The only thing I wonder about is why people still get surprised about this. It’s facebook. It’s the company that’s tracking you on basically every website you visit. Even when you’re not logged in.
I think it’s wise to try to be aware of and not reinforce “facebook is creepy and everything they do is evil by default” bias. Any key presses, scrolling, or mouse movements on a website can be sent to the server. This is the nature of the browser. If you don’t want a website to have access to something, don’t type it into the website!
Don’t we already have them? You can send messages to someone with email (which, if you’re all old school like me and use a local client, doesn’t transmit anything until you send it). You can at least choose to build websites that don’t have any facebook integration, like lobsters. You have perhaps somewhat less choice in the sites you visit, though.
I’m entirely sure what you mean by “this sort of thing” or how one could build systems where it’s not possible. Do you want to disable javascript? Cross domain links? img tags?
Companies like Facebook are data companies and it’s against their DNA to discard data, especially if it’s as valuable as this. I went to a Data Science meet-up last month where Facebook presented some of their work for type-ahead predictions and one of the signals is corrections user make in the search box, that means words they entered, than “backspaced” and replaced with the correct term. This is incredibly valuable data that improves the user experience for everyone and you can’t simply buy off-the-shelf.
But is this a draft that you can view later on? I think that is the point, you know about a draft in gmail or wordpress, facebook is essentially just spying on you with no benefit to the user in this case.
Google also allows you to delete those drafts, whereas Facebook provides no interface to do so. If you delete a draft (or other mail), Google says that it may persist in backups for up to 60 days but will not be kept indefinitely on their servers.
Did I miss something in the gif? The quote says it sends back metadata, but the article says it contains the text of the post. I can’t see the text in the debugger though. What’s the actual content of the post?
The url it hits is /structured_suggestion, although I’m not able to get the server to respond with anything meaningful. If one is to discard the “facebook is creepy by default” bias, it seems reasonable to conclude that there’s some useful feature for suggestions based on your posts that can’t easily be implemented in the client (as URL and friend’s name recognition is).
This is not news at all, sadly. It’s been known and covered since at least 2013.
http://www.slate.com/articles/technology/future_tense/2013/12/facebook_self_censorship_what_happens_to_the_posts_you_don_t_publish.html
The only thing I wonder about is why people still get surprised about this. It’s facebook. It’s the company that’s tracking you on basically every website you visit. Even when you’re not logged in.
I think it’s wise to try to be aware of and not reinforce “facebook is creepy and everything they do is evil by default” bias. Any key presses, scrolling, or mouse movements on a website can be sent to the server. This is the nature of the browser. If you don’t want a website to have access to something, don’t type it into the website!
[Comment removed by author]
Don’t we already have them? You can send messages to someone with email (which, if you’re all old school like me and use a local client, doesn’t transmit anything until you send it). You can at least choose to build websites that don’t have any facebook integration, like lobsters. You have perhaps somewhat less choice in the sites you visit, though.
I’m entirely sure what you mean by “this sort of thing” or how one could build systems where it’s not possible. Do you want to disable javascript? Cross domain links? img tags?
Actually, I do block 3rd party js by default, and facebook and google are on a special blacklist for images and css. I love uBlock.
Curious what you mean by this, the-kenny?
“Facebook shadow profile”
Shadow profiles, coupled with the ubiquitous ‘Like’ button.
Companies like Facebook are data companies and it’s against their DNA to discard data, especially if it’s as valuable as this. I went to a Data Science meet-up last month where Facebook presented some of their work for type-ahead predictions and one of the signals is corrections user make in the search box, that means words they entered, than “backspaced” and replaced with the correct term. This is incredibly valuable data that improves the user experience for everyone and you can’t simply buy off-the-shelf.
So does any service which automatically saves drafts, like essentially all email providers.
But is this a draft that you can view later on? I think that is the point, you know about a draft in gmail or wordpress, facebook is essentially just spying on you with no benefit to the user in this case.
Google also allows you to delete those drafts, whereas Facebook provides no interface to do so. If you delete a draft (or other mail), Google says that it may persist in backups for up to 60 days but will not be kept indefinitely on their servers.
Did I miss something in the gif? The quote says it sends back metadata, but the article says it contains the text of the post. I can’t see the text in the debugger though. What’s the actual content of the post?
I just tried this. It sends the full text.
The url it hits is
/structured_suggestion, although I’m not able to get the server to respond with anything meaningful. If one is to discard the “facebook is creepy by default” bias, it seems reasonable to conclude that there’s some useful feature for suggestions based on your posts that can’t easily be implemented in the client (as URL and friend’s name recognition is).At the time the quoted article was written it only sent back metadata, now it sends back the whole text.