Is Dropbox Doing Linux a Favor?

Monday, August 20, 2018

Image via Wikimedia.
There has been much wailing and gnashing of teeth in open source software circles about a recent decision by cloud storage/sync provider Dropbox to confine its Linux support to a single, unencrypted type of file system. As a pro-capitalist user of open-source software, I have found this reaction both annoying and amusing, rooted as it is in a reflexive paranoia, common among open-source enthusiasts, about anything a large company does.

To get the annoyance out of the way: When the news first came out -- in ample time for us to consider whether we could or should continue using Dropbox, as well as look for alternatives -- many articles about the change outright stated or implied in some way that Dropbox was going to quit supporting Linux altogether. This included even some of those, like the one I'm about to excerpt, that had intelligent things to say about the change. This made it harder than it should have been to evaluate some of those alternatives in cases of genuine confusion.

In my use case, unless Dropbox is lying through its teeth -- and the reps who keep popping up all over the tech forums to correct misconceptions are just well-paid trolls -- I should be able to carry on after making some changes to my three Linux machines. And the hue and cry over the lack of encryption is particularly silly coming from a group of people who pride themselves on figuring things out: After the change, my entire hard drive will be encrypted, and anything sensitive I want to have in the Cloud will be stored within a large file that looks like a blob to the rest of the world, but is actually its own encrypted file system. And transfers are encrypted end-to-end by Dropbox as they always have been. (See P.S.)

I'll grant that the changes I have to make will be inconvenient, but I have until November 7 to make them and continue researching alternatives. Thanks for the heads-up, Dropbox.

That said, there may be a silver lining, long-term, to this change, as Jack Wallen of TechRepublic argues:
... I do believe Linux needs that standard desktop distribution, one that can be the basis for all third-party, commercial application support. Keep all of the choices out there, because those choices make Linux great. However, narrow those choices down such that a single, universal, distribution can be put forth to serve as the public-facing front for Linux on the desktop.

Think about this: If Adobe said, "We'll create a native port of Photoshop for Ubuntu Linux, using the unencrypted ext4 filesystem, systemd, GNOME, and the stable kernel", would rising up against that decision (because they chose that combination) do anything but make Adobe regret (or back out of) their decision? Chances are, it would turn them away from supporting Linux for good (not that they will, that was just an obvious example). [bold added]
Dropbox isn't being nearly as restrictive as Adobe in Wallen's hypothetical example, and people are getting upset. And this is the amusing part, because if some of these people were half as clever as they say using Linux makes them, they'd have already shrugged and gotten to work on finding some combination of file system, cloud storage provider, sync service, and collaboration software that works for them. And -- just as they see the value that Dropbox brings with its amazing services and Linux support -- they'd be receptive to the idea that if this company takes the lead (intentionally or by accident) in herding cats enough to make it easier for companies to support Linux, the door might open for other software vendors to support Linux.

Dropbox doesn't owe me some absurd, Edenic computational paradise. It's a company full of people who want to make a living. I'm glad they're going to continue supporting Linux, and in a way that will make it easier for other companies to do so. It is not reasonable to expect every company to employ an army of specialists to keep abreast of every single file system in use by a minority of computer users. Adopting (or in this case, essentially proposing) a standard is a good way to serve this minority and those who work with it. Thanks, again, Dropbox; and that includes for getting me to look around: I learned a few interesting things I might be able to use along the way.

-- CAV

P.S. The crossed out sentence is incorrect. If server-side encryption, which is what Dropbox has, is not enough for your needs (and using a container like I do is not a viable option), you should consider other alternatives.

Updates

8-21-18: Added a P.S. 

No comments: