navels

1) Work covers my Claude and Codex subscriptions. Each is the ~$200/month tier.

2) I have a "live Zoom call audio transcription" setup that uses Silero VAD, Qwen3-ASR-1.7B, Qwen3-4B-Instruct, and pyannote speaker-diarization-community-1.

3) Claude Code in the Desktop UI using remote control environments.

4) Custom orchestrator for having Codex review Claude or vice-versa: https://github.com/navels/neal

5) I have built over time a personal assistant using Claude remote control running on a mac mini, this is likely similar to Dots?

Following up on @killix's comments, which were helpful but he was flagged (presumably for sounding too much like AI).

Thanks for the feedback. I've made a couple of updates:

- Starting with 0.4.0, the reviewer gets the diff of the earlier chunk for any file the current chunk touches again. Also, if a new chunk weakens or removes a test or assertion from an earlier chunk, the reviewer will block it unless the plan says to do so.

- About read-only: The reviewer's tools were already limited by the SDK (no shell or write tools). However, it was still finding MCP servers from my Claude config. I have now blocked those. The docs now explain what is enforced by the system and what is just a prompt instruction.

Thanks!

I've built an orchestrator that solves some of the issues you ran into (although it doesn't do anything about cheating): https://navels.dev/blog/neal/. Features:

- lets you configure different models for planner, coder, and reviewer roles. (e.g., using Claude as an adversarial reviewer against Codex)

- breaks your plan up into reasonable-sized chunks of work with clearly defined success criteria

- runs each chunk of work through a coder / read-only reviewer loop. Once both agents are satisfied, neal moves on to the next chunk. Once everything is complete there is a final pass through the coder / reviewer loop to ensure the implementation satisfies the entire plan.

- resets the coder's context with each chunk of work to prevent context drift, leaving the reviewer's context long-running.

I've been using this CC hook with decent results: https://github.com/gvzdv/claudish-to-english

It uses a local LLM to translate Claude's output.

The migration was not a single, long run of neal (because I was developing/tweaking neal along the way). However I have used it for large-ish implementations and haven't noticed any issues with the reviewer drifting. It is re-prompted each review turn with its adversarial review instructions plus a summary of the run state (completed scopes, open findings, etc). Also, the final review is executed with a fresh context.

lol how many of these things are there. I've been using https://github.com/f-is-h/usage4claude which includes Codex usage (I use Claude and Codex).

Other activites such as the sublime pursuit of truth and beauty . . . aka mathematics ;-)

The rule is based on the symbol, not the color. Ie, there is no green or blue rule.

same parent had said "It's crazy that a language that prioritizes security so highly in it's design itself is only compiled through such insecure methods."

why not?

How weird, tteck's site is great and what is this nonsense about security.

I'm confused about whether/how this is related to https://github.com/tteck/Proxmox/

I also use NginxProxyManager (8 hosts) and I'm not seeing any replies to your post that would explain why caddyserver or traefik provide any benefit over NPM.

Related: fascinating deep dive of the Python GIL by David Beazley from PyCon2010: https://www.youtube.com/watch?v=Obt-vMVdM8s

Pertains to user scripts injected by MV3 extensions. Noticed this when my Tampermonkey extension updated today with this message:

Enabling developer mode will soon become mandatory for running userscripts via Tampermonkey.

Not a big deal for me, I have extension dev mode on all the time, but ymmv.

Some discussion here: https://groups.google.com/a/chromium.org/g/chromium-extensio...

This is my main usage of ChatGPT :-)

Based on the turnout at the Northwest Chocolate Festival last month, I can confirm it has exploded :-)

lol breaking news

What a useless article, especially given the title. Here's all they have to say about the layout:

There’s some dispute over how and why Sholes and Glidden arrived at the QWERTY layout. Some historians have argued that it solved a jamming problem by spacing out the most common letters in English; others, particularly more recent historians, hold that it was designed specifically to help telegraphists avoid common errors when transcribing Morse code. Regardless, after around 30 test models, Sholes and Glidden settled on QWERTY—and changed the world.

The problem of intergalactic communication is a plot point in this fantastic sci-fi novel I am currently reading (https://en.wikipedia.org/wiki/A_Fire_Upon_the_Deep) cannot recommend enough although it doesn't really inform the current conversation :-)

> It was totally dysfunctional, but a lot of people got promoted. All while we didn’t have a single customer!

At Amazon, really? :-)

Banner from the linked forum:

To our valued customers: After careful consideration, we have made the decision to end the Amazon Honeycode beta service, effective February 29, 2024. New customer sign-ups and account plan upgrades are no longer available. Existing customers will be able to use Honeycode and your Honeycode apps as normal (and add team members to your existing account) until February 29, 2024, when the service will be discontinued. After this date, you will no longer be able to use Honeycode or any of the apps you created in Honeycode. To learn more about this change, and how to download your data, visit the Community Discussions.

When I was working in the field it sure as heck wasn't because of practical applications. The mathematics involved is beautiful.

Current setup on my M1 for dev is to use docker with containers for mysql, localstack, rabbitmq, redis, and ruby (debian bullseye), all arm64 images, and then VS Code with the remote container extension. Was pretty rough going when the M1 first came out but is now flawless and lightning fast. Onboarding new devs takes an hour instead of a day.

The Therac-25 disaster should be on this list: https://en.wikipedia.org/wiki/Therac-25