>>4
Github employs user authentication and keeps the user authenticated with a cookie that is transmitted with every request. Without SSL the cookie would be traveling in plain text and could be intercepted and authentication could be faked.
The bottom line is that EVERY site that keeps the user authentication should be fully HTTPS. The only reason some don't is because the computational expense.
Name:
Anonymous2012-10-08 16:12
>>14 `
>80MiB/s AES-128 on shitty 6 year old L2400
>computational expense
SHIGGITYDIGGITY.jpg.gpg
Name:
Anonymous2012-10-08 16:25
>>15
And server side ? How much computational expense ?
The only reason some don't is because the computational expense.
Or it could be that SSL doesn't work with name-based virtual hosting, or that SSL certificates are prohibitively expensive.
Name:
Anonymous2012-10-08 16:32
>>15
Not for the user, genius. If your website is serving thousands of requests per second and flip the default switch to SSL, your CPU usage is going to skyrocket. This is why Facebook only offered SSL as a hidden user preference for so many years, until some guy made a Firefox add-on that lets anybody walk into a coffeeshop and hijack everybody's Facebook account with a single click.
>>23
All the Philistines died thousands of years ago. The current ``Arab" (really just Jews who converted to moon religion) inhabitants of that land stole it just like the Jews.
Name:
Anonymous2012-10-08 17:09
>>27
But it's not their sand! This sand belongs to sand niggers!
>>17 SSL doesn't work with name-based virtual hosting
Is there a reason why nobody ever bothered to create something like STARTTLS for HTTP? Not being able to balance incoming HTTPS connections on the same host is kind of a big deal, isn't it?
Name:
Anonymous2012-10-09 23:35
Because HTTP was conceived by physicists in their spare time
instead of the Lisp brotherhood of the satori like IMAP.