HackerTrans
トップ新着トレンドコメント過去質問紹介求人

johnthedebs

no profile record

コメント

johnthedebs
·3 か月前·議論
I love the idea.

Just the other day, I noticed my thinking was so hijacked by distractions while building something (with AI help) that I started writing in a notebook to stay on track. The last time I'd written in the notebook was 3 years ago; in this case writing stuff down in it really helped to get me unstuck.

I'm excited to imagine workflows that could make computing a more physical activity. Thanks for writing and sharing this.
johnthedebs
·3 か月前·議論
Love to hear it, thanks for sharing. I only wish there were more parenting "tricks" like this that gave similar quality of life improvements.
johnthedebs
·3 か月前·議論
If you saw our baby, you wouldn't be worried at all that he's being underfed. Maybe the opposite. They'll want to eat every 3-4 hours during the day for sure; at night they can and do just sleep through it (same as me or you).

We didn't force anything or ignore him. And you don't have to believe me, but I'd encourage you to research more for yourself if waking up at night to feed a baby is something you're currently dealing with.
johnthedebs
·3 か月前·議論
In our case(s), it was something that required conscious effort. And when we did that... it worked. It honestly didn't seem like it would at first, but then it does.

Again, n=2 for me personally but as I mentioned in my reply to another comment we also had a friend with a "baby who won't sleep" and when they tried it also worked for them.

I don't make a habit of recommending this to people unless I'm close with them, bc I know that some people may take it personally or believe they are an exception. And I'd bet money that there are plenty of exceptions. But I also think they're exceptions rather than the rule. Whenever I've seen parents who believe that their baby can sleep through the night and work towards that goal, they seem to get there pretty quickly.

Edit to add: To put it in engineering terms, I think part of the problem is that you have to escape a local maximum of baby sleep. You may suffer several nights (possibly a couple weeks) that are worse than what you're used to in order to get to a place that's significantly better than what you're used to. When you're already sleep deprived, that can feel like a big hump to get over.
johnthedebs
·3 か月前·議論
To be clear, they don't eat/feed at night either. The baby is in a separate room from us and we don't see or hear him most nights (95%+) between 7pm-6am. He's around 8.5 months old now and this has been the case for 3-4 months, although that percentage was a bit lower at the beginning.

I'm emphasizing it bc many people are surprised by this, but if you know it's possible, you can start to work towards it. My partner's coworker has a ~1 year old who was still waking up (maybe multiple times?) each night to eat. She introduced them to one of those books (the 12-by-12 one) and they were very grateful.
johnthedebs
·3 か月前·議論
FWIW, and understanding that individual babies do differ, most babies can sleep through the night (10-12 hours) by 3-4 months old. Check out the books "Twelve Hours' Sleep by Twelve Weeks Old" or "Precious Little Sleep" for guidance.

In my case where n=2, naps during the day are/were not all that consistent but at night (unless they are very sick or something) the kids sleep.
johnthedebs
·3 か月前·議論
I'll check them out - thanks for the recommendation!
johnthedebs
·3 か月前·議論
For sure! We ended up learning this the hard way.

We moved into this house partly bc it had an extra ~acre of space beyond the main "yard" which was starting to turn into a forest. We cleared it of woody stuff but left some black raspberries, maybe 10 plants?

2 years later, it turned into an impenetrable ~1/4 acre thicket of mostly black raspberries with some wild blackberries and wineberries among them. We paid to have it mostly cleared again, and now we are occasionally mowing whatever is not intentionally planted or mulched.
johnthedebs
·3 か月前·議論
Awesome! Curious to hear if it ripens good fruit for you.

In case it's interesting: people normally grow them from cuttings to make sure that the trees will 1. be female (males have figs, but they're not really edible), 2. hold/ripen fruit without pollination, 3. be true to type, and 4. bear fruit sooner (cuttings can bear fruit the first year under the right circumstances).
johnthedebs
·3 か月前·議論
I got into fig (and since then, more broadly, fruit) cultivation. Figs have a rich history, lots of variety, and there are very active online (and in-person) communities where you can buy or exchange plants and cuttings, advice, and fruit. This grew out of an initial interest in gardening, and the long-term goal is to create a food/fruit forest around our house where me, family, friends, and neighbors can walk around, spend time, and eat the absolute best fruit possible.

So far I've got about 40 fig trees in containers (~30 varieties), am focusing a bit more on blackberries this year (4 varieties that were planted last year), and we also have strawberries, blueberries, raspberries, as well as a more standard annual garden with tomatoes, peppers, cucumbers, etc as well as some wild edibles: mulberries, wineberries, and black raspberries.

There's a lot of interesting angles to this hobby: fruit selection, cultivation, harvesting, pest management (annoying but still interesting), landscape design, etc. Planning cycles are months at a minimum, and but more often you have to keep in mind what you want the landscape and experience to be like years from now.

It makes it more enjoyable to spend time outside doing physical things when the weather is warm, and I mostly take a break from it (or switch to planning) during the winters here.
johnthedebs
·3 か月前·議論
Agree with sibling comment as someone who used Zen for many months, maybe as long as a year or two. It constantly breaks and often stays broken in small but fundamentally important ways, to the point that I just switched back to FF last week and am glad to be off the roller coaster. Before Zen I had tried Arc and left for a lot of the same reasons.

For all of the (valid) criticism against FF, it's still the best available browser that's not just an experiment IMHO.

Edit to add: part of the switch back is that FF now supports, to some degree, all the features I was using Zen for: vertical tabs (needs customization but works well enough), custom search "engines" (ie, shortcuts), split view, not-Chrome
johnthedebs
·3 か月前·議論
Just noting: it's interesting to see the term "lethal trifecta" used here given the relatively recent coinage relating to LLM security: https://simonwillison.net/2025/Jun/16/the-lethal-trifecta/
johnthedebs
·4 か月前·議論
Sometimes Apple's malicious compliance is in service of (or less generously: aligned with) users' interests. I didn't know about the added fees that parent mentioned, so I appreciate them clarifying in this case.
johnthedebs
·4 か月前·議論
Sounds like you're talking about Apple disabling Advanced Data Protection in the UK? https://support.apple.com/en-us/122234

Weird take to shift the blame to Apple for that.
johnthedebs
·4 か月前·議論
I've read many posts and comments at this point that describe LLMs in very reductionist language. Eg, from the article:

> They’re a trillion numbers in a trenchcoat; not logical, in either a machine or a mental sense, but stochastic.

Many of these posts and comments claim that human minds are substantially different ("better" is implied). The evidence is a sort of broad gesturing at explanations of how LLMs are implemented ("math") and how they work ("guess the next word"). And because of these facts, we should treat them in a particular way, or certain things will never happen.

I've been trying to look past the obvious straw man here and to actually think critically about this tech as well as compare it to my own experience and (admittedly very limited) understanding of the human brain.

In more ways than feels comfortable, it seems entirely possible to me that these things actually are or could be really close to the ways that our own minds work.

Our own minds/consciousness are ultimately based on physical processes, I don't think anyone would dispute that. At some point, the physical phenomena in our brains presumably result in the emergent behavior of thinking and consciousness. We have no idea how it works, but it's our lived experience. Why can't that be the case for silicon-based rather than carbon-based processes? How can we say with any certainty that it's not happening elsewhere if we don't know how it works?

Reducing their function to "guessing the next word" sounds an awful lot like what happens when I start talking to someone. I have an idea of what I want to say, but I almost never have a sentence planned out when I start it.

The article puts "thinking" and "hallucination" in scare quotes. But I mean – the way that they appear to think by working through problems with language mirrors my own "thinking" very closely.

It says "They’re not thinking. They’re not hallucinating"; the exercise of figuring out why is left to the reader. If you've ever talked to a 3 or 4 year old, or someone who's tired, you may have had similar experiences re: hallucinations.

These are all pretty surface level examples, but as I use the tools more and learn more about how they work I'm not seeing any significant evidence that counters the examples.

I do think it's probably dangerous and unhealthy to really anthropomorphize AI/LLMs. They're obviously not human even if they're thinking, and they're being made and shaped by companies (and training sets) that exist in a predominantly capitalist world (but then again, I guess we are too).

I assume similar lines of thinking being discussed somewhere, but I haven't found much (and I feel like I'm reading about AI all day). Curious to hears others' thoughts and/or to be pointed to wherever this stuff is being talked about.
johnthedebs
·4 か月前·議論
Besides nitpicking, even your original point isn't even true. You cannot transplant a 100 year old tree (which has not been constrained in size dramatically) and expect it to survive for any reasonable length of time.
johnthedebs
·5 か月前·議論
IMO (re zoom behavior): if you make a new tab, it should add a new tab as normal and stay zoomed in. the tab bar (of the currently zoomed in panel) would still be at the top while zoomed in, and workspaces still appear to the side

if you make a new split (or navigate splits), it would zoom you back out (contract the panel) and just split/navigate the way it normally would
johnthedebs
·5 か月前·議論
Hey, this looks seriously awesome. Love the ideas here, specifically: the programmability (I haven't tried it yet, but had been considering learning tmux partly for this), layered UI, browser w/ api. Looking forward to giving this a spin. Also want to add that I really appreciate Mitchell Hashimoto creating libghostty; it feels like an exciting time to be a terminal user.

Some feedback (since you were asking for it elsewhere in the thread!). Happy to go into more detail about any of these if it's helpful:

- It's not obvious/easy to open browser dev tools (cmd-alt-i didn't work), and when I did find it (right click page -> inspect element) none of the controls were visible but I could see stuff happening when I moved my mouse over the panel

- Would be cool to borrow more of ghostty's behavior:

  - hotkey overrides - I have some things explicitly unmapped / remapped in my ghostty config that conflict with some cmux keybindings and weren't respected

  - command palette (cmd-shift-p) for less-often-used actions + discoverability

  - cmd-z to "zoom in" to a pane is enormously useful imo
johnthedebs
·6 か月前·議論
I think the only people they're calling "dodgy" are the ones offering these AI tools, and not the people using them.
johnthedebs
·7 か月前·議論
Thanks for sharing your process and config! I went on a pretty similar journey, though I've been on macOS the whole time. I've gotta say, Aerospace is one of my favorite pieces of software. It really makes it so much less tedious to move around my computer.

I tried vimium and homerow too, and I liked them, but lately I've been using mouseless more (https://mouseless.click) and overall would recommend it.