Why web standards can cripple the web


Recently Joe Hewitt, a well-known Facebook developer, sent out a series of tweets relating to the current state of the web and why it is evolving so slowly. He makes quite controversial statements such as that web standards actually cripple the web. He envisions the capabilities of open online technologies being just as powerful as native applications and says this won't be possible unless the way in which browsers innovate changes drastically. He essentially says that browsers should be adding their own browser-specific features (which is generally frowned upon by the web development community) and coming up with new ideas that do not fall within the W3C's and ECMA's specifications. The standards bodies should then come afterwords and clean up and structure these ideas and encourage all browsers to support them. Some other controversial statements he makes:

  • Microsoft was doing innovative things with Internet Explorer until developers asked them to stick to the standards.
  • IE 6 was a very innovative browser back in the day.
  • Flash isn't evil. It simply saw a gap that web standards weren't filling and tried to make up for it.
  • Forcing users to use a specific browser to use certain features on site may be better than the "since all browsers don't support it, I am not going to implement support for it" stance.

Again, the statements he makes are very controversial. But they are well substantiated and his arguments are backed up. Although I have never thought of things in this way, to me, what he says makes a lot of sense. Techcrunch did a great job of compiling these tweets here. It's a great read.


Netscape used to exceed standards, too, but of course pretending to cater to them became the fallback position when they blundered so badly with Netscape 4.

The stagnation argument is correct, and this is why:

"Standards" being "official" and as a universal goal is central planning.

And that never, ever works as well as competition and spontaneous order.

The free market always trumps socialism.

Standards committees are bureaucrats attempting some ivory tower guess at What Works Best, and such cannot be correct except by ultra-rare coincidence.

The "experts" thought Beta was better than VHS, but they were wrong. They thought OS/2 was better than Windows, but they were wrong. Breast feeding bad, formula good...wrong. Make everyone wear shoes in public to cut foot fungus...wrong.

The "experts" can almost never weigh what is right/wrong in such a way that they should be able to even encourage, much less force, everyone to put their eggs in one basket.

That standards retard advancement is not so much an opinion, as a mathematical fact. The only way it could be good is if they magically got it perfectly, and they don't even have an effective metric for trying.


1) Ok, see, that's what I said, nest them. (your example made no sense)

2) It works as it should here:

Though admittedly I have quicktime (ok, QT Alternative) installed which changes the rules of how IE6 OBJECTS work just a bit...

It does bomb how you describe in IE 5 though.


Hmmm that got me thinking!

MS innovating?
Yes they used to innovate in their own way, and IE6 was a leap forward for its time, but MS stopped developing it didn't they, and with that innovation stopped as well?

No Flash isn't evil, it's certainly helped the web move forward with video, and interactive apps, but, as Mr Jobs rightly pointed out the other day it will struggle with the evolving touch screen applications unless Adobe can pull a rabbit out the hat. I think Flash has seen it's hey day. It will still be useful for many years to come but will gradually decline.

Forcing users to specific browsers
The whole point of standards is to make the web accessible for all, irrespective of what device, browser or OS they use. I can see his point but once you start fragmenting things you start to infringe on usability and accessibility and alienate people - why should I download, install and use another browser just to see a certain site?

Standards is all about creating a common platform for all. This article reminds me of the Victorian age where every nut and bolt had to be hand made. This was very expensive to do till someone came along and "standardised" nuts and bolts into commonly used sizes. This had the effect of making production of amazing machinery much cheaper because of mass production, and needless to say revolutionsed the industrial age. Where would web be if nuts and bolts were still hand made?


The end result of doing away with web standards will be 1000+ different web browsers each of which requires a different language to be used in order to work in that browser. You only need look back to the days of having to write each page twice to cater for both IE3 and Netscape 3 to see what not following standards can do.

Anyway there is nothing stopping any browser from creating their own proprietary innovations (and some do). The problem is that such innovations then only work for people using that browser. Removing the standards would just increase the part of the page that only works in one browser. At least the standards provide a baseline that will work across all browsers that support the standard.

If anything removing the standard would decrease the use of innovations because those coding the pages would need to write several versions of each page instead of the one they now write and so wouldn't have time to update pages to incorporate new features.


Please let's not destroy web standards, I know HTML5 is turning into an "ugly sister" of a specification but we don't want to make the web go backwards. Browsers DO create their own features (IE8 is an example - their hSlice microformat is probably one of the best innovations MS have created of late and let's not forget how conditional comments have changed the face of styling IE) and proprietary non-standard extensions are easy to produce on a browser by browser basis (the W3C actively endorses that convention). If however we dumped standards and left it as a free for all, you would end up needing every browser installed to cope with websites having that "best viewed in" crap on their pages. We did the "leave the browser makers to their own devices" gig in the 90's and it left us with a browser war that ended up with illegal anti-competitive behaviour and a backlog of splinters in the web... you're meant to learn from history, not repeat it. smile


This seems like a bit of a false distinction to me. Standards or innovation, as if we're forced to make a choice of one but not the other? What's wrong with implementing standards as a minimum baseline, and adding all your innovative extensions on top of those standards. Is designing things in a modular way really such a new idea?



Someone has things backwards in their thinking since standards are a necessary precursor to innovation. Without standards there is no time to consider innovation.

As for developers asking Microsoft to follow the standards - it wasn't a request to stop innovation, it was a request to make sure that the browser can display basic web pages properly before worrying about adding something fancy.

If innovation were a building then standards are the land the building sits on. Remove the land and the building will collapse.


i can think of quite a number of adjectives that would describe exactly how wrong this is, but suffice to say, it's wrong


It might be wrong in general but it is true for the web. Without the standards the web doesn't exist and each separate browser has its own sub-web that it works with. Everyone creating pages would then need to spend so much time creating five or ten separate versions of each page that they'd have to keep them fairly basic. There wouldn't be time to spend on taking advantage of any innovations and so the innovations may as well not exist. If no one has time to actually use any innovations THEN there wouldn't be any point in any browser introducing any more innovations. They'd have to look at coming to some agreement with the other browsers to simplify the process of creating pages that work in more than one so as to reduce the time people creating pages need to waste on creating basic pages that work in all browsers so that those people will have time to think about implementing some nice to have innovations for some browsers.

Since that is what actually did happen with respect to the web it can't be wrong.

If you look at most things you will find that there is a cycle between innovation and standards and that the development of standards forms a platform on which new innovation then occurs.

As another example consider computers. Up until the mid 1080s each company produced their own proprietary computers that were incompatible with other brands. Even where they could run the same operating system that operating system had to be separately written for each individual computer and so the OS writers were spending their time creating versions of the same OS for different computers. When Compaq managed to revese engineer the BIOS from the IBM computers it meant that they could start producing computers that ran the exact same operating system as IBM. That introduced the defacto IBM compatible standard for computer hardware which then allows the OS developers to start spending time on innovations rather than recreating the wheel in different colours. Over the next few years Microsoft was able to move on to create their OS/2 operating system which was a huge innovation compared to the various DOS operating systems that were around at the time. When that OS failed to catch on straight away Microsoft then used a different innovation by combining a graphical interface directly into their DOS operating system and marketed that in such a way as to get enough people to switch to that so that it then became the defacto standard. Because of internal standards where Microsoft made both their operating systems compatible with one another they were then able to further develop what started as OS/2 and about ten years ago finally convinced everyone to switch to using that instead of DOS with the fifth version of that operating system. The seventh version of that standard operating system with even more innovations incorporated into it was recently released. Without the standards developed along the way there would have been no opportunity for the innovations.


once again, nice backpedal, stephen

smiley smiley


I don't know why you keep saying that. It must be something to do with your not understanding the point I am trying to make properly in the first place but understanding better what I mean when I provide a more detailed explanation. Perhaps my original comments are not worded clearly enough and I need to work harder to try to make it clear what I mean at the start so as to not need to have people such as yourself misunderstanding my meaning in the first place.

I'll try harder to say what I mean more clearly the first time. Thankyou for all the times you have pointed out where what I am trying to say isn't clear.


I can't speak for r937, but I would say that your comment is still sort of wrong. One can innovate to their heart's content without a care for standards, and that's pretty much what Netscape & Microsoft did in the 90s.

Did that stop them inventing new technologies or innovating?

Yes, it created all the problems you mentioned, but that didn't actually seem to hold back the tide of new technologies. And it doesn't support the point you made that r937 took issue with.

I would say that you have it exactly backwards: if anything, innovations are a precursor to standards. A swathe of competing technologies emerge and after a bit of natural selection, the winner(s) are chosen and standards are assembled.



Forcing users to use a specific browser to use certain features on site may be better than the "since all browsers don't support it, I am not going to implement support for it" stance.

Nobody's going to force which browser I use. Imagine being blind, having paid more money than you have for a screen reader, written way back in the day and only even works at all with IE6. Yeah, let's see you force them to some other browser that doesn't work with their reader so that other people can get sparklies bs on your web page.

Or, wait, here's a revolutionary idea... let's put the sparkly flashy junk on our pages and whichever browsers are evolved enough to show it, show it... and all the other browsers simply don't, yet can otherwise render the page!

Wow, brilliant. Wish I'd thought of it.

I know some people see the innerwebs as something like cable tv... you don't need it and you only get it if you're rich (or on welfare, I never understood why it works that way). But the internet is slowly becoming a necessity... around here, our banks are closing their brick and mortar buildings and telling their customers to do everything online. Seriously, I need access to my bank. "Best Viewed in Proprietary Browser Who Only Runs On Proprietary OS" is so not happening. Let's not go back to the Bad Old Days of segregation and separate fountains.

That this guy works at Facebook makes total sense though: my Konqueror (DESKTOP) browser always gets sent to some (doesn't work) WAP page, as if it were a mobile.

It's so overloaded with Javascript, instead of having working hyperlinks, everything is javascripted. Everything! You can't sneeze on Facebook unless you have extensive bs overbloated scripts running on your browser, and you know what? The BASIC application/idea of facebook does NOT need to be scripted. Maybe all the silly BS people like to do like giving hearts or seeds or whatever the heck all that is does, but those are games and it makes sense that people need a certain amount of scripting to do silly BS like that. But Facebook itself is terribly written, does not degrade well... have they ever heard of the terms unobtrusive scripting and graceful degredation?? I don't think so.


Fegall, I understand your reasoning but can't say I agree, at least not completely. Starndards are necessary to keep the sanity of our minds instead of writing the same code in 8000 different languages. And, based on those standards, there's innovation.

But I would say that starndards are needed for innovation.


Unfortunately that's exactly what happens... what that results in is a situation where the abandoned ones have to be supported for the basis of compatibility (otherwise you end up with complaints of stuff not working, things being broken and developers constantly having to throw away knowledge - or worse, give up trying to learn until the standard exists, therefore innovation is halted because people refuse to adopt to non-standards and they don't become standards because of a lack of willing to risk adoption. You see it all of the time, where lack of adoption rejects standards. stuck_out_tongue


To me it would seem it's an iterative process.

First you have standards, then there are innovations that get included into the new standard, after which there are new innovations that get included into the even newer standard, etc

So I think without standards there would be no (real) innovations, but without innovations there would be no (new) standards.


I think that probably sums it up best.


Only you don't always begin with standards.



Hm. Something with a chicken and an egg. What was it again? wink