|
|
|
not-quite-working code merge
Monday, July 8 2024
location: 940 feet west of Woodworth Lake, Fulton County, NY
This morning Gretchen made me cup of ice coffee for some reason (it was probably decaf), so I didn't make my usual french press of hot non-decaf coffee. We sat out in the screened-in porch collaboratively playing Spelling Bee until we could find no more words. Then Gretchen packed herself some food and such and went off to spend the next five or six hours down at the dock
I spent several hours down in the basement further merging the two code bases I'd started merging last night into one ESP8266-targeted Arduino sketch. When it seemed to be working acceptably, at least as inverter data parser, I decided to celebrate by taking a huge 22 oz Juice Force Imperial Hazy IPA down to the lake and paddling around atop the big color innertube (the one made just for tubing, as opposed to the black one that might've once been inside a vehicle's tire). Using my hands as paddles, I went south and then east along the lakeshore all the way to the public dock, where I got out and wandered around. Just east of the public dock is a private parcel with lots of posted signs and a rope strung along the boundary. The sign all claim there are cameras that are relaying a live feed somewhere, which seemed like a lie, since there is no cellular connectivity at the lake and what other technology would they be using. So I crossed the boundary to snoop around, perhaps to have a look in a large shed that dates back to the Boy Scout camp. Then I saw a trailcam and got skeeved out, beating a hasty retreat. Trailcams do not produce a live feed, but I didn't want to show up as an image on an SD card either.
After paddling straight across the lake back to our dock, I headed back to the cabin and resumed work on my SolArk Copilot code. In so doing, I made a disturbing discovery: the companion would only log data for a short time after being rebooted and then freeze, logging nothing else. I couldn't figure out what was happening, and my efforts to debug the problem were stymied somewhat by the fact that the ESP8266 only has one full serial port, which it must use to get data from the inverter. ThiÉs meant there was no way to send real-time information to me. I tried various things to make the sketch work more reliably, such as putting in some code so that it would only service web requests after it was done parsing an inverter data packet (since interrupting that process might cause the serial buffer to overflow, thereby losing data). Or perhaps all the reading of serial data, in concert with all the overhead from the many other ESP8266 Remote duties, was causing it to run out of memory. I also tried greatly decreasing the polling rate for the weather logging and remote features (both of which were working great on the copilot). I tinkered with that thing for hours and didn't really make any progress.
Eventually I took a break to go gather some topsoil and plants to further revegetate the landscape along the north foundation wall (where there is still a danger of erosion).
We don't eat much spaghetti when we're traveling, so after we get home, Gretchen is always wanting to have some. She tried to make some tonight, but our only jar of Raos had gone moldy, forcing her to use the cheese part of a pre-packaged vegan macroni & cheese.
The dock today. Gretchen is sitting in the chair on the right.
Click to enlarge.
For linking purposes this article's URL is:
http://asecular.com/blog.php?240708 feedback
previous | next |
