Much of the heavy lifting is done by this simple but powerful attribute, something surprisingly few developers seem to know about. It's been around for quite a while.
This one line was like 90% of the original implementation of Writely (the startup that became Google Docs; source: I was one of the founders).
The other 90% was all the backend code we had to write to properly synchronize edits across different browsers, each with their own bizarre suite of bugs in their contenteditable implementations :-)
Not OP, but common solutions in this space represent the state as conflict-free replicated data types (CRDTs). Some popular browser-based libraries for that are Y.js[0] and Automerge[1].
I'm not sure exactly what you mean by "dynamic text box" but it was just a contenteditable div. There have been at least two complete rewrites since I was involved, nowadays I believe it's a canvas with all of the editing, formatting, layout, and rendering done "by hand" in JavaScript.
The kind of text box used in php forums ... called text area I think, and this would be hidden every time the focus went away and an HTML based layout presented in its place. This seemed to be clunky but I thought that was what made writely possible. ContentEditable is such a breath of fresh air had I ever known about it. I wonder if IE6 supported that.
Yeah, it's pretty amazing. Back in 2016 I was still doing a lot of custom sites for small/mid businesses, and realized that almost all of this could be done with WP or something except (1) WP styling is really annoying and I don't want to maintain that, (2) I didn't want to rely on plugin functionality and upgrading WP, and (3) I wanted people with different levels of access to be able to edit different parts of the same page, which is basically impossible in most CMSs. So I basically wrote a CMS from scratch that relied on nothing but contenteditable and TinyMCE, and handed those tags to the appropriate portions (with back-end checks on what people could or could not edit before it was committed, obviously). But the point is, once a client (or their employees) are logged in, they get the right to edit the portions of their own pages that they have credentials for. The CMS puts a nice little dashed border around the parts they can edit, and all of their fonts and colors are saved in TinyMCE preferences so they don't need to hunt around or muck with any HTML (although they can, if they want to).
This ends up feeling rather magical to the clients, because they don't need to go to an editing page to make changes before seeing what it will look like. They just literally edit the content in place on the page, and hit save to deploy it live.
Most of these sites are still in operation after 5-10 years and require almost zero maintenance on my end.
In about 2009, the university I worked at adopted one of those enterprise CMS thingamajigs, but this one was called RedDot, and the central UX conceit was a red dot next to every piece of content, which transformed it into an edit control when clicked...
Under the hood it was about as shit as every other enterprise CMS of the era, but our users fucking loved it.
Hah. Never heard of that, but I can see how it would be a total love-fest for users.
Most small site owners, such as remain these days - restuarants, small companies selling furniture, lawyers, whatever - they already have their third party sales platforms. They just want to be able to edit their own web content from time to time, without any learning curve.
I never tried to commercialize my CMS, it wouldn't even make sense to do so, since it's geared toward people who want to pay $200/year for hosting once their site is up and running. (I did charge them well for building, though). But the whole thing is maybe literally 5000 lines of PHP code and a few hundred lines of Javascript and CSS. It's just structured very cleverly to read/write pages in mysql, and it has a few bells and whistles like drag/drop photo uploading in place. Barely competent PHP coders could (and have) adapted it to their own needs after I left clients. One really doesn't need the bloated structure of WP to do this stuff, just a very skeletal frame.
I had the same bright idea, but it doesn't work Firefox.
Turns out that it creates two representations of the PDF content - one is text, which has 0 opacity and another is prerendered image of the same layered on top of it. The latter is what's displayed. So while you can enable contenteditable and get the editing cursor, no changes you make will be shown. Oh, well.
Reads like “oh, here’s this amazing thing, it’s called the wheel and the it’s been around for quite a while”. I’m pretty certain I’ve been working with this 20 years ago.
ContentEditable first introduced in ie 5.5 in 2000, available in all major browsers in 2007.
I like it a lot, and I see from the source you're coding in vanilla js, which is awesome. It's underrated how much of an interactive web app you can build without using any libraries or frameworks nowadays.
I actually also really like the idea of building simple/local web apps in a single HTML page. A while ago my child and I visited a science museum and played with a simple stop motion animator. I decided to make a web app just like it and built it as a single HTML page, copied it to her laptop and she can use it without internet. I built it using React and a bundler though, because I'm too lazy to write vanilla js. It just bundles into a single HTML file. Maybe I'll post a Show HN some time.
really, just like we had a "you don't need jquery for that" we need a "you don't need react (or vue, or svelte, or whatever else you are thinking) for that."
plain, vanilla js and some decent skills will do nearly almost everything in a cleaner and more maintainable way.
after decades of maintaining other people's stuff, dumb design by smart people beats all the other combinations.
Just learn it right. I'm tired of people fighting it because they think they can learn it by intuition and then create monstrosities that are easily solved by the language they refused to understand properly in the first place.
Consider changing the behavior so that when the page is opened in the browser from the local disk, then it opens edit mode automatically, but when it's published to someone's website and accessed on the open web (check if the address is http/https), then it's a normal, read-only view expected from most pages.
Of course:
1. there should be a way to override this (a query string param or magic URL fragment) so you can look at it in edit mode, too
2. there should be a blanket exception for certain domains/URL prefixes—defaulting to keepworking.github.io (or <https://keepworking.github.io/nash/>)—so it activates edit mode on that site, too, which means that you personally don't have to do anything special to make the demo work (best if this exception is user-modifiable, so e.g. someone else can get the same behavior if they want without using the trick from #1)
I wish browsers had better support for local web apps. If local files can access persistent storage easily, this can open it up a lot of opportunities for quick and easy GUI apps. Basically the opposite of electron.
This is cool. It's funny how we are (or maybe just I am) so used writing the JS parts of websites outside of the actual HTML that my first thought upon seeing this was "wow it's crazy that this can be done with just HTML". And then looking at the source and seeing all the JS in plain old script tags made me remember that that's a thing that you can do.
VIM is an HTML page? Or JetBrains editors render an HTML page?
I know that some people code in an HTML page rendered in Electron, that would be Atom and VS Code and family, but that's far enough from the majority of people that declaring "our code editor" is disingenuous.
It seems that the order of most used technologies on SO is JS, SQL, HTML, CSS, and only then Python. We skip back to TypeScript and then Bash before finally getting to Java, C#, and C++.
I daresay that self-selected people who read, and sign up for, and fill out a survey at Stack Overflow does not represent the industry as a whole.
It's interesting, I also do not like the implication that we are all using vscode and feel slightly offended that someone would assume that I do. (The original statement is worded ambiguously and can be understood in this way.)
Anyway, I find it remarkable that a piece of software that I don't even use can trigger such a reaction in me, from a perfectly benign and innocent statement about the prevalence of HTML.
That said, I'm genuinely interested in understanding why you felt that way, is using VSCode considered something to be ashamed of or something, is it because it's a Microsoft product?
PS: I have a strong positive bias towards VSCode due to personal history, so just trying to understand the developer perspective :)
I know, and it's nothing that needs an apology from you. Sometimes people (like me) feel offended over nothing, and it's more interesting for them to question why that is, instead of blamig the other party. That's not to say the reverse can't be true, just that it's tempting and too easy to project one's own offendedness if one isn't careful.
But back to the interesting question of what's up with vscode.
It's a very subjective thing for me, which probably explains the emotional reaction. I'm aware that vscode must be a very solid and practical tool for so many people to use it. But it gives me a negative gut feeling, mostly for the following reasons:
Using what is essentially a browser to edit text feels wrong and bloaty, not unlike going on a cruise ship just so you can eat at a restaurant.
It's made by Microsoft. I'm old enough to remember when they were not ashamed of being openly nasty, how they tried to lock us all into their own walled garden, and the glimpse Internet Explorer gave us into what that would look like. It's hard to trust this company, even if they have taken on a more benign appearance.
So, it just feels very wrong to use a weirdly bloaty browser-editor that a company with a track record of being a Big Bad Influence gives away for free. (I have similar misgivings about Chrome, which does not help vscode.)
Therefore, I prefer dedicated editors and IDEs made by people whose main goal is to make just such tools, and who I can pay for (as a purchase or donation). In summary: vscode works well, but using it makes me feel icky, and there are similar or better alternatives available to me that don't give me such a feeling.
So if there is a public statement that can be read to imply that I'm somehow happily using vscode feels a little like getting mud thrown at me, if that makes any sense, and there's a strong urge to clean up the record and, by extension, myself.
Yeah, I also think it's weird how that works. It's things like this that make me understand better the secret inner workings of people interacting with other people. :)
Ok got it, I think I can relate to your feelings if someone assumed that I use TypeScript or React ;)
I got the courage to ask because I felt that you sounded open for a bit of self-introspection. Thanks for taking the time to describe something that's difficult to describe!
When I joined Microsoft in 2007, Visual Studio was at a state that it didn't even have syntax highlighting for Javascript. Many years later, I got so excited when Microsoft hired Erich Gamma from IBM's Eclipse IDE project (long story) because I knew what was coming. VSCode was originally called Monaco editor, built on top of WinJS library, and was just a little web component. Over the years, it changed significantly to become what it is today.
From your perspective of VSCode being a "Microsoft product", I totally get what you're saying, but VSCode has been one of the few exceptions for me for the reasons explained, and more.
Thanks for providing your perspective and a bit of backstory. It's interesting for me to hear from someone with such a different point of view. It certainly gives more nuance to the topic.
> even if they have taken on a more benign appearance.
The 'nasty' MS hasn't gone away. However their negative actions tend to be swept under the rug as somewhat confined to big biz and govt. Look at their headlines from the last five years for details.
As a side joke regarding vim; I have a theory that my exposure to typewriters as a kid prevented me from adopting vim as an editor :)
Pressing the letter keys for anything other than typing feels so weird to me. My brain might still be thinking of the SHIFT key as a physical modifier that literally shifts the typewriter letters from uppercase to lowercase position.
I admire fast vim users, it's a joy to watch them code.
I can relate to this. I also used typewriters, and my first computer was a mac plus. vim felt so weird and alien to me.
Then I decided to really give it a try and let go of everything I had learned. Now I'm a daily neovim user. kickstart.nvim is really what made it possible to use as a daily coding tool.
Not necessarily a thing you should do if you want any kind of separability or version control. But definitely a thing you can do.
What blows my mind is that people use frameworks where they include stuff that's not even JS in their HTML, stuff that needs external JS to attach to it or hydrate it. That feels like the worst of both worlds.
I definitely come from an opposite place where JS is king and I don't have any files with an ".html" extension exept for "index.html". And I don't have any HTML in my JS, either. Anything that's a partial template is ".htm" and gets hydrated before it gets added to the DOM.
Love this. The proliferation of software that imposes artificial limitations and puts features that can be (and have been) easily performed locally, behind subscriptions is quite frustrating.
I have been working on a free markdown editor that works entirely in the browser and can edit local files through the new filesystem access api (available in chromium only browsers).
I think something underrated but cool about this is that it's kind of like a virus. It saves its own JS source code in every file it saves, so that file can go on to make other files. Or maybe that's not cool? I guess I could think of some nefarious ways to implant a little malicious script in a standalone I send someone, that doesn't activate until they send it to someone else who was a target...
Nice work!
I’ve built a research prototype called Webstrates (webstrates.net) and recently created a local-first version: MyWebstrates (https://github.com/Webstrates/MyWebstrates).
I just tested your note on it, and it works nicely.
Here’s what I did:
I copied the HTML of Nash, went to https://my.webstrates.net, created a new blank webstrate, opened Developer Tools, and replaced the entire DOM with the Nash HTML. Now the content persists across reloads.
To enable real-time collaboration, run this in the console:
I love this. I made something similar, even more bare-bones, for myself a few years ago. I still use it literally everyday: https://tabnotes.page (just type a note)
just if i need to type something quickly, to somewhere with effectively zero startup time, just quickly pop a new tab and get a thought down. it's easy for me to lose a thought when im in the middle of something complex, so "time-to-written-down" is kinda the only thing that matters to me
The HTML part of it perhaps but tiddlywiki is also about connecting ideas.
I think this focuses much more on the standalone HTML part of it.
In the same vein, one could also distribute P2P apps via a standalone HTML file by hooking it up via WebTorrent and using their STUN servers. Or at least, I know that it's possible in the past. Just an example, standalone HTML files can be fully fledged apps is my point.
Tried it in mobile safari.
When i i highlight a word and then use one of the font controls, i expect the word to remain highlighted.
Also i can't seem to remove formatting like bold/italics by selecting it again.
Nice. I do backend web dev, and it's sometimes useful to have a lightweight testing frontend that's more flexible than Swagger. I've had a lot of success vibe coding my way through these frontends with Claude by rapidly iterating on a single file containing HTML + CSS + JS. It's quite fast to get something good enough working, and putting everything in a single HTML file format keeps it simple enough for me, a guy who doesn't do frontend, to mostly understand what's going on. "All in one" HTML file is a good trick :)
UX suggestion. You don't need the "as read only" save options if when you do a normal save the page tells you if it has been changed. That'll get you down to 2 menu options and make me think less.
The "as read only" options produce an HTML document that lacks the editor features - when you open that file from your computer, you'll see static content (although it just changes a couple of HTML attributes and leaves all the JavaScript in, inactive). The normal options re-create the entire page - editor, content and all.
Damn, I was working on the same thing but I got caught on implementing LaTeX-compatable <math> support and bitmap/svg diagram drawing. (https://files.catbox.moe/25hxc3.zip) Also I used emscripten to embed a PNG encoder into the browser
The UI of this is a lot better though, for markup-style tasks.
I love this sort of thing! I have a mood boarding app I wrote that functions the same way, packing itself and its state into a single HTML file on save.
I am also creating a Log viewer in a single HTML file (like Splunk, but a bit smaller). Pure JavaScript with no dependencies makes it runnable and integratable everywhere, which is nice.
Something you might be interested in is that journalctl (systemd's logger) can output real time logs in a format for server-sent-events (journalctl -o json-sse) so with a minimal bash http wrapper you can get realtime logs from most linux systems into your app.
Indeed, that one would be more descriptive. At first I thought that the project is a joke about pretend-discovering that you can create a web page without tools other than a text editor, and without embedding external resources, similar to vanilla-js.com. I wonder whether "note" is readily understood as "a document edited in a WYSIWYG editor" by some, possibly because it is called that way in some software.
I haven’t tried sticking an image in it yet, but I drafted a blog post in it already this morning.
Very fun. Nice and lightweight. I like it.
I may submit a PR for you - I have some ideas about how I could make this more personally useful. Saving off files as Markdown is one idea I had. Would solve my “no pretty markdown editor” woes, and also allow me to chuck output from this directly into my existing GitHub Pages blog.
Hello, very cool app. I have been making apps like this using my tool here:
https://domsy.io
It's pretty cool how quickly and easily I can generate little static apps like this for ad hoc use cases. I have made a weight tracker, expense tracker, prototypes for work, cards for my wife, slides for work, etc.
I've done a similar tool a few years ago, with "contenteditable". It's great. Especially for non-tech people who want to quickly edit articles on a website. (that was an experiment).
The problem I had: when people copy/paste HTML from external sources, after a few edits, a few months later, you see that the HTML is just a horrible mix of various mixed up tags, the structure is lost, you have
<div><div><div><div><p><div>
everywhere and it's impossible to clean it in WYSIWYG.
OP, did you find a satisfying solution to this problem? A good compromise for sanitizing of copy/paste content? TL;DR: not remove the images, but clean the
I'm curious, somehow I remebered as JS not getting executed when opening a html file from filesystem, but clearly that isn't the case since this works. Did this change? Maybe I'm just misremembering
If the JS is embedded within the HTML file it should’ve always worked, but when the HTML loads the JS file from somewhere else (that isn’t a relative path or a full URL that’s hosted somewhere else) it would usually fail, thus it wouldn’t work without running it through some kind of server that resolves the path
script type="module" doesn't work from file systems. To my great annoyance. Other than that, I'm quite certain JS has worked from file systems since it existed.
Ok, after clicking around for 10 minutes I think I figured out what this is: it is a single-page WYSIWYG HTML editor with the ability to export either everything including itself, or just the HTML without the editor.
How is this useful?
For quick writing and sharing of HTML files, but not the links, but the files themselves.
The idea is you would write everything in the webpage, like. Google Doc. Then you can download that web page and share it with anyone anywhere and nearly everyone will be able to read it because it will open in a browser
Would I be able to store and edit this type of note on Google Drive? I like to keep documents in the cloud so I can access and edit them from anywhere, like Google Docs. But I don't think Drive can actually open html files, despite working entirely inside the browser.
This is neat! Reminds me of the good old days of single-page websites. Now, if only it could automatically back itself up to the blockchain... just kidding (mostly).
If I write a url like news.google.com and select it, then click you linkurl button It should use my text. I just had to re-enter the link. So if I highlight something please accept that as a link, even if I didn't put https:// at the front.
There's a lot of value in these kinds of things to manage your own bookmarks if you choose to without having to expose them to Google and a Google doc or something. Of course you risk losing your file.
Whoever downvoted me, care to give any constructive critisism about what exactly deserved the downvote? I'm not meaning that in an arsey way, I'm just a sucker for constructive feedback.
Much of the heavy lifting is done by this simple but powerful attribute, something surprisingly few developers seem to know about. It's been around for quite a while.
This one line was like 90% of the original implementation of Writely (the startup that became Google Docs; source: I was one of the founders).
The other 90% was all the backend code we had to write to properly synchronize edits across different browsers, each with their own bizarre suite of bugs in their contenteditable implementations :-)
This is a truly a great insight from how simple things can create industry giants like google docs.
I believe that a lot of problems can be converted into synchronization problems in browsers.
Are there any general synchronization libraries / applications that you suggest within browser / outside browser?
Thanks in advance.
Not OP, but common solutions in this space represent the state as conflict-free replicated data types (CRDTs). Some popular browser-based libraries for that are Y.js[0] and Automerge[1].
[0]: https://yjs.dev/ [1]: https://automerge.org/
By the way, for those wanting to do this but in Rust, there is https://crates.io/crates/yrs and https://crates.io/crates/automerge
And Loro [1], relatively new player which recently hit 1.0, with solid performance and some features the others don’t have
[1]: https://loro.dev/
Amazing, thanks for the link
What features does Loro have that others lack?
I thought it was a dynamic text box souped up with all sorts of tricks and semantics to make it appear like a rich text editor
I'm not sure exactly what you mean by "dynamic text box" but it was just a contenteditable div. There have been at least two complete rewrites since I was involved, nowadays I believe it's a canvas with all of the editing, formatting, layout, and rendering done "by hand" in JavaScript.
The kind of text box used in php forums ... called text area I think, and this would be hidden every time the focus went away and an HTML based layout presented in its place. This seemed to be clunky but I thought that was what made writely possible. ContentEditable is such a breath of fresh air had I ever known about it. I wonder if IE6 supported that.
Yeah, it's pretty amazing. Back in 2016 I was still doing a lot of custom sites for small/mid businesses, and realized that almost all of this could be done with WP or something except (1) WP styling is really annoying and I don't want to maintain that, (2) I didn't want to rely on plugin functionality and upgrading WP, and (3) I wanted people with different levels of access to be able to edit different parts of the same page, which is basically impossible in most CMSs. So I basically wrote a CMS from scratch that relied on nothing but contenteditable and TinyMCE, and handed those tags to the appropriate portions (with back-end checks on what people could or could not edit before it was committed, obviously). But the point is, once a client (or their employees) are logged in, they get the right to edit the portions of their own pages that they have credentials for. The CMS puts a nice little dashed border around the parts they can edit, and all of their fonts and colors are saved in TinyMCE preferences so they don't need to hunt around or muck with any HTML (although they can, if they want to).
This ends up feeling rather magical to the clients, because they don't need to go to an editing page to make changes before seeing what it will look like. They just literally edit the content in place on the page, and hit save to deploy it live.
Most of these sites are still in operation after 5-10 years and require almost zero maintenance on my end.
In about 2009, the university I worked at adopted one of those enterprise CMS thingamajigs, but this one was called RedDot, and the central UX conceit was a red dot next to every piece of content, which transformed it into an edit control when clicked...
Under the hood it was about as shit as every other enterprise CMS of the era, but our users fucking loved it.
Hah. Never heard of that, but I can see how it would be a total love-fest for users.
Most small site owners, such as remain these days - restuarants, small companies selling furniture, lawyers, whatever - they already have their third party sales platforms. They just want to be able to edit their own web content from time to time, without any learning curve.
I never tried to commercialize my CMS, it wouldn't even make sense to do so, since it's geared toward people who want to pay $200/year for hosting once their site is up and running. (I did charge them well for building, though). But the whole thing is maybe literally 5000 lines of PHP code and a few hundred lines of Javascript and CSS. It's just structured very cleverly to read/write pages in mysql, and it has a few bells and whistles like drag/drop photo uploading in place. Barely competent PHP coders could (and have) adapted it to their own needs after I left clients. One really doesn't need the bloated structure of WP to do this stuff, just a very skeletal frame.
A couple remaining examples of my CMS in action: https://thebarkerlounge.com https://vickmanassociates.com
Ah, RedDot - that’s a throwback if I ever heard one!
I made a live demo of the concept a few weeks ago; it's such as simple but powerful hidden feature of HTML:
https://news.ycombinator.com/item?id=43179649
I love this. Once made a pdf editor by converting a doc to HTML then throwing in contenteditable="true" in the top-most div lol
I had the same bright idea, but it doesn't work Firefox.
Turns out that it creates two representations of the PDF content - one is text, which has 0 opacity and another is prerendered image of the same layered on top of it. The latter is what's displayed. So while you can enable contenteditable and get the editing cursor, no changes you make will be shown. Oh, well.
What do/did you use to convert pdf to html?
Reads like “oh, here’s this amazing thing, it’s called the wheel and the it’s been around for quite a while”. I’m pretty certain I’ve been working with this 20 years ago.
ContentEditable first introduced in ie 5.5 in 2000, available in all major browsers in 2007.
I like it a lot, and I see from the source you're coding in vanilla js, which is awesome. It's underrated how much of an interactive web app you can build without using any libraries or frameworks nowadays.
I actually also really like the idea of building simple/local web apps in a single HTML page. A while ago my child and I visited a science museum and played with a simple stop motion animator. I decided to make a web app just like it and built it as a single HTML page, copied it to her laptop and she can use it without internet. I built it using React and a bundler though, because I'm too lazy to write vanilla js. It just bundles into a single HTML file. Maybe I'll post a Show HN some time.
really, just like we had a "you don't need jquery for that" we need a "you don't need react (or vue, or svelte, or whatever else you are thinking) for that."
plain, vanilla js and some decent skills will do nearly almost everything in a cleaner and more maintainable way.
after decades of maintaining other people's stuff, dumb design by smart people beats all the other combinations.
The downside, of course, is needing to code in JavaScript.
Just learn it right. I'm tired of people fighting it because they think they can learn it by intuition and then create monstrosities that are easily solved by the language they refused to understand properly in the first place.
As opposed to? You can use typescript and compile too. And/or still bundle with esbuild
I like this about quarto. You can create a standalone html file and share with people.
Great work.
Consider changing the behavior so that when the page is opened in the browser from the local disk, then it opens edit mode automatically, but when it's published to someone's website and accessed on the open web (check if the address is http/https), then it's a normal, read-only view expected from most pages.
Of course:
1. there should be a way to override this (a query string param or magic URL fragment) so you can look at it in edit mode, too
2. there should be a blanket exception for certain domains/URL prefixes—defaulting to keepworking.github.io (or <https://keepworking.github.io/nash/>)—so it activates edit mode on that site, too, which means that you personally don't have to do anything special to make the demo work (best if this exception is user-modifiable, so e.g. someone else can get the same behavior if they want without using the trick from #1)
I actually really like the idea of websites that allow you to edit the content and save it locally... IMO, this is brilliant and should stay as is.
This was actually the intent of some early versions of the WWW.
You're not really grokking either my comment or how the existing read-only mode actually works right now.
You could make this self contained and truly portable with "real" save using https://rpdillon.net/redbean-tiddlywiki-saver.html
I built something like this a while back, you can see it at https://original.littr.me
Yes, please make it offer to save a file. That gets us much closer to the capability of having private bookmarks if we want to
Nice job! There was a previous discussion about this idea with a mention of TiddlyWiki:
https://news.ycombinator.com/item?id=43179649
One of my favorites in this space is Feather Wiki: https://feather.wiki/
Yes, before I wrote the post on HN, I shared it with my local community and they told me about the same two features.
I think those two services are more appropriate for the name "Note".
For my project, I should have used Page or Document for the project name.
I wish browsers had better support for local web apps. If local files can access persistent storage easily, this can open it up a lot of opportunities for quick and easy GUI apps. Basically the opposite of electron.
How file can access anything? It is just data. It can not do anything.
Executables are files too.
Then it is really a runtime that have an access. I.e. it depends if process have an access to a file from which user is executing a bash script.
Runtimes and operating systems consist of files, too.
This is cool. It's funny how we are (or maybe just I am) so used writing the JS parts of websites outside of the actual HTML that my first thought upon seeing this was "wow it's crazy that this can be done with just HTML". And then looking at the source and seeing all the JS in plain old script tags made me remember that that's a thing that you can do.
HTML is underrated. Most of the screens we look at daily, including our code editor are HTML pages.
VIM is an HTML page? Or JetBrains editors render an HTML page?
I know that some people code in an HTML page rendered in Electron, that would be Atom and VS Code and family, but that's far enough from the majority of people that declaring "our code editor" is disingenuous.
See my other comment in this thread. It's 3/4 devs according to SO.
Call me a neckbeard, but 3/4 of devs are not representative of 3/4 of devs in the industry in my opinion. And especially not of 3/4 of devs on HN.
According to: https://survey.stackoverflow.co/2024/technology#most-popular...
It seems that the order of most used technologies on SO is JS, SQL, HTML, CSS, and only then Python. We skip back to TypeScript and then Bash before finally getting to Java, C#, and C++.
I daresay that self-selected people who read, and sign up for, and fill out a survey at Stack Overflow does not represent the industry as a whole.
Eh, no?
My Code editor is NeoVim or it's IntelliJ
But there is no HTML page involved at all.
What I want to hint at: You're extrapolating the usage of VSCode.
According to the latest stack overflow dev survey 74% of respondents use vs code. I think it's a pretty safe statement.
It's interesting, I also do not like the implication that we are all using vscode and feel slightly offended that someone would assume that I do. (The original statement is worded ambiguously and can be understood in this way.)
Anyway, I find it remarkable that a piece of software that I don't even use can trigger such a reaction in me, from a perfectly benign and innocent statement about the prevalence of HTML.
Sorry I didn't mean to offend anyone.
That said, I'm genuinely interested in understanding why you felt that way, is using VSCode considered something to be ashamed of or something, is it because it's a Microsoft product?
PS: I have a strong positive bias towards VSCode due to personal history, so just trying to understand the developer perspective :)
> Sorry I didn't mean to offend anyone.
I know, and it's nothing that needs an apology from you. Sometimes people (like me) feel offended over nothing, and it's more interesting for them to question why that is, instead of blamig the other party. That's not to say the reverse can't be true, just that it's tempting and too easy to project one's own offendedness if one isn't careful.
But back to the interesting question of what's up with vscode.
It's a very subjective thing for me, which probably explains the emotional reaction. I'm aware that vscode must be a very solid and practical tool for so many people to use it. But it gives me a negative gut feeling, mostly for the following reasons:
Using what is essentially a browser to edit text feels wrong and bloaty, not unlike going on a cruise ship just so you can eat at a restaurant.
It's made by Microsoft. I'm old enough to remember when they were not ashamed of being openly nasty, how they tried to lock us all into their own walled garden, and the glimpse Internet Explorer gave us into what that would look like. It's hard to trust this company, even if they have taken on a more benign appearance.
So, it just feels very wrong to use a weirdly bloaty browser-editor that a company with a track record of being a Big Bad Influence gives away for free. (I have similar misgivings about Chrome, which does not help vscode.)
Therefore, I prefer dedicated editors and IDEs made by people whose main goal is to make just such tools, and who I can pay for (as a purchase or donation). In summary: vscode works well, but using it makes me feel icky, and there are similar or better alternatives available to me that don't give me such a feeling.
So if there is a public statement that can be read to imply that I'm somehow happily using vscode feels a little like getting mud thrown at me, if that makes any sense, and there's a strong urge to clean up the record and, by extension, myself.
Yeah, I also think it's weird how that works. It's things like this that make me understand better the secret inner workings of people interacting with other people. :)
Ok got it, I think I can relate to your feelings if someone assumed that I use TypeScript or React ;)
I got the courage to ask because I felt that you sounded open for a bit of self-introspection. Thanks for taking the time to describe something that's difficult to describe!
When I joined Microsoft in 2007, Visual Studio was at a state that it didn't even have syntax highlighting for Javascript. Many years later, I got so excited when Microsoft hired Erich Gamma from IBM's Eclipse IDE project (long story) because I knew what was coming. VSCode was originally called Monaco editor, built on top of WinJS library, and was just a little web component. Over the years, it changed significantly to become what it is today.
From your perspective of VSCode being a "Microsoft product", I totally get what you're saying, but VSCode has been one of the few exceptions for me for the reasons explained, and more.
Thanks for providing your perspective and a bit of backstory. It's interesting for me to hear from someone with such a different point of view. It certainly gives more nuance to the topic.
> even if they have taken on a more benign appearance.
The 'nasty' MS hasn't gone away. However their negative actions tend to be swept under the rug as somewhat confined to big biz and govt. Look at their headlines from the last five years for details.
As a side joke regarding vim; I have a theory that my exposure to typewriters as a kid prevented me from adopting vim as an editor :)
Pressing the letter keys for anything other than typing feels so weird to me. My brain might still be thinking of the SHIFT key as a physical modifier that literally shifts the typewriter letters from uppercase to lowercase position.
I admire fast vim users, it's a joy to watch them code.
I can relate to this. I also used typewriters, and my first computer was a mac plus. vim felt so weird and alien to me.
Then I decided to really give it a try and let go of everything I had learned. Now I'm a daily neovim user. kickstart.nvim is really what made it possible to use as a daily coding tool.
Pretty sure IntelliJ contains a bunch of bundled webviews for various things.
Also, extrapolating usage of vscode seems pretty fair since they said "most"
Not necessarily a thing you should do if you want any kind of separability or version control. But definitely a thing you can do.
What blows my mind is that people use frameworks where they include stuff that's not even JS in their HTML, stuff that needs external JS to attach to it or hydrate it. That feels like the worst of both worlds.
I definitely come from an opposite place where JS is king and I don't have any files with an ".html" extension exept for "index.html". And I don't have any HTML in my JS, either. Anything that's a partial template is ".htm" and gets hydrated before it gets added to the DOM.
Suggestion: It should warn when you have edited it and are about to close the page without saving. Consider adding an `onbeforeunload` handler.
This webpage has a warning on close right now but that's a generally good suggestion.
I like this kind of simple tooling - I built a simple custom “homepage”/dashboard the other day: https://github.com/rcarmo/onepage-by-spec
Just updated it to "save" as well.
Love this. The proliferation of software that imposes artificial limitations and puts features that can be (and have been) easily performed locally, behind subscriptions is quite frustrating.
I have been working on a free markdown editor that works entirely in the browser and can edit local files through the new filesystem access api (available in chromium only browsers).
https://zareith.github.io/chillmd/
Ahh, good old contenteditable.
This is probably the simplest method, put straight in the address bar:
data:text/html, <html contenteditable>
You'll need a bit longer URI in order to support unicode.
Related Hacker News discussions:
https://news.ycombinator.com/item?id=6005295
https://news.ycombinator.com/item?id=5135194
I think something underrated but cool about this is that it's kind of like a virus. It saves its own JS source code in every file it saves, so that file can go on to make other files. Or maybe that's not cool? I guess I could think of some nefarious ways to implant a little malicious script in a standalone I send someone, that doesn't activate until they send it to someone else who was a target...
recently i tried to write a local wiki with tiddlywiki and failed lamentably, ending up writing directly an html file *
Nash appears to simplify my job, thanks!
* plus this for an easy TOC https://codepen.io/cgurski/pen/qBrNrPo
Nice work! I’ve built a research prototype called Webstrates (webstrates.net) and recently created a local-first version: MyWebstrates (https://github.com/Webstrates/MyWebstrates).
I just tested your note on it, and it works nicely.
Here’s what I did: I copied the HTML of Nash, went to https://my.webstrates.net, created a new blank webstrate, opened Developer Tools, and replaced the entire DOM with the Nash HTML. Now the content persists across reloads.
To enable real-time collaboration, run this in the console:
webstrate.addSyncServer('sync.webstrates.net')
Then share the URL: https://my.webstrates.net/?s/<document-hash>@sync.webstrates...
Now it’s live-editable with others!
Webstrates works by storing and syncing changes to the DOM.
With the chance of killing my sync server, here's a Nash note on MyWebstrates: https://my.webstrates.net/?s/41W5owzLg94wDQAin4yJXGnFyKWN@sy...
Wow, this is really cool.
I should remember it later and study more.
I came to say exactly this. I was sucked right into this. I am looking forward to reading more thoroughly and understanding it better.
I love this. I made something similar, even more bare-bones, for myself a few years ago. I still use it literally everyday: https://tabnotes.page (just type a note)
I remember this and used it for a while! Great idea.
Just one note in local storage? What would you use it for?
just if i need to type something quickly, to somewhere with effectively zero startup time, just quickly pop a new tab and get a thought down. it's easy for me to lose a thought when im in the middle of something complex, so "time-to-written-down" is kinda the only thing that matters to me
Is this similar to TiddlyWiki https://tiddlywiki.com/ ?
Certainly a similar idea.
The HTML part of it perhaps but tiddlywiki is also about connecting ideas.
I think this focuses much more on the standalone HTML part of it.
In the same vein, one could also distribute P2P apps via a standalone HTML file by hooking it up via WebTorrent and using their STUN servers. Or at least, I know that it's possible in the past. Just an example, standalone HTML files can be fully fledged apps is my point.
Tried it in mobile safari. When i i highlight a word and then use one of the font controls, i expect the word to remain highlighted. Also i can't seem to remove formatting like bold/italics by selecting it again.
This is quite cool, but I am not sure I understand the difference between 'Save' and 'Share'. They both seem to just save a file.
Save uses an a href (link) with download attribute (https://developer.mozilla.org/en-US/docs/Web/HTML/Element/a#...), which will usually trigger a normal file download.
Share uses Navigator.share (https://developer.mozilla.org/en-US/docs/Web/API/Navigator/s...), which should invoke the browser’s “share” menu if it has one.
I actually was hoping to find something like this, but that uses local storage as a backend so it can be plugged into my reframe [1].
It would be nice to have a reasonably nice text editor for notes and whatnot everytime I open a new tab.
[1] https://reframe.canine.sh/czhu12
Great idea, however for some idea ctrl + z (undo) does not work.
HTML: the only markup language that we ever needed.
All it was missing was a self-contained WYSIWYG editor for it.
Nice. I do backend web dev, and it's sometimes useful to have a lightweight testing frontend that's more flexible than Swagger. I've had a lot of success vibe coding my way through these frontends with Claude by rapidly iterating on a single file containing HTML + CSS + JS. It's quite fast to get something good enough working, and putting everything in a single HTML file format keeps it simple enough for me, a guy who doesn't do frontend, to mostly understand what's going on. "All in one" HTML file is a good trick :)
Noice!
UX suggestion. You don't need the "as read only" save options if when you do a normal save the page tells you if it has been changed. That'll get you down to 2 menu options and make me think less.
The "as read only" options produce an HTML document that lacks the editor features - when you open that file from your computer, you'll see static content (although it just changes a couple of HTML attributes and leaves all the JavaScript in, inactive). The normal options re-create the entire page - editor, content and all.
But why is that useful? Just do what Confluence does. Read only, then click a pencil to edit.
Damn, I was working on the same thing but I got caught on implementing LaTeX-compatable <math> support and bitmap/svg diagram drawing. (https://files.catbox.moe/25hxc3.zip) Also I used emscripten to embed a PNG encoder into the browser
The UI of this is a lot better though, for markup-style tasks.
Thanks for making this. Sometimes I want to just write a quick note and serve as a static html file and this is very useful for that purpose.
Would the use case here be like a dirt simple splash page / redirect / under construction page?
a simple download page for beta software in a closed group, a birthday invite with all the relevant info
I love this sort of thing! I have a mood boarding app I wrote that functions the same way, packing itself and its state into a single HTML file on save.
here in case anyone wants to check it out
https://thot-experiment.github.io/big_mood/
Does it make a readable HTML if Javascript was to be disabled?
edit: yup it does. Can use Ctrl + S to save via browser too instead of the save icon. Nice work.
I wanted to make/have something like this [1] and got claude to get something close as a PoC.
https://news.ycombinator.com/item?id=42464740
I am also creating a Log viewer in a single HTML file (like Splunk, but a bit smaller). Pure JavaScript with no dependencies makes it runnable and integratable everywhere, which is nice.
https://github.com/dobin/SemiDataSieve
Something you might be interested in is that journalctl (systemd's logger) can output real time logs in a format for server-sent-events (journalctl -o json-sse) so with a minimal bash http wrapper you can get realtime logs from most linux systems into your app.
ooo~ Remarkably similiar to nullboard https://nullboard.io/preview https://github.com/apankrat/nullboard
Supports Ctrl-z!
Hey :)! That's my name: https://nashkabbara.com
Calling it a "note" is a bit confusing. I'd call it a self-modifying editable single-file web app.
Indeed, that one would be more descriptive. At first I thought that the project is a joke about pretend-discovering that you can create a web page without tools other than a text editor, and without embedding external resources, similar to vanilla-js.com. I wonder whether "note" is readily understood as "a document edited in a WYSIWYG editor" by some, possibly because it is called that way in some software.
It's because I didn't think hard enough when I named it. I should have described it as a document with a built-in editor.
The wysiwyg buttons don't seem to function on Android Chrome mobile, at least for me.
First you need to SELECT text.
I really like this, for two reasons:
1) I’ve been getting sick of Jekyll and using it to write, because I find all the work of managing images tedious.
2) my first name is Nash, so this feels like it was made for me. :)
I hope this was a good gift for you. ;)
I haven’t tried sticking an image in it yet, but I drafted a blog post in it already this morning.
Very fun. Nice and lightweight. I like it.
I may submit a PR for you - I have some ideas about how I could make this more personally useful. Saving off files as Markdown is one idea I had. Would solve my “no pretty markdown editor” woes, and also allow me to chuck output from this directly into my existing GitHub Pages blog.
Hello, very cool app. I have been making apps like this using my tool here: https://domsy.io
It's pretty cool how quickly and easily I can generate little static apps like this for ad hoc use cases. I have made a weight tracker, expense tracker, prototypes for work, cards for my wife, slides for work, etc.
For example, this slide show app: https://domsy.io/share/644305ab-d36b-40a9-80e7-f0b52abaa18b
I import it in domsy.io and give AI a text dump of everything I need, it uses the js in that html to convert to slides that I can download to pdf.
For something very similar — but with markdown, less features, and minimal UI — I made https://kvak.io
Is this local or on your site? It doesn't seem similar if it online...
Does this support Chrome's filesystem API?
Good idea, that's available on many browsers:
https://developer.mozilla.org/en-US/docs/Web/API/FileSystem
The GP is talking about a different API:
https://developer.mozilla.org/en-US/docs/Web/API/File_System...
This is one that Firefox refuses to implement (a personal gripe for me)
for some reason firefox wasnt saving my changes when i saving as html-only, but saving as complete web page it did.
This is excellent. I wonder how hard it would be to add functionally for lists and tables. Tables would likely be difficult.
Great work.
This is impressive! I like how you make this without the use of frameworks. Nice job!
I've done a similar tool a few years ago, with "contenteditable". It's great. Especially for non-tech people who want to quickly edit articles on a website. (that was an experiment).
The problem I had: when people copy/paste HTML from external sources, after a few edits, a few months later, you see that the HTML is just a horrible mix of various mixed up tags, the structure is lost, you have
everywhere and it's impossible to clean it in WYSIWYG.OP, did you find a satisfying solution to this problem? A good compromise for sanitizing of copy/paste content? TL;DR: not remove the images, but clean the
stuff when copy/pasting?Very cool. Did you use an AI tool to help write it? If so, which tool and which LLM model?
That pizza pic at the top. Are those shrimps on it? Any good?
Yes it has shirimp topping, I really liked it.
This uses HTML, CSS and JavaScript... but all in one file.
That's right. Exactly!
I wanted the document to be edited without a separate editor.
I'm curious, somehow I remebered as JS not getting executed when opening a html file from filesystem, but clearly that isn't the case since this works. Did this change? Maybe I'm just misremembering
If the JS is embedded within the HTML file it should’ve always worked, but when the HTML loads the JS file from somewhere else (that isn’t a relative path or a full URL that’s hosted somewhere else) it would usually fail, thus it wouldn’t work without running it through some kind of server that resolves the path
script type="module" doesn't work from file systems. To my great annoyance. Other than that, I'm quite certain JS has worked from file systems since it existed.
These days, local html seems to be running javascripts.
If you send html to Telegram, the preview will not run js when the recipient opens it. I think Android and IOS work differently.
You are definitely misremembering, unless you are talking about external JS files on ancient MSIE versions I think.
It’s a nice idea.. good job.
I still don't understand what is this even though I read all you wrote both here and on the web page.
Ok, after clicking around for 10 minutes I think I figured out what this is: it is a single-page WYSIWYG HTML editor with the ability to export either everything including itself, or just the HTML without the editor.
How is this useful? For quick writing and sharing of HTML files, but not the links, but the files themselves.
Very interesting,thank you.
The idea is you would write everything in the webpage, like. Google Doc. Then you can download that web page and share it with anyone anywhere and nearly everyone will be able to read it because it will open in a browser
Would I be able to store and edit this type of note on Google Drive? I like to keep documents in the cloud so I can access and edit them from anywhere, like Google Docs. But I don't think Drive can actually open html files, despite working entirely inside the browser.
On my dropbox, I was able to modify and share it on the web or apps. It was not saved.
When I opened html on the file app on my iPhone, js didn't run, so I could only see it
I don't think it's perfect in many ways.
By 'note' I guess they mean rich text editor.
Excellent little tool. Thank you for making it
This is neat! Reminds me of the good old days of single-page websites. Now, if only it could automatically back itself up to the blockchain... just kidding (mostly).
I'm clearly missing something, despite the minimalist explanation.
So, this is my bug report. I'm not going to explain more, because this is a minimalist bug report.
Looks Great ad a freeform note taking tool / blogging tool
I would like to understand more how this works -- where the data is stored esp if I were to self host it.
This is so good. Thank you for this.
If I write a url like news.google.com and select it, then click you linkurl button It should use my text. I just had to re-enter the link. So if I highlight something please accept that as a link, even if I didn't put https:// at the front.
There's a lot of value in these kinds of things to manage your own bookmarks if you choose to without having to expose them to Google and a Google doc or something. Of course you risk losing your file.
Wow! thats an interesting one
That's neat.
Love this.
[dead]
[dead]
[dead]
Whoever downvoted me, care to give any constructive critisism about what exactly deserved the downvote? I'm not meaning that in an arsey way, I'm just a sucker for constructive feedback.
Cool.