Crazy idea - let's just make HEAD requests for everything, and use the returned content-type to decide how to display it.
Even crazier idea - let's do this from the clientas they're composing the message. Then, to make a parse request, you need to pass along the returned content-type.
-
(old issue starts here)
Certain links (like dropbox links) end in a url-encoded filename, which means they end in .png.
This results in them being fetched as images, and the clients attempt to display them.
Smarter thing to do here is to make a HEAD request to make sure the type is actually image/gif before pushing an image message into the feed.
Crazy idea - let's just make HEAD requests for everything, and use the returned content-type to decide how to display it.
Even crazier idea - let's do this from the client as they're composing the message. Then, to make a parse request, you need to pass along the returned content-type.
-
(old issue starts here)
Certain links (like dropbox links) end in a url-encoded filename, which means they end in .png.
This results in them being fetched as images, and the clients attempt to display them.
Smarter thing to do here is to make a HEAD request to make sure the type is actually image/gif before pushing an image message into the feed.