Synching is Dead: Moving From Replication to Flow

I had a major headache occur, as the result of moving onto a new Mac. All of a sudden the sync stopped working for calendars: I couldn’t get calendar info onto my n95 from iCal. I would select the ‘sync calendar’ option for the n95 (or other devices) in iSync, and I would get the message that some iCal compliant app needs to be running for the calendar sync to proceed. I would check the box, iCal would open, and… nothing.

Of course, the setup I was using was a second order kludge: I was syncing with iCalendar only as a means to get Google Calendar info onto my phone. All the iCalendar information was being pulled — through read-only subscription — from various Google calendars, plus one Dopplr calendar.

I wasted a Sunday morning at the Apple Genius Bar, fooling around, but really getting no where. The Genius there tried a bunch of stuff — moving preferences around, deleting things, etc. — but nothgin fixed it. At a certain point he advised me to drag my Library folder to the desktop, and reboot. Mac OS X will reinitialized the Library stuff necessary to run, and then I could start dragging back things I might need. Sounded like Windows to me.

One good thing came out of this near-total waste of time: I discovered that Google has a mobile web interface for Google Calendar. In a real sense this frees me from the entire apparatus of syncing. For the first time I have access to the location and description fields of my calendar entries, which were lost in the iCalendar/iSync migration. Even better, the location info leads to integration with Google maps on the phone (although the map display is tiny, and doesn’t really take advantage of the available screen space very well. Still, I was able to use it to navigate in Mountain View yesterday, after getting off the Caltrain (no, it’s not the Calendar Train)).

However, I still want an offline version of the calendar for the laptop and the phone, since sometimes I don’t even have wireless available for the phone (like on airplanes or subway tunnels).

But this episode is one baby step away from synching of static, highly structured and managed information, toward something else. I was handwaving at some folks at Plaxo yesterday about this idea: tht we are moving away from a mindset where we want information normalized and synched all over the place. I don’t really. I am not eager to have all manner of contacts in all the systems I use replicas of each other. Or calendars.

Consider Dopplr: I use that calendar just for travel, and I import it into Google Calendar to avoid duplication of effort. But I certainly don’t want to cram all sorts of calendar entries into Dopplr, even if it was possible.

I don’t want all the contacts in the world on my phone: just the ones I want to call. Adding more makes everything slower.

So the basic premise of blind repication just doesn’t work for me.

What I really want is to push ‘parts’ — bits of information that contain enough metadata to distinguish them from other sorts of information — and to have semi-smart apps that pull these “parts” out of the general brownian motion of the emerging web of traffic.

So, instead of syncing, I was to drop parts into the flow, and have apps that sniff for calendar entries, invitations, contact updates, whatever. Instead of this traffic being concealed in some sort of machine-to-machine communication, the traffic can flow past in plain site. I can see that John has a new office phone number, and some clever Facebook plugin or Plaxo capability could fish out that factoid and act on it. But I also can see it float by, which may cause me to erase the old entry from my little black book.

But, of course, these smartish, parts-sniffing apps don’t exist yet, except in hard-coded ways, like Pownce or Facebook events.