Spanky's Place

Wherein I explain what IBM needs to do to make XPINC live up to it's potential

Read this before you go any further. I'll wait.....

Now allow me to make one thing perfectly clear:

thick clients are not going away.

There has been a whole lot of fuss over the last 18 months about "social", "cloud" and "web based" solutions. XPages technology (which is extremely powerful and cool) seems to have been caught up with a lot of this hyperbole, which is a shame because it generates an unhelpful association with these other things.

XPages are web based -but only because CSS, JavaScript, HTML, and other required client delivery and presentation technologies are bundled and integrated into web browsers. The server is like a honey badger; very badass and doesn't care about the client. It processes GET, PUT, and POST requests -nothing more. I'm not trying to minimize the awesome power of XPages here, the processing stuff that goes on and the things you can do are absolutely incredible. But with regard to communication with the client it is an HTTP processing server -period.

Getting back to my point about thick clients, they are going to be around for a long time. The primary reason for this is the same reason they evolved in the first place -local data access. The pro thin client, pro cloud marketing crowd just don't seem to grasp that "always connected" is a myth. There are a bunch of other ancillary reasons for the thick client to stay around (security, encryption, processing speed, etc), but the most important one is local data access.

When the wire breaks or the cloud goes away, people still need to get things done.

Which brings us to XPINC. A critically important business need for running XPages in the Notes Client seems to have been forgotten: getting work done -even when disconnected. Accessing an XPages application on a Domino server somewhere from your Notes Client is fine and dandy, but what about when you don't have access to the server? The Notes Client has it's own internal XPages server (a GREAT start BTW); but until the Notes Client / Local XPages server combination is fully capable it is really nothing more than a fun playground.

In Sean's blog post (you did read it, right?) he talks about the troubles with opening an XPage from another database. The real crux of the problem isn't so much how you get the data as it is accessing the database itself. There is the rub. We need to be able to define RULES for accessing other databases -how to find it, where to find it, and under what conditions should we even look for it.

Dynamic determination of remote data sources is the single greatest XPINC miss

Until we we can programmatically determine at run-time what data sources to use, and where to find them XPINC will languish in the realm of "could-have-been" computer stuff -a world occupied by the likes of the NeXT computer, OS/2, and Second Life.

I'm not trying to belittle the work done so far. XPINC really has come a long way and is getting better with each release -but it is not good enough yet. I want to encourage IBM to keep pushing, keep making this better; because I really want to use XPINC. My backlog of potential projects waiting for a fully functional, fully featured XPINC implementation is huge. It is large enough to keep me billable for many, many years. But until such an implementation is delivered this backlog will remain nothing more than "potential".

Potential doesn't pay the bills.