4 ms·

It's here now: https://www.digitaldrex.com/single-post/2014/07/08/Why-I-left-Google-for-a-Bitcoin-Startup https://www.digitaldrex.com/single-post/2014/07/08/Why

It's here now: https://www.digitaldrex.com/single-post/2014/07/08/Why-I-left-Google-for-a-Bitcoin-Startup https://www.digitaldrex.com/single-post/2014/07/08/Why-I-lef...

fastball 9y ago
The Medium post probably has better longevity: https://medium.com/@DigitalDrex/why-i-left-google-for-a-bitcoin-startup-f5b7a8406751 https://medium.com/@DigitalDrex/why-i-left-google-for-a-bitc...
quotemstr 9y ago
Wouldn't it be nice if there were some "Previously-Known-By" HTTP header or meta tag? This way, upon seeing an NXDOMAIN or a 404, browsers could ask $SEARCHENGINE, "Hey. Find me any page advertising that it was previously known by $URL", and browsers could then offer to redirect the user to that new location.
- reificator 9y ago
  Going the other direction is fairly easy, if you retain control over the original page, you can help people/search engines find the new page via the following: * Meta refresh tags: `<meta http-equiv="refresh" content="0; url=http://example.com/" http://example.com/" />` * HTTP Status codes `301` and `302` * Canonical link tags: `<link rel="canonical" href="http://www.example.com/"/>` http://www.example.com/"/>`
  - quotemstr 9y ago
    Of course the other direction is easy. I'm talking about COME FROM, not GO TO. :-)
    - reificator 9y ago
      Agreed. However, in this case they still own the domain, so GOTO is the way to go. In fact that's the rub. If they still have access to the original page's location, then GOTO makes sense. If they don't have access, then how do you know you're getting what you hoped for? What happens if I say I used to be https://google.com https://google.com, and then through some disaster the Google homepage is inaccessible for 3 seconds? That's a lot of users that just filtered down to my potentially malicious site, no matter how low the conversion rate. If the web was still a network of documents, I think something like what you're suggesting would be a lot more viable, particularly with hashing and signatures helping do the heavy lifting. But as an application delivery platform that still looks and pretends to be a network of documents, suddenly that carries a lot more risk.
      - quotemstr 9y ago
        I thought about that. Making the "COME FROM" redirect a deliberate, explicit user action might help. ("Google.com is down. h4x0rz.ru claims to be its successor. Go here? [Yes] [No]".)

It's here now: https://www.digitaldrex.com/single-post/2014/07/08/Why-I-left-Google-for-a-Bitcoin-Startup https://www.digitaldrex.com/single-post/2014/07/08/Why-I-lef...

fastball 9y ago

The Medium post probably has better longevity: https://medium.com/@DigitalDrex/why-i-left-google-for-a-bitcoin-startup-f5b7a8406751 https://medium.com/@DigitalDrex/why-i-left-google-for-a-bitc...

quotemstr 9y ago

Wouldn't it be nice if there were some "Previously-Known-By" HTTP header or meta tag? This way, upon seeing an NXDOMAIN or a 404, browsers could ask $SEARCHENGINE, "Hey. Find me any page advertising that it was previously known by $URL", and browsers could then offer to redirect the user to that new location.

reificator 9y ago
Going the other direction is fairly easy, if you retain control over the original page, you can help people/search engines find the new page via the following: * Meta refresh tags: `<meta http-equiv="refresh" content="0; url=http://example.com/" http://example.com/" />` * HTTP Status codes `301` and `302` * Canonical link tags: `<link rel="canonical" href="http://www.example.com/"/>` http://www.example.com/"/>`
- quotemstr 9y ago
  Of course the other direction is easy. I'm talking about COME FROM, not GO TO. :-)
  - reificator 9y ago
    Agreed. However, in this case they still own the domain, so GOTO is the way to go. In fact that's the rub. If they still have access to the original page's location, then GOTO makes sense. If they don't have access, then how do you know you're getting what you hoped for? What happens if I say I used to be https://google.com https://google.com, and then through some disaster the Google homepage is inaccessible for 3 seconds? That's a lot of users that just filtered down to my potentially malicious site, no matter how low the conversion rate. If the web was still a network of documents, I think something like what you're suggesting would be a lot more viable, particularly with hashing and signatures helping do the heavy lifting. But as an application delivery platform that still looks and pretends to be a network of documents, suddenly that carries a lot more risk.
    - quotemstr 9y ago
      I thought about that. Making the "COME FROM" redirect a deliberate, explicit user action might help. ("Google.com is down. h4x0rz.ru claims to be its successor. Go here? [Yes] [No]".)