Nick Pezza

Im so stoked to see the 4k photos artemis ii takes of all the decepticons as it goes around the moon

@bradgessler Rails has a built in postgres actioncable adapter that uses notify/listen but it has a limit of 8kb. So you need more than that SolidCable uses polling to get around it.

Does anybody know why the solid stack doesn’t use Postgres NOTIFY/LISTEN? Is there problems with it in Postgres or is it for compatibility reasons, like MySQL doesn’t have it?

@Shpigford @maybe Any plans to open source synth or detangle since they are shutting down?

Maybe is making a pivot to B2B tl;dr @maybe is pivoting towards B2B financial forecasting + scenario planning The Problem We’re about 6,000 paying customers short of breaking even, with only around 200 paying customers currently, most of whom joined during the beta phases when the cost was significantly lower. We likely need more than 6,000 when you consider that many resources will need to scale up to meet the needs of that many more customers. The reality is that building a B2C personal finance platform is not only technically very challenging, but incredibly slow to grow, and we can’t tackle those challenges while creating additional features people might be willing to pay for within the next 10-12 months (our current runway). I no longer believe a B2C personal finance app is our best bet for survival going forward. The market’s needs are too fragmented, and the feature set is too far from becoming valuable for more affluent customers. I believe it either has to be completely bootstrapped for years or have $10’s of millions in funding to sustain and pour into growth. We’ve got roughly $400,000 left in the bank. So what do we do with that? The Solution We sunset the current version of Maybe. Again. No more giant B2C personal finance app. The economics just don’t make sense for us. We’re pivoting to B2B financial forecasting & scenario planning tools. Specifically, we're focusing on natural-language, generative-UI + UX for creating reports and answering questions. All business financial tools lean heavily on "power", which means lots of formulas, lots of complex interfaces, and spreadsheet inputs...this is why they go up-market, because onboarding requires teaching how to use the platform. This type of tooling typically requires someone with a background in finance to even begin to use them. But what if the UI were a simple text input? “Forecast runway if we added one new employee in 3 months at a salary of $150,000”. Not a chat input, but natural-language, generative UIs. It's something we’ve actually been working on over the past few months anyways and we’re now going to apply that to a business context. Anyone on a team will be able to use it because, to obtain the data they want, they simply write it out. Thank you, AI. The focus is on iterating quickly for a business customer base that will happily spend money to make money. Our thesis here is that the core must be helpful and generate revenue from day one and we’ll be focused on what I call hyperbuilding with that day-one-revenue coming by the end of August. FAQ • Why not just keep slow-and-steady working on Maybe v2? — We simply can’t afford to. While there are certainly other personal finance tools out there that are bootstrapped, they’ve spent years slowly building. We don’t have that luxury. • What about going AI + mobile first? — I still think that’s the best move for any new B2C personal finance app and it was something we were heavily pursuing, but we don’t have the runway to pull it off. • So this is a move to B2B? — Exactly. We’re now a B2B software company building a new B2B product. • What happens to Maybe v2, Synth and Detangle? — All shut down within 6-8 weeks to save on costs and extend runway. • What happens to Maybe’s OSS repo? — It will continue on for those that want to maintain it, but we won’t be putting resources towards it as a company in the short term. • What about paying subscribers? — We'll stop monthly billing and refund everyone's most recent payment. • How are we qualified to build this new B2B product? — Most of my 20-year professional career has been in B2B. Baremetrics was squarely a business financial app. Travis also works with many businesses on a CFO level, so he has unique perspectives and connections there. Our lead engineer spent years in FP&A at a large financial institution. In addition, all of us have been neck-deep in building financial tooling for many years now. • What’s the timeline here? — First B2B dollar within a month. We’re moving at hyperspeed. • What’s the first/core feature? — Runway forecasting & scenario planning. I know for many investors and customers, this will be both a surprise and a disappointment, as what interested you in the first place was a fresh take on personal finance tools. I'm sorry we won't be able to deliver on that. The road to success has a staggering number of forks in it. However, I'm personally extremely excited to get back to the B2B space, given my history there, and the team as a whole has a renewed sense of purpose and focus with this move with an optimism about success in a way that we've just not had in a long time. An annoying number of updates coming soon. 🙂

@juliknl Check out github.com/basecamp/kamal…

This is absolutely the missing bit for me with Kamal (and with my hodgepodge bash script that I use instead of Kamal... because I don't want the damn registries)

With Rails 8’s #SolidCable, real-time WebSocket updates are simpler than ever. Use your existing database, drop Redis as a dependency, streamline your stack, cut costs, and maintain high performance. @typecraft_dev shows you how: youtu.be/bOlrHbhLzZE?si…

@igor_alexandrov You should try using rubylang/ruby:3.4.1-jammy as the base image. It's basically a drop in replacement but is about 75mb smaller then the docker.io/library/ruby version.

#Rails has an almost perfect Dockerfile out of the box. I decided that it will be good to explain every line, every optimization applied to it for everybody. Meet the first article in a series – image size optimizations. igor.works/blog/rails-doc… #rails #docker #ruby

Welp the day has come that the company I worked for got sunset so I am on the hunt for a new job. If anyone happens to know of any great Rails shops(that use turbo 🤞) feel free to reach out! DMs are open. RTs for reach appreciated.

@sama Being able to set the default search engine in safari to chatgpt

what would you like openai to build/fix in 2025?

@geetfun github.com/npezza93/activ… 😉

If you read the comments here, there’s a business idea with product market fit. A simple error tracking app that can be self hosted. Think plausible for analytics. But for error tracking and make it dead dead simpler than things newrelic, or even sentry self hosted.

They somehow outdid the dual falcon landing. Absolutely bananas

Released a new version of ActiveInsights. Gave it a fresh coat of paint to be similar to the new error pages and rubyonrails.org. Before vs After

@rosapolis Thanks Rosa!

Amazing work by Nick here! 🙇🏻‍♀️🔝

@tenderlove ♥️💚💙🩶🧡

Solid keynote #RailsWorld

@fractaledmind Awwww yeaaaaaa

DHH shouts out @pezza_n for his work on Solid Cable in his #RailsWorld keynote! Look at those perf numbers.

My Rails World keynote bingo card

@fractaledmind @roger I’ll add a benchmark comparison to readme to show the difference!

@roger There is a pre-existing PH adapter for Action Cable using its unique NOTIFY feature. Using Solid Cable with PG will _likely_ (I haven’t benchmarked it) be slower than the existing PG Action Cable Adapter. I would recommend just using that

Just saw Solid Cable was merged but the support for Postgres seems to be with warnings. Is Postgres not the most commonly used db for Rails? $hit am I an outlier?

@boramacz Hell yeah 🤘

@roger However I will rewrite the note to make it more clear

@roger Author here: it’s not a warning just a note to make people aware that pg has its own adapter that uses notify which is more performant since it doesn’t have to use polling. While also pointing out that the pg adapter does have a limitation which is why sc is tested against pg