The extremely rickety automation I built to download and rename logfiles collapsed (again) sometime after last Thursday. The failure only came to light this morning when I tried to run Bise, the little analytics program that gives me a high-level view of what kinds of visits my sites are welcoming. Because of the way the server rotates my log files out of existence, it is quite difficult, though not impossible, to recover after more than a couple of days. For now, then, I have given up on stats until I can get home and sort things out. The problem is how to avoid such mishaps in the future.
Having done the heavy lifting to analyse Q1’s transport data, the only hard part today was to remember which bits did what. To that extent, the whole assembly is still manual till it hurts, which honestly isn’t much.
Changes from last time include:
As we head off for a couple of weeks away from home, catsitting for a dear friend, I thought I would get a jump on my usual tardiness and try to get my laptop in order and fully functional. As this would also be the first time using this laptop to do a bit of coding — at least that’s the plan — I also had to install all that stuff. Luckily, I had made a note of Chris Amico’s guide which, along with the bits I have learned about environments, made the whole thing a lot less stressful than it might have been.
Finally found time to summarise the first quarter of data that Overland collects for me.
The biggest surprise is the amount of time spent on the bicycle as opposed to in the car. Some of that difference was caused by the absence of the car for about three weeks, the result of some thieves helpin...
Professor Von Explaino kindly steered me towards ChatGPT to solve the problem I was having building a Python gui. His solution worked, for that part of the problem, and made me realise that I ought to ensure that I ask ChatGPT as well as simple search. Having then got my gui up on screen, happy to receive files, for the next series of errors I asked ChatGPT myself.
Horror!
