Isn't the metric that you've used "bugs per commit ~ per new line of code" going to miss the issue?
All code is technical debt.
If rsync releases used to have 500 lines changed and 5 bugs in and AI-powered rsync releases have 50000 lines and 500 bugs, it's the same bugs/line but much worse experience for the user?
I've not looked into the details of this case and I do use AI assistance coding at work but in my experience, the problem is that it's too easy to write lots of code and therefore hard to review the huge volumes of code and this analysis will ignore that?
edit: actually your table shows there weren't unusually large numbers of commits in this release, so perhaps my initial skepticism shows a bias I have?
OpenBSD used to have sqlite in base, but the code churn rate was too high to review. This was well before the recent LLM craze, so a human (perhaps not a normal one, though) already sufficies to generate too many changes for others to check for errors.
In fact, Hamas are almost certainly not being starved at all. They have the guns, and they will make sure they get food. They have shown again and again they are quite willing to shoot Palestinian civilians for just being slightly inconvenient to them. They certainly won't mind shooting some for food.
I don't notice "disjointed communities" I just look at the posts from people I follow without knowing which server they are on. I'm aware you can see a list of posts on your local server but I imagine people on most instances (unless the instance has a strong theme like for people from a particular location) never use it.
Watched the video and super excited to see where tech like this ends up in a few years. Was on the quest to find a split keyboard last week and ran across this. Ended up with a UHK, but man I thought hard about buying one (pair) of these. I was just worried about dust, dirt and overall longevity. Seems like something that would need to be worked on every month or two. The kit is reasonable and might be a fun (next) project.
Would be interested to hear about anyone's experience that owns one of these. How do they hold up? How long did it take to get used to using? Would you recommend to a friend?
I'm not aware of any production version of Chordite (that this design is based on) as of now; but in the wider family of chorded keyboards, the one with a production version that I'm aware of is the Twiddler.
I was going to say I have a old twiddler 2 (wired usb). There was this program I used to learn called "twidor" that was like a type tutor but had graphic that showed you the chords. Really helpful. I didn't see anything like that in the github repository linked in the video. I guess they are up to twiddler 4 now. I read the linked chordite page and I agree that a problem with the twiddler is you are kind of trying to hold the thing steady so you can chord with the same fingers you chording with.
I don't have a program at the moment. The layout I currently use is in the codebase, starts here: https://github.com/akavel/clawtype/blob/96980f68427eb1089112... Personally, I just edited this layout description into a more compact form and printed it on a sheet of A4 paper as a cheatsheet. I do intend to make this aspect more scriptable, but didn't get to it yet.
For learning, I personally just try to slowly code simple hobby things in vim, with the cheatsheet in the other hand... I tried to make the layout relatively intuitive wherever I could. I also patiently went through all the possible combinations of presses, and tried to group them into categories of easy<->medium<->hard<->impossible, then tried to put esp. the more frequently used keys on the easier combinations/chords.
I made a Google calendar program for the Inkpad 6color eInk display. Because of the small number of displays I doubt there is anyone else using my code and that's ok.
It's Arduino based and it turns out the iCal format has a lot more complexity than I'd guessed so there are still bugs/incorrectly shows some events but it mostly works for me. I work on it very intermittently over a few years and find it quite cathartic.
All code is technical debt.
If rsync releases used to have 500 lines changed and 5 bugs in and AI-powered rsync releases have 50000 lines and 500 bugs, it's the same bugs/line but much worse experience for the user?
I've not looked into the details of this case and I do use AI assistance coding at work but in my experience, the problem is that it's too easy to write lots of code and therefore hard to review the huge volumes of code and this analysis will ignore that?
edit: actually your table shows there weren't unusually large numbers of commits in this release, so perhaps my initial skepticism shows a bias I have?