wilj

It has a pagespeed insights score of 55 and noticeably sluggish on my m3 max.

And what's with the throwaway account for this one comment? Is this becoming reddit with throwaway shills now?

I would like to subscribe to your newsletter.

Compaction.

ChatGPT desktop this morning lost a chat thread while I was actively working in it. I asked Astra to find the lost session, and next thing I know it's prompting for full computer control to drive Finder. It's just jsonl files on disk, not hard to read normally.

Negative feedback filed and ChatGPT uninstalled.

runpod.io is essentially this. You can rent the hardware for cheap in small time slices. I do this whenever I need to do a lot of embeddings, fast. I have an agent skill that will estimate the optimum hardware to reserve for the time/price constraints of the job, and you can spin up temporary inference for cheap via their API as well.

+1 for code mode. It's a game changer for runtime performance, flexibility of orchestrating lots of tool calls with complex logic, and all sorts of other goodies.

I'm in the process of switching all my personal stuff to a self-hosted fork of cloudflare-os right now. It's taking a lot of rearchitecting how my stuff works to fit within the cloudflare "no local files" paradigm, but for now I've got a container gatekeeper they can drive and they can check repos out in it.

I've been consistently unable to get it to tell me anything at all about Deepseek v4 Pro, it keeps insisting it is beyond its cutoff date, but it knows all about R1.

If it's routing to different models on the backend, it's either pinned for the user, or they're all really old.

I'm looking forward to trying it, but not willing to upgrade to Tahoe, so I'd appreciate it for sure!

Similar here. I use Claude Code and Codex for developing specs, but then have Deepseek v4 Pro implement because it just seems to actually do what it's told, given enough tries. And Reasonix + Deepseek is mostly cache hits when I let it grind on it for a long time, and I can have a huge fleet of agents with very specific focus for code review and rule enforcement mostly sharing cache.

> I know open models have gotten quite good in many tasks such as coding or composition, but are there any that can access the internet and retrieve data like ChatGPT, Claude, etc can?

The things you describe are just tool calling, they're a feature of whatever harness you use. Use OpenCode, pi.dev, or maki.sh with any of the open models.

> I do have to admit I have recently begun wishing I could pay five dollars a month for a "just answer the fucking question" plan that would give me results without the guardrails and without the constant simpering and ego-stroking. I keep finding myself going a quick evaluation of "is it faster for me to skim search results myself or to construct an elaborate narrative to make an AI give me a real answer".

You can do most of this with some system prompts added to whatever agent you're using. You can do it from the settings on the claude/chatgpt websites too. (minus the no-guardrails thing)

There are a number of places you can get an online prescription for a CPAP/APAP, and they're not that expensive to buy outright.

I was able to get a new prescription for < $100, and bought a new APAP for about $600, no new sleep study required.

This is exactly what stops me from using them for real work. I love their free tier for my hobby stuff.

My first time using NT 4, I was setting up a bunch of machines that needed babysitting, and Space Cadet Pinball got me through a lot of long, boring nights.

I've thought back throughout my career to how lucky I was, it kept me from going crazy. Thank you!

I just cancelled my Gemini subscription yesterday. I have a big private fork of OpenCode, and I did it the wrong way to start with, so I couldn't pull from upstream.

So I put together a plan for refactoring it, step by step, with tests, etc. After literally 8 solid days of fighting with Gemini 3 Pro, I still couldn't pull it off.

I gave GPT 5.5 a chance with the same prompt, plans, and repo. I'm not sure how long it took, but when I checked in on it a few hours later it was done. All tests passed, everything exactly how I'd asked, and better (it made some improvements).

This is awesome! I have a slight case of tritanopia in one eye and it was neat to see the difference. My boundary is bluer by 59% in one eye and 87% in the other. It tracks with what I would have expected.

I've got to do some cleanup before sharing (yay vibe coding) but the big things I've changed so far:

1) Curated a set of models I like and heavily optimized all possible settings, per agent role and even per skill (had to really replumb a lot of stuff to get it as granular as I liked)

2) Ported from sqlite to postgresql, with heavily extended schema. I generate embeddings for everything, so every aspect of my stack is a knowledge graph that can be vector searched. Integrated with a memory MCP server and auditing tools so I can trace anything that happens in the stack/cluster back to an agent action and even thinking that was related to the action. It really helps refine stuff.

3) Tight integration of Gitea server, k3s with RBAC (agents get their own permissions in the cluster), every user workspace is a pod running opencode web UI behind Gitea oauth2.

4) Codified structure of `/projects/<monorepo>/<subrepos>` with simpler browserso non-technical family members can manage their work easier (agents handle all the management and there are sidecars handling all gitops transparent to the user)

5) Transparent failover across providers with cooldown by making model definitions linked lists in the config, so I can use a handful of subscriptions that offer my favorite models, and fail over from one to the next as I hit quota/rate limits. This has really cut my bill down lately, along with skipping OpenRouter for my favorite models and going direct to Alibaba and Xiaomi so I can tailor caching and stuff exactly how I want.

6) Integrated filebrowser, a fork of the Milkdown Crepe markdown editor, and codemirror editor so I don't even need an IDE anymore. I just work entirely from OpenCode web UI on whatever device is nearest at the moment. I added support for using Gemma 4 local on CPU from my phone yesterday while waiting in line at a store yesterday.

Those are the big ones off the top of my head. Im sure there's more. I've probably made a few hundred other changes, it just evolves as I go.

My biggest problem with CC as a harness is that I can't trust "Plan" mode. Long running sessions frequently start bypassing plan mode and executing, updating files and stuff, without permission, while still in plan mode. And the only recovery seems to be to quit and reload CC.

Right now my solution is to run CC in tmux and keep a 2nd CC pane with /loop watching the first pane and killing CC if it detects plan mode being bypassed. Burning tokens to work around a bug.

I'm 3 weeks into switching from CC to OpenCode, and in some ways it is far superior to CC right out of the box, and I've maybe burned $200 in tokens to make a private fork that is my ultimate development and personal agent platform. Totally worth it.

Still use CC at work because team standards, but I'd take my OpenCode stack over it any day.

It varies from project to project, but applications benefit a lot more than libraries. When I de-lib a normal express app it might add a few hundred lines of code and a few thousand new tests, but if I de-lib an library then depends on how ancient it is. The older the library is, the higher the chances that most of what it needs is built-in to the standard library.

I literally have a Claude Code skill called "/delib" that takes takes in any nodejs project/library and converts it to a dependency-less project only using the standard library.

It started as a what-if joke, but it's turned out to be amazing. So yeah, npmjs.com is just reference site for me now, and node_modules stays tiny.

And the output is honestly superior. I end up with smaller projects, clean code, and a huge suite of property-based tests from the refactor process. And it's fully automatic.

Sorry for late reply, but yeah that's how my workflow looks, but I'm also more just leaning on MiMo V2 Pro now, it's fast, and cheap enough. And I'm using OpenCode.

I'm about 2 days into transitioning, using MiMo V2 Pro in place of Opus and MiniMax M2.7 in place of Sonnet.

I'm finding that the extra "hand holding" that MiMo and MiniMax need isn't really "extra." The Anthropic models happily agree to a plan and then do something else entirely way too often.

With MiMo and MiniMax I'm just spreading the attention throughout the day instead of big spikes of frustration figuring out where Claude went off the rails.

I use the same setup with i3wm (Regolith desktop) and love it.

This seems to have a lot in common with our restorative justice program: https://www.connectionfirst.org/

We operate as a diversion program with our local department of juvenile justice, and I would definitely describe the process as transformative.

From the financial perspective alone, our local law enforcement has seen a 6:1 return on investment on restorative practices. The value of all the changed lives is immeasurable.

Keycloak-as-a-Service

I can build it for you if you like.