Live data from Hacker News

Protect your shed

dylanbutler.dev

81–89 of 89 posts

Re: Protect your shed

#81
post #80

Earlier quoted context omitted.

Psychological in the sense that my kids will need to psychologically accept that they now live under a bridge?

Callous, but that’s your fault for building a life that requires tech money to maintain. I don’t get the point of comments like yours, just to make one feel bad for escaping the golden handcuffs. There are plenty of people that have children and live decent lives earning less than $200k/year + benefits.

I'm making under 200k now. If I got out of software I'm pretty sure it would still go down.

Sorry for making you feel bad.

Re: Protect your shed

#82

Earlier quoted context omitted.

Can we have a call about your dream? I am in a similar boat. My email is in my profile, and my comment history exists. No worries if this is a bit too forward. It just seems fun to brainstorm about a dream like this and we may have some complementary experiences.

Can I get on this action too? This sounds like a good idea.

Feel free to email me :)

Re: Protect your shed

#83
Really good read. I think you really opened eyes here for myself and others. Different fields, different paths but all so impactful. Exactly the kind of perspective that helps you reset and get back on track.

Re: Protect your shed

#85
This post make me feel that working on your own projects is not a waste of time. It helps you more than you realize. Either by bringing you joy or experience, just keep working on your "shed" if you enjoy that.

Re: Protect your shed

#86
My personal projects are the only think keeping me sane on this Brave New World of AI coding and the fact that we, AI skeptics, are all stuck in a prisoner dillema we're so acutely aware of, we find no other way than being the fearless pioneers of all that shit.

I still have lots of reservations against AI generated code, even wrote a couple articles about them that I never published, but decided that fighting against the flow and the mania is useless.

But, I don't want my kids to have student debt, so, I will just smile and nod along during the day, being even an enthusiastic leader of AI initiatives, taking care not to be the cassandra that asks where are the evaluation harness, the safeguards, as this is frequently frowned upon by the people who gets enamorated with a demo and wants to promete it to production yesterday. The people who believe that if you just spit a bunch of csvs for an LLM it will work like a contraint solver.

My personal projects bring me the crafstman satisfaction I increasingly can't find on my day job.

Re: Protect your shed

#87
post #76

Earlier quoted context omitted.

I have a professional 'homelab' and a personal 'homelab'. You're 100% right, they can be a time sink. The important bit is to make sure the time is setup not 'maintenance' time. The trick is twofold: if it isn't 'declare and deploy' don't run it. If it isn't in your backup/restore pipeline don't run it. Pfsense and Home assistant are huge pains in the ass. Everything else is easy breezy. Proxmox/pbs/truenas/talos/lin…

Out of curiosity, what makes Home Assistant a pain?

I would guess I/O. Your normal containers need network access and that's it. HA, depending on your setup, might want Bluetooth, a USB zigbee dongle, z-wave, etc etc

Re: Protect your shed

#88
post #76

Earlier quoted context omitted.

Out of curiosity, what makes Home Assistant a pain?

I would guess I/O. Your normal containers need network access and that's it. HA, depending on your setup, might want Bluetooth, a USB zigbee dongle, z-wave, etc etc

No, the I/O passthroughs are fine. Proxmox and HA are fairly great at keeping them stable. I have passthrough for WiFi/BT/USB zigbee/USB thread Of course it pins you to a singular proxmox host, not benefitting from proxmox HA, but that's the way the cookie crumbles with h/w.

pfsense and home-assistant both claim to be declarative configs, which is technically true. However the config files are not well or effectively documented, and where there is documentation it typically relates to the GUI which diverges significantly in arrangement. Their configs are declarative in that they declare the way their internal processes are configured, not in the way that they should interact and appear to other services (networking people will find that statement very confusing).

Both are effectively "Operating Systems" within operating systems, starting/stopping/configuring/managing other programs, home assistant is doing this to the nth degree. When you start them it is very hard to determine when they have actually started - particularly the bits you care about. Getting errors and logs out of them is painful. Updating configs and restarting has multiple routes, the longest of which is very long.

Both are reasonable ways to get to grips with the problem areas they solve for; they are not optimal however.

Re: Protect your shed

#89
post #76

Earlier quoted context omitted.

I have a professional 'homelab' and a personal 'homelab'. You're 100% right, they can be a time sink. The important bit is to make sure the time is setup not 'maintenance' time. The trick is twofold: if it isn't 'declare and deploy' don't run it. If it isn't in your backup/restore pipeline don't run it. Pfsense and Home assistant are huge pains in the ass. Everything else is easy breezy. Proxmox/pbs/truenas/talos/lin…

Out of curiosity, what makes Home Assistant a pain?

My other comment goes into more detail relating specifically to my earlier comment. But a niggle with Home Assistant in particular:

It appears to be a system designed to be debugging adjacent, with every intent to frustrate debugging.

Post reply on HN