3 ms·

My mailbox opens and closes for my mailman to collect outgoing mail and deposit incoming mail. But anyone can open it. That doesn't mean I want them to, or that

by Zanni 8y ago

My mailbox opens and closes for my mailman to collect outgoing mail and deposit incoming mail. But anyone can open it. That doesn't mean I want them to, or that they are allowed to. But if my mailbox doesn't want to allow access to private information, then why does it open for unauthorized individuals? Because physically securing it would be a pain in the ass, most people are honest, and if I can keep my mail safe through force of law and social contract, that's easier for everyone, including legitimate users of my mailbox (myself and my mailman).

landryraccoon 8y ago
Your argument holds for mailboxes because it is not a common use case of mailboxes that their owners want complete strangers to check as often as possible because they've left something they want taken. A better real world analogy is a bulletin board on campus or a wooden power pole. Lets suppose that it is super common that people staple flyers to power poles, with the expectation that people will read them as they pass by. Your analogy would claim that if I staple a letter to the power pole, expecting that only my friend that I told about the letter should read it, that passers-by are doing something unseemly by reading it, while being surrounded by want ads and for sale flyers that people do want read. Websites are nothing like mailboxes. The vast majority of websites would prefer that as many people as possible read their contents as much as possible. Email would be a better analogy.
kd0amg 8y ago
A request is communication with certain semantic content, which pulling on a mailbox handle lacks. There is no general understanding among people nor specific agreement between you and some other party that pulling on your mailbox handle is how to ask you for access to your correspondence. This is not the case for HTTP. A network protocol is an agreement about the meaning of certain clusters of bytes sent over a network. When someone operates an HTTP server, a reasonable person could conclude that they take HTTP messages to mean what HTTP says they mean. A lot of cases get more interesting because there is also something generally understood to mean, "Please don't access the following resources by automated scraping, independently of whether my server decides to grant those requests."

by Zanni 8y ago

landryraccoon 8y ago

Your argument holds for mailboxes because it is not a common use case of mailboxes that their owners want complete strangers to check as often as possible because they've left something they want taken. A better real world analogy is a bulletin board on campus or a wooden power pole. Lets suppose that it is super common that people staple flyers to power poles, with the expectation that people will read them as they pass by. Your analogy would claim that if I staple a letter to the power pole, expecting that only my friend that I told about the letter should read it, that passers-by are doing something unseemly by reading it, while being surrounded by want ads and for sale flyers that people do want read. Websites are nothing like mailboxes. The vast majority of websites would prefer that as many people as possible read their contents as much as possible. Email would be a better analogy.

kd0amg 8y ago

A request is communication with certain semantic content, which pulling on a mailbox handle lacks. There is no general understanding among people nor specific agreement between you and some other party that pulling on your mailbox handle is how to ask you for access to your correspondence. This is not the case for HTTP. A network protocol is an agreement about the meaning of certain clusters of bytes sent over a network. When someone operates an HTTP server, a reasonable person could conclude that they take HTTP messages to mean what HTTP says they mean. A lot of cases get more interesting because there is also something generally understood to mean, "Please don't access the following resources by automated scraping, independently of whether my server decides to grant those requests."