HN Debrief

Decker, a platform that builds on the legacy of Hypercard and classic macOS

  • Developer Tools
  • Programming
  • Open Source
  • Gaming
  • Productivity

Decker is a modern platform that borrows heavily from HyperCard, Apple’s old card-and-stack system for building interactive software without a full conventional app toolchain. It leans into 1-bit graphics and classic Mac interface patterns, has its own scripting language called Lil, supports custom widgets and data structures, and can export work as things like single-file HTML. That puts it in the same family as older end-user tools like HyperCard, FileMaker, Access, LiveCode, and Delphi-style rapid app builders.

If your team keeps reaching for spreadsheets, ad hoc SaaS, or fragile internal web forms, there is still unmet demand for tools that let non-specialists build polished local apps fast. Watch this space less as a retro curiosity and more as evidence that simple, self-contained software remains a real product niche.

Discussion mood

Warm and enthusiastic, with a strong nostalgia streak, but grounded in a real sense that modern software lost something important. People liked Decker as both a living HyperCard descendant and a critique of today’s bloated web and SaaS workflow, though some doubted the retro aesthetic can carry it beyond hobby, game, and personal-use niches.

Key insights

  1. 01

    Why HyperCard felt like real software

    It gave non-programmers a path to make interfaces that immediately matched the operating system’s visual language. That closed a gap modern tools often reopen. You can script business logic later, but first you need something that already looks legitimate on screen. HyperCard and later Visual Basic nailed that in a way terminals, modern GUI toolkits, and bare web canvases usually do not.

    If you are building internal tools for non-developers, polish cannot be an afterthought. A product that ships useful defaults for layout, controls, and visual fit will beat a more powerful system that starts from a blank canvas.

      Attribution:
    • analog31 #1
    • zeroq #1
  2. 02

    The web took years to catch up

    Modern frontend developers take event handlers, reusable components, and interactive documents for granted, but HyperCard had those ideas in a much simpler package by the mid-1990s. The web won because it was networked, not because it was a better local authoring environment. That reframes Decker as a reminder that software history already solved part of the UX problem and then forgot it.

    Do not assume today’s stack is the natural best form for lightweight app creation. For local-first tools or controlled distribution, you can often strip away layers and still keep the interaction model users need.

      Attribution:
    • scelerat #1
  3. 03

    Why CRUD moved backwards for end users

    A small database app sounds trivial until you try to rebuild what Access or FileMaker gave office workers out of the box. Table creation, navigable forms, reports, mail merge, and acceptable-looking screens used to come almost for free on one machine. Replacing that with SQLite plus a web UI is straightforward for engineers and wildly more complex for everyone else.

    When someone in your company is still running a workflow from Excel or an ancient Access file, treat it as a product requirement, not user stubbornness. The replacement has to beat the old stack on setup, presentation, and reporting, not just on architecture.

      Attribution:
    • whartung #1
    • swiftcoder #1
    • mbreese #1
  4. 04

    Decker already has a real output niche

    The project is not just a tribute page for HyperCard fans. People are using it for personal apps, zines, visual novels, browser-shareable artifacts, and at least one shipped Steam game. The self-contained HTML export is especially important because it gives creators a dead-simple distribution path that keeps the tool’s low-friction feel intact.

    If you are evaluating creator tools, look for narrow formats with easy export and sharing before chasing general-purpose platforms. A strong publish path can create a durable niche even when the tool is opinionated or visually unusual.

      Attribution:
    • RodgerTheGreat #1
    • huseyinl #1
  5. 05

    Decker has a data model beyond widgets

    The interesting technical point is that Decker does not force all state into UI controls. It exposes "reified data" through contraptions, can treat cards as records or object-like units addressed by events, and can encode structured data in fields. That makes it more than a toy presentation builder. It is a small application environment with its own way of making data tangible.

    If you dismiss tools like this as just drag-and-drop UI builders, you miss where they become extensible. Check whether the system has composable data objects and conventions for behavior, not just forms and buttons.

      Attribution:
    • RodgerTheGreat #1

Against the grain

  1. 01

    Retro charm can block present-day usefulness

    The sharpest pushback was that a HyperCard revival only matters if it can produce software people will actually use now, not just software that reminds builders of 1990. From that angle, Decker risks landing as a beautifully limited sandbox. If a project cannot cross from personal delight to practical deployment, the nostalgia becomes the whole product.

    If you are building in a retro-inspired niche, decide early whether you want authenticity or broad adoption. The design choices that make enthusiasts love it may also keep mainstream users from taking it seriously.

      Attribution:
    • donohoe #1
  2. 02

    The 1-bit look narrows the audience

    Several people argued that black-and-white aesthetics are not just a style choice but a ceiling. HyperCard itself was hurt when color failed to become first-class, and modern users expect theme flexibility for presentations, business tools, and client-facing work. Myst only gets cited as a counterexample because it reportedly relied on a modified HyperCard setup rather than vanilla limitations.

    Visual constraint can be a brand, but it also defines your market. If you want wider use, themeability and richer presentation options need to be core features, not add-ons.

      Attribution:
    • InsideOutSanta #1
    • andrelaszlo #1
    • pasquinelli #1
    • brightstar18 #1
  3. 03

    LLMs may become the new HyperCard

    A few commenters suggested the old end-user programming niche has already been absorbed by SaaS and is now being swallowed again by LLM coding agents. The argument is not that these systems are better designed. It is that they are becoming the default way non-specialists get one-off software made. That shifts the competition for tools like Decker from Access or LiveCode to conversational software generation.

    If you are betting on end-user creation tools, assume users will compare them to AI copilots, not just legacy RAD tools. Strong direct manipulation and predictable local behavior are the advantages you need to make explicit.

      Attribution:
    • Exoristos #1
    • NegativeLatency #1

In plain english

1-bit graphics ↩
Graphics that use only two colors, typically black and white, because each pixel stores just one bit of information.
Access ↩
Microsoft Access, a desktop database tool for building forms, queries, reports, and small applications.
Delphi ↩
A rapid application development tool and programming language environment known for building desktop applications with drag-and-drop UI design.
FileMaker ↩
A long-running database application platform used to build custom forms, reports, and small business apps.
HyperCard ↩
A discontinued Apple tool from the late 1980s and 1990s that let users build interactive card-based applications with simple scripting.
itch.io ↩
An online marketplace and hosting platform focused on indie games and experimental interactive projects.
Lil ↩
The scripting language built into Decker for adding logic and interactivity.
LiveCode ↩
A programming environment inspired by HyperCard that uses a high-level scripting language and visual interface builder.
LLM ↩
Large language model, a type of AI model trained to predict and generate text and code.
SaaS ↩
Software as a Service, software delivered over the internet by a provider rather than installed and run locally.
SQLite ↩
A lightweight embedded database stored in a single file, often used inside applications without a separate database server.
Steam ↩
A digital distribution platform for PC games and software.

Reference links

Decker docs and examples

Decker ecosystem and distribution

Related tools and historical context

  • tldraw offline
    Offered as a more agentic modern tool with some overlap in spirit with HyperCard-style creation.
  • MacBasic folklore page
    Referenced to explain how HyperCard’s scripting made it a practical substitute for Apple’s unreleased MacBasic.
  • LiveCode on Wikipedia
    Named as a HyperCard descendant that some still use for personal applications.
  • Myst on Wikipedia
    Used in the discussion about HyperCard’s role in game history and how Myst handled its visual limits.