15 ms·

Why did Alan Kay say the Internet was done well, but the Web was by amateurs?

102 points by yitchelle 14y ago

hp50g 14y ago
I agree entirely with this. the world wide web lacks any "engineering" at all. It just happened and is as a result, a bit of a mess. Even standards such as html5 don't get anywhere near making something coherent and consistent.
- BerislavLopac 14y ago
  Actually, the underlying foundations of WWW -- primarily HTTP and hypertext -- are excellently designed from the beginning. The problem is in fact that they were designed too well -- it was too easy to build a good-enough content that was easily accessible by a lot of users, and the real harm was done not by the amateurs, but by the "wrong" kind of professionals -- graphic designers and developers from other kinds of the field (databases, system, financials) who failed to understand the capabilities and specifics of this new platform. EDIT: Removed assertion that Web was implemented excellently from the beginning; I understand that HTTP 1.0 was inefficient, and there were other early issues. But the design was sound, and the implementation improved quickly.
  - hp50g 14y ago
    I disagree with the first bit. HTTP is an ambiguous, overcomplicated mess. I regularly have to wedge myself into the stack and deal with things like cookie directives, pipelining, persistent connections and utterly broken caching semantics etc. It's just horrible from end to end. HyperText itself is fundamentally a well engineered concept (I mean it worked fine for HyperCard etc) but to base the public facing WWW on SGML was just a plain horrible idea. For a number of years I actually preferred Gopher, WAIS and Usenet. I still do now when I think about it for the sheer simplicity and the fact it's designed to push indexes and information to rather than to ooze marketoid vomit. Agree with the "wrong kind of professionals" statement though.
    - dasil003 14y ago
      I see where both of you are coming from, but I can't help but notice the number of competing network protocols and formats that HTTP/HTML defeated in the market. Did the web succeed despite its technical failings or because of them?
      - dredmorbius 14y ago
        Possibly a mix. The fact that the protocols were free an unencumbered, and reference implementations of both server and client software were provided helped immensely as well. Any idiot could come along and either pick up working pieces or modify them. And they did. Tim O'Reilly's had a few things to say about watching the Web take off. Its primary competition was closed systems: either entirely closed networks, or proprietary protocols, or both. He wasn't willing to bet his company on any such thing. When the Web emerged, it was clear to him that it was the solution he'd been looking for.
  - pjmlp 14y ago
    I used to appreciate the initial web and happily changed from desktop based application development to web based ones. Nowadays, after so many scars from Web applications projects trying to bend the technology to mimic as close as possible the desktop with broken abstractions and browser issues, I gladly take desktop consulting projects over web ones.
    - hp50g 14y ago
      That's a similar path to the one I took and I thoroughly regret it.
qompiler 14y ago
HTML5 is being designed by "professionals" and I have never seen such a mess. Functionality is being introduced that shouldn't even be part of HTML, just look at the amount of new input types.
- Ygg2 14y ago
  At risk of downvotes I must ask why? New input are added probably to bring them more in line with system inputs of the operating systems.
  - qompiler 14y ago
    HTML should be a markup language, CSS should be the styling language and JavaScript should be used to handle matters such as input, processing and output. The principle of separation of concerns is very simple and elegant. In the long run it makes sure you don't get a mess of a framework in which things can be done in 50 ways, of which 25 work only in browser X and the other 25 behave differently in each browser.
    - Ygg2 14y ago
      I do somewhat agree, but how would Javascript be made to create system looking controls (i.e. Unity date controls on Ubuntu, Windows like on Windows 7, etc.), without adding some extensions for DOM? I personally dislike that CSS is turning into a miniature programming language, but I assume this is the way forward until something way better comes along and simply replaces HTML, CSS, JS. I mean its better to say width - 10 pixel than to make JS for each silly snipet like that.
      - pjmlp 14y ago
        By stopping trying to make a document technology bend into an application stack and use native applications instead? The more web projects I work on, the more I like desktop technologies.
        14y ago
ronilan 14y ago
He is right. But that's why the web is so wonderful - it is clearly man-made.
- jeremyswank 14y ago
  Agree. Beyond that, I'd say that it is more of an expression of culture than it is an expression of technology. (Carefully noting that these cannot be easily separated, anyway.) "The Web" as currently practiced is more of a craft (and at best, and art) than it is a technical exercise. Certainly, is inelegant and even ugly in places: much like the tool we humans (maybe should) value most: natural language. Natural language is fraught with inconsistencies and wtf's we have kind of made it up as we've gone along (standards academies notwithstanding). Think of how lovely an eflorescence it is of the workings of the mind, and how well it serves one of our deepest aspirations: to connect with another thinking being and share our thoughts.
fulafel 14y ago
Top answer attributes difference to web pioneers being "not standards people", wtf? professional standards committee attendees aren't exactly known for technical excellence.
liotier 14y ago
The whole reason why the HTML world took off while the existing technologies lingered is that amateurs could grab whatever they could and run with it - incompetence be damned ! Innovation often happens when amateurs find accessible technology and bend it in strange ways that make the professional cringe. Professional are then needed for security, scaling, cost reductions and all the things that require maturity but instead of innovating they would have refined the state of the art to new and still stagnating levels of perfection. Hurray for amateur enthusiasts who do interesting new things because no one told them it was the wrong way to do it !
- qw 14y ago
  I agree. Many underestimate how the lack of strictness helped early users. You could start with plain text and add a few <br> or maybe a <p> (wIthout end tags) to create paragraphs. Anyone with minimal knowledge could start sharing content. If they had to pass any kind of validation, many would not have bothered. Yes it was a mess, but it helped us get enough information to reach critical levels. I would rather have a web with messy HTML than FTP servers filled with Word documents or a proprietary AOL / MSN
  - KMag 14y ago
    The problem of broken markup would have been better solved on the server side, rather than the client side. As a general principle, pushing denormalized data as far to the edges of the system as possible minimizes complexity of the system as a whole.
- AlisdairO 14y ago
  Agreed! The thing about the web is that it was for amateurs. Things like separation of content and style are great for engineers creating larger-scale maintainable sites, but they're absolutely not helpful for new users. Being able to take a text file and add a couple of tags around some text you want to emphasise is a positive advantage for such individuals. It rewards small experiments. Understanding CSS is, by contrast, quite the undertaking. Hypertext had been around for quite some time before the web, and there's still some academics I know who seem to scratch their heads at the success of such an simplistic, impure form. To me, its simplicity is a reason for its success, while projects like Xanadu make a fundamental error in not considering enough the importance of barrier to entry.
jiggy2011 14y ago
The technologies that could be called "the internet" solve more well defined problems than the technologies that could be called "the web". TCP/IP for example is a description of an electronic locomotive network. It ferries packets around. What is in the packets? Who knows? Who cares? What about privacy, authentication, authenticity? Not our problem! The web on the other hand is a mishmash of adhoc parts, HTTP,HTTPS,HTML,Websock,DOM etc. Are we a client/server document protocol? Are we a distributed runtime? Do we guarantee secure communication? Perhaps. Are we stateless? Technically, but state tracking is what people use us for! We're shipping around a turing complete language everywhere, is this an integral part of the web? Or just something we slapped on top for good measure?
- czr80 14y ago
  It's interesting to note that choosing not to worry about those things ("privacy, authentication, authenticity") were very much design choices - perhaps the most consequential design choices ever?
  - DanBC 14y ago
    ...and it's a shame that the WWW is good enough for people to attempt to kludge something on top of it, and continually fail, while not being bad enough for anyone to succeed in launching alternatives that fix those problems. See also SMTP which has flooded the Internet with huge amounts of garbage traffic but which is simple and works well enough that any alternatives fail.
    - Silhouette 14y ago
      See also SMTP which has flooded the Internet with huge amounts of garbage traffic but which is simple and works well enough that any alternatives fail. I agree with your main point, but I wanted to add that there are plenty of alternatives to sending e-mails using SMTP these days. They mostly go by names like "Facebook" or "Google+" and run over protocols like HTTPS talking to a centralised service provider, and they solve the fundamental problem -- letting people send text and picture messages to friends/colleagues for viewing at their own convenience -- in a different and mostly incompatible way.
lukego 14y ago
I saw an interesting talk recently by internet visionary Scott Shenker and I wondered if he was suggesting the opposite: that the internet is a mess and they need to learn from software people. http://www.youtube.com/watch?v=YHeyuD89n1Y http://www.youtube.com/watch?v=YHeyuD89n1Y
dasil003 14y ago
The critique is a bit unfair because the Internet had clear design goals and criteria with which to apply rigorous engineering practice. The web by contrast was a essentially a document sharing and linking protocol and document format. It looks bad mostly because it outgrew it's design, but at the same time if it had been designed correctly for its eventual use it probably never would have taken off to begin with. Granted, it has fundamental warts as well, but I ascribe to that to the relative uncertainties of designing a higher level application that has at once more constraints and unanticipated usage. I don't know, perhaps Kay is right, being that he is 1000 times the engineer I am, but the web just seems such an amorphous thing compared to TCP/IP.
fpp 14y ago
The full quote from Alan Kay's interview with Dr Dobbs last year: ( http://www.drdobbs.com/architecture-and-design/interview-with-alan-kay/240003442 http://www.drdobbs.com/architecture-and-design/interview-wit... ) "...The Internet was done so well that most people think of it as a natural resource like the Pacific Ocean, rather than something that was man-made. When was the last time a technology with a scale like that was so error-free? The Web, in comparison, is a joke. The Web was done by amateurs..."
Lerc 14y ago
There seems to be rather a lot of consensus that the web is a mess but it is what we have. Have there been any serious attempts at producing a solution that is better designed that does not simply add new things to the spec. An argument can be made that if a better system existed it would not be able to compete because of inertia, but surely it would be worth at least making the attempt. I have scribbled notes and designs for years now exploring ideas of what could be done with a clean slate and hindsight that was unavailable to early browser designers. I can't be alone in this. Who else has explored the possibilities of what browsers could be?
- Ygg2 14y ago
  This was some time ago on HN http://axrproject.org/ http://axrproject.org/
  - Lerc 14y ago
    That does indeed look interesting. I shall take a look. I certainly hope there are more projects like this out there. There are many approaches that might be better or worse which might not be apparent until someone attempts it.
- lmm 14y ago
  XHTML2 (now abandoned) was a serious effort in that direction (while sticking quite close to existing HTML) - consistent syntax and rules for handling it. Honestly I don't think we'd see much change to what browsers can do - like turing-complete languages, everything is possible on the web, just not necessarily as easy as it should be. A lot of parts of the web I view as "assembly languages" - it would be possible to use better alternatives to HTML, CSS, Javascript and perhaps even HTTP, but it's not like we handle any of those directly. So no-one should care except a few library authors - and the libraries have mostly been written now.
atirip 14y ago
Well, in my book FTP is The Absolute Worst Protocol Ever Invented :-)
- vidarh 14y ago
  It's high on the list, certainly... I might be inclined to hate more on the SOAP designers these days, but FTP does very well on the "idiotic complexity in few pages" metric.
- protomyth 14y ago
  I think anyone who has ever had to edit firewall rules would agree with you.
- nonamegiven 14y ago
  From a readability or operational perspective? Yes, it was done much too early to take advantage of eventually emerged common ways of doing things (I would never guess in ftp, there's no rhyme and little reason), but files seem to go back and forth pretty reliably once you've properly asked for what you want.
xradionut 14y ago
Having been involved in computers since the 70's, the internet before WWW, and the web since, I come to the conclusion that the Web succeeded because it was barely good enough to work for amateurs, but no better. As hypertext it sucks. As a application platform it sucks too. But for allowing free exchange of information, it's been pretty damn good for such a crufty piece of tech.
trotsky 14y ago
http://1997.webhistory.org/www.lists/www-talk.1993q1/0182.html http://1997.webhistory.org/www.lists/www-talk.1993q1/0182.ht... http://1997.webhistory.org/www.lists/www-talk.1993q1/0197.html http://1997.webhistory.org/www.lists/www-talk.1993q1/0197.ht... Andressen: I'd like to propose adding this in this way. Berners-Lee: I don't think that's a good idea. Andressen: Oh, I totally agree. By the way, we're doing it anyway.
- rgbrenner 14y ago
  The entire thread, starting with your first link, is really great. Tim Berners-Lee even suggests 2 alternatives for the IMG tag. Thanks for that.
  - podperson 14y ago
    The alternatives are not much better (overloading a tags is more elegant and would probably have eliminated the later mess of needing embed tags and plugins for additional types of media). Even now we have img, audio, video, etc. instead of say a single well-behaved and extensive media tag that could also be used to include content from another page.
    - drzaiusapelord 14y ago
      I wonder if this was a concession to the limited browser code and slow networks at the time. With the img tag you know its going to be a jpg, gif, or png. If it was a media tag then you have no idea what it is. Imagine downloading a 10mb quicktime file or flash object on dial-up so the mime header can be read as you wait for the page to render. Its only recently that anyone has bothered to handle video in the browser. Could web servers back then just send the client the header of a file? I think these guys were working with a lot of ugly limitations that we've only recently overcome. Andreeson recently said that he expected the back and forward buttons to be temporary but no one thought of a good replacement for them. We still haven't.
  - vidarh 14y ago
    I can't really tell if Marc Andreessen were being sarcastic in response to the thread going off towards complexity-land or serious when he wrote this: http://1997.webhistory.org/www.lists/www-talk.1993q1/0209.html http://1997.webhistory.org/www.lists/www-talk.1993q1/0209.ht... > Actually, maybe we should think about a general-purpose procedural graphics language within which we can embed arbitrary hyperlinks attached to icons, images, or text, or anything. Has anyone else seen Intermedia's capabilities wrt this? It's one of their most impressive capabilities, actually. > >Something like a cross between PostScript and CGM might work... actually, maybe we should just use one or the other, and add the extensions we need for the links. Also we'd want to make sure that it's completely editable.
walshemj 14y ago
The internet was done by amateurs ISO was done by professionals :-) Which is a good job other wise we would all be using x.400 and x.500 et-al and you subs would take what your PTT gave you and like it. Of course I would have a cool mail address - but I was root on the UK's ADMD at one point ;-)
grey-area 14y ago
I think the most interesting question with regards to the web is how it compares to other sets of protocols/formats which try to do the same thing, not how it compares to the underlying transport (the internet). Why did the the web (i.e. HTML over http) succeed where so many other document markup formats and protocols had failed in the past - before the web we had Hypertext, SGML, gopher, FTP etc. but the web has overtaken them all (while indebted to all of them of course). HTML was the simplest solution we've yet found to a potentially very complex problem - how to share and read documents in a common format. It's certainly not perfect, but part of the reason for its popularity is this simplicity.
dredmorbius 14y ago
The Titanic was built by professionals. Noah's ark was built by an amateur.
- tincholio 14y ago
  With the difference that the Titanic actually existed
  - dredmorbius 14y ago
    The allegory is nonetheless useful. There are other more concrete examples. In the technical world, there's the contrast, say of Linux with other non-free OSes, ranging from BeOS to VMS to WinNT. There's Wikipedia vs. Brittanica or Encarta. There's free markets vs. command economies (or both ends vs. a mixed social market economy). Many 17th, 18th, and 19th century advances and institutions in both science and letters were founded significantly on the efforts of amateurs. The deeper point is that the root of "amateur" isn't some connotation of naive, inexperience, or lesser quality, but "amore", literally, love. And love-of-task is a strong motivator for producing high-quality work.
- kps 14y ago
  But Noah merely implemented the spec; the actual design work was done by an infallible God — making it the canonical example of a water-fall process.
jbert 14y ago
Standards bodies ossify over time. In Europe in the 80s/early 90s, X.25 comms, full seven-layer OSI stack and X.400 email was the "real thing", standardised by OSI, mandated by government procurement and implemented by telcos. This was "commercial grade" networking and email, with TCP/IP and SMTP being derided as "academic". The IETF were the scrappy, pragmatic upstart. "We believe in rough consensus and working code". I've not followed many later IETF standardisations, but I did take a passing interest in CAP (calendar access protocol) a few years back. (since there wasn't a good open source calendaring stack (not sure there is now, either, tbh). It seems to me as a distant observer/potential implementor, CAP would be very hard to get started on. It mandated the use of otherwise unused-at-the-time BEEP [http://tools.ietf.org/html/rfc3080 http://tools.ietf.org/html/rfc3080] (for good reasons - the same good reasons OSI mandated their full 7 layer stack...), embedded (unspecified parts or SQL92 (without a grammar!)) and seems to me a daunting prospect. It looked to me as though they were really saying "everyone doing CAP has to really have an SQL backing store with a particular schema and you pretty much have to pass these queries through to it". So I see OSI->IETF->W3C as a progression over time. Not sure where to now... [I'm sorry if I'm harsh to the people who put a lot of effort into CAP. I didn't contribute and so I don't really have the right to criticise, but from a potential implementors pt of view, it seemed daunting. I'm also not tracking any current IETF work, so I know I have a narrow view and am likely wrong in many ways. update: ...and I've just skimmed the final RFC and they do have a grammar for the SQL subset. I recall that that wasn't present in the drafts available at the time.]
- kabdib 14y ago
  A standard should reference working code. Preferably more than one implementation by different parties. Ideally at least one implementation is open source. This doesn't mean that standards based on working code is automatically superior, but it's less likely you'll get absolute turkeys.
pfortuny 14y ago
Pufff, there is a tremendous bias on 'raw functionality' in that statement: the Internet was so well done BUT FOR SECURITY issues, which were never taken into account. This detail has difficulted development to a huge extent. One could say that the Internet was so well done for 'utopian citicens'. Not that it is bad, but we have learnt a lot for the future from those mistakes. I hope we have learnt. The routing protocols, DNS etc (which are essential for the Internet) have so many gaping security holes that today we wonder how naive one can be when designing communication protocols...
- gizmo686 14y ago
  >One could say that the Internet was so well done for 'utopian citicens'. Even for utopian societies, some of the Internet's security problems would still be problems. The main issue is that for the most part, the protocol assumes that information provided by another source is correct. This could clearly be used mallisiously, but it also allows small, innocent mistakes to take down the network. For example, back in 2008, Pakistan decided to block youtube within their country. While this specific example would not happen in a utopian society, there could be valid reasons that a sub-network would want to disable or redirect requests going to a different server. What happened was that one of the ISPs, instead of simply blocking outgoing requests to youtube, advertised itself as youtube. This change propogated through the network, and youtube went down globally. Still, when these issues do come up without any malice, they should be relatively easy and painless to fix.
acdha 14y ago
I read this as Kay being unfamiliar enough with the lower level protocols to assume they're significantly cleaner than the higher level web. The “designed by professionals” era he's talking about still had major problems with security (spoofing is still too easy), reliability and performance which is why there's still new work being done tuning everything for high speed or high packet loss links. Go back just a little further and hostnames were resolved by searching a text file which people had to distribute! Both systems are complex heterogenous systems and have significant backwards compatibility challenges any time you want to fix a wart. It's easy to spot problems, hard to fix them, and as the array of failed competitors to eithet shows it's surprisingly hard to design something equivalent without going through the same learning curve. As a biologist might tell an intelligent design proponent, if you look at one and see genius design you're not looking closely enough.
- noblethrasher 14y ago
  Interestingly enough, Alan Kay, who has an undergaraduate degree in the subject, was inspired by biology (particularly the way that field discovers and organizes knowledge) when he designed Smalltalk.
pfraze 14y ago
Alan Kay was talking about the program model. There's no isolated module below the document namespace, leading to monolithic web apps and a lack of user-configurability. The browser needs to run more than one program at a time to be a serious OS. I imagine he compared the browser pretty directly to Smalltalk.
robomartin 14y ago
Here's an interesting and very relevant TED talk: http://www.ted.com/talks/danny_hillis_the_internet_could_crash_we_need_a_plan_b.html http://www.ted.com/talks/danny_hillis_the_internet_could_cra... The Internet was not conceived to be what it has become. Having had a front row seat to watch it go from academia to what it has become today one can only be in awe of what --I'm gonna say it-- private enterprise, capitalism and the drive to succeed can produce. Yes, it is not perfect. What project at this scale is?
- cma 14y ago
  > private enterprise, capitalism Shouldn't you also include public spending's role; e.g. CERN?
  - robomartin 14y ago
    The explosion of the Internet into every nook and cranny is and was a 100% due to massive private investment at all levels. It was a rush for gold. Fortunes made and lost. EDIT: Strange words replaced. Damn iPad auto-text!
    - cma 14y ago
      The Internet is different from the Web. It is off topic, but if you want to go into that, the Internet as distinct from the Web had an even bigger public kick-start.
      - robomartin 14y ago
        Of course the two are interdependent. And both were grown to what we have today by private enterprise not government. Knowing what I know I'll venture to say that the public sector contribution to building either the web or the Internet is just-about microscopic when compared to private investment in both time and money. Yes, there was a kick-start phase, but it was nothing compared to what came later. Furthermore, it was done without intent, purpose or even an idea of what it could become. It was simply a communications network for the military at first and for a select few universities later. In the TED video I linked the presenter is holding a small book that contained every single person using the Internet at the time. What's interesting is that he registered the third Internet domain, so he was there from the very start. I was peripherally involved with a couple of interesting private efforts in the early days of the gold rush. One of them was a group led by Phil Anschutz. He owned the right-of-way along major railway routes. They trenches the heck out of those routes to lay in massive amounts of dark fiber. A huge investment in the future that I am sure payed off handsomely.
AnthonyMouse 14y ago
It seems like people are talking about a bunch of stuff like the ability of amateurs to create a web page easily and that sort of thing. Maybe some of that stuff helped. But I don't think those were the big reasons. Look at how the web works: 1) Web pages work on all relevant platforms. If someone creates a new browser, most existing web pages will render on it, and failures are in the nature of bug fixes rather than complete rewrites. Compare this with e.g. new operating systems and C programs. 2) If you change the web page, the update is available immediately to everyone. Compare this to the morass of outdated, incompatible and vulnerable software on your typical Windows box. 3) All the code effectively runs in a sandbox. Users generally don't have to worry about clicking on a link and having it send virus spam to everyone in their contact list. Those are some huge advantages. Advantages that more than make up for the fact that web development is a huge pile of kludges built against each other that only hasn't been discarded and replaced entirely as a result of the massive amount of mutual interdependence. The primary competitor to the web was Java. But Java was hampered by horrible mismanagement on the part of Sun and repeated malicious attempts to destroy it on the part of Microsoft, so the web won. Which is probably unfortunate. Imagine a web where Java (or Go) rather than Javascript was the primary language.
martinced 14y ago
The Web and all the technologies that come with it is a special kind of hell: a Rube Goldberg machine whose level of complexity and crazyness is hard to match. The associated mediocrity is kinda jaw-dropping too: it's 2013 and the main "engine" used to run all these client-side apps in the browser is still... Single-threaded. Seriously. I'm not saying "language" because JavaScript in itself has so many warts that there are several technologies out there who try to dodge the language altogether and directly generate JavaScript source code... Even Douglas Crockford himself is criticizing JavaScript a lot. But it's not just JS... The madness simply never stops: HTML, (X)HTML, CSS (oh, that one...), JS. And all the incomplete and overly complicated "standards" and, of course, all the competing implementation. I'm always amazed that we're in 2013 and a lot of the demo page posted on HN still cannot serve videos correctly to people who don't have Flash in their browsers. That's the current state of affair: most web devs still aren't able to serve various video formats to please everyone. Why isn't this something that comes for free with the webapp server? Is it rocket-science to encode in different formats and serve the correct one? Apparently it is. So imagine for more complicated things... To me the most fascinating is that hardly anyone steps back and wonders "Why do we have such an incredible mess?". But that's what we have and a big part of software development is now for the Web and the trend ain't stopping anytime soon: be it server-side or client-side, it's a browsers world nowadays. So let's stop bitching, let's learn the warts and ins and outs of these uber-shitty technologies and let's get back to work ; )
- robomartin 14y ago
  > "Why do we have such an incredible mess?" Because this wasn't (and couldn't be) centrally planned. It has very much evolved in a fashion analogous to the theory of biological evolution. Lots of experiments of different sizes and with different levels of success. We have yet to identify a more powerful problem solving paradigm. Given enough time evolution by natural selection can produce systems of incredible complexity that actually work pretty well. Not without warts, and, yes, you have to be patient, but it works.
- pornel 14y ago
  Video is not a technical problem, but a legal/political/economic one (technically good codecs covered by patents & corporations unwilling to adopt to other codecs).
andrewbinstock 14y ago
I did the interview with Alan. Let me give the larger context. The interview was done shortly after Kay had shown a presentation done 50 years ago of one of the very first GUI programs. He wondered then and later in this interview why GUIs had essentially not evolved much in the intervening period, despite the huge jumps in technology in other areas. With me, he discussed how UIs were uninviting and mystifying to people who don't know how to use them (children and seniors). In addition, he felt the Web provided very little that would foster the individual's learning on his/her own terms. So while I suspect he'd agree with many of the comments posted here about the Web, he was not primarily speaking about protocols and bits and bytes, but by the end result: a presentation quality that only serves up what a TV can do, rather than deeply engaging individuals.

102 points by yitchelle 14y ago

hp50g 14y ago

I agree entirely with this. the world wide web lacks any "engineering" at all. It just happened and is as a result, a bit of a mess. Even standards such as html5 don't get anywhere near making something coherent and consistent.

BerislavLopac 14y ago
Actually, the underlying foundations of WWW -- primarily HTTP and hypertext -- are excellently designed from the beginning. The problem is in fact that they were designed too well -- it was too easy to build a good-enough content that was easily accessible by a lot of users, and the real harm was done not by the amateurs, but by the "wrong" kind of professionals -- graphic designers and developers from other kinds of the field (databases, system, financials) who failed to understand the capabilities and specifics of this new platform. EDIT: Removed assertion that Web was implemented excellently from the beginning; I understand that HTTP 1.0 was inefficient, and there were other early issues. But the design was sound, and the implementation improved quickly.
- hp50g 14y ago
  I disagree with the first bit. HTTP is an ambiguous, overcomplicated mess. I regularly have to wedge myself into the stack and deal with things like cookie directives, pipelining, persistent connections and utterly broken caching semantics etc. It's just horrible from end to end. HyperText itself is fundamentally a well engineered concept (I mean it worked fine for HyperCard etc) but to base the public facing WWW on SGML was just a plain horrible idea. For a number of years I actually preferred Gopher, WAIS and Usenet. I still do now when I think about it for the sheer simplicity and the fact it's designed to push indexes and information to rather than to ooze marketoid vomit. Agree with the "wrong kind of professionals" statement though.
  - dasil003 14y ago
    I see where both of you are coming from, but I can't help but notice the number of competing network protocols and formats that HTTP/HTML defeated in the market. Did the web succeed despite its technical failings or because of them?
    - dredmorbius 14y ago
      Possibly a mix. The fact that the protocols were free an unencumbered, and reference implementations of both server and client software were provided helped immensely as well. Any idiot could come along and either pick up working pieces or modify them. And they did. Tim O'Reilly's had a few things to say about watching the Web take off. Its primary competition was closed systems: either entirely closed networks, or proprietary protocols, or both. He wasn't willing to bet his company on any such thing. When the Web emerged, it was clear to him that it was the solution he'd been looking for.
- pjmlp 14y ago
  I used to appreciate the initial web and happily changed from desktop based application development to web based ones. Nowadays, after so many scars from Web applications projects trying to bend the technology to mimic as close as possible the desktop with broken abstractions and browser issues, I gladly take desktop consulting projects over web ones.
  - hp50g 14y ago
    That's a similar path to the one I took and I thoroughly regret it.

qompiler 14y ago

HTML5 is being designed by "professionals" and I have never seen such a mess. Functionality is being introduced that shouldn't even be part of HTML, just look at the amount of new input types.

Ygg2 14y ago
At risk of downvotes I must ask why? New input are added probably to bring them more in line with system inputs of the operating systems.
- qompiler 14y ago
  HTML should be a markup language, CSS should be the styling language and JavaScript should be used to handle matters such as input, processing and output. The principle of separation of concerns is very simple and elegant. In the long run it makes sure you don't get a mess of a framework in which things can be done in 50 ways, of which 25 work only in browser X and the other 25 behave differently in each browser.
  - Ygg2 14y ago
    I do somewhat agree, but how would Javascript be made to create system looking controls (i.e. Unity date controls on Ubuntu, Windows like on Windows 7, etc.), without adding some extensions for DOM? I personally dislike that CSS is turning into a miniature programming language, but I assume this is the way forward until something way better comes along and simply replaces HTML, CSS, JS. I mean its better to say width - 10 pixel than to make JS for each silly snipet like that.
    - pjmlp 14y ago
      By stopping trying to make a document technology bend into an application stack and use native applications instead? The more web projects I work on, the more I like desktop technologies.
      - 14y ago

ronilan 14y ago

He is right. But that's why the web is so wonderful - it is clearly man-made.

jeremyswank 14y ago
Agree. Beyond that, I'd say that it is more of an expression of culture than it is an expression of technology. (Carefully noting that these cannot be easily separated, anyway.) "The Web" as currently practiced is more of a craft (and at best, and art) than it is a technical exercise. Certainly, is inelegant and even ugly in places: much like the tool we humans (maybe should) value most: natural language. Natural language is fraught with inconsistencies and wtf's we have kind of made it up as we've gone along (standards academies notwithstanding). Think of how lovely an eflorescence it is of the workings of the mind, and how well it serves one of our deepest aspirations: to connect with another thinking being and share our thoughts.

fulafel 14y ago

Top answer attributes difference to web pioneers being "not standards people", wtf? professional standards committee attendees aren't exactly known for technical excellence.

liotier 14y ago

The whole reason why the HTML world took off while the existing technologies lingered is that amateurs could grab whatever they could and run with it - incompetence be damned ! Innovation often happens when amateurs find accessible technology and bend it in strange ways that make the professional cringe. Professional are then needed for security, scaling, cost reductions and all the things that require maturity but instead of innovating they would have refined the state of the art to new and still stagnating levels of perfection. Hurray for amateur enthusiasts who do interesting new things because no one told them it was the wrong way to do it !

qw 14y ago
I agree. Many underestimate how the lack of strictness helped early users. You could start with plain text and add a few <br> or maybe a <p> (wIthout end tags) to create paragraphs. Anyone with minimal knowledge could start sharing content. If they had to pass any kind of validation, many would not have bothered. Yes it was a mess, but it helped us get enough information to reach critical levels. I would rather have a web with messy HTML than FTP servers filled with Word documents or a proprietary AOL / MSN
- KMag 14y ago
  The problem of broken markup would have been better solved on the server side, rather than the client side. As a general principle, pushing denormalized data as far to the edges of the system as possible minimizes complexity of the system as a whole.
AlisdairO 14y ago
Agreed! The thing about the web is that it was for amateurs. Things like separation of content and style are great for engineers creating larger-scale maintainable sites, but they're absolutely not helpful for new users. Being able to take a text file and add a couple of tags around some text you want to emphasise is a positive advantage for such individuals. It rewards small experiments. Understanding CSS is, by contrast, quite the undertaking. Hypertext had been around for quite some time before the web, and there's still some academics I know who seem to scratch their heads at the success of such an simplistic, impure form. To me, its simplicity is a reason for its success, while projects like Xanadu make a fundamental error in not considering enough the importance of barrier to entry.

jiggy2011 14y ago

The technologies that could be called "the internet" solve more well defined problems than the technologies that could be called "the web". TCP/IP for example is a description of an electronic locomotive network. It ferries packets around. What is in the packets? Who knows? Who cares? What about privacy, authentication, authenticity? Not our problem! The web on the other hand is a mishmash of adhoc parts, HTTP,HTTPS,HTML,Websock,DOM etc. Are we a client/server document protocol? Are we a distributed runtime? Do we guarantee secure communication? Perhaps. Are we stateless? Technically, but state tracking is what people use us for! We're shipping around a turing complete language everywhere, is this an integral part of the web? Or just something we slapped on top for good measure?

czr80 14y ago
It's interesting to note that choosing not to worry about those things ("privacy, authentication, authenticity") were very much design choices - perhaps the most consequential design choices ever?
- DanBC 14y ago
  ...and it's a shame that the WWW is good enough for people to attempt to kludge something on top of it, and continually fail, while not being bad enough for anyone to succeed in launching alternatives that fix those problems. See also SMTP which has flooded the Internet with huge amounts of garbage traffic but which is simple and works well enough that any alternatives fail.
  - Silhouette 14y ago
    See also SMTP which has flooded the Internet with huge amounts of garbage traffic but which is simple and works well enough that any alternatives fail. I agree with your main point, but I wanted to add that there are plenty of alternatives to sending e-mails using SMTP these days. They mostly go by names like "Facebook" or "Google+" and run over protocols like HTTPS talking to a centralised service provider, and they solve the fundamental problem -- letting people send text and picture messages to friends/colleagues for viewing at their own convenience -- in a different and mostly incompatible way.

lukego 14y ago

I saw an interesting talk recently by internet visionary Scott Shenker and I wondered if he was suggesting the opposite: that the internet is a mess and they need to learn from software people. http://www.youtube.com/watch?v=YHeyuD89n1Y http://www.youtube.com/watch?v=YHeyuD89n1Y

dasil003 14y ago

The critique is a bit unfair because the Internet had clear design goals and criteria with which to apply rigorous engineering practice. The web by contrast was a essentially a document sharing and linking protocol and document format. It looks bad mostly because it outgrew it's design, but at the same time if it had been designed correctly for its eventual use it probably never would have taken off to begin with. Granted, it has fundamental warts as well, but I ascribe to that to the relative uncertainties of designing a higher level application that has at once more constraints and unanticipated usage. I don't know, perhaps Kay is right, being that he is 1000 times the engineer I am, but the web just seems such an amorphous thing compared to TCP/IP.

fpp 14y ago

The full quote from Alan Kay's interview with Dr Dobbs last year: ( http://www.drdobbs.com/architecture-and-design/interview-with-alan-kay/240003442 http://www.drdobbs.com/architecture-and-design/interview-wit... ) "...The Internet was done so well that most people think of it as a natural resource like the Pacific Ocean, rather than something that was man-made. When was the last time a technology with a scale like that was so error-free? The Web, in comparison, is a joke. The Web was done by amateurs..."

Lerc 14y ago

There seems to be rather a lot of consensus that the web is a mess but it is what we have. Have there been any serious attempts at producing a solution that is better designed that does not simply add new things to the spec. An argument can be made that if a better system existed it would not be able to compete because of inertia, but surely it would be worth at least making the attempt. I have scribbled notes and designs for years now exploring ideas of what could be done with a clean slate and hindsight that was unavailable to early browser designers. I can't be alone in this. Who else has explored the possibilities of what browsers could be?

Ygg2 14y ago
This was some time ago on HN http://axrproject.org/ http://axrproject.org/
- Lerc 14y ago
  That does indeed look interesting. I shall take a look. I certainly hope there are more projects like this out there. There are many approaches that might be better or worse which might not be apparent until someone attempts it.
lmm 14y ago
XHTML2 (now abandoned) was a serious effort in that direction (while sticking quite close to existing HTML) - consistent syntax and rules for handling it. Honestly I don't think we'd see much change to what browsers can do - like turing-complete languages, everything is possible on the web, just not necessarily as easy as it should be. A lot of parts of the web I view as "assembly languages" - it would be possible to use better alternatives to HTML, CSS, Javascript and perhaps even HTTP, but it's not like we handle any of those directly. So no-one should care except a few library authors - and the libraries have mostly been written now.

atirip 14y ago

Well, in my book FTP is The Absolute Worst Protocol Ever Invented :-)

vidarh 14y ago
It's high on the list, certainly... I might be inclined to hate more on the SOAP designers these days, but FTP does very well on the "idiotic complexity in few pages" metric.
protomyth 14y ago
I think anyone who has ever had to edit firewall rules would agree with you.
nonamegiven 14y ago
From a readability or operational perspective? Yes, it was done much too early to take advantage of eventually emerged common ways of doing things (I would never guess in ftp, there's no rhyme and little reason), but files seem to go back and forth pretty reliably once you've properly asked for what you want.

xradionut 14y ago

Having been involved in computers since the 70's, the internet before WWW, and the web since, I come to the conclusion that the Web succeeded because it was barely good enough to work for amateurs, but no better. As hypertext it sucks. As a application platform it sucks too. But for allowing free exchange of information, it's been pretty damn good for such a crufty piece of tech.

trotsky 14y ago

http://1997.webhistory.org/www.lists/www-talk.1993q1/0182.html http://1997.webhistory.org/www.lists/www-talk.1993q1/0182.ht... http://1997.webhistory.org/www.lists/www-talk.1993q1/0197.html http://1997.webhistory.org/www.lists/www-talk.1993q1/0197.ht... Andressen: I'd like to propose adding this in this way. Berners-Lee: I don't think that's a good idea. Andressen: Oh, I totally agree. By the way, we're doing it anyway.

rgbrenner 14y ago
The entire thread, starting with your first link, is really great. Tim Berners-Lee even suggests 2 alternatives for the IMG tag. Thanks for that.
- podperson 14y ago
  The alternatives are not much better (overloading a tags is more elegant and would probably have eliminated the later mess of needing embed tags and plugins for additional types of media). Even now we have img, audio, video, etc. instead of say a single well-behaved and extensive media tag that could also be used to include content from another page.
  - drzaiusapelord 14y ago
    I wonder if this was a concession to the limited browser code and slow networks at the time. With the img tag you know its going to be a jpg, gif, or png. If it was a media tag then you have no idea what it is. Imagine downloading a 10mb quicktime file or flash object on dial-up so the mime header can be read as you wait for the page to render. Its only recently that anyone has bothered to handle video in the browser. Could web servers back then just send the client the header of a file? I think these guys were working with a lot of ugly limitations that we've only recently overcome. Andreeson recently said that he expected the back and forward buttons to be temporary but no one thought of a good replacement for them. We still haven't.
- vidarh 14y ago
  I can't really tell if Marc Andreessen were being sarcastic in response to the thread going off towards complexity-land or serious when he wrote this: http://1997.webhistory.org/www.lists/www-talk.1993q1/0209.html http://1997.webhistory.org/www.lists/www-talk.1993q1/0209.ht... > Actually, maybe we should think about a general-purpose procedural graphics language within which we can embed arbitrary hyperlinks attached to icons, images, or text, or anything. Has anyone else seen Intermedia's capabilities wrt this? It's one of their most impressive capabilities, actually. > >Something like a cross between PostScript and CGM might work... actually, maybe we should just use one or the other, and add the extensions we need for the links. Also we'd want to make sure that it's completely editable.

walshemj 14y ago

The internet was done by amateurs ISO was done by professionals :-) Which is a good job other wise we would all be using x.400 and x.500 et-al and you subs would take what your PTT gave you and like it. Of course I would have a cool mail address - but I was root on the UK's ADMD at one point ;-)

grey-area 14y ago

I think the most interesting question with regards to the web is how it compares to other sets of protocols/formats which try to do the same thing, not how it compares to the underlying transport (the internet). Why did the the web (i.e. HTML over http) succeed where so many other document markup formats and protocols had failed in the past - before the web we had Hypertext, SGML, gopher, FTP etc. but the web has overtaken them all (while indebted to all of them of course). HTML was the simplest solution we've yet found to a potentially very complex problem - how to share and read documents in a common format. It's certainly not perfect, but part of the reason for its popularity is this simplicity.

dredmorbius 14y ago

The Titanic was built by professionals. Noah's ark was built by an amateur.

tincholio 14y ago
With the difference that the Titanic actually existed
- dredmorbius 14y ago
  The allegory is nonetheless useful. There are other more concrete examples. In the technical world, there's the contrast, say of Linux with other non-free OSes, ranging from BeOS to VMS to WinNT. There's Wikipedia vs. Brittanica or Encarta. There's free markets vs. command economies (or both ends vs. a mixed social market economy). Many 17th, 18th, and 19th century advances and institutions in both science and letters were founded significantly on the efforts of amateurs. The deeper point is that the root of "amateur" isn't some connotation of naive, inexperience, or lesser quality, but "amore", literally, love. And love-of-task is a strong motivator for producing high-quality work.
kps 14y ago
But Noah merely implemented the spec; the actual design work was done by an infallible God — making it the canonical example of a water-fall process.

jbert 14y ago

Standards bodies ossify over time. In Europe in the 80s/early 90s, X.25 comms, full seven-layer OSI stack and X.400 email was the "real thing", standardised by OSI, mandated by government procurement and implemented by telcos. This was "commercial grade" networking and email, with TCP/IP and SMTP being derided as "academic". The IETF were the scrappy, pragmatic upstart. "We believe in rough consensus and working code". I've not followed many later IETF standardisations, but I did take a passing interest in CAP (calendar access protocol) a few years back. (since there wasn't a good open source calendaring stack (not sure there is now, either, tbh). It seems to me as a distant observer/potential implementor, CAP would be very hard to get started on. It mandated the use of otherwise unused-at-the-time BEEP [http://tools.ietf.org/html/rfc3080 http://tools.ietf.org/html/rfc3080] (for good reasons - the same good reasons OSI mandated their full 7 layer stack...), embedded (unspecified parts or SQL92 (without a grammar!)) and seems to me a daunting prospect. It looked to me as though they were really saying "everyone doing CAP has to really have an SQL backing store with a particular schema and you pretty much have to pass these queries through to it". So I see OSI->IETF->W3C as a progression over time. Not sure where to now... [I'm sorry if I'm harsh to the people who put a lot of effort into CAP. I didn't contribute and so I don't really have the right to criticise, but from a potential implementors pt of view, it seemed daunting. I'm also not tracking any current IETF work, so I know I have a narrow view and am likely wrong in many ways. update: ...and I've just skimmed the final RFC and they do have a grammar for the SQL subset. I recall that that wasn't present in the drafts available at the time.]

kabdib 14y ago
A standard should reference working code. Preferably more than one implementation by different parties. Ideally at least one implementation is open source. This doesn't mean that standards based on working code is automatically superior, but it's less likely you'll get absolute turkeys.

pfortuny 14y ago

Pufff, there is a tremendous bias on 'raw functionality' in that statement: the Internet was so well done BUT FOR SECURITY issues, which were never taken into account. This detail has difficulted development to a huge extent. One could say that the Internet was so well done for 'utopian citicens'. Not that it is bad, but we have learnt a lot for the future from those mistakes. I hope we have learnt. The routing protocols, DNS etc (which are essential for the Internet) have so many gaping security holes that today we wonder how naive one can be when designing communication protocols...

gizmo686 14y ago
>One could say that the Internet was so well done for 'utopian citicens'. Even for utopian societies, some of the Internet's security problems would still be problems. The main issue is that for the most part, the protocol assumes that information provided by another source is correct. This could clearly be used mallisiously, but it also allows small, innocent mistakes to take down the network. For example, back in 2008, Pakistan decided to block youtube within their country. While this specific example would not happen in a utopian society, there could be valid reasons that a sub-network would want to disable or redirect requests going to a different server. What happened was that one of the ISPs, instead of simply blocking outgoing requests to youtube, advertised itself as youtube. This change propogated through the network, and youtube went down globally. Still, when these issues do come up without any malice, they should be relatively easy and painless to fix.

acdha 14y ago

I read this as Kay being unfamiliar enough with the lower level protocols to assume they're significantly cleaner than the higher level web. The “designed by professionals” era he's talking about still had major problems with security (spoofing is still too easy), reliability and performance which is why there's still new work being done tuning everything for high speed or high packet loss links. Go back just a little further and hostnames were resolved by searching a text file which people had to distribute! Both systems are complex heterogenous systems and have significant backwards compatibility challenges any time you want to fix a wart. It's easy to spot problems, hard to fix them, and as the array of failed competitors to eithet shows it's surprisingly hard to design something equivalent without going through the same learning curve. As a biologist might tell an intelligent design proponent, if you look at one and see genius design you're not looking closely enough.

noblethrasher 14y ago
Interestingly enough, Alan Kay, who has an undergaraduate degree in the subject, was inspired by biology (particularly the way that field discovers and organizes knowledge) when he designed Smalltalk.

pfraze 14y ago

Alan Kay was talking about the program model. There's no isolated module below the document namespace, leading to monolithic web apps and a lack of user-configurability. The browser needs to run more than one program at a time to be a serious OS. I imagine he compared the browser pretty directly to Smalltalk.

robomartin 14y ago

Here's an interesting and very relevant TED talk: http://www.ted.com/talks/danny_hillis_the_internet_could_crash_we_need_a_plan_b.html http://www.ted.com/talks/danny_hillis_the_internet_could_cra... The Internet was not conceived to be what it has become. Having had a front row seat to watch it go from academia to what it has become today one can only be in awe of what --I'm gonna say it-- private enterprise, capitalism and the drive to succeed can produce. Yes, it is not perfect. What project at this scale is?

cma 14y ago
> private enterprise, capitalism Shouldn't you also include public spending's role; e.g. CERN?
- robomartin 14y ago
  The explosion of the Internet into every nook and cranny is and was a 100% due to massive private investment at all levels. It was a rush for gold. Fortunes made and lost. EDIT: Strange words replaced. Damn iPad auto-text!
  - cma 14y ago
    The Internet is different from the Web. It is off topic, but if you want to go into that, the Internet as distinct from the Web had an even bigger public kick-start.
    - robomartin 14y ago
      Of course the two are interdependent. And both were grown to what we have today by private enterprise not government. Knowing what I know I'll venture to say that the public sector contribution to building either the web or the Internet is just-about microscopic when compared to private investment in both time and money. Yes, there was a kick-start phase, but it was nothing compared to what came later. Furthermore, it was done without intent, purpose or even an idea of what it could become. It was simply a communications network for the military at first and for a select few universities later. In the TED video I linked the presenter is holding a small book that contained every single person using the Internet at the time. What's interesting is that he registered the third Internet domain, so he was there from the very start. I was peripherally involved with a couple of interesting private efforts in the early days of the gold rush. One of them was a group led by Phil Anschutz. He owned the right-of-way along major railway routes. They trenches the heck out of those routes to lay in massive amounts of dark fiber. A huge investment in the future that I am sure payed off handsomely.

AnthonyMouse 14y ago

It seems like people are talking about a bunch of stuff like the ability of amateurs to create a web page easily and that sort of thing. Maybe some of that stuff helped. But I don't think those were the big reasons. Look at how the web works: 1) Web pages work on all relevant platforms. If someone creates a new browser, most existing web pages will render on it, and failures are in the nature of bug fixes rather than complete rewrites. Compare this with e.g. new operating systems and C programs. 2) If you change the web page, the update is available immediately to everyone. Compare this to the morass of outdated, incompatible and vulnerable software on your typical Windows box. 3) All the code effectively runs in a sandbox. Users generally don't have to worry about clicking on a link and having it send virus spam to everyone in their contact list. Those are some huge advantages. Advantages that more than make up for the fact that web development is a huge pile of kludges built against each other that only hasn't been discarded and replaced entirely as a result of the massive amount of mutual interdependence. The primary competitor to the web was Java. But Java was hampered by horrible mismanagement on the part of Sun and repeated malicious attempts to destroy it on the part of Microsoft, so the web won. Which is probably unfortunate. Imagine a web where Java (or Go) rather than Javascript was the primary language.

martinced 14y ago

The Web and all the technologies that come with it is a special kind of hell: a Rube Goldberg machine whose level of complexity and crazyness is hard to match. The associated mediocrity is kinda jaw-dropping too: it's 2013 and the main "engine" used to run all these client-side apps in the browser is still... Single-threaded. Seriously. I'm not saying "language" because JavaScript in itself has so many warts that there are several technologies out there who try to dodge the language altogether and directly generate JavaScript source code... Even Douglas Crockford himself is criticizing JavaScript a lot. But it's not just JS... The madness simply never stops: HTML, (X)HTML, CSS (oh, that one...), JS. And all the incomplete and overly complicated "standards" and, of course, all the competing implementation. I'm always amazed that we're in 2013 and a lot of the demo page posted on HN still cannot serve videos correctly to people who don't have Flash in their browsers. That's the current state of affair: most web devs still aren't able to serve various video formats to please everyone. Why isn't this something that comes for free with the webapp server? Is it rocket-science to encode in different formats and serve the correct one? Apparently it is. So imagine for more complicated things... To me the most fascinating is that hardly anyone steps back and wonders "Why do we have such an incredible mess?". But that's what we have and a big part of software development is now for the Web and the trend ain't stopping anytime soon: be it server-side or client-side, it's a browsers world nowadays. So let's stop bitching, let's learn the warts and ins and outs of these uber-shitty technologies and let's get back to work ; )

robomartin 14y ago
> "Why do we have such an incredible mess?" Because this wasn't (and couldn't be) centrally planned. It has very much evolved in a fashion analogous to the theory of biological evolution. Lots of experiments of different sizes and with different levels of success. We have yet to identify a more powerful problem solving paradigm. Given enough time evolution by natural selection can produce systems of incredible complexity that actually work pretty well. Not without warts, and, yes, you have to be patient, but it works.
pornel 14y ago
Video is not a technical problem, but a legal/political/economic one (technically good codecs covered by patents & corporations unwilling to adopt to other codecs).

andrewbinstock 14y ago

I did the interview with Alan. Let me give the larger context. The interview was done shortly after Kay had shown a presentation done 50 years ago of one of the very first GUI programs. He wondered then and later in this interview why GUIs had essentially not evolved much in the intervening period, despite the huge jumps in technology in other areas. With me, he discussed how UIs were uninviting and mystifying to people who don't know how to use them (children and seniors). In addition, he felt the Web provided very little that would foster the individual's learning on his/her own terms. So while I suspect he'd agree with many of the comments posted here about the Web, he was not primarily speaking about protocols and bits and bytes, but by the end result: a presentation quality that only serves up what a TV can do, rather than deeply engaging individuals.