Ryan Burnette

@ibuildthecloud @_beyondcode I use proxmox to create lxc containers that are isolated development environments

@_beyondcode @ryanburnette proxmox is for VMs. That's what I know it for. So what does that mean for isolated development environments?

Let's make isolated developer environments a real thing. We failed like 20 times to do this but who's up for another round?

@ollama Your platform is slow to the point of being unusable for me again. What gives?

@AdamWhitcroft Start with a Hamilton Field Khaki Automatic.

I'm sick of my Apple Watch. People into mechanical watches, what's a good intro watch? Looking for something good quality and timeless. Help.

Two films. One big screen. Back in theaters, May 13th, for one week only.

@RhysSullivan Use containers

the people who think this stops claude from being able to read your envs

@wagslane You can be kind and strong simultaneously on this. Hang in there.

I literally can't win with pronouns on Boot dev, it's insane. We have people cancel over NOT using "they" in lesson text and people cancelling over us USING "they". > In one of your lessons you refer to a single person named "Logan" using the third person plural pronoun. > > At this point without assurances that the developers intend to correct the course material to correctly use the grammatically correct generic masculine as the correct generic pronoun, I have no interest in financially supporting the intentional cultural vandalism of my personal linguistic heritage Here's your refund, you won't be missed. I DON'T CARE about this. I don't care. WHY DO PEOPLE CARE. These are not real people in the lesson text. They're genderless EXAMPLES. Sometimes we say he/she, sometimes we say they, it depends on context and which maintainer happened to update it last. wwhwhhwyyyyy dooooo weee careeee sooo much. get a life.

@Zain_Wania @wagslane opencode is the harness, tui and web are interfaces the harness offers

@wagslane the harness or the tui? tui is far superior, harness i would highly disagree Auto mode changed my life, not sure if parellel subagents have hit yet in opencode. anthropic plugins feel better to install (marketplace). just to name a few

I would use Anthropic models for coding if I could safely use a max sub in OpenCode But I can't so its OpenAI for me for now yes, the harness matters that much to me

@wagslane @dino11 @AnthropicAI not cool

@dino11 It's not opencode that's the problem. It's anthropic banning subscription users that are using opencode

@wagslane I prefer open code too. Using Ollama right now. Disappointed to hear about this.

@dhh I use telebit

Omarchy has a neat built-in SSH tunneling port forwarder called fip. So you can start three web services on a remote box, forward those ports there, and you'll be able to access them all in the secure context of localhost. Super nice for remote dev! github.com/basecamp/omarc…

An economy is built on the production and exchange of goods and services. Money is the medium of exchange that makes this possible. When individuals accumulate money without contributing to real production or exchange, they dilute the economic power of those who do. Wealth becomes detached from output, reducing what productive participation is actually worth.

@_beyondcode @jmorgan @ollama @coolaj86 Yes, ollama is essentially unusable overnight

@ryanburnette @jmorgan @ollama @coolaj86 Has it gone from excellent to unusable overnight for you too?

@ollama cloud seems *really* slow today on glm-5 and glm-5.1 I'm switching back to hand coding because even simple tasks are taking waaaaaay too long. Are the resources over-provisioned? Is the pricing model not profitable?

@ollama your product has become unusable in the last 24-48 hours, please don’t treat your users like this, add a wait list, do something, please

@_beyondcode @jmorgan @ollama @coolaj86 Not unusable but performance is an issue

I’ve been working through some relativity scenarios and keep arriving at what feels like an unresolved question around energy accounting and received intensity. I want to know how current models address this, whether it’s been tested, or where my reasoning breaks down. Scenario 1: Two identical logs are ignited simultaneously. I travel with one on a ship moving in a large arc around the other at constant distance, then return. Due to time dilation I experience 15 minutes while the stationary log burns completely. My log is only partially consumed. Here is the tension I can’t resolve: the stationary log radiated its entire lifetime of energy during my 15 minutes of proper time, while I maintained constant distance from it. The log beside me only radiated 15 minutes worth of energy. If both logs appeared at equal luminosity to me throughout the journey, the energy accounting seems inconsistent. The stationary log released substantially more total energy during my proper time. Should I not have perceived a higher photon flux from it? Or conversely, was I receiving a reduced flux per unit of my proper time precisely because time dilation caused me to sample a smaller fraction of its total emission rate? Scenario 2: A simpler version of the same question. If I am stationary relative to a burning log and then begin moving in a circular path around it at constant radius, do current models predict a change in received photon flux or perceived luminosity purely as a result of that transverse motion? Not a change due to distance. Not purely a frequency shift. But a change in the rate at which photons are intercepted per unit of proper time — essentially a flux reduction resulting from time dilation induced by angular velocity. Core question: Do current models predict a measurable reduction in received photon flux from transverse motion at constant distance, distinct from and in addition to the transverse Doppler frequency shift? Does the energy accounting in Scenario 1 resolve cleanly somewhere I’m not seeing? And has either of these effects been directly tested or observed?

@_beyondcode @jmorgan @ollama @_beyondcode @coolaj86 I called this. They’ll aggressively over provision and increase the price so we’ll have to find an alternative.

@jmorgan @ollama What about putting new users on a waitlist or switching to an invite system (like gmail and other famous products)? I know this will come off harsh, but: It always sucks as a customer when a SaaS product you chose decides to choose growth over the people who are already paying.

@wagslane I am certain that I write and communicate poorly and AI does a better job than me. 🤦

The benchmark you’re describing can’t really be known without external measurement — an engagement and comprehension study comparing your writing to AI-generated content would be genuinely interesting. Interesting enough that I’d pay to see it. My own experience: AI with minimal guidance produces mediocre output. But when I write something first and use AI for targeted refinement with strict context, the result is noticeably better. I did use AI to help write this, actually — which maybe proves the point.

just dawned on me that its possible my benchmark is bad maybe I'm the bad writer and AI is trying to help