One of my ma­jor prompts for think­ing about the es­sen­tial qual­i­ties of open source has been the Google Fonts project, which claims to be a li­brary of open-source fonts. I de­sign type. I’m in­volved with open-source projects. I was in­trigued by the con­cept. There’s no rea­son open-source fonts can’t ex­ist. But on closer in­spec­tion, Google Fonts are not “open source” in any mean­ing­ful sense.

I wrote some pre­vi­ous com­ments (here and here) about Google Fonts as part of a dis­cus­sion about Ro­boto, the new An­droid font. What fol­lows is a more de­tailed crit­i­cism of Google Fonts, but it’s also a call for Google to do bet­ter. They can and they should.

How do I know they can? In the en­gi­neer­ing arena, Google has shown it­self to be a good cit­i­zen of the open-source world. For in­stance: Python. For seven years, Google em­ployed Python’s cre­ator and benev­o­lent dic­ta­tor, Guido van Rossum. Ac­cord­ing to Guido, he got “to spend half [his] time on Python, no strings at­tached[.]” Google has adopted Python as its “main script­ing lan­guage.” Google has re­leased Python source code and li­braries. Be­yond Python, Google main­tains a whole blog de­voted to its open-source ac­tiv­i­ties. Even the Google Fonts team has re­leased open-source code for read­ing and edit­ing fonts.

But what you won’t find on Google’s open-source blog, nor on its Open Source Pro­grams web­site, is any men­tion of the Google Fonts them­selves. This is odd, be­cause ac­cord­ing to Google, “All the fonts are Open Source.” Re­ally? The omis­sion sug­gests that Google might have doubts whether these fonts qual­ify as open source.

I’m con­vinced that they don’t. To make the case, I’ll step through my seven es­sen­tial qual­i­ties of open source (in­clud­ing the di­lu­tion–re­al­ity di­chotomy in­tro­duced in that ar­ti­cle). I’ll ex­plain how Google Fonts falls short, and what Google could do to im­prove the program.

Es­sen­tial qual­ity #1
Di­lu­tion: Open source arises from a spirit of free­dom and co­op­er­a­tion.
Re­al­ity: Open source arises from a spirit of cap­i­tal­ist competition.

It’s no co­in­ci­dence that Google launched its web­font project in May 2010, the same pe­riod that other com­pet­ing ser­vices (like Type­kit and Web­type) were also get­ting off the ground. It’s also no co­in­ci­dence that this ef­fort has emerged dur­ing Google’s cam­paign to be­come more plat­form-ori­ented rather than ap­pli­ca­tion-ori­ented, build­ing out An­droid, Chrome OS, and Google Docs. The Droid and Ro­boto font projects are the clear­est ex­am­ples of plat­form-ori­ented fonts.

So let’s agree that Google is pur­su­ing Google Fonts pri­mar­ily be­cause it’s in Google’s com­pet­i­tive in­ter­ests to do so. It’s not al­tru­ism. And these fonts are “free” only in the triv­ial sense that Google doesn’t charge us to use them. But Google al­ways finds other ways to con­vert our at­ten­tion into rev­enue, ei­ther di­rectly (sell­ing ads) or in­di­rectly (dis­trib­ut­ing open-source code). These busi­nesses are lu­cra­tive for Google be­cause our col­lec­tive at­ten­tion is eco­nom­i­cally valu­able. That’s why Google is a $200 bil­lion company.

My com­plaint with the Google Fonts project isn’t that Google is ben­e­fit­ing from it. There’s noth­ing wrong with mak­ing money from open source. Many com­pa­nies do. It’s that Google is try­ing to pre­tend oth­er­wise by fly­ing the flag of free­dom and sharing:

This is disin­gen­u­ous. Google doesn’t care about “bar­ri­ers to mak­ing great web­sites.” Google cares about Google. It’s also con­trary to Google’s usual pol­icy of not be­ing cagey about how it makes its money. When Google puts an ad on a page, there’s no guile about what it is or why it’s there. Why is Google play­ing its cards close to the vest in this in­stance? Which brings us to the next point:

Es­sen­tial qual­ity #2
Di­lu­tion: Open-source de­vel­op­ers work for free.
Re­al­ity: Open-source de­vel­op­ers are paid.

The idea that open-source con­trib­u­tors work for free is one of the most per­sis­tent and in­sid­i­ous myths of open source. In­sid­i­ous be­cause it’s the pre­ferred tool of com­pa­nies who want to har­vest the ben­e­fits of open source with­out as­sum­ing its burdens.

Here, it’s not in Google’s in­ter­est to re­veal the bor­ing re­al­ity—Google ben­e­fits from Google Fonts—be­cause it would dampen the en­thu­si­asm of the de­sign­ers who con­tribute fonts for lit­tle to no money.

In the de­sign world, there’s a well-known swin­dle where a pres­ti­gious but stingy client says “I wish I could pay you, but I don’t have the bud­get. How about you let me use your work for free? I know it’ll be great ex­po­sure for you, and lead to pay­ing work.” In truth, it’s not, and it won’t. Rather, it’s just good-na­tured grift­ing, an ex­ploita­tion of the weaker by the stronger. Nev­er­the­less, it works, be­cause there will al­ways be de­sign­ers hun­gry enough to be­lieve that they don’t have other choices.

Google Fonts uses a vari­a­tion of this toxic line as part of its pitch to po­ten­tial open-source font contributors:

Here, Google is us­ing the lure of ex­po­sure to “mil­lions of peo­ple” as an in­duce­ment to get de­sign­ers to con­tribute their time and their work for less than its mar­ket value, all for the plea­sure of be­ing an open-source contributor.

To be fair, over the past year, Google has been mak­ing more of an ef­fort to pay de­sign­ers. One Google Fonts con­trac­tor told me that they were of­fer­ing $500‑3000 per font. Let’s sup­pose Google pays you in the mid­dle of that range ($1750) for each style in a four-style font fam­ily, or $7000 to­tal. Let’s also sup­pose you spent three months on that project. Do you think that’s com­pa­ra­ble to what a web de­signer em­ployed by Google would get paid, in­clud­ing ben­e­fits, health care, stock op­tions, etc. in the same three-month pe­riod? Clearly not.

How can Google do bet­ter? One model is to pay mar­ket rates for type-de­sign ser­vices and then re­lease the re­sults un­der an open li­cense. This was the model for the Droid fonts, made by Mono­type Imag­ing. It would be an im­prove­ment, but not es­pe­cially open-source in spirit.

The bet­ter op­tion would be for Google to em­brace the open-source model more whole­heart­edly. Google doesn’t pay com­mu­nity en­gi­neers to work on open-source code. So Google also shouldn’t pay com­mu­nity type de­sign­ers to work on open-source fonts. In­stead, Google should pro­vide those type de­sign­ers the other ben­e­fits of work­ing on an open-source project. (Keep reading.)

Es­sen­tial qual­ity #3
Di­lu­tion: Open source makes things free.
Re­al­ity: Open source re­de­fines what is valuable.

In the open-source model, the point of mak­ing soft­ware code free is not to de­stroy the mar­ket­place for soft­ware, but rather to shift the value else­where. Mak­ing one thing free with­out mak­ing some­thing else more valu­able misses the point.

Google has like­wise missed the point by re­leas­ing hun­dreds of fonts for free with­out cre­at­ing cor­re­spond­ing value else­where. In that re­gard, Google Fonts is not part of the lin­eage of open-source projects, but rather file-shar­ing projects like 1001 Free Fonts, which are guided by the prin­ci­ple “Here’s a bunch of use­less crap that you’ll like be­cause it’s free.”

You might counter by point­ing out that the Google Fonts team has cre­ated new value by re­leas­ing open-source code for work­ing with web­fonts. True, but I con­sider the code-writ­ing part of Google Fonts to be dis­tinct from the font-cre­ation part. Google has a good han­dle on how to work on open-source soft­ware. What Google is miss­ing is a sim­i­lar ra­tio­nale for fonts.

How can Google do bet­ter? By putting for­ward a co­her­ent con­cept of how Google Fonts can re­de­fine value in the font mar­ket. The world al­ready has plenty of use­less free fonts. Google Fonts is blaz­ing no trails there.

Google has mul­ti­ple op­tions. I think that Google Fonts could be an in­ter­est­ing train­ing op­por­tu­nity for as­pir­ing type de­sign­ers who can’t drop every­thing and move to Berlin or New York (or other ur­ban cen­ter) to learn the trade. In this way, it would be sim­i­lar to how open-source soft­ware projects pro­vide op­por­tu­ni­ties for en­gi­neers with­out a strong ré­sumé to learn and con­tribute. But the value of this mech­a­nism de­pends en­tirely on the next point:

Es­sen­tial qual­ity #4
Di­lu­tion: Open source has no bar­ri­ers to par­tic­i­pa­tion.
Re­al­ity: Open source re­lies on meritocracies.

I al­luded to it above, but this is the mo­ment to fully con­front the un­avoid­able truth: mea­sured by pro­fes­sional stan­dards, the av­er­age Google Font is just aw­ful. Some are bet­ter than oth­ers, but nearly all fall prey to at least one fa­tal flaw of be­ing ugly, in­com­plete, poorly drawn, poorly spaced, am­a­teur­ish, or just un­us­able. And I don’t say that to crit­i­cize the de­sign­ers them­selves. They’re en­thu­si­as­tic about type de­sign. But that en­thu­si­asm should be chan­neled into im­prov­ing their skills and mak­ing bet­ter fonts. Right now, it’s not. And Google isn’t helping.

If you think it’s un­fair to com­pare Google Fonts to pro­fes­sional fonts, sorry, but that’s the open-source way. (See also Es­sen­tial Qual­ity #1.) The open-source op­tion is only rel­e­vant if it can com­pete with the qual­ity of the pro­pri­etary ver­sion. Does Google re­lease tons of low-qual­ity open-source code into the world? No. It lives up to pro­fes­sional open-source stan­dards. Google should adopt anal­o­gous stan­dards for fonts.

How can Google do bet­ter? This one’s easy: fewer fonts, higher qual­ity. The need for new fonts is never in doubt. Time passes. Tech­nol­ogy evolves. Re­quire­ments change. In the last gen­er­a­tion, com­pa­nies like Ap­ple and Mi­crosoft in­vested large sums in mak­ing the sys­tem fonts that mil­lions of peo­ple have re­lied on for 20 years. Google has the op­por­tu­nity to step into that role, if it chooses.

But mak­ing qual­ity fonts re­quires qual­ity de­sign­ers. Open source does not mean “open to all.” It means “open to all who can work to the nec­es­sary stan­dard.” If Google wants qual­ity, Google Fonts can­not re­main open to any­one who wants to give a font away for free.

Es­sen­tial qual­ity #5
Di­lu­tion: Open source is de­mo­c­ra­tic.
Re­al­ity: Open source re­lies on benev­o­lent dictators.

This one’s also easy: as­sum­ing that Google de­vel­oped fewer fonts, each project should be led by an ex­pe­ri­enced, re­spected, pro­fes­sional type de­signer. This would be anal­o­gous to open-source soft­ware, where projects are led by ex­pe­ri­enced, re­spected, pro­fes­sional soft­ware en­gi­neers. Google should pay these pro­fes­sional type de­sign­ers mar­ket rates to as­sume these roles. They would be­come the ar­biters of which con­tri­bu­tions get in­cluded and which don’t.

Three results:

1. Google would get pro­fes­sional-qual­ity fonts at a lower cost than pro­pri­etary development.

2. Com­mu­nity type de­sign­ers would get to work with pro­fes­sional type de­sign­ers (which is the best way of im­prov­ing skills).

3. Google would be re­leas­ing gen­uine open-source as­sets into the world, that would be good enough to in­spire more de­vel­op­ment. For in­stance, the freely avail­able font Char­ter—de­signed by the es­teemed Matthew Carter in 1987—be­came the ba­sis of Charis SIL. Have any cur­rent Google Fonts been sim­i­larly adapted?

“But un­der this scheme, far fewer de­sign­ers will have their con­tri­bu­tions used.” Yes. That’s the point. Benev­o­lent dic­ta­tors get to pick the wor­thi­est con­tri­bu­tions. Open source is not open-mic night. Those who want to re­lease a font for free will al­ways have plenty of op­tions. It’s not true that any­one is en­ti­tled to par­tic­i­pate in an open-source project. If you dis­agree, please get me Guido van Rossum on the phone, be­cause I want him to add Klin­gon com­mands to the core syn­tax of Python. And an ASCII uni­corn to every source file.

Es­sen­tial qual­ity #6
Di­lu­tion: An open-source project can have one de­vel­oper.
Re­al­ity: An open-source project re­quires mul­ti­ple developers.

Most Google Fonts are the work of one to three de­sign­ers, work­ing in iso­la­tion from other de­sign­ers. Maybe Google does some rudi­men­tary qual­ity con­trol—it’s hard to tell from the re­sults. But in most cases, the font is done when the de­signer says it’s done. The font does not have to meet any ex­ter­nal standards.

With mul­ti­ple de­sign­ers work­ing un­der a benev­o­lent dic­ta­tor, there would be a com­pe­ti­tion of de­sign ideas. Type de­sign­ers who were lazy or care­less would quickly find that there was no room for their work. They would ei­ther im­prove their work, thereby be­com­ing bet­ter de­sign­ers, or re­treat. Ei­ther way, the project would benefit.

Es­sen­tial qual­ity #7
Di­lu­tion: A soft­ware project can be open-sourced at any time.
Re­al­ity: Open source is part of the project’s DNA or it’s not.

In the­ory, there’s no rea­son open-source fonts can’t ex­ist, and can’t be good. I’ll as­sume that Google sin­cerely wants to make fonts that have open source in their DNA, and not just “An­other 1001 Free Fonts.”

But if that’s the case, Google has to change its ap­proach. I said be­fore that “Google has a great en­gi­neer­ing cul­ture, a weak de­sign cul­ture, and no dis­cernible taste.” I stand by that. To get dif­fer­ent re­sults, Google will need to ap­proach open-source fonts in a way that plays to its strengths (en­gi­neer­ing) and avoids its weak­nesses (de­sign and taste). Fo­cus­ing on a smaller num­ber of fonts and hir­ing benev­o­lent dic­ta­tors from the pro­fes­sional type-de­sign in­dus­try would be good first steps. Even Mi­crosoft was able to over­come its taste deficits to make Ver­dana, which is now in the Mu­seum of Mod­ern Art.

More broadly, Google should con­sider that its in­ter­ests as a par­tic­i­pant in the font world are par­al­lel to its in­ter­ests in the soft­ware world. Re­gard­less of whether your project is open or pro­pri­etary, the re­sults de­pend on the tal­ent of the peo­ple build­ing it. De­vel­op­ing the tal­ent pool al­ways pays div­i­dends. But that can’t hap­pen by merely mak­ing fonts free. It re­quires a more thor­ough and thought­ful ap­proach. Google is pos­si­bly ca­pa­ble of that.

It’s also pos­si­ble that Google, with its cul­tural bias to­ward en­gi­neer­ing, sim­ply doesn’t ac­knowl­edge that type-de­sign skill ex­ists and has value. If that’s so, Google Fonts will re­main the Costco of ty­pog­ra­phy: al­ways get­ting big­ger, never get­ting bet­ter. If that ends up be­ing in Google’s best in­ter­ests, fine. But please, Google—don’t call it open source. You know bet­ter. So do we.

First, to give Google some credit: I be­lieve the Google Fonts project has been im­por­tant in get­ting web users & de­vel­op­ers com­fort­able with web­fonts. When this ar­ti­cle was first writ­ten, most ma­jor com­mer­cial web­sites were still us­ing com­bi­na­tions of Geor­gia and Ar­ial. Web­fonts were rare. To­day, the ra­tio has re­versed. Web­fonts have be­come part of the main­stream web.

Are a lot of the web­fonts we see to­day cruddy free fonts? Yes. And are many of those from Google Fonts? Also yes. But as a type de­signer, I would much rather have peo­ple us­ing cruddy free fonts than sys­tem fonts. Why? Be­cause it re­quires them to over­come their rest­ing in­er­tia—“why should I use web­fonts when I’ve got Geor­gia?” That first step is a big one. Even­tu­ally, some of these peo­ple will get sick of their cruddy web­fonts, and then they’ll be in the mar­ket for bet­ter ones. Which they won’t find at Google Fonts.

Like many of to­day’s tech com­pa­nies, Google likes to por­tray its most mediocre ac­com­plish­ments as works of rev­o­lu­tion­ary ge­nius. (PR peo­ple are cheaper than en­gi­neers, af­ter all.) My con­cern dur­ing the Google Fonts surge in 2012 was that Google might try to re­de­fine the font mar­ket around its own low stan­dards, much as Ama­zon has de­fined the e-book mar­ket with the dread­ful Kindle.

For­tu­nately, that didn’t hap­pen. The Kin­dle is a key project for Ama­zon. By con­trast, fonts have al­ways been a mi­nor sideshow for Google. Over the next few years, Google Fonts con­tin­ued to pump out dozens and dozens of pretty bad fonts. (It wasn’t un­com­mon to see ar­ti­cles with ti­tles like “Google Web­fonts that Don’t Suck”.)

Maybe the worst thing that hap­pened to Google Fonts was that their core premise—namely, that quan­tity mat­ters more than qual­ity—was de­bunked. In 2012, Adobe re­leased Source Sans and Source Code, two open source fonts that were ac­tu­ally quite good. The next year, Mozilla re­leased Fira, which I thought was one of the best fonts of 2013.

Still, it re­mains un­clear whether a font can be mean­ing­fully “open source” in a sense be­yond hav­ing an open li­cense. At a pre­sen­ta­tion at Type­Con 2013 af­ter the re­lease of Source Sans and Source Code, Adobe’s de­sign­ers mourn­fully won­dered “why aren’t more peo­ple con­tribut­ing?” I was there—on the best day, elic­it­ing sym­pa­thy for a $45 bil­lion com­pany is dif­fi­cult. But the sys­temic an­swer might be that the au­di­ence for most open-source soft­ware tools is other soft­ware de­vel­op­ers, who at least have the ca­pac­ity to con­tribute. Whereas with fonts, the au­di­ence of end users in­cludes few (if any) type designers.

At some point, I was told that my cri­tique of Google Fonts had led to changes in their de­sign process to be more open and in­clu­sive. Yeah, right, what­ever. Look, Google is a se­cre­tive cor­po­rate be­he­moth. Google’s en­gage­ment with open source is partly a the­atri­cal ges­ture, as cyn­i­cal as its candy-col­ored geo­met­ric logo. The point of dis­cussing Google Fonts was not to help Google. (As if they care what I think.) Rather, it was to en­cour­age type de­sign­ers who were think­ing about par­tic­i­pat­ing in the project to crit­i­cally re­flect on the costs & ben­e­fits. Be­cause ul­ti­mately, Google is into open source like Mc­Don­ald’s is into an­i­mal welfare.

—Matthew But­t­er­ick
Jan­u­ary 2012 & No­vem­ber 2015