linsomniac

TheOutdoorBoys youtube channel mentioned something about special batteries for cold temps. -10c for many of his hiking and camping trips isn't unusual, and (as I understand it) he did weekly videos with Go Pros for a decade. He probably knows a thing or two.

This comment is WAY more interesting than the article it is a comment on.

My concern there is that the friction is too high: it wouldn't be something that family members would use in casual conversation, and could also be written off by the AI: "They won't give me my phone", "I lost my phone", etc...

"Did you see that game last night" replied to with "The Visitors are my favorite team" is something you could say instead of "hi" as a family.

A year ago I promoted the idea among my wife's family that we should establish a sign/countersign system for the family and use it regularly so that in the event of something like this we could positively identify a legit request.

Would have come in handy when our niece was traveling in Asia and asked for money a few times, but in this case it wasn't a scam.

I got no traction with it, which I was a bit surprised by because one of the family members works at the Puzzle Palace.

People need to stop yelling "Slop!" all the time

People are applying that label to anything that has the taint of AI on it. I get why.

I realize there are a lot of reasons why people hate AI, but one that is frequently cited is that it's dehumanizing. It's kind of ridiculous to see those same sort of people call something "slop" when I've put tens of hours into guiding the AI tooling. Talk about dehumanizing!

Reminder: Serious people have been predicting a market crash "within the next 3 months" for 3 years now. In that time, the "market" has gone up around 70% (66%-86% depending on the what part you are looking at).

A friend of mine and I go out to lunch every 3 months and talk about, among other things, investing. We've made a trope of it, calling out the people who are predicting an imminent market crash every time we have lunch.

I'm not saying that it doesn't look like it's going to crash, but I'll also say that there's also a very sizeable downside potential for getting out of the market.

As another reply said, the consensus I read is that you really shouldn't memorize the table of "-.- is K", if you do you'll likely need to "unlearn" it to get any level of speed. The best thing is not really think about the dots and dashes but train on the sound; "dah-di-dah" = K.

It's a lot like how when you read you almost never read every letter, you look at the entire word and that turns into an idea in your mind.

A few months ago I had Claude build a morse code trainer web app, it was pretty fun but my interest waned: https://morsetrainer.linsomniac.com/

I've been throwing together some web-playable retro arcade games:

Rally-X: https://linsomniac.github.io/rally-xy/

Tempest: https://linsomniac.github.io/teapot/

Dig-Dug: https://linsomniac.github.io/digger/

And not an arcade game, but a multi-player throwback to a multiplayer shooter game my team used to play called nSnipes: https://github.com/linsomniac/isnipes

iSnipes does require downloading and running a server, the others you just play on the web.

apt-cacher-ultra: To help reduce the impact of future DDoSes of Ubuntu. Just released 1.0 yesterday after working on it a couple months.

As that DDoS was going on I realized that some of our dev and staging processes were impacted by it, and that apt-cacher-ng was doing nothing to help us.

apt-cacher-ultra snapshots the repo meta-data after verifying it, and only promotes it if the metadata all checks out. Additionally, it can optionally keep a list of "hot" packages, and can include those in the snapshot calculation.

Additionally, apt-cacher-ng would regularly choke and require some handholding. I'm hoping -ultra resolves that as well.

https://github.com/linsomniac/apt-cacher-ultra

I wanted the thing 30 years had taught him

Unpopular answer that the author seems to be dismissing: Maybe the thing that 30 years has taught this guy is that the LLMs can answer the question better than he can. Or that he can't give a substantive answer without doing research into it with an LLM.

LMGTFY

I mostly saw LMGTFY used when the question was the sort of thing that a person would have to research but that google results had a high chance of getting with "I'm feeling lucky".

If you've already done a bunch of research, and already asked the LLMs, when someone says "Honestly, ask Claude", you should be able to come back with what results you got to your question and what you need clarification on.

I've been doing programming and sys admin for 40 years. When I run a coworkers question through the AI tooling and talk through the answer with them, it's because my 40 years of experience tells me that's the next step.

I've been saying for months that any time you say something about the capabilities of the AI tooling you should say what exact model, agent, and effort level you're using. Because I've taken a few of those "AI can't do X" and had AI do a fine job at them.

I have Sol 5.6 ultra working on Thrust right now.

A couple years ago I decided to stop maintaining a .vimrc (after ~35 years), and started using LunarVim and later AstroVim with as few customizations as I could live with.

But then around 2 months ago I decided to switch to NixOS and there wasn't a Nix way I could find to use Astro. So I had Opus build me a vim setup for NixOS that included the batteries I wanted in my setup. I gave it a paragraph description and it built something that has been a joy to use.

^G in Claude and Codex will drop you into $EDITOR to edit your prompt. I use that all the time.

"vimdiff" is a really great way to review code changes side-by-side.

I'm also a long-time Claude Code user here, though the last 3 weeks I've been doing loops having claude use codex to review until they reach consensus; uses tons of tokens but the result is really good.

I'm trying Codex as my primary the last day or so, because I'm at 98% use and reset in 3 days on Claude. I'm worried about a lot of our skills and CLAUDE.mds and the like getting lost unless I migrate them, but otherwise codex seems to be working great.

Has anyone tried teaching the agents to use jj or gitbutler-cli to do their ideas of lightweight branches? I keep feeling like that would be a huge win for the multi-branch development I'm doing much more of with agents, but I haven't yet run that experiment.

Kind of shooting themselves in the foot here. In the process of getting all my Fable use in, I'm also at 68% on overall week limit and 5 days left. So not only am I likely to be using OpenAI much more heavily this week, I'm going to be doing it while being slightly annoyed at Anthropic.

Man, this "Fable" has been rough. Last week we get access to it again, then they reset usage after a day or two with no warning. Lots of "missed" Fable time because I was pacing it for a reset 5 days down the line. Use up all my Fable time last week, resets Sunday and I don't have enough work projects to burn up the tokens by today, so I burn them on some toy and side projects, which I wouldn't have done if I knew I was going to have it until Sunday. Now I'm at 100% and they give out more access.

It would have been WAY more useful for them to announce the extension, you know, yesterday. This is basically the worst time for them to announce it. Bunch of goobers, who thought this would be a good idea?

For me, that was the big selling point. If you aren't working with multiple agents at once, I'm not sure you'd pick Herdr.

Yes, it seems that it does though I have not yet tried it: https://herdr.dev/docs/configuration/#worktrees

The primary benefit I've gotten over just a straight tmux session is that there is a collapsable left "tab" bar that shows you your different workspaces, which you can relabel, and below that is a list of the agents you are running (claude code, codex, etc) along with their status (idle, blocked, working).

So I will start a workspace for each different thing I'm working on, label it "Studio Shed Packet", "Teapot game", "Mux experiment". Then in each one I run a Claude Code. Then I can see the status of my Claude Codes just by glancing at the sidebar, rather than having to switch between screens to see what is waiting next.

I've been using it around a week so far.

I've been using Herdr for running my AI agents and it has been really nice. I like that I can reconnect to it from multiple sessions (I have one up in a window on my desktop, and when I ssh in from my "after hours" laptop I can also attach to it there and continue one if I'm at my son's Dr appt or the like. I can also attach to it natively from my Mac over SSH transport as well as resuming a remote wezterm connection that is running herdr.

A few small downsides: I can't copy/paste in wezterm using the keyboard/vim keys because it is constantly drawing the screen and unselects my selection. The mouse drag in herdr works very well though. It'd also be nice if you could rebind key mappings in the UI, because I still haven't rebound the keys and am using the mouse.

I love that people can just bang stuff into existence now.

I need to see if it can build me a fabulous todo app.

4-5 years ago I spent around a year refining a paper-and-pen daily planner, and I went 100% into it and liked it. The paper planner though had a non-ideal workflow, and so I wanted to look into some automation of it. I got a Kindle Scribe hoping they might grow its capabilities. It had a great "it's like paper" experience, but no automation or tools or apps seemed to come along. So then I tried a fairly high end Boox, I got it and then used it only a couple weeks and then stopped doing that daily planner process entirely.

But if Fable could build a Boox or reMarkable app, that might be a big win.

I've tried task warrior, and todoist, and in the end I just keep going back to a straight text file I edit.

FYI: I've never had any problems with that. Back in the mid-2000s I was running the WiFi for 1000+ people at a time at conferences for 3-4 years, and basically had no complaints.

It is basically always better to run more APs at lower power in the areas where you need coverage, than to boost the power. Especially today with the radio spectrum being so congested.

Despite this, I could expect 3-5 people to hunt me down at PyCon when I was running the wireless to tell me that I had misconfigured the wifi because it was set to low power. More reports of that than reports of wifi not working, IIRC. ;-)

(I was running the wireless because the people we paid do to the wifi would just set up one or two APs and crank the power)

A human would probably ask clarifying questions

You say that, and that's a possibility, and maybe in your experience that is true. But I'd say I've found actual humans will frequently use existing internal models of something unless they have a reason to believe something has changed to invalidate that. We even have a well worn joke about what happens when you assume.

you could for example be involved with multiple teams of different sizes that are subject to different rules

Just to clarify, my experience with the memory of these tools doesn't provide any data (for or against) how well/poorly the memory may work with someone who is involved with multiple teams.

In the "AI 2027" article/story, they're talking about next generation models "using 100 times the compute used to train GPT4. Here's a video covering that part of it https://youtu.be/5KVDDfAkRgc?si=TR8XDQmye6O6kkiT&t=169 (the 100x is at 4:38, this video starts where he starts talking compute of GPT3).

So my guess is that what is driving the need is that (speculated) 100x the next major models.

I like the memory system, in general. For reference I'm using mostly Opus 4.8 + Max effort. It will often pull things out of memory that are relevant. Like I'll ask it to come up with a few options I should consider for, say, a self-hosted OIDC provider and it'll say things like "Considering the size of your operations team, this might be a better fit because of X and Y".

Now, I'll agree that this is probably the sort of thing I should put in the CLAUDE.md, but in this case it wasn't on my radar to put that in my CLAUDE.md, so it was nice that it surfaced that.

It does sometimes go awry though. Today I was asking about a problem I was having authenticating, and it said "you may be running into this trusted proxy setting because you put your apps behind an haproxy". That is true of 95% of our apps, so it was worth mentioning, but in this case it was not so I had to correct it. But, I'm glad it mentioned it because if we did have it proxied it could have saved me a lot of time.

My use case is I have them installed and let Claude decide when to use them. Looks like for my recent sessions it has been using superpowers:test-driven-development 5%, and superpowers:subagent-driven... 1%. I haven't really been working on new projects this past week though which seems to be where they fire off the most, in particular the "writing a plan" one.

That's why I have my LLM sprinkle some into "my" writing...