Skip to Content

A Clean Start for the Web

How do we make the web fun, participatory, and good?

The we­b’s evo­lu­tion over the last decade has mir­rored the American econ­omy. All of the es­sen­tial in­di­ca­tors are go­ing up and to the right,” a steady stream of fun­da­men­tal ad­vances re­as­sure use that there is progress,” but the ac­tual ex­pe­ri­ence and ef­fects for in­di­vid­u­als stag­nates or re­gresses.

tmcw is ap­proach­ing the con­tem­po­rary prob­lem of tech­ni­cal debt, the re­sult when our sys­tem is so com­plex and fucked that it’s start­ing to col­lapse and there’s noth­ing we can do about.

The plat­form side is what changed last week, when Mozilla laid off 250 em­ploy­ees and in­di­cated that it would af­fect Firefox de­vel­op­ment … The real win­ner is not just Chrome, but Chrome’s en­gine. One code­base, KHTML, split into WebKit (Safari), and Blink (Chrome, Microsoft Edge, Opera, etc.)

This a text­book mono­cul­ture. In one sense, it’s a vic­tory for col­lab­o­ra­tion be­cause no­body’s wasting time’ on com­pet­ing im­ple­men­ta­tions and web de­vel­op­ers can ex­pect the same fea­tures and bugs across dif­fer­ent browsers. But in a deeper way, it threat­ens one of the ba­sic prin­ci­ples of how the web has evolved.

Tmcw iden­ti­fies that our browsers our now a mono­cul­ture, which is an in­ter­est­ing prob­lem to have be­cause 1) there are some clear in­di­ca­tions that a mono­cul­ture by tech­ni­cal be­he­moths is bad but it’s only an in­ter­est­ing be­cause 2) browsers should just be all im­ple­ment­ing the spec any­way. The spec is an in­ter­est­ing prob­lem, since it also is just a mono­cul­ture dom­i­nated by tech­ni­cal be­he­moths.

Organizations like the WHATWG, W3C, and IETF have been col­lab­o­ra­tion spaces for in­de­pen­dent de­vel­op­ers, cor­po­ra­tions, and aca­d­e­mics to dis­cuss po­ten­tial new fea­tures of the web.

At least the spec pre­tends or tries to be open and com­mu­nity ori­ented.

Chrome has been mov­ing ex­tremely fast, adding new spec­i­fi­ca­tions and ideas at a star­tling rate, and it’s be­com­ing one of the hard­est pieces of soft­ware to repli­cate.

The prob­lem is that Google dom­i­nates the spec, and is gain­ing mo­nop­oly on the spec.

Not only is it nearly im­pos­si­ble to build a new browser from scratch, once you have one the on­go­ing cost of keep­ing up with stan­dards re­quires a full team of ex­perts.

This is prob­a­bly the point and the goal of the hos­tile takeover of the spec. The cre­ation of a plat­form that be­comes a mo­nop­oly on a mar­ket is a hall­mark tech com­pany busi­ness model.

The web has had about 25 years to grow, few op­por­tu­ni­ties to shrink, and is now sur­rounded by an ex­tremely short-sighted cul­ture that is an out­growth of eco­nomic and ca­reer short-ter­mism.

Haha, Im not sure I have any in­sight here but its funny

For folks who just want to cre­ate a web page, who don’t want to en­ter an in­dus­try, there’s a baf­fling ar­ray of tech­niques, but all the sim­plest, prob­a­bly-best ones are stig­ma­tized. It’s eas­ier to stum­ble into build­ing your re­sume in React with GraphQL than it is to type some HTML in Notepad.

Tmcw iden­ti­fies that there is a prob­lem for cre­ators in that mak­ing web shit is out­landishly dif­fi­cult.

The end re­sult is that I no longer ex­pect pages to be fast, even with uBlock in­stalled in Firefox and a good lo­cal fiber in­ter­net provider.

Tmcw in­den­ti­fies that there is a prob­lem for con­sumers in that our tech­ni­cal debt has re­sulted in a hor­ri­ble ex­pe­ri­ence.

I don’t want to lay all of the blame at those web de­vel­op­ers … Segment of­fered a big page of on/​off switches with hun­dreds of data providers & ad-tech com­pa­nies on it. And, sure, enough, some folks closer to the busi­ness side started click­ing all those but­tons.

This is an in­ter­est­ing way of throw­ing the busi­ness team un­der the bus for mak­ing choices that are ob­vi­ous to them and re­mov­ing our own cul­pa­bil­ity in build­ing these sys­tems.

How do we make the web fun, par­tic­i­pa­tory, and good?

There is Tmcw’s the­sis: which ob­vi­ously has a ton of value judg­ment wrapped up in it.

the cur­rent web which has ba­si­cally thrown away se­man­tic HTML as an idea.

Is that true or is that just a dig at the cur­rent dom­i­nant process and sys­tem?

Then there’s the application web”. This started as server ap­pli­ca­tions…

We’ve seen Tmcw dig at the front-end space be­fore.

The prob­lem is that the document web” is of­ten plagued by ap­pli­ca­tion char­ac­ter­is­tics - it’s the JavaScript and an­i­ma­tions and com­plex­ity that makes your av­er­age news­pa­per web­site an un­mit­i­gated dis­as­ter. Where doc­u­ment web­sites adopt ap­pli­ca­tion pat­terns they of­ten ac­ci­den­tally sac­ri­fice ac­ces­si­bil­ity, per­for­mance, and ma­chine read­abil­ity.

I feel this is clas­sic Tmcw get off my lawn” crank­i­ness. Rich Harris ad­dresses this: It's a spectrum, and the interesting stuff happens in the middle. There's magic in that grey area. I still believe the web has the potential to be a fundamentally more expressive medium than its predecessors ( is on a similar wavelength to that notion)

The ap­pli­ca­tion web uses JSX, not HTML, and would like that in the browser it­self, or Svelte, in­stead of JavaScript, and would like that too.

Haha our boy is com­ing for Rich! I’m not I un­der­stand what the dig here is … he’s mad at a lan­guage that has a com­pile step? Isn’t that what like … most ap­pli­ca­tions use? Rust, C, what­ever, they all com­pile down to a na­tive run­time lan­guage.

When I read blog posts from traditional web de­vel­op­ers’ who are mad that HTML & CSS aren’t enough any­more and that every­thing is com­pli­cated — I think this is largely that the ap­pli­ca­tion stack for build­ing web­sites has re­placed the doc­u­ment stack in a lot of places.

I mean, that’s a de­cent point. The prob­lem is that we’ve cre­ated a dis­con­nect be­tween what we want to build, why we won’t to build it, and how we build it. We use in­sane tool­chains to solve sim­ple prob­lems, and not all tools are for the same thing.

The ap­peal of so­cial net­works is partly be­cause they let us cre­ate doc­u­ments with­out think­ing about web tech­nol­ogy, and they pro­vide guar­an­tees around per­for­mance, ac­ces­si­bil­ity, and pol­ish that oth­er­wise would take up our time.

This seems like a de­cent point well made. Those of us who want to re­duce tech be­he­moth mono­cul­ture ab­solutely need to reckon with the real value that so­cial net­works of­fer to users.

stan­dards like RSS and ser­vices like Instapaper show that pleas­ing for­mat­ting and dis­tri­b­u­tion can be done at the plat­form level and be pro­vided on top of ex­ist­ing vanilla web­sites.

Thats a de­cent proof of con­cept the prob­lem, as il­lus­trated above, is around the tools and sys­tems we’ve built for our­selves to make things easier”.

Okay, so let’s say we’re cre­at­ing a new doc­u­ment web … First, you need a min­i­mal, stan­dard­ized markup lan­guage for send­ing doc­u­ments around. … Then, you need a browser … Could you build a pure-Mark­down-brows­ing browser that goes straight through this pipeline? Maybe?

This is a pretty in­ter­est­ing idea, but sounds like it’s ab­solutely a sub­set of the cur­rent sys­tem which is railed against above. Maybe the idea is that it can grow in­de­pen­dently from be­ing a sub­set to be­ing some­thing more.

The prob­lem with imag­in­ing the ap­pli­ca­tion web is that it’s pretty ex­pan­sive.

Yeah no shit tmcw. Im not sure how any of this gets around the prob­lem out­lined above: hos­tile takeover of a pub­lic spec by pri­vate tech be­he­moths.

The worse the Mac App Store’ and Windows App Store’ and App Store’ and Play Store’ get, the big­ger a cut those mo­nop­o­lies de­mand, the more it costs to be a Mac or Windows de­vel­oper, the more that ap­pli­ca­tions get pushed to the web. Sure, some ap­pli­ca­tions are bet­ter on the web. But a lot are just there be­cause it’s the only place left where you can eas­ily, cheaply, and freely share or sell a prod­uct.

Creating an App Web” could so eas­ily just col­o­nize that space with those same mod­els of mono­cul­ture.

There are a lot of other ways to look at and solve this prob­lem. I think it is a prob­lem, for every­one ex­cept Google.

Right, this is the goal for Google. And cre­at­ing a new set of specs is not go­ing to solve that prob­lem, be­cause that prob­lem is a re­flec­tion of our cur­rent so­ci­ety and what we value and how we work.

The idea of a web browser be­ing some­thing we can com­pre­hend, of a web page be­ing some­thing that more peo­ple can make, feels ex­cit­ing to me.

This is a good goal, and I en­dorse it! Our medium tho is that of the tech­ni­cal mono­cul­ture, the baked in as­sump­tions that a small group has used to cre­ate this new world. The is­sue is with the hard­ware and the phys­i­cal things and the hu­man so­ci­ety that sur­rounds and makes those things pos­si­ble.

What if we could start over?

Tmcw is call­ing for a rev­o­lu­tion of the web with­out be­ing so ex­plicit about it. The idea he is pre­sent­ing is that there needs to be a messy, de­mo­c­ra­tic, so­cial­ist ap­proach to our specs that cre­ate more room for the in­di­vid­ual, and that in em­pow­er­ing the in­di­vid­ual with rights also cre­ates a num­ber of re­spon­si­bil­i­ties to the so­ci­ety that they must up­hold. This is a so­cial­ist ar­gu­ment for the web, and I wholly agree with it. I think that he needs to be ex­plicit about what he’s ac­tu­ally call­ing for and imag­ine what the real world reper­cus­sions of the po­si­tion would be, and how we need to make the same call of our so­ci­ety for the same ends.