Our minds on media.

Musings on the effects of media on cognition.

Desktop/Cloud Hybrid Software Will Win

There have been a lot of technology pundits discussing the demise of the desktop—primarily arguing that the desktop is going to get sucked in to the browser. And there has been a lot of conversation about switching from the desktop to “the cloud”—the idea of the network as the computer. In a funny comment in that Wired article I just linked to, Clay Shirky is quoted as saying that when Thomas Watson estimated that the world only needed five computers, his estimate was off by four. It rings true because it is a simple and funny observation, but this new view of the network as the computer is a binary view, problematic because as software engineers still tend to do, the solution takes the user into account second and not first. A user-first outlook for most software demands of it that it be a desktop-cloud hybrid—with good reason. And a desktop-cloud hybrid won’t suck the OS in the browser, it will suck the browser into all the apps that a user has. I want to point out two real successes in this regard first, and then look at gaps in the current software offerings out there.


I think the runaway winner of this particular concept category—the desktop/cloud hybrid—has got to be iTunes. On the desktop, it is the ultimate organizer for your music (and video) files. It helps you move from atom media (the CD) to new media (digital). And there are several ways for it to interoperate with your home stereo system (and iPod for mobile use). To paraphrase, iTunes has great local functionaliy. And as far as it’s cloud functionality is concerned, it catalogs and names albums and tracks through CDDB. It allows you to listen to steaming music through its radio function or other iTunes users on the local network; as well as download podcasts. And it lets you outright purchase music from the online store, which in itself is a pretty savvy application. So, iTunes also functions exceptionally in your local cloud and the big cloud.

DRM and monopoly arguments aside, part of iTunes’ success is that it is available to you on multiple computers and multiple devices.1 It’s on the desktop and it’s in the cloud. In fact, it’s more than that. It is local to the computer you are on at the moment, it lets you access your music elsewhere in your private cloud ((your home desktop, work desktop, laptop, ipod, iphone, etc.)) and it operates in the public cloud. And with special regard to the iTunes store, I think it is imperative for software developers to note that iTunes has taken the reverse strategy of many online efforts; that is, rather than move its application to the browser, it has moved the browser into the application.


The guys at Newsgator get this too. NetNewsWire (for those that don’t know) is a really nice RSS feed reader for the Mac OS that does a couple of really important things that I’ve already noted here. One, you can set up NetNewsWire on multiple computers (work, home) and they synchronize to each other. To boot, if you’re away from your computers (or in my case, on a PC), you can hit your feeds through any web browser. And it’s interesting to note that NetNewsWire has also moved the browser into its own structure. While you can set up your preferences so that NetNewsWire opens your favorite browser to let you read a full article on a web page, the software will also allow you to open the article’s web page right inside the NetNewsWire display pane. On a large monitor (in my opinion) this is the way to go. It operates locally, letting you read flagged items even if you’re off the net, it operates on your private cloud by synchronizing among your own machines and devices, and it operates in the big cloud.


Duh. I won’t speak too much to this example accept to point out that unless you’ve switched to the IMAP standard then your email isn’t quite yet the local/cloud hybrid it could be. This is why I primarily think that IMAP is really the standard of the future for email.

Investigating the Gap

Of course, a lot of Web 2.0 applications aren’t things that we need access to all the time.2 I love del.icio.us but it’s a fact that if I can’t get to any pages on the web then I likely don’t need to get to the bookmarks for those pages either. I can’t think of a local use (on my own computer, off the net) for something like Facebook either. These more social kinds of Web 2.0 sites don’t seem to gain value from being anything other than a browser dependent app. But I would still bet that the local/cloud hybrid app that did figure out why you need it locally would beat out the cloud-only apps after a while.

It strikes me that productivty apps—specifically where a user is generating a piece of content for their own use—is really where the hybrid model becomes an imperative. The simplest example I can think of is to do lists. For the Mac there’s a beautiful application called Things that lets you track all your tasks—on one computer. It doesn’t synch with other versions of itself over the net on your other machines, and it doesn’t have an alternate web interface. And web sites like Remember the Milk, Tada, and Todoist all have the problem that they can only be accessed throgh the web and can’t operate locally off the net and then “catch up” later.3 Until either of those desktop/cloud applications accomplish that, for all their sorting and priotizing and color-coding capabilities, they just don’t beat a notebook in my pocket.

Google Apps is great and I keep much of my writing there for convenient access on multiple computers and platforms, but at the end of the day, I have to transfer the writing to some kind of program that will give me the format options that I need. It’s not terribly inconvient but it’s not effecient either. The winner, in my book, will be the word processor that can sync with multiple copies of itself on multiple platforms and still make the text available to the user online when no copy of the desktop app is available.4

And just off the top of my head, another sure winner in this hybrid category would be a merger of Delicious Monster and Amazon—two great tastes that could taste great together.

Going Forward, and Really Forward

It’s just a simple fact that as any technology development gets more complex, the probability of a malfunction increases as well. The idea of always-on internet connections is flawed significantly in this way. Software and data that exists only in the cloud is going to be completely unavailable some of the time. Software that exists only on the desktop is going to squirrel data away in locations that we can’t get to, or in formats we can’t access elsewhere in the cloud. But software that takes the user into account first will make sure that data is available on or off the net, privately and/or publicly, and in standard formats. Software that takes a user’s point of view first will win (in an open competition).

When you think about a really wild augmented-reality future, where Facebook profiles appear over people’s heads on your personal HUD and Google Maps data is laid out over the real world, the value of hybrid local/cloud software becomes much more obvious.5 In the case of a world like that, falling off the network for a moment due to some hiccup in the datastream would result in a literal kind of blindness. It would appear as if half of your visual conception of the world had vanished, and even if it was for only a moment it would have to disconcerting—like losing your real vision for only a second or two would be. Local/cloud synchronization solves this problem.

This kind of data blindness is exactly what results when a user’s personal data doesn’t permeate through the multiple layers of software (and privacy) that an individual user regularly encounters. Users need their content on their computer, on their other computers, on other’s computers, on their platform and on other platforms, on their cell phone, on their iPod, on their TV, and on and on. The idea of cloud computing only partly solves that problem in that there’s only one big cloud. Users really need their own cloud. They need it for senitive data that they need to get on multiple machines. They need it for privacy. And the synchronization of data both locally and in a big cloud backpack ensures the safety of their data. I want my mp3s. But it’s also nice to stream them from my desktop to my laptop or stereo. And a big cloud backup wouldn’t hurt either.

Our content has been moving towards standardization for a while, i.e. mp3s and XML, and the standardization is getting better all the time, but the application interfaces have some catching up to do to really take advantage of that standardization.

  1. I am NO fan of the DRM that Apple is forced to use and have limited my own purchases from the iTunes store to iTunes Plus songs. And I used AppleMacSoft DRM Converter to get rid of the DRM on my older purchases. 

  2. I define Web 2.0 sites primarily as user-generated databases with specialized UIs for a data category. 

  3. By catch up, I mean synchronize with their online database counterparts—the apparent goal of Google Gears

  4. For the record, Scrivener is the closest on this track with its use of bundled text files and multiple export options. Being able to synch those text files with other machines I have and, say, a WordPress blog, would cinch it as the best writing application. 

  5. For the two best fiction examples of a future like this, see Vernor Vinge’s Rainbow’s End and Cory Doctorow’s Down and Out in the Magic Kingdom—I’ll link to them in a little bit when Librarything.com is up and running again. 

« Previously: