Because if there are valid ones, they may impact users... It's important to do due diligence (but this takes time to validate them).

A lot of things seem important in software, but we need to prioritize and compromise based on resources available. Based on what you've said so far, it seems to me that this project isn't giving you enough resources to invest in this particular problem.

That's the attitude I have with my software projects.

Well these are well written security vulnerabilities with reproduction steps. It's hard to tell if it's an AI discovering or a user using AI to find issues. But suddenly, I'm having an influx of issues where-as for the past 5 years, I received maybe 5. Just this month, I've been hit with 5 low effort vulnerabilities (all very small, unlikely to expose anything of value).

But it's very hard to maintain these in addition to the release work.

If it has steps to reproduce, you give it to your coding agent to "fix [bug] using TDD". If it can't make a test it wasn't reproducible.

I don't get what you're saying. You're frustrated that Open Source projects were used to build these AIs and that OS devs (or devs in general) are paying to use AI.

Then you say you had money that you used to donate(?) to OS and have cut that because of the frustration?

Open source just means sharing the source code for people to learn off or have the ability to customize on their own. I don't think there is any need to be frustrated about that (now if it was copyright/private of course).

> Open source just means sharing the source code for people to learn off or have the ability to customize on their own.

Yes people, not corporations. The point is there a licenses to be respected that weren't.

Model training pretty clearly falls under fair use.

We could fix that, but it requires a political will to change the law.

This has not been determined in courts and your willingness to speak so confidently about it speaks volumes.

The closest we've come to a court decision on this so far has been the Anthropic case, which did indeed find that training on unlicensed data falls under fair use: https://www.documentcloud.org/documents/25982181-authors-v-a...

> To summarize the analysis that now follows, the use of the books at issue to train Claude and its precursors was exceedingly transformative and was a fair use under Section 107 of the Copyright Act. And, the digitization of the books purchased in print form by Anthropic was also a fair use but not for the same reason as applies to the training copies. Instead, it was a fair use because all Anthropic did was replace the print copies it had purchased for its central library with more convenient space-saving and searchable digital copies for its central library — without adding new copies, creating new works, or redistributing existing copies.

If you look carefully model training is a very good relicensing exercise of your code

Just an FYI, your site text is not readable. It's a white fade background with white text.

Are you trying to read the sample chapters? What browser are you on? And did you try toggling dark/light mode?

What does "fast" even mean? I always see people talk about performance and wonder, what kind of applications are they building where they are seeing a massive gap in performance between frameworks?

I'm pretty sure in this late game, all the frameworks are more-or-less the same in terms of render performance. Angular has a lot of cool tricks (with these signals) that allow for only re-rending what changes.

Signals are a privative data structure in Angular, hence core. Signal-based forms are part of the Forms module. You aren't using forms, you don't get the overhead.

primitive

Really excited for this. I've been dying to use signal-forms and resources since they were experimental. Once I got on the signal train, I could never go back and having to use RxJS for forms became a major pain point.

Could you say more about signals? Is it are all analogous to, say, game engine signals paradigms (eg Godot) - components at any depth emit signals and any other component can subscribe? Or something totally different?

What is rough in the ecosystem? I haven't had any issues finding packages. Most packages have been keeping up with the signal trends as well.

Thanks for sharing, these are awesome ways to explore your data.

One thing that is so painful with Stripe is the Disputes because no matter how much evidence I show, even emails with the user claiming they did it because of X, Y, Z. ToS not upheld, etc, the customer always wins.

For me, I do a cheap subscription (4$/mon, first month 2$) and one dispute costs me like 20-30$. So that one person wipes a ton of profit from me. I always try to refund them (but you can't refund a customer with a dispute in effect).

Stripe is great to get going, but has a lot of painful points.

Does Stripe have any published stats on the ratios? Did any merchant ever won such a Stripe dispute?

I’d be curious too. Previous experience at a SaaS was that we never won disputes and it was too time consuming on top, so we just ate the dispute fees and refund.

I always thought things are easier with a physical product where you have a 3rd party like DHL that proves delivery was made. But at least in my tiny sample space, that’s not enough to win the dispute.