September 24th, 2001


Sunday Dinner, Sep-23-2001

Since Monica and Dani had another dinner engagement, they were coming late. So we decided to be very non-kosher and had Bacon-Cheeseburger Potato Pie.

The Bacon-Cheeseburger Potato Pie is basically meatloaf formed into the shape of a pie crust, filled with mashed potatoes and topped with tomato, bacon, and green onions. It was pretty darn tasty. But: the meatloaf crust involves ketchup, and the only ketchup we had on hand was the green ketchup. It turns out that green is a very unappetizing color for meat.

We ate late, because we were waiting for Mike and Deanna, and Mike had his Sunday school class first.

After dinner, we played Starfarers of Catan, which Deanna had brought, while Lori made dessert. Starfarers of Catan seemed like a pretty good game. It was extremely high on toy value, with lots of funky parts. We didn't finish the game; I would happily play it another time.

Dessert was this wonderful apple-pecan caramel stuff. (I'll edit this entry to give the exact title.) It was extremely tasty. Mmm.

Krispy Kreme

On Friday night, we got together with Laura and Willie for dinner and a drive up to Cranberry for Krispy Kreme donuts.

We had good Italian food at Don Parmesan's, enough of it that we were unwilling to go for donuts right away. So we went to Borders for a bit; we bought a children's book Lori needed ("Millie's Purple Plastic Purse") and a CD of jazz versions of spirituals and hymns by Charlie Haden and Hank Jones.

Then we went off to Krispy Kreme. The hot fresh donuts really were qualitatively different from donuts I'd had before, and they were very nice. We bought two dozen with Laura and Willie--one dozen was the right amount for us.

Oy Vey

I didn't mention this bug when I discovered it on Friday:

Resetting statistics for a ProfileSet is a time-consuming operation. The changes Nick and I made last summer helped a lot, but there's still still significant I/O of opening and closing databases, and I/O is slow. Resetting statistics consumes at least a few seconds.

On Friday, I discovered that the annotation engine code was resetting statistics every time it created a filter, instead of doing it only once--and an analysis tree will have a few dozen filters.

Today, I started the day by changing that code to reset statistics lazily, so it resets only once. It should speed up the loading phase dramatically.

I Grumble About Code Yet Again

This repeated use of only the first thing returned by an iterator is evidence of poor design.
If there can only be one, don't use an iterator.
If there could be more than one, do your thing for each of them.