I was in the first ~100 users to sign up for tangled after seeing that they were roadmapping stacked patch style pull request review and being generally impressed with how quickly they were building and shipping features all off the excitement from a budding community in the world of atproto. I am very excited to see where this goes and how they continue shipping.
There are a lot of opportunities here that could be levered to offer much better experiences and fundamental control and long term viability if they continue to execute this way.
It is simply hilarious to make grown adults visit a website called tender love making dot com (a sexual reference) to read a very specific and niche blog about technology.
I can't seem to authenticate because it had a localhost URL in the address when I was first authenticating. Now my work laptop has blocked the site completely to sign in or anything at all due to the activity of rapidly trying to bypass a site with no certificate because of that pass-on or something. Bummer, this computer has some more juice vs mine!
Might be an issue with "Sign In With Apple" if no one else has reported.
Hey, thanks for trying to get it working, this is unexpected! The reason we have the localhost in the callback is so that we can write the api key to ~/.omnara/credentials.json, so for future use, you would just use that api key. We temporarily spin up an endpoint on your computer during auth time so that we can send a request to it with the api key.
The workaround is to just login to omnara.com, click on your email at the bottom left, click on "API Keys", and then create an API Key named CLI Key.
Then copy that key, and make a file at ~/.omnara/credentials.json, and the contents should be:
{
"write_key": "<YOUR-API-KEY>"
}
Then you can run omnara, and it shouldn't ask you for authentication again. Let me know if this works out!
Their support has been very responsive with me, even if it takes a second to get a reply. I cannot believe that they would not reverse this in good faith if you proceed to not utilize the credits or whatever.
If they don't address this reasonably, it will be a very bad image for their brand! It is a simple problem, and an honest mistake that functionally should cost them nothing.
My customer service experience with them has sometimes required turnaround time, but never been helpful or unreasonable.
Their support bot responded with:
“Thanks! We’re transitioning your question to one of our human support agents for further assistance. You don’t need to keep this window open — we’ll email you once an agent replies. Typical wait time is around 2 business days.” Hoping this gets resolved soon.
What do you mean exactly? If you need a notification engine, reaching for a pubsub implementation is very easy with phoenix’s popularity and quite battle tested. I’ve implemented notifications at scale a few times in the ecosystem. What problems are you encountering that you don’t feel you have a tool in the shed to work with in this case?
I wrote this with some help from Claude Opus 4, researching how well it could construct things in a cited-only manner. I wrote a good bit of it myself and had Claude edit and do the deep digging. I found the deep dive to be very HN in nature so I figured I would share here and see if anyone else enjoyed it.
The modern developments are intriguing and they previously tied deeper into emotional and property value levels with precedent, but have taken a more modern approach based around environmentalism efforts that parallel the original lumber and forestry intents from things like free planting being legal but the cutting of another person's timber being treble damaged to disincentivize. It's all quite interesting. And with America's predisposition to single family homes with large yards (versus more dense urban cores with multifamily much more common in the rest of the world) there is a large well of potential disputes to flow to influence the law with real dollar risks of civil liability as things go onwards.
I have been using LLMs to write Elixir full time at work for months now. AMA, if anyone is interested.
I sometimes have different thoughts of approach than Zach, but this post really resonates with me. I've been in Elixir full time for over 10 years and would love to see an evolution in its adoption fueled by this.
I'm curious what your setup looks like. Do you use an IDE like Cursor or Windsurf, an editor like Zed or more directly an agent like Claude Code?
Have you used tidewave.ai? The demo from Jose looks fun, but I've yet to play with it.
What use-cases have LLMs shined in for you? I've really enjoyed using it to reduce the learning curve, eg to use Svelte and LiveSvelte on a small side-project.
(I've been using Rails for 20 years, Erlang & Elixir for ~10, but spend more time on the product side nowadays.)
> Do you use an IDE like Cursor or Windsurf, an editor like Zed or more directly an agent like Claude Code?
At work, we use both Windsurf and Cursor.
At home, I also use Claude Code.
> Have you used tidewave.ai?
My coworkers have and are impressed, I can't speak to it.
> What use-cases have LLMs shined in for you?
Well, I managed to literally have it do all my work for 3 weeks. I didn't write a single line of code. That was pretty cool. I didn't even cherry-pick work best for the LLM. It was my normal flow.
I also use Claude code as an interactive tutor. I will have it implement something, break it into logical commits, then in each one break a few pieces and write tests for them and have me learn by fixing the broken tests.
There are a lot of opportunities here that could be levered to offer much better experiences and fundamental control and long term viability if they continue to execute this way.