I don’t know if he remembers it, but last year at OSCON Ubuntu’s Jeff Waugh and I got the chance to meet and discuss Ubuntu, GNOME, and a variety of other topics relating at least tangentially to the Linux desktop. One of the things that I felt was an obstacle then, as I still feel now, is migration.
I seriously doubt that Mark Pilgrim had any particular difficulty with that particular task (apart from the obvious proprietary formats problems), nor do I suspect that Tim Bray will. Ted Leung, likewise, should his technical concerns be satisfactorily addressed, would unquestionably be able to migrate his information from one platform to another. But what about the average user?
Forget, for just a second, questions of application compatability, hardware support, and UI polish: let’s think about more basic considerations. How, for example, would a user get his or her POP3 email from Outlook Express transitioned over to a Linux machine? How about something as simple as their IE bookmarks? Cookies? Documents? Pictures? Music? Instant Messaging client settings? Even if all the users do is browse the internet and send email, there are settings on their machines that are important and will need to be migrated.
I find it interesting that most usability studies seem to gloss over this topic. The Novell sponsored efforts at BetterDesktop.org, for example, include use cases such as “Task: Upload your photos to your website” and “Task: Create a playlist” without – as nearly as I can determine – asking the question of how the photos or music got onto the filesystem in the first place.
Now to be fair, it must be said that Windows is not much more adept at this than is Linux – I’ve felt this pain migrating several family members and friends to new Windows machines in the past (you don’t know migration pain until you’ve been asked to port Mahjong scores to a new machine). OS X, I’m told, actually has it down fairly well if you stick to their platform; you can theoretically connect two OS X machines and have everything ported over for you with little difficulty.
But I think that ultimately, if Linux has ambitions beyond the niche market of catering to the technical elite, it must do more than its competitors in this area. The low effort, tactical approach would be to increase mainstream awareness of synchronization tools such as Google’s Browser Sync, or push for greater use of software as a service offerings, because they in their own ways make the individual platform less important. But the better approach, I think, would be to create a tool that would reach into a common informational stores on competing platforms (AIM, Firefox, iCalendar, iChat, iMail, iPhoto, IE, MSN Messenger, Outlook Express, Thunderbird, Yahoo Messenger, etc), extract the settings and application data that can be extracted, export that to some sort of file archive, and have a tool available on the Linux end to import that archive and populate the free alternatives with the data and settings.
Non-trivial, I know. Decidedly so. And there will be types of data that simply cannot be migrated, due to the nature of the protocols they are stored in. But there’s certainly a lot that could be done.
I think that competing with OS X and Windows will require Linux to a.) innovate in areas those platforms ignore, and b.) lower its barriers to entry. If migration gets to the “just works” plateau, things could get fairly interesting.