[mmahemoff]

Something I don't get about Dropbox API: Why can't I just throw URLs at it?

If a user wants to save <thing at URL>, I have to pull it to my server and push it into Dropbox. Dropbox could just have grabbed it directly and in many cases, won't even have to thanks, to system-wide caching.

(This is based on the last time I looked at the API; maybe this is supported now, but I don't think so.)

[timdorr]

You're describing the Saver Drop-In: https://www.dropbox.com/developers/dropins/saver

[natch]

Can the Saver Drop-In be used with arbitrary third-party URLs? To me this is what was being described, so if not, then no, it's not describing the Saver Drop-In.

[rpearl]

Yes, you can provide arbitrary URLs to the Saver and our (Dropbox's) servers will download them, assuming they are publicly available.

(Note: I am an engineer at Dropbox and worked on this product.)

[natch]

Thanks. Off topic but slightly related if I'm going to rely on your service as a developer: as an engineer at Dropbox, can you tell me why I am seeing random photos that are not mine in my dropbox Photos folder? And does this mean that other people get to see my random photos? Edit: only on iOS. And maybe it's not a "folder" per se, it's the second tab from the left at the bottom of the screen.

[gkurtz1]

That photos tab shows you photos from anywhere in your Dropbox account, so these photos may be from shared folders you're a member of. (Likewise, other people can only see your photos if you shared them with them.) Another thing that may happen is someone "camera uploaded" their photos to your account (e.g., by plugging into your computer and choosing to import photos when Dropbox asked.) If you're still not sure where they're coming from, you should open a ticket: https://www.dropbox.com/support/s/92/3201098/c/2

[natch]

Thanks, very informative answer. I think they must be from a certain shared folder I have access to, which would explain it. Quite a relief, actually.

[mmahemoff]

Great to know they got around to this and did so in a way that leverages the existing HTTP caching protocols. Thanks!

[JosephRedfern]

How would system-wide caching work unless the HTTP server dropbox requested the file from sent a checksum of the file being requested? Just because something has the same URL doesn't mean that it's the same file at the other end (http://wordpress.org/latest.tar.gz for instance)

[varenc]

It obey's HTTP caching controls. Checkout the end of the docs: https://www.dropbox.com/developers/dropins/saver

[JosephRedfern]

I see - but the typical dropbox deduplication methods wouldn't work here (at least not in terms of bandwidth usage), would they?