First TBL accepts DRM, now he eschews HTTPS because it changes the URI protocol and breaks the old http link?

Who are you and what did you do with the inventor of the WWW? Redirects are an easy solution to deprecated http links. And I still wish you understood DRM. http://www.theguardian.com/technology/blog/2013/mar/12/tim-b...

This sounds like a tempest-teapot scenario because he doesnt seem to know about http/2 which includes the feature TBL is ranting about. Instead, he comes off as fingerwagging at folks for doing something to stop-gap fix the web because the http RFC didnt include security from the beginning. Adding tls to http/1.x is impractical, utopian mythology because there's too much deployed to ever change until http/2 becomes widely deployed.

Instead of ranting, downplaying the EFF and others for improving the situation or wishing for the impossible, TBL missed the opportunity to support adoption of http/2, which might not be perfect (PHK mentioned rushed process and protocol complexity) but it has been touted to supersede http/1, https and spdy.

btw: there's also unauthenticated opportunistic encryption like tcpcrypt which works via server and client plugins.

https://en.wikipedia.org/wiki/HTTP/2#Encryption

https://tools.ietf.org/html/rfc7540

So if I understand this right, Tim's main issue is that "s" is just going to lead users to think "oh, it's secure, I'm good right?" when really web security is a continuum, even on the level of certs - the browser doesn't meaningfully differentiate between domain validation, identity validation, and extended validation (EV), and less so on the technical characteristics of the cert (signing algo, ciphers chosen, etc.)

It's not that encryption is bad, it's that the https method encourages lazy thinking.

Call me pessimist, but I think that any method which relies on "user education" is doomed from the start. Any UX designer can tell you that users do not read error messages and just click by them ASAP to get back to what they were doing before the computer so rudely interrupted them.

No, he only gets to the point in one of the final paragraphs:

  The problem is of course that moving things from http: space into 
  https space, whether or not you keep the rest of the URI the same, 
  breaks any links to. Put simply, the HTTPS Everywhere campaign 
  taken at face value completely breaks the web. 

Of course, in reality, anyone who moves to HSTS also adds a rewrite rule that redirects incoming http vistors to the https URI. In fact, Let's Encrypt even does it automatically: https://github.com/letsencrypt/letsencrypt/blob/master/letse...

His complaint is completely spurious.

In a world where people disable their antivirus because a piece of malware told them to, a red broken padlock or even a scary error message will not help a whole lot.