About the future of the web.

Posted 15 Mar 2008 at 19:04 UTC by DeepNorth Share This

The following questions were posed by someone on LinkedIn and I answer them here. I think they are timely, interesting and important: "About the future of the web: what do you miss, what do you hate? 1. What would you like to change on the web? 2. What would you really want to keep? 3. What are the technological chances for internet? 4. What are the threats?"

Somebody on LinkedIn posed the following questions:

1. What would you like to change on the web?

2. What would you really want to keep?

3. What are the technological chances for internet?

4. What are the threats?

I had something to say about it, so I answered the questions. Ironically, if you read this, the response form was too fragile to accept my answer. Since I thought it was thought provoking and might be of interest here, I decided post here and link to it. As always: Flame On!

1. What would you like to change on the web?

I would like the web to be more robust. It is much too fragile. The transport (TCP/IP) is quite robust, standards compliant, etc. The stuff coming into browsers is beyond broken. It is most annoying that sites depend upon all kinds of arcane plugins and invasive client side scripting. I am constantly given the choice of either not viewing the site or allowing it to do what it likes in my browser. To work around these problems, we need a trust system (PKI or similar) that is inexpensive enough to be practical for use by everyone. We could then simply allow sites we can trust to go on about their business while blocking sites that cannot be trusted.

The web is crazy-slow. So are all the components. I have a 5Mbit connection and should see pages instantly appear. Instead, I can count five seconds between clicks half the time. The obvious thing to do is to create a dense programming language that server and client understand. This has been done a few times (Rebol comes to mind), but the purveyors all want to 'own' that network. The marketplace has rightly rejected that.

Java/AJAX/Web 2, etc. weenies would like to insist that they have the answer to the above. However, their solutions are frighteningly fragile, slow to the point of dysfunction, and by the time anyone finishes one of these projects the standard has been changed so old code does not work. It is to laugh.

Speaking of changing standards, somebody has to get the whole LAMP environment under control. Versions of Apache, MySQL, PHP, Perl, Subversion, CMS systems like Joomla, Wikis of various stripes and many other applications keep breaking compatibility. They do this, in most cases, without letting the other code know. This is a chain. If I run Subversion with Apache (as many do); I should not be facing an upgrade to Apache to install a new version of Subversion. What actually *did* happen at one point was that the current version of Apache did NOT work with the current version of Subversion and people (like me) had to use unsanctioned builds of Apache modules to make this critical system work. [Subversion is a popular source-code control system. It is very bad if it breaks.]

I am becoming increasingly frustrated with the fact that everyone and his brother seems to think I want to be on his mailing list and I want to sign up for his site so I can view a page. What is the deal with that? Even large companies like IBM and Microsoft who should know better keep sending me stuff that I DO NOT WANT. They do this by making me opt into a mailing list whose name should be something like: 'a tiny number of important notifications that crucially affect your life, plus a TON of stuff we KNOW you would not accept otherwise'.

Once upon a time, people used 'handles' with wild abandon. They could post anonymously. I liked that. Sure, we had to put up with nonsense at times, but we would see stuff that could only be posted under a 'handle'. People could keep their privacy. The mechanisms (or at least the underlying technology) exist to allow people to both keep their privacy and still be 'known valid'. The people resisting this are mostly motivated by self-interest. They use your information for their personal gain. If they made it possible for you to withhold that information, they would lose money. They will not do it voluntarily. Informed 'netizens' should insist upon it.

BTW -- this is just a personal preference, but I would like the web to be less 'noisy' with cleaner pages that do not have their entire content on the first page along with moving graphics and flashing letters.

Bottom line:

I would like the web to stop breaking. It breaks more now than it ever did. I would like the web to be MUCH more efficient. I would like the web to be MUCH more secure. I would like the web to be MUCH more convenient. I would like the web to be MUCH more private. I would like the web to be more pleasant.

2. What would you really want to keep?

I REALLY, REALLY want to keep and enhance rights to freedom of speech, privacy, security of the person, etc. I also want to keep the ethic of freely sharing information so that people can collaborate without paying EvilBogusIPRightsCo for the pleasure.

I also 'rilly rilly', want to keep the robust nature of the transport and network 'ethos'. If we continue on that path, I expect we will have a virtually indestructible network.

I want to keep and accelerate convergence of media such as voice, music, video, etc. One network for all!

3. What are the technological chances for internet?

I am not sure what the above question means, but:

Opportunities The network could converge completely so that all forms of EM bandwidth such as radio, television, IR, etc and all forms of transport such as wire, network cabling, fiber-optic, etc. carry IP traffic. Breaking the EM spectrum into small enough cells could make transparent, ubiquitous, robust, high-speed transport possible. It would eventually mean you do not even have to carry a phone if you wish. You could just speak your request into the air and the network would make it happen.


"I need to get to Union Station" -- voice from the sky says, "I'm a cab nearby, would you like me to pick you up?"

"Joe, what was the name of that restaurant you mentioned?" -- Voice from the sky says -- "Joe said it was called Blue-Sky grill. It is five blocks away. The cabby passing by has said he will drop you there for free."

If the question means, "what are the odds it will go away"; I would say zero. Within a decade or two, the entire world will be so integrated with the network that shutting the network down would shut everything else down as well.

4. What are the threats?

The threats are very grave right now. As we speak, forces are gathering to gain complete control of the network. They have some probability of success. Unlike other times in history, once a despot takes control, there is no way to unseat them. With a wave of a hand, the ruler could shut off water, food, electricity, gas, oil, train service, highways and perhaps even the air you breathe. Oh, to make a point, they probably would leave communications on so that other potential revolutionaries can see what they are up against, but face it -- they could sure shut down your communications.

Unless we are vigilant, the system will be configured in such a way that the 'bad guys' will have complete control. Thus far, with the DMCA, the Patriot Act, Black Box Voting (and associated 'irregularities'), Carnivore, take-down notices, unlawful copyright extensions, unlawful software patents, the RIAA, Google censoring searches (for China), 'Trusted' Root certificates only owned by entities with fundamental conflicts of interest (and by the LAST people I would trust with my wallet), uh -- etc. I would say that we have decidedly NOT been vigilant.

As a tactical matter, current vested interests -- phone companies, network companies, media companies, etc -- all bandwidth 'owners' are doing their level best to confound convergence and/or extend their control over converged networks. It is a mess, it is bad, it is costly and it benefits only a vanishingly small number of people at enormous expense to the rest of us. Convergence will happen, but its ownership and control could become a very ugly thing indeed. Meantime, a convergence that would take about 18 months start to finish to the benefit of everyone is taking literally decades.

Unless the first part of this section did not make it clear, there is a very real physical threat to health and safety if control of networks fall into the wrong hands. Eventually, everything will be tied into the network. Something as mundane as a light switch could become deadly.

There is also a rather odd negative effect happening that will be accelerated with the converging network. People are losing their literacy. It is happening piece-meal, so it does not seem so alarming. Widespread literacy in languages such as Latin and ancient Greek is vanishing. Artisan skills are vanishing. The generations born in the 20th century are dying off. As they do, their skills, and much of their knowledge will go with them. In a world where computer interaction is done with spoken languages, gestures and even implicitly without communication at all, what will happen to keyboarding, handwriting or reading?

There is a danger that as machinery (amplified to ungodly heights by network effects) begins to take everything over that our skills will atrophy. There is, in turn, a danger that people will become more specialized to the detriment of general knowledge and skills. How much social cohesion is afforded by common understanding of language, arts and science? What will happen socially as people come to have less in common with each other and more in common with their prosthetics?

There is a very real danger that as our machines become more powerful and we more dependant upon them that we will eventually become slaves to the machines. Arguably, my wife's Blackberry is the senior partner in their relationship already. It tells her where to be and when to be there and she does it. It has more success with that than I do.

The thing that most irritates me, posted 16 Mar 2008 at 00:12 UTC by Omnifarious » (Journeyer)

The Slashdot effect should not be able to happen. It should be possible for random people to post pictures without having to have a hosting service. So much of the high volume data going over the web is stuff that doesn't change but it always comes from the server you get the web page from unless someone is using akamai or something like it. That is so broken and to me reveals a major flaw in HTTP.

email your wife, posted 28 Mar 2008 at 14:52 UTC by lkcl » (Master)

you should send instructions to your wife on what she should be doing. ironically, however, in bed, she ... hmmm.... nooo, she _can_ keep hold of the blackberry ha ha

description of the problem, posted 28 Mar 2008 at 15:05 UTC by lkcl » (Master)

deep, hi,

you have described what the problem is - yet have not outlined possible solutions. i've been thinking about - and discussing - exactly the issues you describe, for quite some time. i'll see if i can get permission to post a recent discussion from the person who initiated it. also i'll write up a reply later.

separate article, posted 29 Mar 2008 at 18:58 UTC by lkcl » (Master)


after some discussion and thought, i'll write up a separate article. in the mean-time, top 10 linux desktop hurdles again "states a problem" but does not offer any answers.

this is extremely common in the free software world - problems are stated again and again and, whilst people "get on with their project", there's nobody outlining a cohesive strategic direction which makes their project worthwhile - useful - to _real_ people and i mean to people as a "whole" not "individuals".

by that, i mean that it's all very well that apache serves up web pages: it does this extremely well. however, that still leaves the problem of creating a single-point-of-failure server which, if it's up - and accessible - that's fine, but if there's any failures along the way, some people don't get "served".

so my article will outline the necessary paradigm shift which really does have to be made - and mainstream - rather than the "bitty" approach that it is today.

Separate Articles, posted 6 Apr 2008 at 16:22 UTC by DeepNorth » (Journeyer)

I think separate articles would be productive. It *is* a rule of thumb not to go whimpering to management about problems without suggested solutions.

In my defense:

1) I was only asked to state answers to a given set of questions.

2) In some cases, the solution is implied by the problem. Like the old joke -- "It hurts when I do that. Well, then stop doing that." For instance, I say that we will have problems and have had problems flowing from a lack of vigilance. Well, it's a little glib, but I guess you could say the prescriptive is to 'get vigilant'.

I have TONS to say about possible solutions. Unfortunately, my time is limited and there really is a lot to say. I had the stuff above already written and thought it might be 'of interest' to some members of the Advogato community.

I hope I'm not sounding defensive. I don't mean to, but the lack of time is making my response clumsy. I appreciate the comments, honest.

One thing that I am actually working on is a way to make it possible to take 'CA powers' out of the hands where they currently reside. I find it bothersome that the web is fragile. However, I am (ironically), more concerned with it becoming unstoppable while being controlled by entities that injure us.

That vigilance thing is frustrating. The main impediments are not technical, they are social, political and economic. I would LOVE to get behind someone who could and would champion the cause. I do, when I am able, donate money to the EFF, etc. However, I am already a little out there for a man with a family that depends upon his small business. Making big political waves is iffy for a family man. Sorry about that.

I think there is ample room for others to write articles addressing various of the issues raised. I, for one, would read whatever material people have with interest. If and when I have the time, I will address certain particulars where I feel my input would be helpful.

Thanks for listening!

Outrageous Flame Provoking Post?, posted 6 Apr 2008 at 22:28 UTC by Chicago » (Journeyer)

The "Web" as what? as in HTTP? or as in the idea of the International IP network?

From my point of view, certain applications for the Web where developed by academics who couldn't see them being misused. For example, SMTP gateways blindly forwarding on mail in the correct direction, on the surface looks like a fantastic idea, but then lead to the "workingness" of Spam relays.

Wouldn't mail be fantastic if mail from a particular domain could only be sent from certain IP addresses listed on that domain's Record? Wouldn't it be fantastic if the rules about domain name registration where tightened? When domains expire, they should be available for re-use.. Why can *anyone* register *any* domain anyways? Why is it possible to fake the origin of packets? (Yes I am aware of the performance impact I am suggesting here).

Why was there the invention of the .com domain name? I mean, if you're a company in the UK, it should be enough to have a .co.uk and not get looked down upon because there's some other company, five thousand miles away who do nothing to do with area of specialty has taken the .com.

I am also beginning to opt-out of Email. I fully understand the <user>DeepNorth</user> for not wanting to have to provide email addresses for signing up to web pages. Oh Don't get me wrong, I'm not a privacy nut. I fully subscribe to the idea that when you do something (visit a website) that the thing you're doing it to has the right to know "who" you are (be that an semi-traceable identifier of some kind)... so I'm saying using things like OpenId, and MS Passport etc to allow cross-site authentication of WHO you are without everyone having the possibility of contacting you.

I don't believe in the overuse of the privacy argument. I used to, I don't know what changed (other then me living in the Real World for a bit). Lets say I run a shop. If you walk into my shop, I look you up and down, and make some mental notes about you. I might make a note on a tally pad marking that you've walked in. I might make another one showing you bought something, or if you didn't. Hell, I might even TIME how long you where in my shop. Nothing invasive here, just standard market research. And whats so different about doing this on the Web? All of a sudden, me tracking this information is looked at as if us "shop keepers" are perverts. How dare we track that you came in on three different occasions before buying.

What am I suggesting?

Complete virtual revolution. Destruction of the DNS system, and replacing with a new concept which allows authentication, and a method of identification where both individuals and organizations have [one or more] virtual identities which are used when you are on the network using whatever service (HTTP (or a replacement of)/SMTP (or a replacement of) etc, but these identities are managed (somehow). Where there is little or no social status associated with the identities (no upper and lower class identity), where cyber squatting is useless.

Should this be P2P? Should this be centrally controlled? I don't know.

No, I don't have the technical answers for any of what I've suggested. But this was a hypothetical bordering on sci-fi question, and we're all friends right?

p.s. I also call for the destruction of Google. I don't believe their "Do No Evil". They're just biding their till 100% of the population can not do their job without them before suddenly all the TV's in the world show a bald man stroking a cat demanding Antarctica and 100 million dollars...

p.p.s. I don't know why the cat would want Antarctica, but they're evil too. They're behind Googles evil plan. Them looking "cute" is just a population control thing they've got going.

google "do no evil", posted 6 Apr 2008 at 22:40 UTC by lkcl » (Master)

chicago, there is no concept of "evil" or "good" - there is only "energy" and "lack of energy". where "energy" is usually in the stored form of "money".

google's ethos of "do no evil" is unfortunately in the setting of the "corporate-led" world. this leads them to have to "protect" themselves from other pathological "corporations". and that means that they have to buy up all the free software talent that they can possibly get their hands on - so that nobody else can use that valuable resource.

i don't call for the destruction of google (yet). what i _would_ call for is their transformation - into a company that helps takes on the next global challenge: organising the world's knowledge. (preferably without shoving advertising at people, but there you go...)

organising the world's information is easy: it's a matter of scaling up some simple algorithms. but organising _information_ doesn't help: we're _deluged_ with "information", thank you. organising _knowledge_ is a much more interesting - and useful - task.

yes, mr and mr google, that means you get off your behinds, stop discussing what type of entrees you're going to have served on the company ocean liner, and do something!

ok - it's too much to ask you to read our humble rants, i'll have to organise the world's knowledge myself mwahahahah.

apologies!, posted 6 Apr 2008 at 22:41 UTC by lkcl » (Master)

i'm so sorry, everybody, my cat came in and wandered across the keyboard - god knows how it knew how to type such complete crap, there ho hum, never mind the monkeys typing shakespeare, be afraid, be very afraid of your white cat!

Tech Fusion Outline, posted 9 Apr 2008 at 01:43 UTC by lkcl » (Master)

organising the world's knowledge. solutions for the issues you raise, chicago, are outlined in the article. each of the solutions requires a great deal more expansion. i've outlined each solution in roughly dependency-order, moving on to the high-level solutions towards the end of the article.


who's going to do all this programming? there's many man-decades of integration work required - far too much for the usual calls i get of "well if you're so smart you can do it yourself" to be _remotely_ practical.

i have some ideas - i'd prefer to hear from people who'd like to make this happen.

done!, posted 16 Apr 2008 at 03:37 UTC by lkcl » (Master)

singularity of computing

duh., posted 16 Apr 2008 at 03:38 UTC by lkcl » (Master)

posted to wrong article, duh. sorry!

lkcl, posted 18 Apr 2008 at 01:21 UTC by DeepNorth » (Journeyer)

I apologize for taking a while to get back to you. Ironically, I was doing the write-up for somebody else's grant-funded research project, which addresses many of the issues you raise.

Somewhat ironically, I had to stop my own work into related stuff to do so.

For the record, I like the fact that you really put your should into it. This stuff (reporting the news) is a thankless task and you usually get punished for it. I'm glad you stepped up to the plate.

Now, off to comment on YOUR article. I confess I have a few nits to complain about, but I have some extremely important agreements with you. In fact, I feel so strongly about these issues I have written about them and am building software for them.

cool!, posted 18 Apr 2008 at 18:09 UTC by lkcl » (Master)

good man! :)

it'll happen.

we could really do with a formal structure - a "social business" (see prof. yunus "creating a world without poverty") - to take care of these things. a charter etc. etc.

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!

Share this page