Posts tagged: ping

All posts with the tag "ping"

89 posts latest post 2026-11-01
Publishing rhythm
Nov 2026 | 2 posts

small teams still work

Listening to the latest raising an agent podcast the last section was talking about small teams being effective and other people telling amp,

rough translations from memory

thats cute you can just talk to everyone in slack and just ship things without pr, that doesnt work on a gigantic team

...

build your tools

Whatever tool is getting in between you and delivering products, agents can just make it if you ask them.

No one knows what an agent is

Friendly reminder, something like 1% of developers are using agents with frontier models. Meaning almost no one in the world knows what the heck they are.

Ventoy ready

I got my live boot ready with ventoy a few days ago…. I should have backed up $HOME first

Google like a boomer

I now Google like a boomer, full sentence structure, its surpassed keyword search

Personal software is better

If gippity can vomit out software 80% of what you made I aint using it anymore, I get way more value from asking the magic clanker to make it for me that I do trying to figure out how to use yours.

Start your own learning platform

If you want to git gud you need to have some projects of your own. And it does not need to look like a @technotim thumbnail.

Ask your friends if they have any unused pc’s, hard drives, anything, other peoples trash makes a surprisingly great homelab. This is not just your starting point but where you keep an eye out for new nodes that you can join to your cluster.

No friends, look for any refurbished enterprise mini pcs. These work great for low power compute nodes. They do not handle a lot of disk, you will need...

...

Open binaries

binaries are open. what kind of s*** are we going to find cars, planes and trains?

What to do while you wait for your agent

Why is this still a question? You start another one, thats how this works now. I often have 23 agents up and going, maybe 3-5 actively cooking. A bunch of them are done that im reviewing the final statement, maybe i want to make a note still, or need to review the output, or debating a follow up.

Im not saying its the most healthy, productive, or best. But its what all of the people I talk to that actually build things are doing

this question came up on syntaxfm

...

typesafe jev any good

Just got access to typesafe Jev very interesting new paradigm of responding to structured queries with structured responses VERY FAST, It will be interesting to see how this changes developer tooling moving forward.

Are the ai labs really slowing down?

Does anyone believe the labs when they say they are slowing down? I think a majority of people believe they ALL should. I think it’s arguably impossible to slow them all down, so is there really any benefit to only slowing down the ones that you can regulate and let the others catch up?

Touch dishwasher controls

Touch dishwasher controls that don’t work with wet hands should be illegal, I’ll be starting mine with dried on s**** now.

politics in linux

IDK if I agree with DHH on politics, have not gone looking for his views, don’t really care to. I can 100% get behind leaving politics out of work, foss, linux. Until he starts dropping that omarchy money on some non-oss linux manner I’m with it. I don’t think there is a world in which you can completely disassociate yourself with everyone that has a different viewpoint than yourself. There will always be things that simply don’t align with your lived experiences and that is okay.

when did openai bring back the hourly limit

Here I am cranking away with lunamax tokenmaxxing on my $20 plus sub trying to keep 3 running around the clock to burn those precious tokens when open ai comes back and re-introduces the 5 hour limit, I just got used to rocking only a weekly limit.

second brain slop is dumb

The second brain hype train of people looking at fancy graphs in obsidian of the content their clanker wrote for them is top tier cringe. If you actually want to get better at anything you need to actually put some effort in with your first brain.

agentic engineering takes discipline

Agents allow so much scope creep it is harder to say no than it is to say yes, and let the model just pop in a new feature for you alongside your current PR with almost no effort on your behalf, stacking up so much validation work for you to do that it becomes very difficult to merge. We are all learning this discipline right now.