Talk giv­en by Matthew But­t­er­ick
at TYPO San Fran­cis­co, 11 April 2013
mb@mb­type.com

The Bomb in the Garden

Pho­to © 2013 Am­ber Gre­go­ry

It’s now or nev­er for the web. The web is a medi­um for cre­ators, in­clud­ing de­sign­ers. But af­ter 20 years, the web still has no cul­ture of de­sign ex­cel­lence. Why is that? Be­cause de­sign ex­cel­lence is in­hib­it­ed by two struc­tur­al flaws in the web. First flaw: the web is good at mak­ing in­for­ma­tion free, but ter­ri­ble at mak­ing it ex­pen­sive. So the web has had to rely large­ly on an ad­ver­tis­ing econ­o­my, which is weak­en­ing un­der the strain. Sec­ond flaw: the process of adopt­ing and en­forc­ing web stan­dards, as led by the W3C, is hope­less­ly bro­ken. Ev­i­dence of both these flaws can be seen in a) the low de­sign qual­i­ty across the web, and b) the speed with which pub­lish­ers, de­vel­op­ers, and read­ers are mi­grat­ing away from the web, and to­ward app plat­forms and me­dia plat­forms. This ev­i­dence strong­ly sug­gests that the web is on its way to be­com­ing a sec­ond-class plat­form. To ad­dress these flaws, I pro­pose that the W3C be dis­band­ed, and that the lead­er­ship of the web be re­or­ga­nized around open-source soft­ware prin­ci­ples. I also en­cour­age de­sign­ers to ad­vo­cate for a bet­ter web, lest they find them­selves con­fined to a shrink­ing ter­ri­to­ry of pos­si­bil­i­ties.


Thank you to the TYPO con­fer­ence for hav­ing me. It feels like I’ve sort of com­plet­ed a cir­cuit, be­cause I spoke at the first TYPO con­fer­ence, back in 1995, in Berlin.

At the time, I had just moved to San Fran­cis­co to be part of the first wave of web de­sign­ers. If that sounds like an ob­vi­ous thing to do, it wasn’t, be­cause un­like to­day, there weren’t re­al­ly jobs for web de­sign­ers. The busi­ness plan was sort of let’s show up and see what hap­pens. And re­al­ly, to most of my friends and fam­i­ly, the web was kind of weird and mys­te­ri­ous. They said to me, in the nicest pos­si­ble way, “But­t­er­ick, you’re an id­iot. Why would you do this?”

I was a type de­sign­er at the time. But I felt strong­ly that I need­ed to go work on the web. For two rea­sons.

The first rea­son: be­cause the web was pri­mar­i­ly ty­pog­ra­phy. It was ugly-as-fuck ty­pog­ra­phy. But it was ty­pog­ra­phy. These are ac­tu­al slides from my 1995 pre­sen­ta­tion. (I held onto them be­cause I knew I’d need them this year.) But I could see that the web was go­ing to need ty­pog­ra­phers. Still does.

The oth­er thing is that the web was re­al­ly a blank slate. No one re­al­ly knew what to do with it. We didn’t have trends then like mo­bile or so­cial, these grooves for peo­ple to trav­el in. So the web was open to any cre­ative peo­ple who could look at this thing that was void and with­out form and see the pos­si­bil­i­ties.

So now it’s 2013. Now the web is ac­tu­al­ly 20 years old. And ob­vi­ous­ly, in a lot of ways, the web has been suc­cess­ful.

But here’s what I’ve been think­ing about re­cent­ly. I’ve been think­ing about how we’ve en­tered this phase where the cre­ative spir­it that’s al­ways been es­sen­tial to the core of the web is start­ing to get pushed out. And that’s what I’m go­ing to talk about to­day.

And my main goal to­day is not to con­vince you of every point I’m go­ing to make. Lord knows, you’re go­ing to dis­agree with a few. Rather, my goal is to get you to take this prob­lem se­ri­ous­ly, as de­sign­ers. Be­cause as de­sign­ers, this prob­lem af­fects all of us. It af­fects the work that we’re al­lowed to do with the web. It af­fects our ca­reers. And it’s not some­one else’s prob­lem to solve. It’s our prob­lem. And it’s our prob­lem now. So I feel like we should all be think­ing about how we want to fix it. And it’s a hard prob­lem. But de­sign­ers, for the past 20 years, have al­ways had an im­por­tant role in both crit­i­ciz­ing the web, and mov­ing it for­ward. So we should feel like it’s some­thing we can do.

So in fact, let’s start with that idea:

It’s difficult, but necessary, to criticize the technology industry.

It’s strange­ly dif­fi­cult, right? In this coun­try, you can pick on the en­er­gy in­dus­try, you can pick on biotech, health care—god knows you can pick on lawyers. But when you start say­ing things about tech­nol­o­gy com­pa­nies, it’s like you punched a baby in the face. Peo­ple don’t like that. Why is that? Why does tech­nol­o­gy get more slack?

In one sense, I feel like maybe it’s be­cause here in the US, we’ve al­ways been much bet­ter at cre­at­ing new things than pre­serv­ing the old ones. So we have to tell our­selves this myth about the in­fal­li­bil­i­ty of tech­nol­o­gy and progress. Be­cause it plays to our strengths. That’s what we’re good at.

But the tech­nol­o­gy in­dus­try in par­tic­u­lar re­al­ly spins this myth into this self-serv­ing idea that con­sum­ing tech­nol­o­gy has no con­se­quences. We don’t usu­al­ly fall for that. When an Amer­i­can cor­po­ra­tion tells you “hey, just have all you want of our prod­uct, please, go ahead”—do good things usu­al­ly hap­pen? No, nev­er.

For in­stance, Mc­Don­ald’s. Their food looks de­li­cious. Their food is de­li­cious. Their food is sci­en­tif­i­cal­ly en­gi­neered to be de­li­cious. But we know that if we eat all the Mc­Don­ald’s food we want, we’re go­ing to get a big fat ass. So we have a sense of the con­se­quences.

Where­as with tech­nol­o­gy, I don’t think we’re as skep­ti­cal. But we should be. Be­cause tech­nol­o­gy com­pa­nies sure love to hitch a ride on these hu­man virtues. Look at what Ap­ple’s done here. They’ve tak­en this Mc­Don­ald’s slo­gan and they’ve made it even creepi­er. In­stead of I’m lovin’ it, it’s iPhone 5. Lov­ing it is easy. That’s why so many peo­ple do.

I’m think­ing—[hold­ing up iPhone]—I’m go­ing to have to break up with this iPhone. Be­cause it’s a lit­tle more emo­tion­al­ly needy than I was look­ing for. [Ad­dress­ing iPhone:] And now I know about you and all the oth­er peo­ple! Get out of here! [iPhone is thrown off stage.]

Or think about Face­book and its mis­sion state­ment. Face­book’s mis­sion—I hope you know—is “to make the world more open and con­nect­ed.” Did you ever won­der what Face­book means by “open and con­nect­ed”? Well, they’ve told us.

They mean “big­ger, more me­dia-rich ad for­mats.” They also mean “re­al­ly rich things like big pic­tures.” I don’t know about you, but when I think about “open and con­nect­ed,” I don’t think about re­al­ly big ads. But that’s what they mean. So we should be crit­i­cal about what these com­pa­nies want from us, and what they’re of­fer­ing in re­turn.

Es­pe­cial­ly as de­sign­ers. Be­cause part of our job is to tell the truth. If you think that your job is to put blue rec­tan­gles to­geth­er with Hel­veti­ca, that’s cool. But to me, that’s not enough. To me,

Solving problems is the lowest form of design.

Be­cause de­sign wants more from us. It wants our hu­man­i­ty. It wants our op­ti­mism. It wants our hon­esty. It wants our ideas for what a bet­ter world looks like. Some days, those are small ideas. Some days, those are big ideas. And the web in­cludes both kinds.

I’ve some­times heard that it’s too soon to crit­i­cize the web. “We need pa­tience. It’s get­ting bet­ter. These things take time.” Here’s the thing: I re­al­ly do feel like—

After 20 years of patience, more patience is not the answer.

And not only has the web got­ten 20 years, it’s got­ten bil­lions and bil­lions of dol­lars of in­vest­ment cap­i­tal and hu­man time. If any tech­nol­o­gy prod­uct should be sub­ject to high ex­pec­ta­tions, it’s the web.

And any­how, what’s wrong with im­pa­tience? We’re im­pa­tient about all kinds of oth­er tech­nol­o­gy. We’ve got to stand in line, out­side the big cube, for the iPhone. So to me, I feel like, geez—the web is a more con­se­quen­tial tech­nol­o­gy than the iPhone. We should re­al­ly be more im­pa­tient, not less.

Call me a ro­man­tic, but I take it as a giv­en that—

The web is primarily a medium for creators.

That’s what orig­i­nal­ly drew me to the web. It’s what orig­i­nal­ly drew many oth­ers to the web.

But it’s also part of the orig­i­nal mis­sion of the web. Here’s the very first web page, which was post­ed by Mr. Tim Bern­ers-Lee, who’s cred­it­ed with be­ing the in­ven­tor of the web, over at CERN in Switzer­land. Look how he de­scribed the web: as an “in­for­ma­tion re­trieval ini­tia­tive aim­ing to give uni­ver­sal ac­cess to a large uni­verse of doc­u­ments.”

Ob­vi­ous­ly tech­nolo­gies go on to have dif­fer­ent uses than their in­ven­tors ex­pect. But I think that prin­ci­ple still has cur­ren­cy to­day. And maybe that’s just my bias. Be­cause I think we should have “uni­ver­sal ac­cess to a large uni­verse of doc­u­ments.” I think we should low­er the bar­ri­ers to cre­at­ing, low­er the bar­ri­ers to learn­ing. Be­cause every time we do, we win. Hu­man­i­ty gets smarter. Our lives get bet­ter. And that’s why the writ­ten word is by far the most con­se­quen­tial hu­man in­ven­tion. It’s been the key to all these oth­er im­prove­ments in our cul­ture. And the web is the next man­i­fes­ta­tion of the writ­ten word. But more than that, it re­al­ly should be the best man­i­fes­ta­tion. It’s got so much go­ing for it.

And that comes back to the oth­er thing that I said drew me to the web: that it was a ty­po­graph­ic medi­um. It was true then. It’s true now. Much of the web is made of ty­pog­ra­phy and de­sign. But de­spite this, the web isn’t the best man­i­fes­ta­tion of the writ­ten word. In fact, quite the op­po­site: the web re­al­ly doesn’t have a cul­ture of de­sign ex­cel­lence.

That’s not to say that de­sign ex­cel­lence doesn’t ex­ist in places. It does. Here’s Frank Chimero’s on­line book, The Shape of De­sign. I think this is pret­ty much per­fect, as a web-based book goes. But it’s an out­lier. A de­sign cul­ture doesn’t mean out­liers. A cul­ture is a habit. A tra­di­tion. A stan­dard of qual­i­ty. There aren’t that many ex­am­ples like this.

Now, you may say “hey, but the web has got­ten so much bet­ter look­ing over 20 years.” And that’s true. But on the oth­er hand, I don’t re­al­ly feel like that’s the right bench­mark, un­less you think that the high­est role of de­sign is to make things pret­ty. I don’t.

I think of de­sign ex­cel­lence as a prin­ci­ple. A prin­ci­ple that asks this:

Are you maximizing the possibilities of the medium?

That’s what it should mean. Be­cause oth­er­wise it’s too easy to con­grat­u­late our­selves for do­ing noth­ing. Be­cause tools & tech­nolo­gies are al­ways get­ting bet­ter. They ex­pand the pos­si­bil­i­ties for us. So we have to ask our­selves: are we keep­ing up?

You can de­cide for your­self. Let’s go to But­t­er­ick’s first an­nu­al quest for de­sign ex­cel­lence on the web.

I’m go­ing to show you a few web­sites. Be­cause I was think­ing to my­self: if I want­ed to find some de­sign ex­cel­lence on the web, where should I go? And I thought, what’s a type of or­ga­ni­za­tion that has a com­mit­ment to de­sign, and de­sign­ers? And I thought, well, news­pa­pers do. A lot of the world’s news­pa­pers are very well de­signed pub­li­ca­tions. We know they have de­sign­ers on staff. We know they have bud­get for de­sign. OK, so let’s go look at their web­sites.

That’s the Los An­ge­les Times. My lo­cal news­pa­per.

The San Fran­cis­co Chron­i­cle. The SF Gate. That’s pret­ty bad.

The Wash­ing­ton Post. I want you to no­tice one big thing here. Not only are all these web­sites aw­ful, they’re all aw­ful in ex­act­ly the same way. We’ve just seen three sites that pret­ty much have the same pat­tern: a clut­tered tool­bar on the top, and then they’ve got the ban­ner ad, they’ve got the ar­ti­cle on the left side, and then they’ve got more ads and links and oth­er junk in the right col­umn.

Let’s keep go­ing. The New York Times. Right? I think we’ve seen this.

The Guardian, in the UK. A lit­tle vari­a­tion there, a few more links, but still ba­si­cal­ly the same.

Le Monde, from France.

And fi­nal­ly, Die Zeit.

A friend of mine said “But­t­er­ick, you’re be­ing sil­ly. Be­cause news­pa­pers, they’re di­nosaurs. They’re go­ing out of busi­ness. Why would you look to them for de­sign ex­cel­lence?” And I said OK. So who would have mon­ey for de­sign ex­cel­lence? Be­cause news­pa­pers don’t. I thought, I’m go­ing to look at the most pop­u­lar mag­a­zines in Amer­i­ca by cir­cu­la­tion. Be­cause I fig­ure they must have the mon­ey to do some­thing pret­ty good, right?

All right, so there’s the AARP Jour­nal. Do you feel like this is fa­mil­iar at all?

Game In­former. Did you know these were the most pop­u­lar mag­a­zines in Amer­i­ca? Game In­former, to­tal­ly aw­ful.

Bet­ter Homes & Gar­dens. I mean, this is—I don’t even know what that’s about.

Read­er’s Di­gest. Look, they even use a cou­ple web­fonts. But ba­si­cal­ly the lay­out is the same.

Good House­keep­ing. Oh my god.

Fam­i­ly Cir­cle. You had no idea. I mean, this is just—look at all the ads on these pages, it’s as­ton­ish­ing.

Na­tion­al Ge­o­graph­ic. This is amaz­ing. They’ve ac­tu­al­ly tripled the num­ber of tool­bars they fit into the top. And they’ve got not just one logo in the cor­ner—they’ve got two lo­gos! That is so web­tas­tic! You’ve got to have the logo in the up­per left cor­ner. I’d like to have my name dan­gling around the up­per left of my head as I walk through life: But­t­er­ick.

Peo­ple mag­a­zine. Aw­ful. Though con­grat­u­la­tions, Anna Chlum­sky.

Woman’s Day.

Time. Even Time. Some nice peo­ple have worked on Time. The web­site is shit.

All right, so you might say I’m be­ing fool­ish here. Those are just pop­u­lar mag­a­zines. What do they know about de­sign? And so I thought, how about this. I’ll look at the mag­a­zines that were nom­i­nat­ed for Na­tion­al Mag­a­zine Awards in the de­sign cat­e­go­ry in the last five years. What do you think we’ll find?

Hmm. Bloomberg Busi­ness­week.

GQ. Oh my god.

In­ter­view. OK, they’re shak­ing it up a lit­tle bit. Still pret­ty bad.

Wired. My god, Wired. It looks like Game In­former. I mean, what’s go­ing on here?

New York mag­a­zine.

OH JE­SUS! That’s hideous! What is that? Es­quire? These are ac­tu­al­ly good look­ing pub­li­ca­tions, a lot of them. You’ve seen them. But what the hell hap­pens when they go on the web? They just lose their minds. They give up.

And Good mag­a­zine. Again, all these print pub­li­ca­tions were nom­i­nat­ed for de­sign awards. But they go on the web and they do this.

So I say to my­self, you know what—I’m ex­pect­ing too much. What if I looked at pub­li­ca­tions about de­sign for de­sign­ers? That must be a place where I could find some de­sign ex­cel­lence.

Whoa, OK. Nice Hel­veti­ca there. Thank you, Eye mag­a­zine.

And the AIGA. That’s again, pret­ty dull. These are go­ing to go from best to worst. That was the high mo­ment. Eye and AIGA. Now it’s just go­ing to get more and more bor­ing.



Ugh. It’s like look­ing at blogs.

I’m sor­ry, De­sign Ob­serv­er. I like the con­tent on your site. But this was the ty­pog­ra­phy of 1998. You’re us­ing 6 point Ver­dana?

Com­mu­ni­ca­tion Arts. Sor­ry, con­fer­ence spon­sor, I don’t like your web­site. It’s no good.

Why is this so dis­ap­point­ing? It’s dis­ap­point­ing be­cause we want to like these pub­li­ca­tions and these or­ga­ni­za­tions. But they’re hurt­ing de­sign! They’re hurt­ing de­sign, be­cause they’re set­ting a bad ex­am­ple. They’re just fol­low­ing what every­one else is do­ing.

And I even found out­side, be­fore this speech, this mag­a­zine. This is the Com­mu­ni­ca­tion Arts ty­pog­ra­phy an­nu­al. I’m not go­ing to pick on them. I just want to note, though, in the pur­suit of ex­cel­lence—16 type­faces in here are sin­gled out for recog­ni­tion. And do you know how many web­sites are sin­gled out for ty­po­graph­ic recog­ni­tion? One. One web­site. Af­ter 20 years, one web­site gets to be in the an­nu­al. Pret­ty sad.

That’s my ev­i­dence. I’d be hap­py to look at yours.

So, the cul­ture of de­sign ex­cel­lence. Why don’t we have one? And here, I think, is where the plot thick­ens. Be­cause I’m go­ing to con­tend to you that there are two se­ri­ous struc­tur­al flaws in the web. And here’s the biggest one:

The web is much better at making information free than making it expensive.

Some­one’s al­ready tweet­ing—“But­t­er­ick is an id­iot. He doesn’t know that in­for­ma­tion wants to be free.” You know, I have heard that. But I also know that 99.99% of peo­ple who men­tion this line for­get to talk about the first and last parts of it.

“What? There’s a first and last part?” Yeah, yeah. The whole line goes like this:

“In­for­ma­tion wants to be ex­pen­sive, be­cause it’s so valu­able … On the oth­er hand, in­for­ma­tion wants to be free, be­cause the cost of get­ting it out is get­ting low­er … So you have these two fight­ing against each oth­er.”

This was said by a guy named Stew­art Brand, way back in 1984.

So what’s the mes­sage here? In­for­ma­tion wants to be free? No, that’s not the mes­sage. The mes­sage is that there are two forces in ten­sion. And the chal­lenge is how to bal­ance the forces.

The web, how­ev­er, has nev­er bal­anced these forces. The web has al­ways been great for mak­ing in­for­ma­tion free, and ter­ri­ble at charg­ing for it. And that’s a tech­no­log­i­cal flaw that’s ex­ist­ed since the be­gin­ning of the web. So from ear­ly on, folks took this lemon and tried to make web lemon­ade by latch­ing onto the be­lief that ex­po­sure mat­tered more than mon­ey.

Does this sound fa­mil­iar, de­sign­ers? “This project’s go­ing to be great ex­po­sure.” It’s nev­er true. It’s nev­er been true. It’s nev­er been true on the web. But it be­came one of the web’s core re­li­gious be­liefs.

The prob­lem, of course, is that in­for­ma­tion wasn’t ac­tu­al­ly free. It’s just that no one want­ed to pay for it. But some­body had to. So we’ve end­ed up with what?

We’ve ended up with a web dominated by advertising.

Again, I know that pay­walls ex­ist. There’s like three or four of them that ac­tu­al­ly work. Those are out­liers. Here are the 53 top US sites, as ranked by Alexa. (I left off the porn.) How many of these sites do you think rely on ad­ver­tis­ing rev­enue?

The an­swer: 80% of them.

And it’s not that all ad­ver­tis­ing is bad. As with that quo­ta­tion we saw be­fore—it’s a ques­tion of bal­ance. Be­cause ad­ver­tis­ing isn’t cre­ation. It’s mar­ket­ing. It’s sell­ing. And it’s not in­com­pat­i­ble with the web. I can be­lieve that ad­ver­tis­ing has a role to play in sub­si­diz­ing cer­tain kinds of web con­tent.

But ad­ver­tis­ing de­pends on a healthy cre­ator habi­tat. Cre­ators have to be mak­ing stuff that peo­ple want to pay at­ten­tion to. And if you de­stroy the cre­ator habi­tat, what hap­pens to ad­ver­tis­ing? You de­stroy that too. Then there’s noth­ing. So even if you like ad­ver­tis­ing, it’s still re­al­ly im­por­tant for the web to take care of the cre­ators.

More­over, when so much of the web de­pends on ads, you also start to get un­in­tend­ed con­se­quences. Like city news­pa­pers—your dear old San Fran­cis­co Chron­i­cle. Call them di­nosaurs. They are. Say that they de­serve to go out of busi­ness. Maybe they do. But news­pa­pers like this have tra­di­tion­al­ly per­formed an im­por­tant watch­dog func­tion, keep­ing lo­cal busi­ness­es and politi­cians hon­est. Be­cause that’s part of the cul­ture of news jour­nal­ism. It’s not part of the cul­ture of the Huff­in­g­ton Freak­ing Post. And as these news­rooms die, we’ll in­evitably—it’s like a law-of-grav­i­ty guar­an­tee—we’ll in­evitably see a rise in pri­vate & pub­lic cor­rup­tion. And when that hap­pens, in a very real way, we’ll have trad­ed good gov­ern­ment for ban­ner ads.

And it’s hap­pen­ing here. Look at this.

The Chron­i­cle’s cir­cu­la­tion is also off about half in re­cent his­to­ry. I mean, that pa­per is go­ing down the drain. And then, you start to see things like this—

Isn’t that crazy? A tech­nol­o­gy com­pa­ny’s be­ing paid with tax dol­lars to stay in San Fran­cis­co? That just bog­gles my mind. That’s a tens-of-mil­lions of dol­lars ben­e­fit to them. So we’re just go­ing to see more of this.

We might find a tiny sil­ver lin­ing in all this if the web’s ad­ver­tis­ing econ­o­my were ac­tu­al­ly healthy. But it’s not. In fact,

The web’s advertising economy is weakening.

The prob­lem is that as every­one has bet on ads—80% of these sites, right?—ads have got­ten less valu­able. As ads have got­ten less valu­able, you have to gen­er­ate more click­ing and more ads.

That’s why every­thing is a slideshow now! You know how it works: You’re do­ing your job, and you come across a link like this:

It’s like, oh shit—I have to click on it. I like how the link “goes vi­ral”—don’t be the last per­son to not make it go vi­ral! You’ve got to click on it. And it comes out of some kind of crazy link gen­er­a­tor that knows what peo­ple will click on. This one doesn’t even have biki­nis in it.

But of course you click on it.

And oh god, there’s the cute dog. I mean, they’re giv­ing you what you want­ed: the dog, and the bird, and the bun­nies.

But then you scroll down, and it’s like oh no, there’s a 73-pic­ture slideshow. There’s videos of danc­ing dogs. How can you say no? You’re there.

And this, my friends—this is the the death spi­ral. It’s an adorably cute death spi­ral. But a death spi­ral nev­er­the­less. Be­cause the val­ue of this web page is not de­ter­mined by what’s on it, but rather its abil­i­ty to at­tract clicks from across the web, and then gen­er­ate some more clicks. And that’s true of the next page in line. And the next one. And the next one. So the in­cen­tive is not re­al­ly to hold read­er at­ten­tion, but to man­u­fac­ture dis­trac­tion.

And this does cre­ate a real prob­lem for de­sign­ers. Be­cause—

The advertising economics of the web inhibit design excellence.

Be­cause a web page view is like any oth­er prod­uct. You have to spend mon­ey to man­u­fac­ture it—you have to build the site, main­tain the servers, and so on. And then you can make mon­ey from it—for in­stance, by putting ads on it. But you have to make more mon­ey than you spend, oth­er­wise you’ll go broke.

So one of the biggest costs of a web page is the cost of what? Cre­at­ing & im­ple­ment­ing the de­sign. So as ad rev­enue goes down on the web, all the costs have to come down. And that in­cludes the cost of de­sign.

Go back to the dogs. The only real de­sign con­sid­er­a­tion on this page is how you get the read­er to click again. If you’re the Huff­in­g­ton Post, how much will you spend on de­sign that makes the con­tent read­able and ap­peal­ing? You’re go­ing to spend as lit­tle as pos­si­ble. Be­cause it’s ir­rel­e­vant. This page, and so many oth­ers like it, are just ad-de­liv­ery ve­hi­cles.

So, de­sign­ers—I know, your head’s spin­ning—we still might be able to man­age de­sign ex­cel­lence, if it were pos­si­ble to cre­ate ex­cel­lent web de­sign cheap­ly. Be­cause we’re go­ing to have to do it cheap­ly, with the ad­ver­tis­ing push­ing costs down. This was sup­posed to be one of the virtues, I think, of web stan­dards. Right? But as any­one who has de­vel­oped big web­sites with web stan­dards finds out, you dis­cov­er a ter­ri­ble truth—

Web standards aren’t standard.

I was say­ing be­fore that the web has these two big struc­tur­al flaws. The first, that it’s bad at mak­ing in­for­ma­tion ex­pen­sive. But this is the sec­ond.

And I know—some­body’s al­ready tweet­ing: “But­t­er­ick’s an id­iot. He doesn’t like web stan­dards.” No. I like stan­dards. I think the web should have stan­dards. I even think the web stan­dards that we have are bet­ter than none at all. Be­cause they do make easy things easy.

But as a de­sign­er, my job is not to do easy things. My job is to do hard things. And that’s where web stan­dards cre­ate mis­ery.

And the mis­ery ex­ists be­cause of the W3C—the World Wide Web Con­sor­tium. That’s the or­ga­ni­za­tion that su­per­vis­es web stan­dards. They’ve been so slow over the last 20 years at adopt­ing stan­dards. And way too le­nient in en­forc­ing them. And so we have these prob­lems get­ting cre­at­ed at the stan­dards lay­er, and they just kind of flow down­stream, like tox­ic waste, down to the HTML and CSS lay­er. Where de­sign­ers work. And then it be­comes our prob­lem.

For in­stance, what you’re look­ing at here. I re­cent­ly learned on a project that this is the cur­rent pre­ferred way of cre­at­ing col­or gra­di­ents in CSS. Web stan­dards were sup­posed to elim­i­nate the com­plex­i­ty of brows­er-spe­cif­ic markup. But in prac­tice, you still need to use all these crazy hacks and workarounds and overtag­ging to get re­sults that are ac­tu­al­ly stan­dard. Here, you’ve got to write a line for sev­en dif­fer­ent browsers.

Stan­dards were also sup­posed to re­duce test­ing com­plex­i­ty. For in­stance, when you make a PDF, does it even oc­cur to you that you should test it on five dif­fer­ent plat­forms and 10 dif­fer­ent PDF read­ers? No. Be­cause PDF is an ISO stan­dard. It just works. Where­as for web­sites, you nev­er re­al­ly fin­ish test­ing. You just give up.

So this brings us to the big con­se­quence, which is that—

Nonstandard web standards also inhibit design excellence.

Be­cause for us, as de­sign­ers, these non­stan­dard stan­dards—they lim­it the things we can ac­com­plish. And they also waste our time, by mak­ing us fix all this stuff that was nev­er sup­posed to be our prob­lem in the first place.

Now, I can get an­gry about this, and some­one’s al­ways like “yeah, but But­t­er­ick, that’s just the web, man. That’s just the way things are.” And I say, you know what? I think it’s time to lose that at­ti­tude. For one big rea­son:

The web is in danger of becoming a second-class platform.

Be­cause over the last few years, it’s been rapid­ly los­ing ground to app plat­forms like iOS and An­droid. Los­ing ground to me­dia plat­forms, like the Kin­dle.

For in­stance, if you ever use your iPad to browse the web, you’re see­ing more of this. Are you see­ing these? It’s like “you can’t use our web­site, but you can see our ad for the app.” And I like how all these ads are show­ing you an iPad that’s hav­ing a much bet­ter time than your iPad. This per­son has nicer fin­ger­nails. They’ve got a cool­er cup of cof­fee. Oh, I need to get that freak­ing app! I need to stop us­ing the web!

Why are pub­lish­ers grav­i­tat­ing to­ward apps? For two rea­sons. Go back to those two flaws.

For one thing, apps make it re­al­ly easy for in­for­ma­tion to be ex­pen­sive. They make it easy to col­lect mon­ey.

The oth­er thing is that the tech­ni­cal stan­dards are to­tal­ly stan­dard. That’s es­pe­cial­ly true with Ap­ple and iOS. It’s a very smooth de­vel­op­er kit. And as for test­ing—the ben­e­fit of de­vel­op­ing for iOS is that there’s a huge num­ber of de­vices in cir­cu­la­tion, but there aren’t that many types. So the test­ing bur­den is rel­a­tive­ly man­age­able. It’s easy to see now why apps—if you want­ed to make mon­ey—why apps would be a more ap­peal­ing de­vel­op­ment plat­form.

The web didn’t have to com­pete with this in the past. Right? But it does now. So the ques­tion emerges: are the habits of the last 20 years the right ones for the next 20 years?

To me, the an­swer is clear­ly no. Ei­ther the web is go­ing to have to adapt to to­day’s mar­ket­place, or it’s just go­ing to get re­strict­ed to this small­er and small­er box. And like I said at the be­gin­ning, you as de­sign­ers, you may think “well, ad­ver­tis­ing, tech­nol­o­gy, what­ev­er.” But it is your prob­lem. Be­cause you’re in that box. As the pos­si­bil­i­ties of the web shrink, so does web de­sign, and so do all the things that you can do in your ca­reer.

So what are we go­ing to do about this?

Well, I told you be­fore what I think the biggest flaw of the web is: that it can’t make in­for­ma­tion ex­pen­sive. So I think we need to start there.

We need to make it much easier for information to be expensive.

There are dif­fer­ent ways to do it. But this func­tion needs to in­te­grat­ed with the web at a low lev­el. And I know as de­sign­ers, we don’t re­al­ly con­trol this. But we can ad­vo­cate for it.

Even the in­ven­tor of the web and the cur­rent di­rec­tor of the W3C, Tim Bern­ers-Lee, agrees with this: he said—I love this—

“We need to find a whole lot of new busi­ness mod­els.” OK, that’s good, man. What else? He says, “I think we should de­vel­op new pay­ment pro­to­cols, so that when you’re us­ing a web brows­er, it’s a lot eas­i­er to pay for things.”

I love that, Tim. Did you say that 15 years ago? No. Well, did you say it five years ago? No. When did you say it?

You said it three months ago?

Oh man. I mean, has he just been hold­ing out on us? This is the guy—he’s not the boss of the web, but he’s an in­flu­en­tial voice in set­ting the agen­da for what the web stan­dards are.

So, I wouldn’t be the first per­son, by far, to crit­i­cize the W3C for be­ing slow and out of touch. But we’ve en­tered this phase where the web re­al­ly does need to move faster. And the W3C doesn’t do fast. So re­spect­ful­ly, but quite se­ri­ous­ly, I sug­gest: let’s—

Disband the W3C.

I mean, why not? And not in anger. Let’s thank them for all that they’ve done to get the web to this point. But we’re ready to take off the train­ing wheels. And now that the web has com­pe­ti­tion, we re­al­ly have no choice. The costs of de­lay are get­ting more se­vere. Think about those ads pop­ping up for apps—“use this, in­stead of us­ing the web.”

And if you wor­ry, as some do, that the al­ter­na­tive to no W3C is chaos, or if you wor­ry that the al­ter­na­tive is a web ruled by Google and Mi­crosoft and Ap­ple—I don’t think so.

I think the al­ter­na­tive is a web that’s or­ga­nized en­tire­ly as a set of open-source soft­ware projects. And we didn’t have that choice 15–20 years ago, at the be­gin­ning of the web. Be­cause open source hadn’t quite ar­rived as a way of do­ing things. But it has now. In fact, it need­ed the web to get it go­ing. But we’ve seen plen­ty of tech­nolo­gies that are in­te­gral to the web—Lin­ux, Apache, Perl, Python, Word­Press—all of these tech­nolo­gies have suc­ceed­ed with­out a W3C-style su­per­vis­ing au­thor­i­ty. In fact, they’ve all evolved a lot faster than the web has. So the web needs to catch up now. It can do that by learn­ing some­thing from these projects. The W3C was a great mod­el at one time. But it’s not any­more.

While we’re get­ting that web rev­o­lu­tion planned out, I think we should also, as de­sign­ers—

Refine what we mean by standards-based web design.

Be­cause stan­dards-based de­sign has of­ten seemed to mean: let’s start with stan­dards, but then, when it’s time to make it work in all those browsers that don’t sup­port stan­dards cor­rect­ly, we’re just go­ing to put in all the patch­es and hacks and every­thing.

This is what econ­o­mists would call moral haz­ard. Moral haz­ard means that rather than pun­ish­ing the bad browsers for their bad be­hav­ior, you’re in­dulging them. You’re ac­com­mo­dat­ing them. So what do they do? Do they stop? No, they keep do­ing it. And you get the op­po­site of what you want. In­stead of the bad browsers com­ing into com­pli­ance, you get 10 years of your ca­reer spent fig­ur­ing out what In­ter­net Ex­plor­er 6 can do. That’s a ter­ri­ble use of a de­sign­er’s ca­reer, in my opin­ion.

I think we also have this ques­tion of whether the tra­di­tion­al web-stan­dards view of the world is some­what ob­so­lete. Be­cause it emerged about 15 years ago, an era where we most­ly used the web on desk­top com­put­ers. There were only three or so browsers that any­one cared about. And now we have all these oth­er web-en­abled de­vices and so on.

I don’t think it’s prac­ti­cal to tell de­sign­ers & de­vel­op­ers that they should only stick to web stan­dards, and ig­nore the oth­er in­no­va­tions of these de­vices. But on the oth­er hand, it’s not prac­ti­cal to think that any stan­dards-mak­ing process is go­ing to move fast enough to cov­er them. And that’s re­al­ly why apps have been drink­ing the web’s milk­shake re­cent­ly.

So hav­ing web stan­dards that are tru­ly stan­dard would be a good start.

But be­yond that, I think that stan­dards-based de­sign should de­note a small­er idea. It should fo­cus on a small­er ter­ri­to­ry where it can work well, and work con­sis­tent­ly, rather than try­ing to of­fer it­self as this holis­tic ap­proach to web de­vel­op­ment. Be­cause his­tor­i­cal­ly, it just keeps com­ing up short.

Though some­times I won­der whether de­sign­ers have been over­ly pa­tient with web stan­dards be­cause they don’t to­tal­ly ap­pre­ci­ate how pro­gram­ming stan­dards usu­al­ly work. So de­sign­ers’ ex­pec­ta­tions are low­er than they ought to be.

Which brings me to my next top­ic—what should de­sign­ers know about pro­gram­ming?

And I know that some of you will think this is beat­ing a dead horse. But when we talk about restor­ing cre­ativ­i­ty to the web, and ex­pand­ing pos­si­bil­i­ties, we can’t avoid the fact that just like the web is a ty­po­graph­ic medi­um, it’s also a pro­gram­ma­ble medi­um.

And I’m a de­sign­er who ac­tu­al­ly does a lot of pro­gram­ming in my work. So I did read the oth­er 322,000 com­ments about this on the web. I still think there’s a sim­ple and non-dog­mat­ic an­swer, which is this:

You don’t have to learn programming, but don’t knock it till you try it.

Be­cause there are good things about it. I’ll tell you what I like about pro­gram­ming. First, though, I want to clear up a mis­con­cep­tion, that I gath­ered as I read the 322,000 com­ments.

HTML and CSS—those are not pro­gram­ming lan­guages. Those are data-en­try for­mats. So if you as a de­sign­er have ever looked at HTML and CSS and said, “you know what, I would rather drink bleach than learn how to do that,” then I think you’ve got the mak­ings of a great pro­gram­mer. Be­cause no­body wants to work with that stuff. Pro­gram­mers hate te­dious work of any kind. I hate HTML and CSS. That’s why I like pro­gram­ming. Be­cause the point of pro­gram­ming is to rep­re­sent ideas at a high­er lev­el. Which I think is very much some­thing that we like to do, as de­sign­ers. Rep­re­sent ideas. We can do that in the pro­gram­ming lan­guage. Then we can let the com­put­er han­dle the bor­ing non­sense.

All right, so why do I like pro­gram­ming? Here are three rea­sons.

One, like I said—it can elim­i­nate a lot of bor­ing work. That gives me more time for the fun parts of the project.

Two, pro­gram­ming ac­tu­al­ly ex­pands the scope of what I’m ca­pa­ble of as a de­sign­er. For in­stance, here’s a web­site I did for my font Con­course. My goal with this—these are all web pages—I want­ed to do some­thing I hadn’t seen be­fore, which was to do web ty­pog­ra­phy, us­ing web­fonts, and make it as close as I could to a PDF, in the sense of be­ing rich and de­tailed. I also made it ed­itable. I also made it scal­able.

But the idea of try­ing to do this by hand—I just would’ve quit in de­spair. I would’ve been in tears, scratch­ing my eyes out. So rather than try to do it by hand, I wrote some Python code that gen­er­at­ed all the HTML and CSS.

And the third rea­son I re­al­ly like pro­gram­ming, it’s kind of an un­der­rat­ed thing. But once you know a lit­tle bit about what can be ac­com­plished with pro­gram­ming and with code, you start to ex­pect more from your­self, and from the peo­ple you work with, and from the medi­um you work with, the web. Even if you don’t know a whole lot about pro­gram­ming, you un­der­stand very quick­ly that there’s very lit­tle that’s tru­ly im­pos­si­ble. So when you have con­ver­sa­tions with oth­er de­sign­ers, and oth­er pro­gram­mers, you re­al­ize that when peo­ple say “that can’t be done” or “that’s im­pos­si­ble,” what they’re re­al­ly say­ing is “I’m just too lazy and stu­pid to fig­ure that out,” or “that’s not my job” or “I don’t want to.” Again, if you’re into that, fine. I like fig­ur­ing out how to do things.

So that’s my pitch for pro­gram­ming. If it’s not for you, that’s fine. But don’t rule it out with­out try­ing it.

I also think that web de­sign­ers would do well to—

Make common cause with book authors.

Be­cause on the web, au­thors have re­al­ly en­dured a lot of the same slings & ar­rows as de­sign­ers. Only worse.

And when I say book, I don’t mean the shit­ty thing that Ama­zon sells on your Kin­dle. I just mean this idea—this place in our cul­ture where deep think­ing can hap­pen. Where a writer can do this think­ing that you can’t do in 140 char­ac­ters, or even 140 blog posts.

Go­ing back to the orig­i­nal premise, I think that if the web is the best man­i­fes­ta­tion of the writ­ten word, then it should also be a great book-pub­lish­ing tool. The web should re­al­ly en­cour­age au­thors to write books. But it doesn’t. It re­al­ly doesn’t. The best you can do on the web is make an ugly book and give it away for free. That’s not good enough for most au­thors.

And it goes back to that prob­lem with mak­ing in­for­ma­tion free. When you re­duce the in­cen­tives to pub­lish, you’re not just im­pov­er­ish­ing au­thors. They’re just find­ing oth­er things to do with their time. It’s like “why am I go­ing to write, and go through the has­sle? I’ll just play Xbox.” So what hap­pens is the books dis­ap­pear. They nev­er come into be­ing. They nev­er get made. Au­thors find some­thing else to do with their time. When smart peo­ple who could be shar­ing their ideas choose not to—that’s a real loss, to all of us. And then when it’s tech­nol­o­gy that’s de­ter­ring them—that’s kind of crazy. So I feel like we’ve got to do bet­ter.

Think about iBooks, on the iPad. To be bet­ter than the web, it doesn’t have to do much. All iBooks has to do is solve these two prob­lems I’ve been talk­ing about—solve the eco­nom­ic prob­lem, solve the tech­nol­o­gy prob­lem. And it does. Here are a cou­ple pages from George Gamow’s won­der­ful book One Two Three … In­fin­i­ty from 1961.

No­tice the parts that are past­ed in. These are ac­tu­al­ly scans from the 1961 print­ing. Some of you prob­a­bly rec­og­nize—that’s let­ter­press print­ing. So what are we find­ing out here? What we’re find­ing out is that in book ty­pog­ra­phy, the iPad is not keep­ing up with the print­ing tech­nol­o­gy of the ’60s. Pret­ty sad, right.

I like the “V+F” in the di­a­gram—it’s as if the per­son who was do­ing this was like “I can han­dle V+F = E+2.” Then they get down to the ta­ble and they’re like “ah, fuck it. I want to hire that print shop from 1961. They did a good job with that. Yeah, that’s nice.” So we’re not do­ing so well with the print­ing techol­o­gy of the ’60s.

But how are we do­ing with the pin­ball tech­nol­o­gy of the ’60s? Here you go. This is also on my iPad. A pin­ball sim­u­la­tor. This is amaz­ing. This is a 1966 pin­ball ta­ble. Ful­ly playable. It’s got sound, col­ors, physics. Look at the up­per right hand cor­ner. They’ve even sim­u­lat­ed this toy mon­key that rings a bell every time you get 100 points. I mean—is this what we want? Pin­ball is win­ning, and books are los­ing? It’s not what I want.

I re­al­ly think that rein­vent­ing the book is a huge­ly im­por­tant project. And we’re nowhere with it. Books on the web should be the best books we’ve ever had. And it’s not like we don’t have role mod­els. Look at Frank Chimero’s book from be­fore. I think if every au­thor could make a book like this, and make mon­ey from it, then every au­thor would be pub­lish­ing on the web. But they can’t. So they don’t. We’re not there yet.

But I don’t say that pes­simisti­cal­ly. I think we can get there. And I’m op­ti­mistic, al­ways—even though I’m pes­simistic about cer­tain tech­nol­o­gy or­ga­ni­za­tions—I’m op­ti­mistic about peo­ple and de­sign­ers chang­ing things for the bet­ter. Be­cause that’s how we got to where we are to­day.

But to­day, the web wants more from us. Web de­sign wants more from us.

Be­cause a lot of you, maybe most of you, are go­ing to spend most of your de­sign ca­reer putting things on screen, and on the web. Not on pa­per. So again, go­ing back to my first point—my ma­jor point to­day is that I hope you feel in­vest­ed in this strug­gle, be­cause what­ev­er hap­pens, it’s go­ing to af­fect you for a long time.

As I said at the be­gin­ning—de­sign­ers have al­ways been vi­tal to the web, in terms of ex­plor­ing its ca­pa­bil­i­ties and shar­ing those pos­si­bil­i­ties. So as we stand here at the 20-year birth­day, let’s not for­get that. We’re not mules. We’re not sheep. We’re not data-en­try drones. We shouldn’t tol­er­ate a web that rel­e­gates us to those roles.

We need to work for the web that we want. To rein­vent. To de­mand bet­ter. It’s not that we’re go­ing to get every­thing. But if we work at it, we’ll get some­thing. And if we don’t work at it, we’ll def­i­nite­ly get noth­ing.

But ei­ther way, we’ll be get­ting the web that we de­serve.

Thank you.

Re­lat­ed

But­t­er­ick’s Prac­ti­cal Ty­pog­ra­phy
My new web-based book. It in­cludes a few more thoughts about what de­sign ex­cel­lence on the web should mean.

Re­vers­ing the Tide of De­clin­ing Ex­pec­ta­tions
Tech­nol­o­gy in­vites de­sign­ers to take short­cuts, but we have to ask more of our­selves.

Re­build­ing the Ty­po­graph­ic So­ci­ety
In its pur­pose and method, ty­pog­ra­phy is in­sep­a­ra­ble from hu­man­i­ty’s most con­se­quen­tial in­ven­tion: the writ­ten word.


Com­ments

I work in book pub­lish­ing and am see­ing the grad­ual de­gra­da­tion of qual­i­ty de­sign on the web mir­rored in the qual­i­ty of e-books, which is hard­ly sur­pris­ing con­sid­er­ing their shared struc­ture and the way browsers and read­ing en­vi­ron­ments sim­i­lar­ly dis­re­gard “stan­dards”. E-ink Kin­dles are our IE6, iBooks our Chrome.

Ar­guably pub­lish­ers are part­ly to blame for this; for years we put out poor­ly proofed and poor­ly for­mat­ted texts into a nascent mar­ket­place. It’s part­ly our fault that read­ers have come to ex­pect less from e-books than they do from print­ed books.

Of course the dif­fer­ence be­tween e-books and the web is that the for­mer is much eas­i­er to charge for, which is both good and bad; we’re in a sit­u­a­tion where the vast ma­jor­i­ty of our books are sold into a read­ing en­vi­ron­ment that has the least sup­port for pre­vail­ing stan­dards. It’s akin to IE6 be­ing the only place you could sell web con­tent.

I guess what I’m re­al­ly con­cerned about is that peo­ple like you and me are in this space where these things re­al­ly mat­ter to us, but the av­er­age web­site user or e-book read­er is now ac­cus­tomed to these re­al­ly low stan­dards of qual­i­ty, and that some peo­ple will nev­er want to pay more/any­thing for some­thing that looks or be­haves bet­ter.

—Eoin No­ble


I also orig­i­nal­ly came from a art back­ground, fol­lowed Comm Arts and all of the in­dus­try stuff and there is ide­al­ism and there is re­al­i­ty. The day to day busi­ness of artists and de­sign­ers does not gen­er­al­ly in­volve a blank slate and a blank check—it is dif­fi­cult to see how many of the prob­lems you showed have much to do with the Web or stan­dards re­lat­ed to as much as they do that busi­ness­es are in­ter­est­ed in mak­ing mon­ey … I think the Web as we know it will evolve sig­nif­i­cant­ly and I have high hopes that giv­en events in the last year or two we are on the verge of a shift where we’re fig­ur­ing that out.

—Bri­an Kardell


I agree with you in a lot of ways. I do not like what has hap­pened to the in­ter­net. The goal of mak­ing rich and in­ter­est­ing de­signs has been re­placed by a vast waste­land of cor­po­rate ad­ver­tis­ing garbage.

—Den­nis Sae­va


The prob­lem with the web and pay­ments is (in my mind) the ghast­ly state on­line pay­ment is in. … Al­though it is de­sign­ers and pro­gram­mers who have to ac­tu­al­ly solve the prob­lem, it is ul­ti­mate­ly a bank­ing prob­lem. And sad­ly bank­ing haven’t es­tab­lished a rep­u­ta­tion of be­ing very pro­gres­sive in re­gards to tech­nol­o­gy late­ly. I strong­ly be­lieve that we’re go­ing to be stuck with bro­ken ad-based rev­enue mod­els un­til the pay­ment prob­lem has been solved.

Re­gard­ing your oth­er pro­pos­al … [m]aybe the so­lu­tion is rapid de­vel­op­ment through a open source mod­el, with a stan­dard­iz­ing or­gan, such as the W3C, cre­at­ing a cer­tain base­line stan­dard/rec­om­men­da­tion. What they do well is pro­vide one cen­tral­ized ref­er­ence.

—Kris­t­ian Kalvå


[T]he main prob­lem is that de­sign­ers are taught that they need to match each oth­er’s de­signs–es­pe­cial­ly on the web … [t]his was the case with me at uni­ver­si­ty. The irony is that I see the most en­joy­able de­signs com­ing from peo­ple who have no for­mal ed­u­ca­tion in de­sign (most­ly de­vel­op­er blogs). It seems as though the un­der­ly­ing prob­lem here is the elec­tron­ic de­sign in­dus­try as a whole: not nec­es­sar­i­ly the W3C. How­ev­er, I agree that the W3C should be dis­band­ed; the web does need to take a more or­gan­ic evo­lu­tion­ary path. Don’t be fooled into think­ing that dis­band­ing the W3C will do any­thing about bad de­sign on the web, on the con­trary, it will like­ly give the bad de­sign­ers more tools with which to shoot them­selves in the foot.

—Jonathan Dick­in­son


De­sign­ers are al­ways put into a sit­u­a­tion where the de­sign is judged by peo­ple that are not qual­i­fied to do so. … [t]he mar­ket will ei­ther prove that good de­sign begets bet­ter prof­its, or more de­sign-ap­pre­ci­at­ing man­agers will suc­ceed and nat­ur­al se­lec­tion will sort out the rest. … I gave up try­ing to make the world pret­ty be­cause every­one that was will­ing to pay me want­ed to make it ugly. Af­ter a life of de­sign, I am chang­ing ca­reers al­to­geth­er.

—John Car­val­ho


The W3C doesn’t “adopt” stan­dards; the mar­ket does. The W3C doesn’t re­al­ly even cre­ate stan­dards for Web browsers; brows­er ven­dors do. The W3C bro­kers the cre­ation of stan­dards by pro­vid­ing a place for brows­er ven­dors and oth­ers to get to­geth­er to reach agree­ment on de­tails of new browsers tech­nolo­gies in such a way as they’re will­ing to ac­tu­al­ly im­ple­ment them. The W3C has zero means for “en­forc­ing” stan­dards for brows­er tech­nolo­gies.

Get­ting agree­ments among im­ple­men­tors is the re­al­ly hard part, and there’s no mag­ic to make the process of reach­ing agree­ments quick, easy, and pain­less. ... Nowhere in Matthew But­t­er­ick’s talk is there a real pro­pos­al for how we could get agree­ments any quick­er or eas­i­er or less painful­ly than we do now by fol­low­ing the cur­rent stan­dards-de­vel­op­ment process.

—Michael Smith (on the W3C blog)


Post­script: about the W3C.
Since I gave this talk I’ve heard from nu­mer­ous de­fend­ers of the W3C, in­clud­ing Michael Smith. The two ma­jor threads: a) the W3C’s job is in­her­ent­ly hard, so we should ex­cuse its mediocre per­for­mance, and b) there are no al­ter­na­tive meth­ods of de­vel­op­ing web stan­dards. There­fore, ap­par­ent­ly, we have no choice but to ac­cept the W3C and its meth­ods.

Are you per­suad­ed? I’m not.

“The W3C has zero means for ‘en­forc­ing’ stan­dards for brows­er tech­nolo­gies.”

Not true. The W3C could refuse to re­new the mem­ber­ship of or­ga­ni­za­tions that took ac­tions con­trary to the spir­it of web stan­dards, in­clud­ing re­peat­ed­ly fail­ing to im­ple­ment them. If the W3C made par­tic­i­pa­tion in the con­sor­tium con­tin­gent on time­ly im­ple­men­ta­tion, mem­bers would ei­ther com­ply or quit.

But it doesn’t. So the W3C es­sen­tial­ly re­wards its mem­bers for un­der­min­ing web stan­dards—an­oth­er ver­sion of the moral haz­ard I men­tioned ear­li­er. For in­stance, it’s crazy that a com­pa­ny like Mi­crosoft can put six em­ploy­ees on the CSS work­ing group to help make stan­dards, and then have hun­dreds of en­gi­neers back at head­quar­ters de­stroy­ing them—year af­ter year af­ter year. And for once, let’s not blame Mi­crosoft. It’s sim­ply be­hav­ing con­sis­tent­ly with its prin­ci­ples. The W3C, how­ev­er, is not.

If you know a W3C staffer, ask them: what ben­e­fit does the W3C get from main­tain­ing the mem­ber­ship of cor­po­ra­tions that have re­peat­ed­ly failed to im­ple­ment stan­dards? I mean, aside from an­nu­al mem­ber­ship fees of about $70,000. (More on that be­low.)

“Get­ting agree­ments among im­ple­men­tors is the re­al­ly hard part, and there’s no mag­ic to make the process of reach­ing agree­ments quick, easy, and pain­less.”

The web doesn’t need a stan­dards process that’s quick, easy, or pain­less. The web needs a stan­dards process that yields stan­dards that are im­por­tant for the con­tin­ued vi­tal­i­ty of the web—for in­stance, those “new pay­ment pro­to­cols” that Tim Bern­ers-Lee men­tioned. The web also needs stan­dards that be­have in a tru­ly stan­dard way.

Right now, we have nei­ther. And we are suf­fer­ing the con­se­quences.

The idea that stan­dards couldn’t hap­pen with­out the W3C to “bro­ker” them is plain­ly false, and sort of out­ra­geous in its moth­er-knows-best my­opia. For decades, peo­ple and com­pa­nies have come to­geth­er on tech stan­dards when it’s been in their ra­tio­nal in­ter­est to do so. It has noth­ing to do with “mag­ic.” Some­times one ven­dor cre­ates a prod­uct that be­comes a de fac­to stan­dard (e.g., Adobe with Post­Script). Some­times mul­ti­ple ven­dors come to­geth­er to de­fine mu­tu­al­ly ben­e­fi­cial stan­dards (e.g., Mi­crosoft & Adobe with Open­Type). Some­times the stan­dards go through a for­mal process (e.g., PDF be­com­ing an ISO stan­dard). Some­times they come from in­di­vid­ual de­vel­op­ers and mi­grate up­ward (e.g., Uni­fied Font Ob­ject). (For­give the ty­pog­ra­phy bias, that’s just an area I hap­pen to know well.)

The W3C would ob­vi­ous­ly pre­fer that we not study this ev­i­dence, as it only con­firms that the web is sad­dled with one of the least ef­fec­tive stan­dards-mak­ing bod­ies in the his­to­ry of dig­i­tal tech­nol­o­gy.

“Nowhere in Matthew But­t­er­ick’s talk is there a real pro­pos­al for how we could [im­prove on] … the cur­rent stan­dards-de­vel­op­ment process.”

Yes, there is a “real pro­pos­al”—open source. Hu­man be­ings aren’t any less con­ten­tious on open-source projects. It’s just that there’s a dif­fer­ent pow­er struc­ture: a “benev­o­lent dic­ta­tor” who is the ul­ti­mate ar­biter. This elim­i­nates the need for con­sen­sus. (In­deed, open-source projects are set up this way in recog­ni­tion of the fact that such con­sen­sus is not worth the cost.) It doesn’t mean oth­ers can’t have a voice—for in­stance, look at the Python PEP sys­tem. Still, these “dic­ta­tors” serve only at the plea­sure of the users, since any­one can sim­ply fork the project and move on. So the dic­ta­tors have an in­cen­tive to make choic­es that are in the best in­ter­est of the user com­mu­ni­ty. There is a healthy bal­ance of pow­er.

“But open source pro­duces soft­ware, not stan­dards.” True in the­o­ry, not in prac­tice. Stan­dards and open source are both so­lu­tions to the peren­ni­al prob­lem of how to spread the costs of tech­nol­o­gy with­out com­pro­mis­ing prof­it op­por­tu­ni­ty. If some­one else solves the prob­lem, you can rely on that so­lu­tion, and fo­cus your ef­fort on oth­er things. With stan­dards, that so­lu­tion comes in the form of a writ­ten spec­i­fi­ca­tion; with open source, it comes as code. But if we’ve learned any­thing in the first 20 years of the web, it’s that the hand that codes the brows­er rules the world. Web stan­dards can be—should be—de­fined by a ref­er­ence im­ple­men­ta­tion. And that im­ple­men­ta­tion should be open source.

Eco­nom­ic in­cen­tives are also im­por­tant. As long as the W3C charges as much as it does for mem­ber­ship, mem­ber­ship will re­main re­strict­ed most­ly to wealthy cor­po­ra­tions. More­over, those cor­po­ra­tions are go­ing to want to pro­tect their in­ter­ests and not get pushed around. We could imag­ine a W3C that re­duced the role of cor­po­rate mem­bers and in­stead was fund­ed by small­er do­na­tions from a larg­er con­stituen­cy—think PBS rather than NBC. Would it yield dif­fer­ent re­sults? I’m cer­tain of it.

But don’t take my word for it—if you want to know where the W3C’s loy­al­ties lie, just ob­serve its re­cent con­duct. First, af­ter many years of keep­ing DRM out of open web stan­dards, the W3C is prepar­ing to adopt a DRM pro­pos­al for stream­ing me­dia draft­ed by three W3C mem­bers—Google, Mi­crosoft, and Net­flix. (Also see the ob­jec­tions from the Elec­tron­ic Fron­tier Foun­da­tion, and nov­el­ist Cory Doc­torow’s cri de coeur.) Mean­while, the W3C has been drag­ging its feet on Do Not Track, which would en­hance in­di­vid­ual pri­va­cy but re­duce the ad rev­enue of W3C mem­bers. Be­cause let’s face it, all those ban­ner ads aren’t go­ing to click on them­selves.

—MB


[Com­ments wel­come—thought­ful ones will be pub­lished]