For the curious, the note taking app Trilium has a "standalone" mode¹, built on top of OPFS² and sqlite OPFS persistence, such that your notes (~web pages here) are stored in sqlite locally, offline-available, optionally syncing with a remote, while being a pure local web app (after the app assets have been downloaded once). If you want to take your whole personal notes/wiki/pkms for offline access on a device you don't own/have install rights, that's the only way I know. I suspect that OP could have gone this route instead had he known about it?
It's not using the File System Access API, on Firefox it uses <input type="file"> for file-picking, and the save just downloads it. It's not synchronizing anything with your filesystem.
I believe you are able to connect to a Google Drive account with a purely client side site via google oauth and then using the Google Drive rest api. There was an article in particular that I liked that walked through it. Let me see if I can find it
Yes but Google OAuth is a nightmare if you want to "publish" your client so that anyone can use it. If only you use it, you can add yourself as a test user. Just know your access token expires super fast and you will have to frequently re-auth. Its crazy annoying.
I don't know, this seems like an idea that maybe works in a specific context being generalized past the point of the original idea making sense.
If the user needs to download a specific application to run this these web apps, then why not just send them that initial application in the first place? Why jump through the hoops of using Capsule when the same hoops can be jumped through to get to the same endpoint?
If this was a near-universally adopted application, then it'd make sense. But it's not, and the closest thing we have to that are browsers... which already do what you're describing?
Bundling data with the application makes sense, but is also only appropriate in pretty narrow circumstances. If I'm willing to ship my data with the web app, then I'll just embed the data in the HTML file. If the expectation is that the user will modify this data, then I don't think I'd want to ship it like this.
Not a bad thing, but it's just another dependency and tied to an OS-specific base software install.
They could work well pre-installed on the Mecha Comet with a bunch of app files ready to launch.
Can the added layers guarantee a new level of security or privacy like zero telemetry or keyjacking? Protection from bad actors is a good thing while making things faster and easier overall. Browser extensions are a security risk, right?
Java applets were a good idea in the sense that so long as you had Java installed, the applet was just sitting there on a webpage. Finding and managing the apps is the problem, and using browser bookmarks is a decent solution to it. The web page should demo the app, and there can be an 'install local' button next to it. But at some point all these local versions need to be updated and then not be messed with like browser extensions can be. FanFare would be a decent name.
Seems pretty useful in an enterprise context to me. I see it as an alternative to 10 million vibe coded apps to dockerize and manage. You can put the runtime on everyone’s machine and provide a web version for really collaborative stuff.
Bonus points for an intelligent way to manage these stored in OneDrive and other cloud folders. Easy to get corruption on SQLite files stored there.
It could be useful for us with home labs for the same reasons. Kind of annoying to deploy docker containers for all the services I build that amount to personalized/skinned trello boards.
Not trying to plug myself, but I’ve been working on a similar project for this exact reason. It doesn’t use special file types. Just HTML and localStorage proxies to sync devices.
It’s self-hostable too. I’ve got a bunch of super niche todo lists and personal apps running on my home server with it.
There are a few things in Capsule to avoid corruptions of the file during syncing. The DELETE journal mode is used and writes are batched into transactions so that updates are always complete changes. And if there is a copy created because of a conflict merging two capsule files is supported.
My take: If your app needs to update and preserve state (using SQLite or any other DB), it is probably not something you want to pass around as a bundled file.
Or at least, it's extremely limiting, compared with hosting it somewhere on the web, which is not that hard to do these days.
Think of the workflow: Any time the state changes, you need to email a new Capsule file to whoever else is using the app. And one would think the state would change at least occasionally, because otherwise there's little reason to use a DB.
Alternatively, you can just host it on the web, the DB state dynamically updates, and it's automatically available to anyone with app access. Isn't that a lot simpler?
Or the copies can go their own way. I could send someone an itinerary and they could change the dates or drop half the stops and add their own notes without needing to send an updated copy. Getting them to install another app just to open it might be a problem, especially if they only want to use it once.
theres still lots of programs where local is the main user. but as time goes on the user may want to portably move it just by emailing the file itself.
i made like a very light clone of notepad++ using it for my macbook.
as a side note: i think they missed an opportunity to call it .pdfv rather than .capsule
You’re glossing over the recurring cost and annoyance of dealing with the hosting and having to secure the data. This way you just share and it’s local.
Sure, if you have to share data between people, your points are valid. But this seems like a way to build apps geared for individual users, like a notes or recipes app.
For most users having to mail a file to someone is way more work than having to just share a link. The cost of cloud storage is almost nothing (since email is mentioned I assume 50MB max per project). A notes app that don't sync between all devices is just a text editor to most.
I do like apps that have sensible export formats though. Having control over my data is important.
Beyond "not working", concurrent editing is likely to silently lose data, or corrupt data, or possibly just fill up your drive with lots of "conflicted copies" that you have to manually reconcile.
When I read about this project, it doesn’t seem like concurrency is much of a concern. I think it’s _supposed_ to be treated like sharing a document with someone else.
If a capsule is a recipe, like in one of the examples, that’s pretty cool. If a capsule is a recipe _manager_, then I think it would get messy.
I miss from the home page a link to github and/or a description of it's license and/or some code of an actual capsule app (what backend language is used? -- I have no clue and cannot find out) and/or does this bundle Node?
While I wish your project success and like the simplicity of Sqlite; I prefer to start with Postgres. PG has many of the benefits (x-platform, free, embeddable, etc) that sqlite has, but also provides an industrial strength db in case you need to scale at some point, or want to use more complicated constructs.
You can get a cheap approximation of this with chromium browsers + Filesystem API and a PWA manifest for that native-ish feel. AI has reignited my interest in building based 100% on browser native features.
If you want to distribute apps to many users this would probably be a better solution. The main benefit of Capsule is the personal document like app. You create a capsule, move it to an USB stick, save it in your cloud storage and it's still the same document. Nothing has changed, not the UI, not the data. The document will stay the same.
If it’s web native, anyone can just use it. If it’s a capsule, they first need to download the capsule app to associate the file extension with the OS and have it launch, right?
That feels like a big barrier, almost like a Java runtime.
In defense of capsule, the Filesystem API is janky and has kind of shit UX. Less sophisticated users would probably prefer the UX of a purpose built tool.
I prefer worse UX for the sake of standards and zero install, but many people would prefer the opposite. I don't think Capsule is intended for the HN crowd.
I love this idea! One of the fallouts from widespread AI adoption that I've seen: They're very good at creating visual artifacts, but if you ever have to provide data in those artifacts, you don't really have a great way to share it without hard coding it.
> One downside with this approach is that multiple people working on it will create different copies. To make it possible to merge different copies of the same file, each data entry has a unique UUID and timestamp.
This was the first thing that popped up in my mind: Changes stemming from two sources and reconciling them. I see that you have a statement about how to handle data entry from different sources, but I don't see exactly how those are reconciled? For instance, if two users have a copy of the .capsule and make changes, then want to share their changes with the other, you have two individual .capsules with different data.
Each document record is saved with UUID and timestamps. When deleting a record the data is removed but the record keeps a tombstone so only the ID remains. For conflicts currently the newest one is used, but there will be a merge like UI in the future to confirm which one to use.
Whats the advantage over plain single html files/apps using colocated css and use the browser/web view local storage? Those could be saved as PWAs or not?
It kinda reminds me of MS Access, which ended up being a pain in the butt for IT departments, but also covered a need within some department with limited resources.
Really curious to see where this goes. Do you plan to open-source the client, or simply make the file specification open when stabilized?
I'm currently 50/50 on open source the full app vs just the spec. Since I'm currently having a strong idea of what features and functionality is still missing I feel that open-sourcing right now would distract to much in this phase. But maybe when the main features are done it would make sense to have more people looking at and improving it.
An idea for a killer feature:
Make it so that auhors can expose their capsules to the internet, and the people whoe wants to use them(lets call them users) can type the name(lets call it address) of the caplsue into the app, and your program will pull that capsule in the the users device..
This is great. Curious: this requires users to have a host app installed on their machine that can read .capsule file right? Won't that be a point of friction for distribution?
Isn't html/css a better distribution mechanism as most computers already have the tech to run them?
You'd think so, but web browsers are downright lobotomizing the ability for local HTML files to run any meaningful javascript code. Want to run a JS file from the same directory? CORS ERROR!!!!
Hence 200MB applications that are complete copies of the Chromium browser to run under 1MB of actual web code.
couldn't you just embed the js directly into the html file though? that would sidestep the CORS error. and since most of the people using a program like this are likely to be using AI to generate their ( likely throwaway ) applets anyway, readability and maintainability for the source doesn't matter as much as it would with human authors so having a separate js file doesn't bring that many advantages anyway.
Sharing HTML files with embedded Javascript is problematic: iOS users can't open the files directly, and have to save the file and then 'open with ... '.
Isn't that the price of a relatively well sandboxed experience by browsers?
I've experienced the same limits building small tools for myself and friends that I wanted to contain in a single HTML file without external dependencies... but I understand why the same project without these restrictions could easily become malware, and ones that could be easily propagated.
I wish the browser would either allow local-only or internet-only, then only make up its mind after the first attempted request. Fetch an image from the internet? No script access for local files. Fetch a JS from elsewhere on the hard drive? No internet access. That could provide the sandbox while still being flexible.
The problem is that you can't save the user data in way that you can click a single file, similar to a Word or Excel file. I you think a about it, a Excel spreadsheet is also just UI + data in a single file.
This is a frustration of mine as well, but I vaguely remember someone showing something on HN where the file could essentially save itself, I think they were using some file APIs for that... sadly I can't find the details anymore. But I keep running into the "HTML file opened locally can't update itself" issue repeatedly now that I build tools for myself with LLM agents.
The File System Access API can do this, and it works in local ("file:///") HTML files, but it's only currently supported in desktop Chromium browsers: https://caniuse.com/native-filesystem-api
However, every browser will let you download a new version of the HTML file and save it over the old one - yes, even from a local HTML file:
I really like the idea. It's so easy to create small tools with AI now, but hard to install them as local apps or share them.
But I'm missing a few relevant features:
1. syncing data and apps (capsules) between my devices. not necessarily over a single SaaS, but maybe p2p or some existing service
2. I think data and apps should be separated, I might want to share an app with someone else without sharing the data
3. Apps should support updates, install the same app again and just replace the code for the existing app, but keep the data. Also share updates either as a file or publish apps to a repo (just a folder of files on a public http server or github repo)
Both keeping data in sync between devices and providing offline capabilities is hard, I know.
This would be great if it evolved in a universal format/spec/feature supported by all major browsers.
Only then you'd have to deal with cross web engine compatibilities. And if capsules need to do something non-trivial then you need to consider all the security aspects of having system, file or network level access. Just like any executable.
Otherwise with custom app to run capsules it feels like something that can be vibe coded with electron/nw.js rather easily.
But there is something compelling in the idea - there is need for an interactive smart document format that you can share around easily.
Kind of like a next gen markdown derivative cross with jupyter notebook where you bundle data, live parametric visualizations, code, audio, video, asciicinema like playback, comments, metadata, file artifacts etc
Decker works similarly to this (without the SQL-lite, but directly in the single flat html file which can be statically hosted or passed around, and includes an editor for itself!) https://github.com/JohnEarnest/Decker
If try to do data management in Excel or need to download an app just to save some recipes, this could be an alternative. For example you can track time and billing with a nice UI without requiring a SaaS subscription. You can try some examples in https://withcapsule.app/preview
This is a really cool and practical idea! Some of my first professional efforts out of college were writing Python mini-apps with Tkinter to replace "abused" Excel spreadsheets. One of my friends working in sales might benefit from something like this, I'll send it his way.
I don't know about the stability/market value of this if you intend to turn this into something that attempts to make money, however. Said friend is currently using stuff like Gemini to generate HTML mini-apps as well; while he's not storing anything in databases, he's able to generate receipts and other things for essentially free.
Indeed. But Lotus Notes actually had the design and data stored independent of one another and you had database templates. You could update your design without breaking the data, or even have entirely different design that works with the existing data model.
This capsule thing seems to store the design itself within the database, which is odd to me. I guess I don't get it at 100%.
I've been doing this with bun lately, single file app with built in SQLite. JS runtime which serves the app with SSR, websockets, and all that good stuff. The binary size is not that small, but that's usually worth the convenience.
With the battle of AI versus human agency raging I have a prediction that local-first desktop apps are going to make a comeback. It's nice to see something like this make it easier.
I agree and would love to see a shift back to what we had circa ~2009-2010: high quality, desktop and platform-specific apps that you could just pay for once (the best variant for everyone being an app that works indefinitely, but only receives updates if you renew your license every year).
The whole shift to SaaS has decimated software quality and has made most tasks a nightmare to complete (or at best, an obstacle course of multiple tools and "upgrades" and other bs).
In preparation for the upcoming [Android lockdown](https://keepandroidopen.org/) I've started vibe-coding some PWAs for apps that I like so I can have a webby version of them with no ads and no Google/iOS lock-in. I like this, it feels like a good direction.
It's depressing that everyone decided Electron was the final form of all applications and that native apps were dead until AI came along, and if native apps come back it will only be as vibe-coded slop. It's like we've just moved from once circle of hell into another.
Also Show HN is supposed to be for posting code that the community can discuss and play with. Plain advertisements and landing pages aren't allowed.
I wish there was a way to see the underlying HTML/JS and schema of these container files. Maybe you can on the desktop app? But it would be cool to see what's happening under the hood.
Do you have any plans to do merging of two versions of the same app? E.g. I send a todo list to someone, they mark some things as done and send it back to me. Will that merge into my copy if I've added some more items or would it open as a "fresh" app?
Merging works already but is a little bit hidden under the data editor function in the Capsule edit menu. Currently the last record update wins, but a merge UI is planed to do manual conflict resolving.
What I would want to see is a version of this that lets me write an app that can write to the DB through an API, and then have a google drive-esque "host this for me" thing.
A focus on local-first really misses the point IMO. Like sure it's good it can be run locally, but really most people have a distribution problem. Be a distribution solution!
The thing I would do involves having the hosting just be available by dropping the capsule into the google drive/dropbox-y thing, and then having it be that you can just copy/paste the file to someone _and it carries over the data and everything by default_.
For a lot of people the "thing" is the app + the data. Sometimes you want to decouple them but a lot of times you don't want to.
I think it would be very rare for someone to want to share the single-file bundle just without data.
Looks useful, but the Mac download shows "“Capsule” is damaged and can’t be opened. You should eject the disk image." and then you realize it could be ripe for malicious payloads as well.
This reminds me a little bit of https://webtil.es/, which can be either bundled standalone files or hosted, and add a special secure execution context.
I remember a million years ago, when I was writing apps with Cordova/Phonegap, I would use Web SQL to do all my storage [1]. This kind of reminds me of that.
I always thought that there should be an easy way to export it, but I don't think that ever materialized.
I like that there are so many haters who don’t get it or don’t think it’s useful and then there are 5 or so very similar projects in various stages of development listed in the comments. It’s giving “what if I don’t like beans?”
i built something similar back in 2013, but it used to keep the persistent state in the url params (obv not a good storage backend), and you would get a super long url to share over facebook haha
Many big companies are starting now to test agentic coding to automate business processes. This could be a nice tool to distribute apps for workflows that aren't big enough to justify the operational overhed of maintaining a traditional web app backend. Nice!
For saving data in the capsule file it would create complexity regarding migration handling when updating the HTML file to add a field to a form for example. Keeping the data schemaless makes it also easier for merging data or importing data from a CSV.
Personally, you lost me at `.capsule`. I assume this is your format? Now, what happenss when you are not around or I don't have access to you when it is needed most?
Sorry I am confused. If you(u1) share a link (html page) with your friend (u2), they fill data/form, then it get saved in u2's desktop. How does u1 see it?
This sounds like a Smalltalk image, just without the self-modifying capability, which is what makes this whole idea of code+data in a single file so appealing IMHO.
Looks amazing, I don't know if this is legit or useful feedback, but I navigated away after reading "Generate & update apps using AI".
I'm so tired of AI everywhere and everything trying to sell to you because of AI this and AI that.
I think that at this point we assume AI will be available if it makes sense for the product, making it a selling point to me is like slapping "No cholesterol!" on lettuce.
The HTML files does not have any permission to access anything on the file system. Ideally the whole app will be running inside a sandbox on desktop, for mobile apps this will be the default anyway.
"PocketBase is an open source backend consisting of embedded database (SQLite) with realtime subscriptions, builtin auth management, convenient dashboard UI and simple REST-ish API. It can be used both as Go framework and as standalone application.
The easiest way to get started is to download the prebuilt minimal PocketBase executable"
This is for local document like apps. Tauri is actually using the OS webview instead of bundling Chromium like Electron. That's why the download is only 10MB for Mac.
The problem would be cross-platform support. In the end the capsule file is the standalone binary similar to a PDF or a Word document, you only need the viewer app to open any capsule file.
you can trivially bake X binaries if you have the viewer app for each platform. This is the normal way to distribute software for cross-platform support.
I might just be hyper-defensive to any sort of vendor lock-in like a "viewer app".
For the curious, the note taking app Trilium has a "standalone" mode¹, built on top of OPFS² and sqlite OPFS persistence, such that your notes (~web pages here) are stored in sqlite locally, offline-available, optionally syncing with a remote, while being a pure local web app (after the app assets have been downloaded once). If you want to take your whole personal notes/wiki/pkms for offline access on a device you don't own/have install rights, that's the only way I know. I suspect that OP could have gone this route instead had he known about it?
¹: https://app.triliumnotes.org/
²: https://developer.mozilla.org/en-US/docs/Web/API/File_System...
³: https://sqlite.org/wasm/doc/trunk/persistence.md#opfs
> but trying to save data required hosting it somewhere
With the File System Access API, webpages nowadays can read and write local files just like desktop applications:
https://developer.chrome.com/docs/capabilities/web-apis/file...
Try this text editor for example:
https://googlechromelabs.github.io/text-editor/
It works nicely on Desktop and Mobile.
*In Chrome.
That file editor seems to work in Firefox. I just tested the save/open functionality.
It's not using the File System Access API, on Firefox it uses <input type="file"> for file-picking, and the save just downloads it. It's not synchronizing anything with your filesystem.
I see. Thanks for enlightening me.
It just downloaded the file on my firefox.
*In browsers that have implemented it
My dream is that you could write to Google Drive or iCloud Drive transparently too if the browser allows you to
I believe you are able to connect to a Google Drive account with a purely client side site via google oauth and then using the Google Drive rest api. There was an article in particular that I liked that walked through it. Let me see if I can find it
Yes but Google OAuth is a nightmare if you want to "publish" your client so that anyone can use it. If only you use it, you can add yourself as a test user. Just know your access token expires super fast and you will have to frequently re-auth. Its crazy annoying.
Even better how about a fully functional VSCode: https://vscode.dev
The File System Access API makes it possible.
I don't know, this seems like an idea that maybe works in a specific context being generalized past the point of the original idea making sense.
If the user needs to download a specific application to run this these web apps, then why not just send them that initial application in the first place? Why jump through the hoops of using Capsule when the same hoops can be jumped through to get to the same endpoint?
If this was a near-universally adopted application, then it'd make sense. But it's not, and the closest thing we have to that are browsers... which already do what you're describing?
Bundling data with the application makes sense, but is also only appropriate in pretty narrow circumstances. If I'm willing to ship my data with the web app, then I'll just embed the data in the HTML file. If the expectation is that the user will modify this data, then I don't think I'd want to ship it like this.
Yep, like Hyperclay. https://news.ycombinator.com/item?id=49690814
They're like lightweight versions of https://github.com/kem-a/AppManager which is for AppImages.
Not a bad thing, but it's just another dependency and tied to an OS-specific base software install.
They could work well pre-installed on the Mecha Comet with a bunch of app files ready to launch.
Can the added layers guarantee a new level of security or privacy like zero telemetry or keyjacking? Protection from bad actors is a good thing while making things faster and easier overall. Browser extensions are a security risk, right?
Java applets were a good idea in the sense that so long as you had Java installed, the applet was just sitting there on a webpage. Finding and managing the apps is the problem, and using browser bookmarks is a decent solution to it. The web page should demo the app, and there can be an 'install local' button next to it. But at some point all these local versions need to be updated and then not be messed with like browser extensions can be. FanFare would be a decent name.
Seems pretty useful in an enterprise context to me. I see it as an alternative to 10 million vibe coded apps to dockerize and manage. You can put the runtime on everyone’s machine and provide a web version for really collaborative stuff.
Bonus points for an intelligent way to manage these stored in OneDrive and other cloud folders. Easy to get corruption on SQLite files stored there.
It could be useful for us with home labs for the same reasons. Kind of annoying to deploy docker containers for all the services I build that amount to personalized/skinned trello boards.
Not trying to plug myself, but I’ve been working on a similar project for this exact reason. It doesn’t use special file types. Just HTML and localStorage proxies to sync devices.
It’s self-hostable too. I’ve got a bunch of super niche todo lists and personal apps running on my home server with it.
https://github.com/momja/Exhibit
There are a few things in Capsule to avoid corruptions of the file during syncing. The DELETE journal mode is used and writes are batched into transactions so that updates are always complete changes. And if there is a copy created because of a conflict merging two capsule files is supported.
My take: If your app needs to update and preserve state (using SQLite or any other DB), it is probably not something you want to pass around as a bundled file.
Or at least, it's extremely limiting, compared with hosting it somewhere on the web, which is not that hard to do these days.
Think of the workflow: Any time the state changes, you need to email a new Capsule file to whoever else is using the app. And one would think the state would change at least occasionally, because otherwise there's little reason to use a DB.
Alternatively, you can just host it on the web, the DB state dynamically updates, and it's automatically available to anyone with app access. Isn't that a lot simpler?
Or the copies can go their own way. I could send someone an itinerary and they could change the dates or drop half the stops and add their own notes without needing to send an updated copy. Getting them to install another app just to open it might be a problem, especially if they only want to use it once.
theres still lots of programs where local is the main user. but as time goes on the user may want to portably move it just by emailing the file itself. i made like a very light clone of notepad++ using it for my macbook. as a side note: i think they missed an opportunity to call it .pdfv rather than .capsule
> Isn't that a lot simpler?
You’re glossing over the recurring cost and annoyance of dealing with the hosting and having to secure the data. This way you just share and it’s local.
Sure, if you have to share data between people, your points are valid. But this seems like a way to build apps geared for individual users, like a notes or recipes app.
For most users having to mail a file to someone is way more work than having to just share a link. The cost of cloud storage is almost nothing (since email is mentioned I assume 50MB max per project). A notes app that don't sync between all devices is just a text editor to most.
I do like apps that have sensible export formats though. Having control over my data is important.
Why not on a shared drive like dropbox?
Concurrent editing wouldn't work, though.
Beyond "not working", concurrent editing is likely to silently lose data, or corrupt data, or possibly just fill up your drive with lots of "conflicted copies" that you have to manually reconcile.
When I read about this project, it doesn’t seem like concurrency is much of a concern. I think it’s _supposed_ to be treated like sharing a document with someone else.
If a capsule is a recipe, like in one of the examples, that’s pretty cool. If a capsule is a recipe _manager_, then I think it would get messy.
I miss from the home page a link to github and/or a description of it's license and/or some code of an actual capsule app (what backend language is used? -- I have no clue and cannot find out) and/or does this bundle Node?
While I wish your project success and like the simplicity of Sqlite; I prefer to start with Postgres. PG has many of the benefits (x-platform, free, embeddable, etc) that sqlite has, but also provides an industrial strength db in case you need to scale at some point, or want to use more complicated constructs.
I've been working on this exact idea, with sqlar as the "format specification". Works in the browser, and desktop + Android using Tauri.
https://github.com/JoshTheDerf/uapp
Demo apps and games: https://thederf.com/uapp/demo/
FYI requests from your demo are being blocked by CORS in Firefox and it never loads.
Thanks for the heads up! Fixed now, was a bug in the service worker.
You can get a cheap approximation of this with chromium browsers + Filesystem API and a PWA manifest for that native-ish feel. AI has reignited my interest in building based 100% on browser native features.
https://developer.mozilla.org/en-US/docs/Web/API/File_System...
If you want to distribute apps to many users this would probably be a better solution. The main benefit of Capsule is the personal document like app. You create a capsule, move it to an USB stick, save it in your cloud storage and it's still the same document. Nothing has changed, not the UI, not the data. The document will stay the same.
If it’s web native, anyone can just use it. If it’s a capsule, they first need to download the capsule app to associate the file extension with the OS and have it launch, right?
That feels like a big barrier, almost like a Java runtime.
In defense of capsule, the Filesystem API is janky and has kind of shit UX. Less sophisticated users would probably prefer the UX of a purpose built tool.
I prefer worse UX for the sake of standards and zero install, but many people would prefer the opposite. I don't think Capsule is intended for the HN crowd.
how is a filesystem supposed to prevent you from doing that?
I love this idea! One of the fallouts from widespread AI adoption that I've seen: They're very good at creating visual artifacts, but if you ever have to provide data in those artifacts, you don't really have a great way to share it without hard coding it.
> One downside with this approach is that multiple people working on it will create different copies. To make it possible to merge different copies of the same file, each data entry has a unique UUID and timestamp.
This was the first thing that popped up in my mind: Changes stemming from two sources and reconciling them. I see that you have a statement about how to handle data entry from different sources, but I don't see exactly how those are reconciled? For instance, if two users have a copy of the .capsule and make changes, then want to share their changes with the other, you have two individual .capsules with different data.
How do you merge them?
Git, html+css+svg and data in json.
Each document record is saved with UUID and timestamps. When deleting a record the data is removed but the record keeps a tombstone so only the ID remains. For conflicts currently the newest one is used, but there will be a merge like UI in the future to confirm which one to use.
Whats the advantage over plain single html files/apps using colocated css and use the browser/web view local storage? Those could be saved as PWAs or not?
It kinda reminds me of MS Access, which ended up being a pain in the butt for IT departments, but also covered a need within some department with limited resources.
Really curious to see where this goes. Do you plan to open-source the client, or simply make the file specification open when stabilized?
I'm currently 50/50 on open source the full app vs just the spec. Since I'm currently having a strong idea of what features and functionality is still missing I feel that open-sourcing right now would distract to much in this phase. But maybe when the main features are done it would make sense to have more people looking at and improving it.
Do what you feel is best for this project and good luck :)
An idea for a killer feature: Make it so that auhors can expose their capsules to the internet, and the people whoe wants to use them(lets call them users) can type the name(lets call it address) of the caplsue into the app, and your program will pull that capsule in the the users device..
Hear me out. Here is the name
PWA : Probably Web Again
nice. we should call it web, like spiders' web all connected.
wow, that sounds revolutionary.
This is great. Curious: this requires users to have a host app installed on their machine that can read .capsule file right? Won't that be a point of friction for distribution?
Isn't html/css a better distribution mechanism as most computers already have the tech to run them?
You'd think so, but web browsers are downright lobotomizing the ability for local HTML files to run any meaningful javascript code. Want to run a JS file from the same directory? CORS ERROR!!!!
Hence 200MB applications that are complete copies of the Chromium browser to run under 1MB of actual web code.
couldn't you just embed the js directly into the html file though? that would sidestep the CORS error. and since most of the people using a program like this are likely to be using AI to generate their ( likely throwaway ) applets anyway, readability and maintainability for the source doesn't matter as much as it would with human authors so having a separate js file doesn't bring that many advantages anyway.
Sharing HTML files with embedded Javascript is problematic: iOS users can't open the files directly, and have to save the file and then 'open with ... '.
Sounds like iOS is problematic.
Isn't that the price of a relatively well sandboxed experience by browsers?
I've experienced the same limits building small tools for myself and friends that I wanted to contain in a single HTML file without external dependencies... but I understand why the same project without these restrictions could easily become malware, and ones that could be easily propagated.
I wish the browser would either allow local-only or internet-only, then only make up its mind after the first attempted request. Fetch an image from the internet? No script access for local files. Fetch a JS from elsewhere on the hard drive? No internet access. That could provide the sandbox while still being flexible.
The problem is that you can't save the user data in way that you can click a single file, similar to a Word or Excel file. I you think a about it, a Excel spreadsheet is also just UI + data in a single file.
This is a frustration of mine as well, but I vaguely remember someone showing something on HN where the file could essentially save itself, I think they were using some file APIs for that... sadly I can't find the details anymore. But I keep running into the "HTML file opened locally can't update itself" issue repeatedly now that I build tools for myself with LLM agents.
Are you thinking of Nash? a standalone editable note as HTML that was featured on HN some time ago. https://keepworking.github.io/nash/
The File System Access API can do this, and it works in local ("file:///") HTML files, but it's only currently supported in desktop Chromium browsers: https://caniuse.com/native-filesystem-api
However, every browser will let you download a new version of the HTML file and save it over the old one - yes, even from a local HTML file:
... but you need the user to do this for every update. So we're back to manual "Save" buttons.This is the approach Bento took I believe (similar idea, but just for decks): https://github.com/nyblnet/bento.
Saw it on HN a few weeks back
Technically that was more MS Access :-)
That reminds me of TiddlyWiki[1], a wiki application in the form of a single self-contained HTML file.
1: https://en.wikipedia.org/wiki/TiddlyWiki
I imagined the capsule website would be able to either convert or directly run/open them
I really like the idea. It's so easy to create small tools with AI now, but hard to install them as local apps or share them.
But I'm missing a few relevant features:
1. syncing data and apps (capsules) between my devices. not necessarily over a single SaaS, but maybe p2p or some existing service
2. I think data and apps should be separated, I might want to share an app with someone else without sharing the data
3. Apps should support updates, install the same app again and just replace the code for the existing app, but keep the data. Also share updates either as a file or publish apps to a repo (just a folder of files on a public http server or github repo)
Both keeping data in sync between devices and providing offline capabilities is hard, I know.
This would be great if it evolved in a universal format/spec/feature supported by all major browsers. Only then you'd have to deal with cross web engine compatibilities. And if capsules need to do something non-trivial then you need to consider all the security aspects of having system, file or network level access. Just like any executable. Otherwise with custom app to run capsules it feels like something that can be vibe coded with electron/nw.js rather easily.
But there is something compelling in the idea - there is need for an interactive smart document format that you can share around easily. Kind of like a next gen markdown derivative cross with jupyter notebook where you bundle data, live parametric visualizations, code, audio, video, asciicinema like playback, comments, metadata, file artifacts etc
Decker works similarly to this (without the SQL-lite, but directly in the single flat html file which can be statically hosted or passed around, and includes an editor for itself!) https://github.com/JohnEarnest/Decker
That's pretty cool. But who is this for? What problem does it actually solve?
How would I know if I need this, vs something else?
If try to do data management in Excel or need to download an app just to save some recipes, this could be an alternative. For example you can track time and billing with a nice UI without requiring a SaaS subscription. You can try some examples in https://withcapsule.app/preview
This is a really cool and practical idea! Some of my first professional efforts out of college were writing Python mini-apps with Tkinter to replace "abused" Excel spreadsheets. One of my friends working in sales might benefit from something like this, I'll send it his way.
I don't know about the stability/market value of this if you intend to turn this into something that attempts to make money, however. Said friend is currently using stuff like Gemini to generate HTML mini-apps as well; while he's not storing anything in databases, he's able to generate receipts and other things for essentially free.
>>Capsule packs your entire app — UI, data, and everything — into a single portable .capsule file.
I'm having Lotus Notes flashbacks.
Indeed. But Lotus Notes actually had the design and data stored independent of one another and you had database templates. You could update your design without breaking the data, or even have entirely different design that works with the existing data model.
This capsule thing seems to store the design itself within the database, which is odd to me. I guess I don't get it at 100%.
This seems ripe for an injection attack. Can it be inspected before running?
I've been doing this with bun lately, single file app with built in SQLite. JS runtime which serves the app with SSR, websockets, and all that good stuff. The binary size is not that small, but that's usually worth the convenience.
With the battle of AI versus human agency raging I have a prediction that local-first desktop apps are going to make a comeback. It's nice to see something like this make it easier.
I agree and would love to see a shift back to what we had circa ~2009-2010: high quality, desktop and platform-specific apps that you could just pay for once (the best variant for everyone being an app that works indefinitely, but only receives updates if you renew your license every year).
The whole shift to SaaS has decimated software quality and has made most tasks a nightmare to complete (or at best, an obstacle course of multiple tools and "upgrades" and other bs).
In preparation for the upcoming [Android lockdown](https://keepandroidopen.org/) I've started vibe-coding some PWAs for apps that I like so I can have a webby version of them with no ads and no Google/iOS lock-in. I like this, it feels like a good direction.
I know that I personally have a handful of personal local only apps under development, that in the past I would have reached for a web solution.
It's depressing that everyone decided Electron was the final form of all applications and that native apps were dead until AI came along, and if native apps come back it will only be as vibe-coded slop. It's like we've just moved from once circle of hell into another.
Also Show HN is supposed to be for posting code that the community can discuss and play with. Plain advertisements and landing pages aren't allowed.
I wish there was a way to see the underlying HTML/JS and schema of these container files. Maybe you can on the desktop app? But it would be cool to see what's happening under the hood.
Do you have any plans to do merging of two versions of the same app? E.g. I send a todo list to someone, they mark some things as done and send it back to me. Will that merge into my copy if I've added some more items or would it open as a "fresh" app?
Merging works already but is a little bit hidden under the data editor function in the Capsule edit menu. Currently the last record update wins, but a merge UI is planed to do manual conflict resolving.
Looks neat. Shuffling artifacts around is more important than ever with all the agentic web sites and apps we are making.
Im finding a lot of success embedding metadata and data in SQLite files for music management.
Having a spec for a single file with a schema, version, metadata and binary blobs is super easy to shuffle around.
Agents crush reading and writing the data and writing tools around the spec.
All of a sudden things like backup, sharing and dedupe are snapping into place for my music library.
More thoughts here: https://deadca7.com/blog/deadca7-open-music-database-specifi...
I feel like this is close to a good idea.
What I would want to see is a version of this that lets me write an app that can write to the DB through an API, and then have a google drive-esque "host this for me" thing.
A focus on local-first really misses the point IMO. Like sure it's good it can be run locally, but really most people have a distribution problem. Be a distribution solution!
The thing I would do involves having the hosting just be available by dropping the capsule into the google drive/dropbox-y thing, and then having it be that you can just copy/paste the file to someone _and it carries over the data and everything by default_.
For a lot of people the "thing" is the app + the data. Sometimes you want to decouple them but a lot of times you don't want to.
I think it would be very rare for someone to want to share the single-file bundle just without data.
Looks useful, but the Mac download shows "“Capsule” is damaged and can’t be opened. You should eject the disk image." and then you realize it could be ripe for malicious payloads as well.
This reminds me a little bit of https://webtil.es/, which can be either bundled standalone files or hosted, and add a special secure execution context.
I remember a million years ago, when I was writing apps with Cordova/Phonegap, I would use Web SQL to do all my storage [1]. This kind of reminds me of that.
I always thought that there should be an easy way to export it, but I don't think that ever materialized.
[1] https://en.wikipedia.org/wiki/Web_SQL_Database
This is great. My initial thought is to create a zettel for my research projects.
I like that there are so many haters who don’t get it or don’t think it’s useful and then there are 5 or so very similar projects in various stages of development listed in the comments. It’s giving “what if I don’t like beans?”
i built something similar back in 2013, but it used to keep the persistent state in the url params (obv not a good storage backend), and you would get a super long url to share over facebook haha
reminds me of https://fzakaria.com/2026/08/23/your-executable-is-a-sqlite-...
Many big companies are starting now to test agentic coding to automate business processes. This could be a nice tool to distribute apps for workflows that aren't big enough to justify the operational overhed of maintaining a traditional web app backend. Nice!
This sounds like the modern take on MS Access.
What is wrong with using IndexedDB?
For saving data in the capsule file it would create complexity regarding migration handling when updating the HTML file to add a field to a form for example. Keeping the data schemaless makes it also easier for merging data or importing data from a CSV.
Doesn't indexedDB have builtin migration mechanism?
>Keeping the data schemaless makes it also easier for merging data or importing data from a CSV.
Not sure how it makes easier. There is always an implicit or explicit schema. IMHO explicit is often better.
Personally, you lost me at `.capsule`. I assume this is your format? Now, what happenss when you are not around or I don't have access to you when it is needed most?
Sorry I am confused. If you(u1) share a link (html page) with your friend (u2), they fill data/form, then it get saved in u2's desktop. How does u1 see it?
this is for local apps
This sounds like a Smalltalk image, just without the self-modifying capability, which is what makes this whole idea of code+data in a single file so appealing IMHO.
Is the file still fairly readable as plain text when opened with any old text editor? Is this asking too much?
if state is best shared, why keep it w/ the code?
You're gonna end up having to build a complex state sync system to a central DB anyways...
Was Capsule written in Capsule?
We are so back. I remember ActiveX in the old days
Do you plan to open source it? Seems like a proprietary app & spec.
This is proprietary. Yuck
Is this project open source?
“Capsule.app” is damaged and can’t be opened. You should move it to the Trash.
It's 1990s Microsoft Access but with AI
So, let’s do Visual Basic again, but this time with LLMs and web stuff.
like downloading exe files from the web. but we should invent something to catch bad actors putting bad stuff in them
my-bookreport-v2-final-final-done.capsule
this is great for the 'make a webapp for me and my family' thing
Looks amazing, I don't know if this is legit or useful feedback, but I navigated away after reading "Generate & update apps using AI".
I'm so tired of AI everywhere and everything trying to sell to you because of AI this and AI that.
I think that at this point we assume AI will be available if it makes sense for the product, making it a selling point to me is like slapping "No cholesterol!" on lettuce.
New way to deliver viruses just dropped! Great idea but it can be abused for sure.
The HTML files does not have any permission to access anything on the file system. Ideally the whole app will be running inside a sandbox on desktop, for mobile apps this will be the default anyway.
"No cloud. No accounts. Just share it." close tab
Can you elaborate a little?
It reads like the whole thing was vibe coded and 0 work was done beyond what the AI spit out.
well that's a first
yes yes we get it, you don't like AI ...
Really really nice
I'm surprised to not see Pocketbase mentioned in this thread at all.
https://pocketbase.io/docs/
"PocketBase is an open source backend consisting of embedded database (SQLite) with realtime subscriptions, builtin auth management, convenient dashboard UI and simple REST-ish API. It can be used both as Go framework and as standalone application.
The easiest way to get started is to download the prebuilt minimal PocketBase executable"
Attackers / Malware Developers: "A document as a web app?, What could possibly go wrong?"
is the idea this would be hosted or run locally (like Electron)?
This is for local document like apps. Tauri is actually using the OS webview instead of bundling Chromium like Electron. That's why the download is only 10MB for Mac.
i like that idea. Anything that improves native OS UI, responsiveness (120hz) and reduces resources is a good idea.
Seems very "packable into a standalone binary" but doesn't seem like that is a planned use case.
The problem would be cross-platform support. In the end the capsule file is the standalone binary similar to a PDF or a Word document, you only need the viewer app to open any capsule file.
you can trivially bake X binaries if you have the viewer app for each platform. This is the normal way to distribute software for cross-platform support.
I might just be hyper-defensive to any sort of vendor lock-in like a "viewer app".
Is this open source ?
love this idea!
Is this a durable object? /s