mathew?

But are licenses such as the GPL and OFL really being widely (mis)used in this way by design­ers (on a pretty large scale, i would say, look­ing at your claims)? Are you really cer­tain that design­ers like myself, are devel­op­ing and pub­lish­ing fonts under these licenses only to “claim the man­tle of open source” and/or just to give us a “mar­ket­ing catch­phrase”? Are you sure about that? How am i doing that exactly? I’d like some evi­dence from you on that claim, as it is quite a claim.

And, i’m still curi­ous if you would include fonts, pub­lished under the OFL, by SIL, in your claim of “aren’t really open source”. As far i know fonts pub­lished by SIL are no more, or, no less, encap­su­lat­ing of your ‘essen­tial qual­i­ties’ as fonts that i pub­lish under the same licence.

Many thanks
–ver­non

Mathew, thanks for more fur­ther clar­i­fi­ca­tion.
As i’ve stated before, my com­ments and views are wholly my own, as is my deci­sion to release fonts under open­source licenses. I’m more than happy to respond per­son­ally to the con­cerns you have about them. Just shoot! I’ll pos­si­bly defend them too.

My view of your ‘essen­tial qual­i­ties for open­source devel­op­ment’ con­cept, is that it clearly reflects some­one with a very ‘top down’ approach. It prob­a­bly explains why i rel­ish an open­source model, and you, maybe less so. To me, your check­list is a pretty typ­i­cal view of some­one not used to work­ing with open­source con­tent & users in their day-to-day work. What are your expe­ri­ences of your con­cept in your own open­source projects? how do your ‘esen­tial qual­i­ties for open­source’ invect your own open­source out­put? There’s no sub­sti­tute for real world, day after day, front­line expe­ri­ence, so I am gen­uinely inter­ested how your con­cept works for you in the real world. Per­son­ally i’d be wor­ried that your check­list would hin­der as many open­source devel­op­ers & users as it might help or attract.

Imho the full breath of open­source mod­els have their place, as does the full breadth of pro­pri­etry mod­els. I bet that devel­op­ers are best to use the work­ing model that they work ‘hap­pi­est with’. Then a project can con­cen­trate on cre­at­ing stuff and users either respond, or they do not. Rewards come, or they do not. You pack up, or you keep going. It’s Dar­win­ian; if it fits, then it sur­vives and repro­duces. Per­son­ally, i adopt a bot­tom up, open­source model; high on action, low on seman­tics. I like the way that works at the moment, i like the way users respond, and i like the rewards. I wouldn’t use that model if it didn’t work for me. With this approach i am also free to change/adapt/reform/abandon if i feel any­thing starts to not work. I’d say that’s another asset of a bot­tom up approach; noth­ing get’s ‘too pre­cious’, there are few over-arching con­cep­tual bur­dens, and you can eas­illy scrap work and start over. Also, because it is such an open model, it allows col­lab­o­ra­tion and input from any direc­tion, from any­body, there’s no ‘do you pass the But­t­er­ick 7 Essen­tial Qual­i­ties test?’. To me your model, would close down that open­ness way too much. As you say, a project is either trully ‘open’ or it is not, the guage of that is sim­ple; can any user, at any time, look at/get/use/adapt my prod­uct and it’s source code. My con­cern is that your model would impede too much on that open­ness for my lik­ing.

cheers
–ver­non

> And how does your work live up to those val­ues?
I have opened all my projects now, in a way that’s easy for any­one to par­tic­i­pate, send sug­ges­tions, edit the source files and post them back, etc:
http://www.impallari.com/projects
I hope this new inter­face will enable wider par­tic­i­pa­tion and col­lab­o­ra­tion.
Also, will be host­ing guest projects for any­one inter­ested.

I am really suprised by the amount of critisim of GWF. Ok, I wasn’t aware of how many things actu­ally work @GWF (so thanks Matthew for this inter­est­ing insight) but still – even though the author makes some good points – my feel­ings about this project are rather pos­i­tive.

I don’t want to argue about the qual­ity of the fonts, too many great design­ers par­tic­i­pate in this dis­cus­sion. But read­ing these com­ments made me feel like Michelan­gelo and Van Gogh are vis­it­ing a small sta­tionery store and com­plain about the poor qual­ity of brushes and can­vasses. I think the fonts @ GWF aren’t ment to be the next cor­po­rate font for Deutsche Bahn or Bank of Amer­ica. But: There are some good entries, which let us take a break from star­ring Ver­dana, Geor­gia and the best font in the world aka Comic Sans.

I also wanted to men­tion what ser­vices like Google Web Fonts, Type­kit & Co have changed for me: I am a web­de­signer, work­ing on rather small projects. Few years ago it was almost impos­si­ble (well, at least very hard) to imple­ment cus­tom fonts on a web­site. There were some options to over­come tech­no­log­i­cal hur­dles (using sIFR, Cufon etc.) but that wasn’t really con­veinient, plus there were no clear licens­ing mod­els for using fonts in the web. I asked directly at the foundries a few times, but i never got an offer afford­able / ratio­nally suit­able for my clients.

Since 2009/2010 things began to change. Sud­denly it was much eas­ier to imple­ment fonts by adding few lines of code (@GWF you just need copy one line of CSS!). The browser-support for fonts became bet­ter and bet­ter. Online typeface-libraries began to offer clearly priced web-licences and detailed instruc­tions how to use them. It’s much eas­ier to present the ben­e­fits of a nice web­font to a client and esti­mate the costs.

Now i use web­fonts on almost every webpro­ject, either GWF, Type­kit or a EOT/WOFF license. Recent changes allowed me to extend my ser­vice port­fo­lio, I can offer bet­ter prod­ucts to my clients. I hope this new enthu­si­asm and inter­est in web typog­ra­phy will also have a pos­i­tive influ­ence on sales @ font­foundries and design­ers.

In my eyes Google plays a part in this process of bring­ing type to the web. And it was really about time we start this jour­ney. Google is doing it their own way (what­ever that means), but hey, they are right to do so, because the *do* some­thing. If you are *doing* some­thing, you can decide how to do it.

This is a very inter­est­ing debate.

I really think that GWF has a lot of good fonts, and some great fonts. It would be nice if Google could be more of a men­tor and guide to some of the unex­pe­ri­enced design­ers, but if they don’t want to do that, design­ers will find other ways to help each other and inspire each other.

What Pablo does on his page is great! Expand­ing “Rale­way” is exactly why these fonts were made Open. I remem­ber some­one expanded the “Chunk Five” font which started with a very restricted set of char­ac­ters. I don’t have the knowl­edge to edit fonts myself, but I do write com­ments or e-mails to design­ers who ask for feed­back, like Ver­non on his blog.

Hi But­t­er­ick,

I think you’re wrong about open source.

I’m an ama­teur coder. I don’t have no uni­ver­sity or any cer­tifi­cate. How­ever I have some open source projects ^^ If you need proofs, here is one of mine: http://code.google.com/p/android-filechooser/

I pushed my pas­sion into the projects. So I don’t care if they are craps to some­one. That’s open source. As like as you have a car, and your friend needs a car, he or she bor­rows it. Yes, if you like it, use it. If you see bugs, feel free to fix. If you don’t like it, that’s ok.

Open source doesn’t need qual­ity. It needs pas­sion. When you have pas­sion, you do the best. That’s “qual­ity”. Have you ever paid Linus Tor­valds a cent? I think he’d never think about that. Do peo­ple see that in the first days of his works, no one thought about nowa­day?

If Google want to buy me some beer, or buy me a heli­copter, that’s fine. I’ll be very happy if it’ll be a heli­copter But if there is noth­ing, then no prob­lem.

If there are libraries which have same func­tion­al­i­ties as mine, and Google pay their founders. That’s ok too. Why? Because when I’ve had the ideas, I’ve thought about doing them, money-has-been-never-being-a-condition.

About Google Web Fonts. I don’t care how many peo­ple in the world can pro­duce “best fonts”. To be hon­est, that sucks.

There-is-no-limitation.

A man com­ing from Africa can pro­duce best fonts (no quo­ta­tion marks). I don’t care if one of them was born in Sil­i­con Val­ley, or was born in Viet­nam, my coun­try — one of the last com­mu­nist coun­tries in the world. I see by-my-own-eyes — the fonts are beau­ti­ful. So it’s ok. Down­load and use! To my stan­dards, they’re the best. No need for Amer­i­can Stan­dard.

If open source needed to be cer­tifi­cated before it is pub­lished to the com­mu­nity, then there would be no Fedora, Man­driva, Ubuntu, or Android…

And, I don’t work with fonts, but you’re wel­come to be a com­mit­ter with them. It’ll have fun

Sin­cerely,
Hai Bison

Nice arti­cle. I really enjoyed read­ing some­thing with Charter/Charis (I can­not dis­tin­guish on screen)

Just one com­ment:
>Real­ity: Open source relies on benev­o­lent dic­ta­tors.

Not really absolutely true. Canon­i­cal is an exam­ple of what you mean, but Debian may be an exam­ple of a dif­fer­ent point, which existed before Ubuntu and it seems to be that it will still be some years around

I just found this thread. My annoy­ance with Google Web Fonts have been echoed here numer­ous times: It’s a dump­ster with hun­dreds of low-quality fonts. There’s no real way to make a request like “Show me only fonts with delta hint­ing” or “Show me only fonts designed for body test”—a strange defi­ciency for the king of search.

I’ve gone through GWF to try and find decent fonts with Delta Hint­ing. Ubuntu is pretty good, but it’s no Ver­dana, and its license is an unusual non-OFL one. Cou­sine is a beau­ti­ful fully delta hinted OFL font, but, alas, it’s a mono­space. The PT series (PT Sans, etc.) is nice; it’s a shame that PT Sans Cap­tion doesn’t have decent delta hint­ing (nor an italic/oblique).

I would much rather have one OFL-licensed font with delta hint­ing that is as read­able on the screen as Ver­dana than 10,000 use­less fonts. Unfor­tu­nately, I haven’t found one yet. And I’ve been look­ing for years.

If you’re look­ing for Delta-hinted fonts (or what­ever other qual­ity stan­dard my be applic­a­ble for your intended appli­ca­tion), per­haps you should con­sider pay­ing a license for a pro­fes­sional font. There is only so much you can expect from fonts that have been designed by peo­ple in their spare time and then made avail­able for free.

Do you give your work away?

Indeed. TANSTAAFL. I’m a soft­ware devel­oper, and there’s a rea­son why my open-source pro­gram doesn’t have DNSSEC. Every­one wants a really great free web font, but no one wants to pay for it.

That said, I did look at the Kick­starter projects and there isn’t one for an OFL fully delta hinted screen font out there. But, real­is­ti­cally, a good delta hinted font will prob­a­bly cost between $50,000 and $100,000 to pay the design­ers to develop—and I don’t think Kick­starter is able to raise that kind of money (Exo got $7,500).

Matt… you’re awe­some. You can talk law, open source prin­ci­ples, typo­graphic excel­lence, python, eco­nom­ics and the col­lec­tive action prob­lem all in one essay AND keep up with the com­ments. WTF? I started talk­ing typog­ra­phy the other day to a fel­low law stu­dent and their eyes glazed over in a mat­ter of sec­onds.… I appre­ci­ate you and peo­ple like you in this world.

Erik Spiek­er­mann. I’m get­ting some­what star struck here. Sorry for that.

I think Sam could pay a license for a pro­fes­sional font with delta hint­ing. Myselft have spend quite a lot over the years. How­ever, when it comes to Open Source projects, like GPL, i may need to ship it with a font with a license that is com­pat­i­ble with GPL.

”There is only so much you can expect from fonts that have been designed by peo­ple in their spare time and then made avail­able for free.
Do you give your work away?”

This is true in most cases. How­ever, there could be dif­fer­ent. An exam­ple is the very soft­ware you are using on your blog.

Word­Press. Lot’s of peo­ple are actu­ally giv­ing away their work and does not get paid for it (not every­one, but many). The result is progress. So, how do they get paid?

They get paid as every­one else work­ing. But maybe not from Automat­tic or some other com­pany involved with Word­Press. Many of them are doing this at their spare time out­side their ordi­nary jobs. The result is some­thing mil­lions of peo­ple are using.

How­ever, a font project is not the same as a project like Word­Press. And it’s hard to imag­ine 200 peo­ple work­ing on a pro­fes­sional font at the same time.

I think com­pa­nies like Google, if they actu­ally are seri­ous with Google Web Fonts should have a strat­egy with more com­mis­sioned work to bring in more great typog­ra­phers to the project. While still be ”gen­er­ous” about what fonts are included in the repo.

That way i believe Google Fonts will be more inclu­sive. If Aunty wants to use crazy font 1 for her web­page about ham­sters. Fine.

If some­one else wants to have fonts with lig­a­tures, more styles than one, delta hint­ing and so on, they would have it.

Will it eat mar­ket shares from the var­i­ous type foundries? I don’t really think so. Type foundries are mostly not mar­ket­ing their busi­ness towards the web any­way. The few that do, will still have their own niche and audi­ence.

Cheers. And keep up the good work (actu­ally read­ing about you in the Swedish mag­a­zine Cap & Design right know).

Christo­pher, almost all the fonts in Google Fonts were com­mis­sioned for it, and the design­ers paid to make them.

[…] always on the make and always look­ing for oppor­tu­ni­ties has some pseudo open-source fonts at http://www.google.com/fonts/ – but ‘mea­sured by pro­fes­sional […]

[…] always on the make and always look­ing for oppor­tu­ni­ties has some pseudo open-source fonts at http://www.google.com/fonts/ – but ‘mea­sured by pro­fes­sional […]