Quill v2 – Rich text editor | Hacker News Reader

Quill v2 – Rich text editor

quilljs.com

453 pointsulrischa96 comments

At my last job I had built a system where users could enter text into a quill textbox (and style it), and then save it into a database in their delta format. Later, a different team member would pull that into an InDesign plugin where we could easily read what styles had been applied to the text in quill, and style that using the styling in InDesign. This was used when the front desk person would enter in a legal notice for a newspaper, and the production team would be the ones who would get that into InDesign for printing. The quill delta format that it uses made this process so much easier than if we were to have been given HTML like most other editors output / save. Glad to see this is still around, its been a few years since I had heard / used quill.

Sounds like an interesting system :)

Release notes: https://github.com/quilljs/quill/releases/tag/v2.0.0 Announcement post: https://slab.com/blog/announcing-quill-2-0/

I've used quill quite a bit. It's easy to integrate, the json delta format works decently well, and is pretty straightforward, but one thing I find myself doing whenever I use it is asking myself, "Should I just be using prosemirror?".

I find Quill to be a great stopgap solution for when I need editing and I need it quick, but in the back of my mind I'm always mentally adding that I'll need a migration plan should that specific feature ever grow in requirements.

I wish we could change the submission to [1], since it really is about the new version. On the hand I sometimes wonder why we can have standardise rich text editor.

[1] https://slab.com/blog/announcing-quill-2-0/

I've been looking at ProseMirror (mostly through tiptap2, which seems to wrangle it effectively) but taking the jump is daunting. I'd be interested in others' experiences.

Tiptap is a great wrapper around ProseMirror, and if you need to dig deep, TipTap lets you do that but the documentation, for both TipTap and PM isn't as good as it could be and the conceptual model for PM is _hard_ to pick up on the go. I've spent several days on it to get close to what we spec-ed, granted, that's probably a tenth of what I would've spent if I had to build it from scratch, but definitely an investment, because getting all the little interactive details and the way that the custom inline/block elements should work just as the user expects is _hard_.

Overall, it's the only tool I found that makes it as easy as possible to create a good writing/editing experience but it's a hard easy.

I've used prosemirror enough to have written custom nodes, commands, and custom code around its collaboration model. I got good results with all of this and I don't know any other platform that could have matched it.

The docs are thoughtful. There's an up-front learning curve to understand the architecture. When doing highly customized things, I referred to the source when needed.

For standard rich text, there are a lot of options. Prosemirror shines when you want to build on it as a platform.

I spent weeks back in 2020 fighting with PM trying to customize behaviors such as bullet points etc. At the time the documentation was sparse and for me progress was extremely hard.

I ended up abandoning the project but wrote a short introduction here which might help you get started.

https://github.com/PierBover/prosemirror-cookbook

Tried Tiptap recently and it was fairly buggy even for basic stuff like lists. Apparently the team behind it went the VC route and raised a few million so ideally they'd be moving faster at this point, but it doesn't seem much improved from when I tried it a few years ago.

That said, text editing in the browser is indeed a nightmare, so it's hard to fault anyone trying to make it easier.

I didn't know there was that much money in writing a rich text editor ...

CKEditor and TinyMCE seem to be doing fine.

Are they? Both have been acquired by the same holding company.

With the release of Quill v2, I went away from TipTap + PM to Quill. My use case is not super complex. I just want a decent editing experience (for trusted, technical people, not average end users) and solid HTML I can save to a db.

Quill has `getSemanticHTML()` which is great for my implementation.

PM is great, but expansive and overpowered for my needs.

TipTap has the equivalent with `getHTML` which is calling a PM function that does the same.

I needed to implement rich text editing at my job. So I looked at many options, quill was among the first one I looked at, but I ended going with ProseMirror. The reason I picked ProseMirror was how low level it was, and the extensive documentation that goes with it (kind of reminds me of D3 really). Since picking it I’ve written a few custom nodes, custom commands, a custom tool bar, integrated with markown + custom markdown plugins, and it all integrates rather neatly with Vue (though with some glue code).

So far I’m very pleased with my choice of ProseMirror.

Am I crazy for having an obnoxiously strong opinion about how return should be handled by text editors?

Plain text return = new line.

Rich text return = new paragraph (shift + return = new line if you want).

This was a big problem for me with earlier versions, unsure how customizable this behaviour is now.

I do appreciate that I probably care too much about this, but I found TipTap worked well.

It should be easy to configure, but I don't think that behavior should be the default since it's less general and only works when you press Return from certain types of blocks.

Shift + Return also isn't available on mobile. And it's annoying having a hidden key combo just to get behavior you want. I seem to regularly battle with Reddit's rich editor that does this.

All to save me from having to press Return twice.

Though I can appreciate the niceness of how clean it is to always keep the user inside a while making them opt (with Shift + Return) into a , I'd yield to UX over cleanliness when it comes to something as finicky as rich text editors.

Totally agree - even that "shift+" support makes me nervous, but I'd probably concede it! The " in the Quill example looks... awkward.

No, you are 100% correct.

If I am reading this correctly then it resembles the behavior of the Confluence/JIRA text input windows.

That seems correct to me!

TipTap is fantastic.

Kudos on the docs: I encourage all open source projects to have a "Why Quill"[1] type page so I can quickly understand the philosophy of the project and how it aligns with my goals.

Even better when they also have "Why Not Quill" type content so you can quickly learn situations where it may not be the best fit. Every project makes tradeoffs and communicating those directly and clearly helps ensure users have a better experience.

[1] https://quilljs.com/docs/why-quill

This is not obvious from the link but I guess this is posted because of v2 released 3 days ago.

I have been working with Quill v1 not long ago and oh boy it is not good. Pain to make tables work with plugins and weird scrolling issues. Hopefully v2 fixes a lot but I'm sticking with TinyMCE as it is superior in term of features.

I like TinyMCE too and swapped out a Quill editor in its favor 6 months ago. But I was annoyed that you needed an API key to use their static JS in your site and disturbed that they rapidly shrunk free tier "use" of rendering an editor a few days back [0]. Out of curiosity, is that not an issue for you?

[0] https://news.ycombinator.com/item?id=40044544

You can self-host without an API key. Unfortunately they just switched to the GPL license, which may or may not work for you.

At first I couldn't figure out why Quill was being posted again, but now I see that a long awaited version 2.0 was released 3 days ago (see https://slab.com/blog/announcing-quill-2-0/ ). Thank you for the heads-up!

Yeah, I think this link and title would be a better entry point into the topic.

An old project I worked on was badly bitten by using Quill. IMO the issue is that they didn't vet the limitations of the editor well enough compared to the project's requirements but the way the documents went on about the extensibility of the library did feel like it was going to cause harm.

It's a great library provided you stick with its limitations (ie nothing too complicated with nested content). A "Why You Should NOT Use Quill?" type section on the website would probably go a long way

Ymmv, but I will never use Quill for a production project again.

We started using Quill 1.x around 2016, I initially liked it a lot. Jason proactively reached out to me when I first tweeted about considering Quill, and he was helpful when I had questions about the implementation.

Fast forward to 2019, we were running into some bugs, Quill 1.x was in maintenance mode, and 2.0 was effectively undocumented [1]. I reached out to Jason on Twitter [2] to get a sense for the timeline, and he basically told me to get lost. (Also his company Slab now blocks me on Twitter, though I imagine he just meant to block me from his personal account.)

Everyone has stressful days, especially when building a startup, so maybe that was just a bad time for me to reach out. My best guess is that they were focused on their own company (fair!) and barely had the time to maintain Quill for their own internal usage, let alone release it as open source. If that's the case, I wish they'd been upfront about it.

5 years later, Quill v2 is finally live. I'm glad to see Slab is still around, and I hope they have the bandwidth to publicly maintain Quill going forward. But if you're thinking about using this in production, I would push on what their commitment is to supporting it, and have a plan for if they stop responding.

(Jason, if you're on this thread, no hard feelings, water under the bridge. Feel free to ignore or chime in with your own perspective.)

[1] https://github.com/quilljs/quill/issues/2435

[2] https://twitter.com/nickbaum/status/1177240103925796865

No experience with Jason, but I have to say I’ve spent waaay too much time reading the Quill code trying to figure out how to patch around weird behavior. Which is frustrating because that’s what a RTE library (vs roll your own) is supposed to be. Not sure about v2, but the v1 code is just not very well architected.

There are basic features we’ve held off on implementing for fear that it will lead to a million little bugs.

The Delta format is the best thing about Quill by far. But I’d advise anyone looking for more than bold/underline/italics to use prose mirror.

Related. Others?

Announcing Quill 1.0 - https://news.ycombinator.com/item?id=12437345 - Sept 2016 (82 comments)

Quill – A cross browser rich text editor with an API - https://news.ycombinator.com/item?id=10446865 - Oct 2015 (47 comments)

Quill – An Open Source Rich Text Editor with an API - https://news.ycombinator.com/item?id=7716376 - May 2014 (60 comments)

This announcement is for version 2, which I think was released a few days ago.

I remember using https://github.com/facebook/lexical for a project a year ago and mostly things worked our of the box.

Any reason to prefer quill?

I have this project[1] that you can run from a GitHub code space, drop a some markdown files in, and push to GitHub pages. Basically trying to be as low code or no code as possible for a blog replacement.

Feedback from non-technical users is markdown is still annoying. Could quill be used in to edit as a wisywig with the output being markdown? That way the user could start the dev server and edit from there without having to touch the vscode file explorer.

[1] https://github.com/easy-astro-blog-creator/easy-astro-blog-c...

This looks perfect for a nascent guide I wanted to write, thank you for creating and sharing it!

Just a small bug report:

On the custom fonts demo, select one of the paragraphs and change its font. Then select that paragraph and another paragraph (which has the original font).

The font selection dropdown shows one of the paragraphs' fonts, and selecting it (to set both paragraphs to have that font) is not possible.

Slightly off topic question:

I want to publish blog on a domain I own -- I currently use Netlify+Hugo and NetlifyCMS but the editor experience (Richtext / Markdown + upload individual images) is very poor. There is no concept of saving drafts either; or ability to edit the hierarchy/URLs of posts after they are published.

I wish this tech got out of my way and simply let me focus on the content -- text, formatting, pasting images -- and one click save/publish.

Is there a simpler WYSIWYG editor that allows us to simply type, format and publish blog posts?

Life is too short to deal with high-maintenance / high-friction / brittle setups.

Trix is simple and easy to use for basic writing like a blog. It’s what Basecamp and HEY both use (it was built by 37signals and is the default in Rails)

https://trix-editor.org/

Try WordPress,

it’s not as bad as you think. Either the hosted version or on a php hosting which are cheap.

Quill or Slate is the question im asking myself these days.

I've used Quill before, but I have Editor.js, Slate, and Tiptap2 on my to-do list to check out, probably for similar reasons as yours.

Problem i got is that i have to be able to support html documents from old tinymce as well as being suitable for use with collaborative editing with something like ys.js. I might just have to bite the bullet and write a html to some other representation parser.

- https://discuss.prosemirror.net/t/prosemirror-for-live-editi... - https://prosemirror.net/examples/collab/#edit-1 - https://docs.yjs.dev/ecosystem/editor-bindings/prosemirror

I think lexical is the future

Relevant discussion on prosemirror forums https://discuss.prosemirror.net/t/differences-between-prosem...

Neither, use prosemirror.

Check out write-down.

Man, I really should revisit whether or not TinyMCE is still the best open source option

TinyMCE just changed their licensing from MIT to GPL, which may impact your decision. Fun fact - TinyMCE and CKEditor are both owned by the same holding company (Tiugo Technologies).

Curious - whats the real life implication of using GPL in this case?

I'm not a lawyer, but it is my understanding that using a GPL library on the frontend triggers the "distrubution to users" clause of GPL. It's a strong copyleft license, turning the entire frontend portion of the application GPL-licensed, which means any user can now request the source code for it. That's an IP risk for any commercial code, and many companies shy away from GPL. Backend use of GPL should be fine since there's no distrubution (AGPL "fixes" this, and such libraries should just be avoided in commercial settings).

I've actually heard this from lawyers before - essentially that if you use a JS library that is GPL and that code gets sent to the user via the browser, that its now considered "distribution".

Yeah, that's how I understand it. Applies to mobile code as well.

Quill is already over 10 years old! I've deployed it in multiple applications with a lot of success.

What a journey. Used it in production when I just needed arbitrary richtext documents in whole.

Customizing has been harder than anyone would like. I became an event sourcing aficionado, initially to support parchment’s delta format. It’s a solid idea, but also required (unless you do backend diffs, which I tried too).

My vanity test: I can’t believe how hard it still is to make a TODO checklist with indented levels. I try like every 18 months, surveying the plugin ecosystem.

It never works well, so far. Editor.js isn’t a one-stop rtf editor, but that was the first thing I noticed “just working.”

That said, I’m upgrading an old prototype to quill v2 today.

I really like to hear about experience with Editor.js. It looks like the most modern editor but seems complicated. Is this true?

What about CKeditor?

I much prefer to CKEditor to TinyMCE, but CKEditor has done a lot around licensing:

https://ckeditor.com/legal/ckeditor-oss-license/

They also modified their last CKEditor 4 open release to check the version and prompt the user to buy the LTS support.

The style would be beautiful if you could remove the borders

https://quilljs.com/playground/snow

Which one is easiest to implement hashtags (predefined list, associated colors) and @ mentions, which insert an <a> tag we can programmatically define. I can compromise on formatting features.

Is there a way to do nested bullet lists on mobile?

Looks promising but I do not see advanced customization examples like a widget or dialogs to insert custom html elements

This isn't clear to me, either, and it also seems from the docs like images, videos, and formulas are the only embeddable content types. However, elsewhere they say "Quill exposes its own document model, a powerful abstraction over the DOM, allowing for extension and customization. The upper limit on the formats and content Quill can support is unlimited. Users have already used it to add embedded slide decks, interactive checklists, and 3D models."

I wish they had a page covering what it takes to create a custom content type for the Quill DOM.

I developed a project that went deep into quill and its capabilities.

It's relatively straightforward to develop little widgets that can plug into the renderer. I did it for several classes of healthcare information, with "smart links" with embedded icons for things like appointments, medications, etc...

Even better, I was able to use vanilla web components for the markup that got embedded, and quill was able to use it just fine.

Was a huge success and worked great.

That's awesome news! Will have to investigate the renderer plugin model.

Is anyone aware of a word processor, text editor, whatever, that can handle HTML definition lists natively?

I'd really like a WYSIWYG editor that's designed to produce HTML documents. Not for web pages, but for documents. HTML has great prose features, but it seems like there's nothing in between Microsoft Word and "now you're building a website".

I guess one of the many Markdown editors is what you're looking for, then you just need to add a build step at the end. Unless you need any of the 'document' features of html that aren't in md, of course.

I agree some of the WYSIWYG markdown editors are the closest thing to what I want. I quite like Obsidian's editor for what it is, which does a pretty good job balancing rendering and letting you edit the actual markdown syntax.

But you're editing markdown, not HTML. And, for example, markdown doesn't have definition lists.

None of the visual markdown editors - the ones that render the text, eg. Typora - seem to support definition lists.

I don't think so. They were pretty specific: "I'd really like a WYSIWYG editor".

There are definitely Markdown editors that are WYSIWYG. In fact, md is a lot 'closer' to WYSIWYG than HTML is.

I think HTML is a pretty decent semantic structure for documents. The WYSIWYG is good for convenience, as is markdown text format.

Probably prosemirror. I don't know that it's built in by default, but it's well within prosemirror's model, so it shouldn't require a large investment to use.

That said, prosemirror's got a pretty steep learning curve.

Nothing is built into ProseMirror except for the text span.

For a definition list, you’d probably want two node types: a definition list, and a definition list item.

And then create toDOM and parseDOM methods to render and parse respectively the html you want.

Isn't the "flat markup" for inline elements in prosemirror kind of unnatural to work on html?

No download link, no mention of RTF. What "rich text editor" is that ?

I just merged a bigger pr last week where I used quillv1. My feature was fairly straight forward - I added a DSL so that users could build custom filter expressions for our trace-data analysis (automotive). I wanted to be tight with my business logic so I added my own lexer, parser and grammar evaluation with neat compiler hints and such - was a joy.

But when it came to the frontend it was infuriating. All I wanted was to hightlight the idex of my "compiler-error" but it was just a mess (every once in a while the text-cursor would jump around, sth I simply couldnt fix entirely). Ontop of that this [issue](https://github.com/quilljs/quill/issues/3806) exists - which forces us to upgrade once primeNg lets us. Anyone knows a good alternative? I am done with quill and would love to use something more stable.

Funny, I'm using quill.js + tribute for a project last week

Is there support for CSS classes in Quill v2?

Is Gutenberg a good alternative these days?

Can anyone compare to Lexical?

Lexical is more of a framework comparable to prosemirror. Learning curve is therefore higher, but IMO way more flexible and better designed than quill.

I'd honestly wondered if Quill had been abandoned!

Great to see a release!

What does “trusted by” actually means? A friend of mine works at Microsoft and they have never heard of this project.

• Is it that they know one person at Microsoft, any random employee among +221k, who uses the editor?

• Or is it that one of the developers works at Microsoft? (which, unless they are Satya Nadella, shouldn’t matter)

• Or is it that their (not so anonymous) telemetry tells them that someone is using the editor from a Microsoft-owned IP address?

• Or are the company logos simply there to mislead potential new users into thinking those companies actually trust the developer(s)?

Quill's an extremely well-known JavaScript rich text editing control, not an "app". Were you confused by the example at the top of the landing page? Your friend might not have heard of it, but I'd be more surprised if Microsoft didn't have significant projects using it because it's so well-established. And having large adopters of an open-source library featured on the landing page is not unusual.

I'm unclear as to what there deserves this kind of accusatory posturing.

> I'm unclear as to what there deserves this kind of accusatory posturing.

The project claims to be "trusted" by one of the largest tech companies in the US. Asking for a definition of "trusted" is extremely reasonable, and calling it "accusatory posturing" is not only inaccurate (the comment itself reads pretty neutrally), but also somewhat emotionally manipulative.

I don't think there was anything "accusatory", it's a legitimate question.

The comment reads like it’s being accusatory to me

Considering the size of Microsoft and the hundreds/thousands of teams building various things, I would guess that most well known open-source projects (like Quill) have been used by Microsoft at some point.

I would be very surprised if no team there had used Quill before.

But there's 0 proof that Microsoft has ever used this. Without evidence, these claims are useless

I believe the implication is that those companies uses the library in their products to power rich text editing

I’ve seen the company I work at show up on sites like this, when I’ve been explicitly told we can’t use whatever the thing is.

My guess is all it takes is 1 person inside Microsoft using it, or even 1 load of the library from a Microsoft owned IP.

No idea why you got downvoted. It's very reasonable to ask about this considering they didn't provide a definition of trusted and there's 0 proof anyone from MS used their product. False claims

For a second there I thought this was https://quill.art/

Reply on news.ycombinator.com