jscheel

Yeah, I definitely agree with your concern about getting sloppy in production. But to be fair... "we'll fix it in post" has been a thing for a while.

I tried to buy new SD cards for my camera. The cards I used to buy at $28 are now $80-120... if you can find them. Another cheaper card I used to buy for $19 is now $46. It's just absolutely insane at this point.

I really don't understand why people are complaining about the AI features. These all mostly seem like solid quality of life enhancements and CGI-like tweaks.

Zeb Evans posts that ClickUp is reducing headcount by 22%.

I feel like we've been working in parallel here :) We are using PyAirbyte (hi aaronsteers) for our users to connect their data sources to our agents. We originally wanted to use the airbyte white-label platform, but the team said that it was being deprecated. I think this really drives home just how crucial it is to have a clear model for accessing your data, and Airbyte has been great at that for quite a while.

Generally speaking, raw-milk cheeses are much safer to consume than fresh raw dairy milk. However, the dairy in question is a garbage company that has been responsible for numerous outbreaks in the past.

I got through this entire article before I realized it was written by someone I worked with back in my agency days. Beth is an awesome designer with a great eye. Nice to see her on the front page here. Now, to the content: I often wonder how much we have lost with our endless quest for minimalism. We can't even make buttons look like buttons anymore. Affordances have become anemic at times. Designers who think and care deeply about functional color theory and usable design should be cherished.

And exactly how many consumer routers are not foreign made?

Jack Dorsey has a habit of explosively increasing headcount. Twitter was so overweight that 80% were eliminated when Musk took over. Block's headcount grew from 3,900 to 12,500 in three years during Covid. Block's stock price has also tumbled from ~$275 to ~$54 since 2022. I think that the severance package is incredibly generous, and the willingness to communicate with those affected is admirable. But I also think that Dorsey is spinning a story to cover up for ZIRP-era mismanagement. AI provides the justification, with the hope that dumping 2x the work on the survivors won't crush them because AI tools will help. The bet may pay off, I'm just skeptical of the justification.

and every single one of them checked "I have searched existing issues and this hasn't been reported yet"

Epic generated approximately $6 billion in revenue in 2024, just for comparison.

I get a flood of these every single day. Because we use SendGrid as a critical part of our product, I have to look for any emails from them pretty closely. It’s gotten impossible to do with all of these phishing attempts. I gotta hand it to them, though, the attempts are excellent.

I’ve been doing some of this through a term on my phone, but it honestly sucks. Other interfaces (telegram, web ui, email) are gonna be much better experiences on your phone.

Hah, I set up basically the same thing on Saturday during a long car ride. Couple of differences: I’m an opencode user and I used a different VPS provider (though I use vultr for other things). It was my first time actually sitting down and using tailscale, which was quite easy to get going. Did everything from my phone, didn’t even have my laptop with me.

They do a lot more nowadays to make it clear to the user what is happening, but it still feels unnatural to me.

The problem is that people just really do not comprehend what the "public" schema means in supabase. My guess is that that they think it means "default" or something along those lines. If you read the supabase documentation, you can clearly see that it says "your database's auto-generated Data API exposes the public schema by default", but to truly understand that, you need to understand what the data api is and how it relies on rls. For people first coming to supabase, they are probably either new devs, or they think of the db as a backend service that has application-layer authentication in front of it.

We are a happy user of DBOS. I’ve been building out a lightweight TUI for managing our DBOS application internally, since we have workflows with tens of thousands of steps. I know the team is working on improving conductor for this use-case, but our internal TUI handles it pretty well for now. Hoping to open source it, but the code is a wreck atm. Anyways, I say all this to say, DBOS has a good client that can communicate with the instance easily, so building out a UI that fits your specific needs should be fairly simple.

oh man, don't even bother with bazel... hermetic python builds are such a mess.

For sure, but most people aren't thinking intentionally about what they are dumping into their context either ;)

For repeating objects of the same structure, yaml will still require each key on each object, whereas this is a hybrid with csv, so it defines the keys once.

Code is also unique in its suitability for agentic grep retrieval, especially when combined with a language server. Code enforces structure, semantics, and consistency in a way that is much easier to navigate than the complexities of natural language.

I recently got a Daisy Pod (https://daisy.audio/product/Daisy-Pod/). It supports a few different runtimes, including a max/msp runtime and a Pure Data wrapper called Plug Data. It's a pretty neat little device, even though the documentation could use some work.

Doing quite a bit of that as well, but I’ve held off moving anything to gpt-5 yet. Guessing it’s a capacity issue right now.

This was my problem with Dagster, too. All the documentation and all the examples encourage you to split items into small discrete tasks. Then you realize that their cloud pricing is absolutely bonkers if you go over the paltry 30k credits… unless you sign up for a meaty annual enterprise contract. Got a $500 bill for something like 13k executions over the limit. That’s less than 45k executions in a month. Just for comparison, our main product’s sidekiq queue processes tens of millions of jobs every single day. Just a silly imbalance. I ended up having to combine a bunch of tasks to the point that I started asking myself why I was even bothering with using it at all.

I've had a Lamy Swift (palladium) for about 8 years now. It's a great pen with a unique clip mechanism. The Lamy rollerball refills are really nice, but they don't have a fine point version, so be ready for thick lines. I actually like the broad tip though. I just checked the Lamy website, and it looks like this pen has basically doubled in price since I bought it... yikes.

k2 on groq is really bad right now. I'm not sure what's causing the problem, but they've said that they are working on a few different issues.

best move is to switch to based-pyright

Not really the same use-case. DuckDB is more for read-heavy analytical uses.

It’s a pretty common issue. Search around and you will find pages and pages of people experiencing the same issues on macrumors, the apple discussion forum, etc. Glad it hasn’t been a problem for you, I envy you!

Oh yeah, I’ve turned every single possible thing off. It still randomly dark-wakes. And then, when it does, it will sometimes get into a doom loop of scanning the network and refusing to release its power assertions, meaning it won’t go back to sleep. Eventually it will hit a thermal emergency and then shut down.