Isn’t interesting that when people say “X cannot be overstated”, that statement in itself is surely an overstatement? And while we’re on the topic, the less said about “…, literally.”, the better.

I wish my Java debugger could catch an AlmostStackOverflowError. Catching the StackOverflowError is nice and all, but when I go to try to see the values of variables in the stack frames, I get errors that I presume are because there’s no stack left to evaluate the values. I’d sorta think that evaluation would be done somewhere outside the overflown stack anyway, but then, I don’t really know a thing about Java debugging internals, I just know when I’m stuck…

Have you ever had the situation where there’s a song A, and a cover version B, where B is quite a bit better than A, but because you’ve heard A a lot more often than B, whenever you hear or think of B, A sorta takes over in your mind, and you get annoyed, so probably the only thing you can do is listen to B a lot of times, even if you don’t really love it _that_ much, just so that you can wipe A out of your mind? c.f. “How Deep is your Love” by the Bee Gees and The Bird and the Bee. See what I mean?

One of the systems I’m working on uses LCDS Data Management and Hibernate to communicate between a database on the server and a Flex client. We’ve, uh, learned a lot. I don’t think I’d do it this way again, were I to choose (‘course, I didn’t really choose in this project, either).

We’ve had a lot of problems due to what is conceptually supposed to be a simple matter, that of lazy loading. Even overlooking the fact that lazy loading is not very transparent on the Flex side, there’s the question of how to distribute those lazy flags. And it’s my contention that in a complex data model with lots of interdependencies and lots of different data use-cases, it is very hard to answer that question. Too many lazy properties and you’re making all sorts of round trips across the internet (and lemme tell you, some of your customers may have ping times way worse than you do when you’re testing these things. Way, worse.). One too few and you’re sucking in the entire database in the midst of one property access.

Ah well, whatever.

The process of working well together to create some useful artifacts can be pretty beautiful. Converging on an understanding, refining the artifacts and passing them around for inspection, making a bit of progress then cementing it in place, growing a collaborative process that endures with or beyond the artifacts, accumulating individual experience, shaping stuff to the world and the world to stuff…