Why is this agent fetching my images?

[aristotle]

According to the logs, some of the images on one of my sites keeps getting fetched by the following agent:

Agent: facebookplatform/1.0 (+http://developers.facebook.com)


Does anyone know what this agent represents and why it is fetching my images?

I did a web search and found where someone said that it relates to the Facebook "Like button". But I don't have Facebook Like buttons or any other Facebook code on my sites, so that doesn't explain it.

[lucy24]

Obvious question: is it coming from a facebook IP, like 69.171. or 66.220-221? (Those are facebookexternalhit, aka hotlinks. There are probably others. I blocked by UA rather than keep playing whack-a-mole with new addresses.)

[wilderness]

If your getting this Facebook bot links, than your also getting standard UA's with a dead-link-Facebbok-referral from other IP's.

Somebody has simply "in-line linked" to one or more of you images and the Facebook bot is checking the link.

99.99% of internet users "believe" there is no wrong in in-line linking, nor wrong in utilizing another's server.
They just don't comprehend the difference in protocols.

[aristotle]

Thanks for your reply, wilderness. But I don't know what you mean by "a dead-link-Facebbok-referral from other IP's".

The site in question does get a lot of image fetches from the facebook-externalhit bot, which I assume are due to hotlinks on Facebook pages. I used to have hotlink protection for this site, but removed it more than a year ago, so the server returns a 200 and gives the images to the bot. So I don't know what you mean by "a dead-link-Facebbok-referral from other IP's" in this case, since the hotlinks on Facebook aren't "dead". Thus I don't know why the other facebook bot [facebookplatform/1.0 (+http://developers.facebook.com] would need to check the links.

[Pfui]

I was wondering if it could be on a par with AppEngine-Google (and its scad of iffy UAs and their actions), with some FB coder testing (and/or goofing up) who knows what?

Searching for "facebookplatform" via the link in the UA yields two main areas:

1.) Site-based doo-dads ranging from 'Social plugins' (buttons a la Like, Feed, Rec, etc.), to authentication and analytics; and

2.) Pretty geeky code doo-dads like SDKs and PHP tools.

ALso, there's at least one FB dev forum thread about the same thing -- facebookplatform/1.0 hitting images -- and it turned out to be related to an app hosted on FB. The thread is circa 2009 and offers lots of ideas but few concrete solutions, sorry. [forum.developers.facebook.net...]

FWIW

[aristotle]

Thanks for the replies. So it could be somebody developing or testing a facebook app which for some unknown reason is fetching images from one of my sites. Anyway, it's probably not worth spending more time investigating or worrying about it, so I'll just let it go. Thanks again.

[lucy24]

99.99% of internet users "believe" there is no wrong in in-line linking, nor wrong in utilizing another's server.
They just don't comprehend the difference in protocols.

They also don't realize that pages loaded with hotlinks can take unbelievably long to load-- up to several minutes on slowish DSL-- as you know if you've ever visited a suspected hotlink. So they're also ###ing their own users.

:: still trying to figure out why 99.99% of my own hotlinkers appear to use the same WYSIWYG editor, requiring at least a 30" monitor for horizontal display (is there maybe an Edit mode that silently resizes everything, and they're too dumb to realize this won't happen magically on the real site?) and, of course, several yards of vertical scrolling ::