I have this 11 year old oddly resistant Pentium laptop and I’m thinking of turning it into a reading/light-programming tool. It used to run great back in the day but modern software has gotten so bloated that it can barely run GNOME with Firefox, so I was thinking of sticking to command line only. Is there anything specific I should look into?
In specific I mainly only want to be able to download and read mdbooks in the terminal, probably using archlinux32 as the OS (or maybe LFS?). Captcha abuse and all that javascript already ruined browsing with Lynx so I have little hopes of actually browsing the web. I also intend to get a new battery as it only lasts 1-2 hours nowadays. Any other 32bit/tty-only customisation guides are also welcome.
A bit of a tangent, but I find old laptops make great home servers and media centres. The nice part with them is that they effectively have a UPS built in, and there’s a keyboard/screen when you need to do any troubleshooting.
On that note, here’s a massive awesome list of potential self-hosted projects and tools one can consider.
nice
One thing to consider there is that CPUs are much more efficient now. So if it is a really old laptop, it might consume quite a low of power.
Yeah, there is that too, but it is going to be idling most of the time and laptop CPUs have been designed to conserve battery for a while now, so they should behave when not under load.
I thought of that too and may still do it in the future or if the screen gets defective. I just changed my mind because I realised I’ll have to spend a lot of time away from home soon and don’t want to take this flimsy, expensive and ultra-thin modern crap that I use for most stuff. I’ll probably repurpose some other older notebook from a relative for that too.
So long as I have some sort of easy way to locally browse documentation and mdbook reader with the tty, and maybe something for github and Overflow Offline if their owner corporations stop ruining things for a second, my productivity will remain basically the same.
Oh didn’t know about overflow online. That’s really nifty. On a random note, would be neat to see if you could train a LoRA with this data for one of the open source language models and use it as a private chatgpt style helper. It’d be pretty cool if you could just ask questions using natural language and it could spit out a relevant SO answer tailored to the situation.