Xcode should ship with a built-in MCP that handles auth when an LLM connects to a project. Notion should have `mcp.notion.so/mcp` available natively, instead of forcing me to download `notion-cli` and manage auth state manually.

Implicit state is the Enemy

The software design model can be compared to an architected plan for a house. High-level plans represent the totality of the house (e.g., a three-dimensional rendering of the house). Lower-level plans provide guidance for constructing each detail (e.g., the plumbing lay). Similarly, the software design model provides a variety of views of the proposed software solution.

including both high-level software architecture and low-level component and algorithm design.

Software architecture choices include specific structural options from possibilities in the design of the software.

I was already aware that images cannot be inserted in the DOM like you would any normal image. If you write <img src="https://my-pod.com/recipes/ramen.jpg">, this will probably fail to render the image. That happens because the image will be private, and the POD can't return its contents without proper authentication.

a complex problem should not ~be regarded immediately in terms of computer instruc- tions, bits, and "logical words," but rather in terms and entities natural to the problem itself, abstracted in some suitable sense

There are two pairs of methods for sending/receiving messages: Object#send and ::receive for when the sender knows the receiver (push); Ractor.yield and Ractor#take for when the receiver knows the sender (pull);

As John Palmer points out in his brilliant posts on Spatial Interfaces and Spatial Software, “Humans are spatial creatures [who] experience most of life in relation to space”.

My UIs are data/store driven. The UI is just a way to visualize the data. Your data could flow through all of of the extensions and the extensions can make decisions (e.g. setting visible to false). Like middlewares in a Connect/Express/Polka app. And the UI doesn't even know about all this, it just updates with the current state and makes sure it's consistent.

At some point software design becomes less about what and more about when.

Write your code so that you can add new ways of interacting with the world later without having to modify anything you've already written.

The WebSocket Protocol is designed on the principle that there should be minimal framing (the only framing that exists is to make the protocol frame-based instead of stream-based and to support a distinction between Unicode text and binary frames). It is expected that metadata would be layered on top of WebSocket by the application Fette & Melnikov Standards Track [Page 9] RFC 6455 The WebSocket Protocol December 2011 layer, in the same way that metadata is layered on top of TCP by the application layer (e.g., HTTP). Conceptually, WebSocket is really just a layer on top of TCP that does the following: o adds a web origin-based security model for browsers o adds an addressing and protocol naming mechanism to support multiple services on one port and multiple host names on one IP address o layers a framing mechanism on top of TCP to get back to the IP packet mechanism that TCP is built on, but without length limits o includes an additional closing handshake in-band that is designed to work in the presence of proxies and other intermediaries Other than that, WebSocket adds nothing. Basically it is intended to be as close to just exposing raw TCP to script as possible given the constraints of the Web. It's also designed in such a way that its servers can share a port with HTTP servers, by having its handshake be a valid HTTP Upgrade request. One could conceptually use other protocols to establish client-server messaging, but the intent of WebSockets is to provide a relatively simple protocol that can coexist with HTTP and deployed HTTP infrastructure (such as proxies) and that is as close to TCP as is safe for use with such infrastructure given security considerations, with targeted additions to simplify usage and keep simple things simple (such as the addition of message semantics).

something called federated wiki which was by ward cunningham if anyone knows the details behind that or how we got these sliding panes in the first place i'm always interested

behaviour should come first and drive our storage requirements.

Worse still is the issue of “service” layers requiring you to basically build your own ORM. To really do a backend-agnostic service layer on top of the Django ORM, you need to replace or abstract away some of its most fundamental and convenient abstractions. For example, most of the commonly-used ORM query methods return either instances of your model classes, or instances of Django’s QuerySet class (which is a kind of chained-API results wrapper around a query). In order to avoid tightly coupling to the structure and API of those Django-specific objects, your service layer needs to translate them into something else — likely generic iterables to replace QuerySet, and some type of “business object” instance to replace model-class instances. Which is a non-trivial amount of work even in patterns like Data Mapper that are designed for this, and even more difficult to do in an Active Record ORM that isn’t.

A second problem is that when you decide to go the “service” route, you are changing the nature of your business. This is related to an argument I bring up occasionally when people tell me they don’t use “frameworks” and never will: what they actually mean, whether they realize it or not, is “we built and now have to maintain and train our developers on our own ad-hoc private framework, on top of whatever our normal business is”. And adopting the service approach essentially means that, whatever your business was previously, now your business is that plus developing and maintaining something close to your own private ORM.

This isn't about writing boilerplate setter properties for each field in the model, but rather about writing methods that encapsulate the point of interaction with the database layer. View code can still inspect any field on the model and perform logic based on that, but it should not modify that data directly. We're ensuring that there is a layer at which we can enforce application-level integrity constraints that exist on top of the integrity constraints that the database provides for us.

For me the diagrams make it easier to talk about what the tests do without getting bogged down by how they do it.

Uber::Option implements the pattern of taking an option, such as a proc, instance method name, or static value, and evaluate it at runtime without knowing the option's implementation.

Instead of one big code pile, activities will gently enforce a clean, standardized way for organizing code.

A leaked snippet of the endpoint architectural design draft document, highly confidential.

Each of the programming language generations aims to provide a higher level of abstraction of the internal computer hardware details, making the language more programmer-friendly, powerful, and versatile.

provide interfaces so you don’t have to think about them

Trailblazer is an architectural pattern that comes with Ruby libraries to implement that pattern.

Trailblazer offers you a new, more intuitive file layout in applications.

Instead of grouping by technology, classes and views are structured by concept, and then by technology. A concept can relate to a model, or can be a completely abstract concern such as invoicing.

While Trailblazer offers you abstraction layers for all aspects of Ruby On Rails, it does not missionize you. Wherever you want, you may fall back to the "Rails Way" with fat models, monolithic controllers, global helpers, etc. This is not a bad thing, but allows you to step-wise introduce Trailblazer's encapsulation in your app without having to rewrite it.

By the same token, marketing or political incentives often push design idioms

Programming to interfaces is at the core of flexible structure.

The more important point comes from a program design perspective. Here, "programming to an interface" means focusing your design on what the code is doing, not how it does it. This is a vital distinction that pushes your design towards correctness and flexibility.

Top level architecture

If you want to implement a form with a superb User Experience, you have to take care of many variables:

However, after solving this problem a few times, a pattern appeared.

Keep the local state isolated.Think about which state is local to a particular UI representation — and don’t hoist that state higher than necessary.

I couldn't land on how I wanted to box primitives. Should I use a getter/setter, or function form like Knockout, or explicit get/set like MobX? These were all ugly.

This library exports a single React Hook, useMethods, which has all the power of useReducer but none of the ceremony that comes with actions and dispatchers.

Domain-driven design separates the model layer “M” of MVC into an application, domain and infrastructure layer. The infrastructure layer is used to retrieve and store data. The domain layer is where the business knowledge or expertise is. The application layer is responsible for coordinating the infrastructure and domain layers to make a useful application. Typically, it would use the infrastructure to obtain the data, consult the domain to see what should be done, and then use the infrastructure again to achieve the results.

The pattern below has become exceptionally useful for me (pun intended). It's clean, can be easily modularized, and the errors are expressive. Within my class I define new errors that inherit from StandardError, and I raise them with messages (for example, the object associated with the error).

There are thousands of to-do list apps out there, in part because no system works perfectly for everyone. I’m not going to say todo.txt is the exception, and that it will work for everyone, because that would be crazy. But todo.txt is the most flexible tool I’ve come across. In part, this is because of the sheer number of clients available, but also because the simplicity lends itself to improvisation.

Middleware works a lot like a decorator. It doesn't alter the original API of the service, but it can augment it with new features and concerns. This has the inherent advantage of allowing all thidparty code to have an influence over the behaviour, state, and UI of a component.

AdNauseam is a work in progress, with new features continually being added, tuned, and, sometimes, deprecated. If a setting no longer appears in the settings page, we have likely found a better means of implementing the design goal.

for example, comments and identifiers

For large-scale software systems, Van Roy believes we need to embrace a self-sufficient style of system design in which systems become self-configuring, healing, adapting, etc.. The system has components as first class entities (specified by closures), that can be manipulated through higher-order programming. Components communicate through message-passing. Named state and transactions support system configuration and maintenance. On top of this, the system itself should be designed as a set of interlocking feedback loops.

how does misrepresentative information make it to the top of the search result pile—and what is missing in the current culture of software design and programming that got us here?

I don't think passing the entire http client is very idiomatic, but what is quite common is to pass the entire "environment" (aka runtime configuration) that your app needs through every function. In your case if the only variant is the URL then you could just pass the URL as the first parameter to get-data. This might seem cumbersome to someone used to OO programming but in functional programming it's quite standard. You might notice when looking at example code, tutorials, open source libraries etc. that almost all code that reads or writes to databases expects the DB connection information (or an entire db object) as a parameter to every single function. Another thing you often see is an entire "env" map being passed around which has all config in it such as endpoint URLs.

Something that I've found helps greatly with testing is that when you have code with lots of nested function calls you should try to refactor it into a flat, top level pipeline rather than a calling each function from inside its parent function. Luckily in clojure this is really easy to do with macros like -> and friends, and once you start embracing this approach you can enter a whole new world of transducers. What I mean by a top level pipeline is that for example instead of writing code like this: (defn remap-keys [data] ...some logic...) (defn process-data [data] (remap-keys (flatten data))) (defn get-data [] (let [data (http/get *url*)] (process-data data))) you should make each step its own pure function which receives and returns data, and join them all together at the top with a threading macro: (defn fetch-data [url] (http/get url)) (defn process-data [data] (flatten data)) (defn remap-keys [data] ...some logic...) (defn get-data [] (-> *url* fetch-data process-data remap-keys)) You code hasn't really changed, but now each function can be tested completely independently of the others, because each one is a pure data transform with no internal calls to any of your other functions. You can use the repl to run each step one at a time and in doing so also capture some mock data to use for your tests! Additionally you could make an e2e tests pipeline which runs the same code as get-data but just starts with a different URL, however I would not advise doing this in most cases, and prefer to pass it as a parameter (or at least dynamic var) when feasible.

One thing Component taught me was to think of the entire system like an Object. Specifically, there is state that needs to be managed. So I suggest you think about -main as initializing your system state. Your system needs an http client, so initialize it before you do anything else

The Moodle project