Tuesday, March 29, 2005

Lessons in technology: PDAs and the national health information infrastructure

Faughnan's Notes: Technology bites

Another example of why computers still don't work; and a comment on some deeper meanings.

Business need: be able to manage personal and corporate tasks and appointments separately. (Rule: don't intermingle.)

Ideal solution: A PDA/Desktop solution that allows one to identify tasks and appointments as being either work or home related and control their desktop synchronization appropriately.

Best available solution: Run two complete 'organizer' environments on my Palm PDA -- one is the native set, another Chapure is KeySuite. Each has its own synchronization. Integration is limited and occurs only on the PDA.

Problem: Even when the data models are consistent (that's another story, too awful to describe here), synchronization is problematic. Take for example today's story. Due to quirks in Outlook/Exchange I need to change the work desktop folder where I store my tasks.
1. Back up tasks in Outlook to a secondary store prior to surgery.
2. Copy all tasks to new folder. (Don't move, copy. I know from past experience moves are usually tougher for synchronization engines to handle than copies.)
3. Delete all tasks.
4. Synchronize and check.
5. Discover:
  • there are two tasks on the PDA KeySuite app that the synchronization engine can't "see". So they can't be dealt with. One is missing completely from the desktop. Not good.
  • when I copied in #2 there was an unrecognized filter applied; so the copy and sync didn't work the way I thought they did
Now #5 is a mixture of user error and bugs, but it's worse than that. I know from past experience:
1. In our setting Outlook/Exchange server are consistently messing up filters. Filter settings get lost or misapplied. It's a longstanding bug.

2. KeySuite has problems sometimes with filters -- it's a quirky bug and not replicable.
Fortunately, from years of experience, I know step 1 is reliable and it did indeed work today. So I'll wipe everything out, I'll get the missing task back somehow, and I'll sync and resync until everything is "clean" (for now).

How many people will handle all this? All those who combine servere geekishness with bull-headed obstinacy. I'd guess about five of us.

So once again -- why did the PDA market really collapse? It wasn't Graffiti that killed the Palm. I'd say it was a one-two punch:
1. Microsoft's FUD and well funded market entry incapacitated the few people at Palm who understood what was needed to provide true "profitable value" (vs. "user perceived value at time of buying decision").

2. Without a laser focus on the fundamentally very tough problem of synchronization and data models, the above was inevitable.
Are there broader lessons? Yes.
  1. We are ver far from having a reliable personal technology infrastructure. We can only manage simplicity, but the user market demands complexity.

  2. The US and UK are making billion dollar bets we can solve data model and synchronization problems in healthcare. But now I'm wandering into my real job, which I rarely mention on this blog ... This will, however, feature in a lecture I'm giving this Thursday.

No comments: