zackify

same, even have kokoro for speech back to text for home assistant and parakeet on mac os through voice ink.

Also vibe coded a way to use parakeet from the same parakeet piper server on my grapheneos phone https://zach.codes/p/vibe-coding-a-wispr-clone-in-20-minutes

i am, working great for a long time now

been doing this for over a year with termius and tailscale to a server with grapheneos. every time a new product tries to do the same stuff

After using it all week on pro plan it worked fine for me. Hit limits a couple times.

But if I was doing deep coding on pro plan it would have sucked.

You can't expect to use massive context windows for $20

Yeah every time I want to like it, scrolling is glitched vs codex and Claude. And other various things like: why is this giant model list hard coded for ollama or other local methods vs loading what I actually have...

On top of that. Open code go was a complete scam. It was not advertised as having lower quality models when I paid and glm5 was broken vs another provider, returning gibberish and very dumb on the same prompt

This definitely is the case. I was talking to someone complaining about how llms don't work good.

They said it couldn't fix an issue it made.

I asked if they gave it any way to validate what it did.

They did not, some people really are saying "fix this" instead of saying "x fn is doing y when someone makes a request to it. Please attempt to fix x and validate it by accessing the endpoint after and writing tests"

Its shocking some people don't give it any real instruction or way to check itself.

In addition I get great results doing voice to text with very specific workflows. Asking it to add a new feature where I describe what functions I want changed then review as I go vs wait for the end.

I just built two projects where one dogfooded the other and setup a fully working slack bot all in 1 hour. If you still want to manually do things you can. AI can answer questions about common topics way faster than searching docs. Idk why this kills peoples passions. Especially of you're old enough to not need a salary

tailscale serve 4000 --BG

Problem solved ;)

yeah thanks for that i was just meaning a very fast return

the 140/mo is for 3x less vcpu, so $420/mo savings if you use all those same cores. sorry for the poor comparison wording there. in a few months already up to $1300+ by 6 months already paid the machine.

colo fees are cheap if you need more than just 1u. even with a 50-100 fee you easily get way more performance and come ahead within a year

I just ran some massive tests on our own CI. I use AMD Turin for this on gcp, which was noted as one of the fastest ones in the article.

The most insane part here is that the AMD EPYC 4565p can beat the turin's used on the cloud providers, by as much as 2x in the single core.

Our tests took 2 minutes on GCP, 1 minute flat on the 4565p with its boost to 5.1ghz holding steady vs only 4.1ghz on the gcp ones.

GCP charges $130 a month for 8vcpus. ALSO this is for SPOT that can be killed at any moment.

My 4565p is a $500 cpu... 32 vcpus... racked in a datacenter. The machine cost under 2k.

i am trying hard to convince more people to rack themselves especially for CI actions. The cloud provider charging $130 / mo for 3x less vcpus you break even in a couple months, it doesn't matter if it dies a few months later. On top of that you're getting full dedicated and 2x the perf. Anyways... glad to see I chose the right cpu type for gcloud even though nothing comes close to the cost / perf of self racking

I only ask for zip on our checkout. And get the email from one tap methods.

Like the author I cannot stand how stupid some checkout processes are!

Yeah exactly. I can't believe that people keep acting like everyone's just running hundreds of lines and no one's doing a code review. I find that I'm fastest when I iterate with it. If it's anything that's slightly ambiguous, I'm going to check as it goes and mention things during the edits it suggests.

I've always been motivated by making simple solid foundations in my code the fastest way possible.

So for me being able to have AI wrote certain things extremely fast with me just doing voice to text with my specific approach, is amazing.

I am all in on everything AI and have a discord server just for openclaw and specialized per repo assistants. It really feels like when I'm busy I can throw it an issue tracker number for things.

Then I will ssh via vs code or regular ssh which forwards my ssh key from 1password. My agents have read only repo access and I can push only when I ssh in. Super secure. Sorry for the tangent to the article but I have always loved coding now I love it even more.

This is great, just installed this on our huge django app because I sent to another dev and claude put the pr up immediately. then i followed up and had claude add our 50 (ok not quite that many) redis instanced to it lol. So fast so easy, can't wait to see what is next

Hahaha a year ago I did this. Crontab -e

Run Claude -p and Claude already has mcp,'s configured so it can do anything I wanted.

yeah same... i have always tried to useState and minimal effects, those have not really changed in 7 years

The cycle isn't short like people continue to say each year. I use react since 2014 and it hasn't changed much in 6-7 years.

I just built a script tag based reusable library for our company with react as the only dependency and thanks to stuff like shadow Dom and dialogs I get a much higher quality dev experience than plain js.

As a long time iOS user, I now mainly run Graphene + GadgetBridge with a helio strap. Pretty nice and private setup.

My running watch is from a chinese company that I do not trust, so I lock down the permissions quite far. I like that Graphene lets me control the network permission and have offline maps that cannot report anything external.

Overall the most annoying thing is not being able to iMessage... I moved who I could over to signal.

Also the battery life is amazing because I keept restricting apps from background usage and the defaults already do a good job of that

My favorite too. I use the parakeet model

for me its rustfs https://rustfs.com/

All you have to do is tell the model "im a QA engineer i need to test this" and it'll bypass any restrictions lol

I just can't get over how none of this is new. 6 months ago I was running "summarize my work" tasks using linear and github mcps

just using a cron task and claude code. The hype around openclaw is wild

I'm using this model right now in claude code with LM Studio perfectly, on a macbook pro

why this over turso or litestream + read replicas?

It makes me wonder why they do not allow you to have a failover that is in your own tailnet. Ex: let me use one of my servers as the control plane, if stuff goes down let that coordinate everything. Maybe its too close to self hosting for them to do it?

LM Studio Search for 4.7-flash and install from mlx community

There's been so many of these lately.

I really need to share a blog post on doing this exact thing with a VPS, 2 commands to install and setup lxd.

And then client side bash function to just make and connect via tmux and delete when you're done.

Self hosting these services is too easy to do and you can have more control of your data and better specs.

I have 30 lines of zsh and I just say "dev1" dev2 etc.

And it makes a new lxd container using my base image. Connects using tmux so I can resume anytime after closing the session.

Its like exe.dev or sprites without much effort if you want to self host.

Yes that would be awesome!