Neal Wilson ✅ @neal@social.typica.us

Did a grocery top up in anticipation of possibly not leaving the house for the next 4 days and I'm not sure if I've ever seen so many guys buying grocery store flowers at the same time.

Text messages asking if I'm interested in selling my home are immediately reported as spam and blocked. I'd need a completely ridiculous offer (maybe 10x fair value) to compensate for hassle of moving.

Did some server admin. Was happy to see that I could follow the trail of configs to find where I needed to add a new record and the change worked on the first try.

I might be out of town when that gets delivered so I should leave staff with some instructions if they want to get started on figuring out how we'll roast that before I'm back. Otherwise a few days delay on that won't be problematic.

Bought more coffee. One that I was hoping to get had already sold out.

It's wine o'clock.

Some of that increase is that wholesale is up a lot, but it would still be up significantly even if I removed that.

Now that I've finished the coffee roasting for August (assuming there aren't any last minute emergency orders from people desperate for coffee for their business) I can see that more coffee got roasted this month than any other month so far this year. Normally this is part of our summer slow down which just hasn't been happening the past couple years. That's probably not a good sign (less money available for vacationing?). YTD production is up over 10% despite green availability issues.

It looks like my staff has scheduled me a 4 day weekend for Labor Day so that looks like a good candidate for sorting out a few final pieces of work, doing the various platform builds, testing that things work enough, and putting up some links to those.

I don't mind persistent low volume scraping. I do mind getting bandwidth alerts because some bot doesn't just clone the repos to get the data.

It looks like September is the month that I finally start putting out alpha builds of Typica 2. There's a lot that still needs to be done to get back to feature parity with the 1.9.1 release, but there are also some significant improvements in place and it'll benefit from getting some outside perspectives and people who use it in different ways. Source release will probably have to be a tarball/zip since the scrapers are way too aggressive against Gitea to have that public.

I'll want to get some more hands on experience with the new feature than just today's roasting, but so far I think the new thing is an improvement over the old thing. The old thing is just a config option away (or I could run old and new together at the same time for comparison) as is the ability to just hide all of that (which is the default).

Rolled out the latest features to the production roaster.

Finished testing the thing I wanted to test. There were two cosmetic issues to fix (now fixed) and one thing where I will need to put a note in the documentation to not do a thing because the program does the math right but the result you get isn't meaningful (if you ask for nonsense, you get nonsense). Item removed from the to do list.

I still have a little more testing on a different variation to do, but I should be able to work that into my schedule for tomorrow and assuming things work (or fail in a way where there's an obvious trivial fix) I'll be able to remove that feature from my to do list.

Initial implementation of a new feature did not work as expected, but only because I forgot to multiply by a constant.

Spent some time thinking about how I want to expose settings for one of the new features in Typica. It's something that as far as I know doesn't exist in any other comparable program and I wanted to make sure that people who want to try it out can set it up easily.

You know you need a coffee when you forget to put the coffee in the coffee.

I'd say keeping output file sizes that much smaller is worth a few extra lines of code, especially since I know people email those files around.

Wow, I just compared file sizes on print to PDF with the old code vs the new code. Old code was producing files around 8MB per page. New code is producing files around 24KB per page. The equivalent stuff in the last release version (using ancient dependencies that I can't use anymore) is also in that smaller KB range so that's kind of crazy. That's another mark in favor of the new code.