infryq

Yesterday we wound up with about eight inches of snow, so I worked from
home. I found another bug in the early stages of the trivia-questions
pipeline (dataset #1), which is unfortunate since it takes a few days to
run. I started collecting documents for another dataset (#2), and finished
a multiply-parallel subgraph builder for a later stage of the same dataset
which now runs fast enough that we should be able to do 6k queries in about
30 hours on one of our 64-core machines. Found and fixed a bug in the type
system inference builder for dataset #3. Really productive wfh day.

In craftland I re-knit a fingering-weight swatch a couple of times with
smaller needles, and got a fabric I like at 27-28 stitches per 4”. That let
me narrow down my pattern options to five, three of which are customfit
(which will do all my fitting and gauge math for me). I just have to
measure the grist of my sample yarn to get the expected yardage, and that
should either clinch it or throw me back to the drawing board. I also wound
the skein of outer hat yarn into a ball and swatched on US 8s. I’m not sure
I like the textured pattern I tried, so I’ll probably put a few more
options in there before I measure and cast on. Really excited for this hat;
it’s going to be so warm and the recipient is both chronically cold and
extremely knitworthy.

I was unusually good about my meal schedule considering I was home all day
and that usually means I forget to eat.