21–30 of 205 posts
The internet is decentralized, the web has never been. HTTP protocol doesn't support it. But it's time to do that. What we need is the possibility of peer-to-peer connections between browsers.
It's the current philosophy of "web apps" and not "hypertext" that isn't decentralization.
Earlier quoted context omitted.
That doesn't imply the title was changed. Shorter slugs are nicer, and there could be some policy involved in keeping them short. It doesn't mean anything.
"reclaim" vs "re-decentralise" The actual article never mentions the term "re-decentralise" at all.
Puh-leez.
Also, isn't this decentralization? Not having the data in one mega-American-cloud? Seems to me that Tim is doing a lot of PR for big companies lately, masked as a benefit for the users, just like he's doing with the whole DRM thing, which he actually says would be beneficial for users.
But let's assume he's not trying to be malicious here, and that he has a point. Here's the thing. Yes, I agree that having every country demand companies to host the data locally is going to make it very hard for innovation to spread, and therefore, progress will slow.
HOWEVER, right now it seems we only have a choice between this, and allowing US to get their hands on all the data. I didn't even see Obama mention anything about NSA not being able to tap the world's fiber cables anymore.
So until US gets serious about not doing shitty stuff like that to the world's users, then I absolutely think all the other countries should try to force companies to host the data locally.
There is one other solution to this, that will allow companies to keep the data wherever they want, and that's encrypting everything by default, and even them not being able to decrypt the data without the user allowing it. So stuff like OTR for chat systems, DarkMail/PGP for e-mail systems, and so on.
Make it so the web is completely trust-less, so other countries don't have to trust Google or the American government to not get their data, because they could be assured there's nothing for them to get other than strongly encrypted data.
Unfortunately, this option isn't even on the table right now with the big companies, and the US government will push against companies trying to do this, too. And the only way to get this option on the table, is for them to think that other countries are going to inevitably force them to store the data locally, and build data-centers locally. Only then they might start preferring this "encrypt everything with the user having the key" solution, as an alternative to storing data in every country or region.
So until that happens, I absolutely support countries demanding data to be stored locally, because I know that minutes before that will begin to happen, US companies and the US government will agree to letting the data be fully encrypted and trustless. But not any sooner than that. So in the end, we'll get what we want, and the Internet will be safe.
The internet is decentralized, the web has never been. HTTP protocol doesn't support it. But it's time to do that. What we need is the possibility of peer-to-peer connections between browsers.
The internet is decentralized, the web has never been. HTTP protocol doesn't support it. But it's time to do that. What we need is the possibility of peer-to-peer connections between browsers.
The internet is decentralized, the web has never been. HTTP protocol doesn't support it. But it's time to do that. What we need is the possibility of peer-to-peer connections between browsers.
Sure HTTP does. You can cache things, sign things, &c. It's the current philosophy of "web apps" and not "hypertext" that isn't decentralization.
The problem is not technical, it's social, political and economical.
The internet is decentralized, the web has never been. HTTP protocol doesn't support it. But it's time to do that. What we need is the possibility of peer-to-peer connections between browsers.
I'm curious, what will that enable?
Why Tim Berners-Lee is wrong about DRM in HTML5 http://craphound.com/?p=4651
Here's a link to the real text: http://www.theguardian.com/technology/blog/2013/mar/12/tim-b...
For ADD-ers: http://youtu.be/hSyfZkVgasI
The story: http://www.archive.org/details/paulotlet
Long PR story short: RPC was already prevalent, that is how you control(led) your stuff remotely. Like Tim did on his day job. Today, if you'd ask for a NeXT-like toy, you'd be denied were you an average Eastern member of CERN (n.b. western equivalent). As a true westerner, TBL asked for it and got one. Put the gopher link address ptr in the reserved field of the text font properties (where things like bold and italics properties are stored) and voilà. Mix in a simplified SGML for markup (as it was customary back then for typefaces and layout on printers). One could also hire cheap disposable students to actually write the web clients and servers to be cross platform (the web's virtue/value).
Would the "web" have just ran on NeXT, it would be long extinct, let alone have taken off. Clicking on text is how you used Oberon or even Genera Document Examiner (perhaps even over the network). It was the spirit of the times. Even Jobs demoed a WYSIWYG client-server application in 5 minutes on the NeXT. What an insult.
On linking and hypertext: all post-war era anglo-saxon stuff is spin (including Vannevar and Tim invented the web myths). The real stuff comes from Belgium, as did Tim's boss, btw.