I did a similar set of evals myself utilising the baseline capabilities that Phoenix (elixir) ships with and then skillified them.
Regularly the skills were not being loaded and thus not utilised. The outputs themselves were fine. This suggested that at some stage through the improvements of the models that baseline AGENTS.md had become redundant.
So I built a thing that looks _very_ similar to this a while back: http://pipie.io/agent-tracker (it doesn't orchestrate)
And what's really fascinating isn't that multiple people have built the same thing is just how convergent all the ideas are. This makes a heap of sense given that they are all vibe-coded and using similar libraries. It's just very real when you see it right in front of your face like this.
I think there's a heap of really interesting stuff when you start linking into hooks which can also create tmux splits and really make the agent the center of an ide which seems like this is on the way to.
This is actually a really weird example of using the Context API. I have no idea why you would want to use it to manage a single switch of state.
The far better examples are providing a "theme" across multiple disparate components, and each of them being able to access the colours, or styles. (You could also just use CSS)
The other good example is being able to set the logged in user into a context and then being able to consume the context if you need access to the logged in user.
I think it makes more sense if you start thinking outside the Toggle component itself.
For passing down the state to the actual button(s), as Kent says, you could just use React.Children.map(). When you think about consuming that state from other components that need to know the toggle state, things get a bit messier, as many of those components might not be direct children of the actual <Toggle />. Maybe the toggling affects something in your navbar, for example.
In that case, the alternative to using context would be a global store, where the state provider sits above all components (think Redux, MobX).
If you're only consuming the toggle state from child / grandchild / sibling etc. components, the alternative typically involves passing down the toggle state from some parent / grandparent component, which means passing this.state.toggled down 2+ levels, instead of simply importing the context's consumer and subscribing where needed.
Context is now a reasonably simple and clean alternative to those options, but this example leaves envisioning the pain alleviated as an exercise for the reader.
Fair enough, I suppose. I still think the additional complexity of having to bring in the context for the save of passing 1 variable through the whole tree is overkill though.
I'm developping a web app with fairly complex forms (imagine form sections, form templating...). The immediate use case of Context that comes to mind would be to pass down the `readonly` prop from the top-level Form component all the way down to the Field component, without having to drill through `FormSection`, `Template`, `TemplateSection` or any other component between the Form and the Field.
In the case of a form, what’s wrong with composing a whole tree of custom “presentation” components in one render? Does this take up too much space?
You don’t always need to pass the prop through each individual component, instead you can render a section (or the whole form) together as one larger context-specific UI component, where everything in the render call has access to the same prop values.
How does this work in the case where the majority of your customers currently pay by direct-debit/cheque? This is certainly the case for us, but we have long been looking for a affordable solution to have them managed with some sort of managed subscription engine
Beer isn't in 'publicly' accessible vending machines any more as far as I can tell. It is however often available in the foyer of a place that normally sells beer (that might be currently shut.)
Ski resorts are my main source of vending-machine beer. Nothing better than cracking a can of Sapporo Classic in your first gondola ride of the day.
only place i saw them in my entire trip to japan was in a hotel. And it was a little underwhelming because there was cheaper and a higher selection of alcohol at the Lawson 5 feet away from the hotel
Didn't Github comment on their blog recently [1] on how they were improving and reducing failure detection times? Doesn't speak well for them if it's first on Hacker News and not their own Status page.
tl;dr; > One of the biggest customer-facing effects of this delay was that status.github.com wasn't set to status red until 00:32am UTC, eight minutes after the site became inaccessible. We consider this to be an unacceptably long delay, and will ensure faster communication to our users in the future.
EDIT: "We're investigating some issues with our databases".
Regularly the skills were not being loaded and thus not utilised. The outputs themselves were fine. This suggested that at some stage through the improvements of the models that baseline AGENTS.md had become redundant.
reply