GPT 5.6 Sol comes out ahead of Kimi K3 on price/task (but not significantly so). You're probably thinking, "Why use Kimi K3? Isn't an open model supposed to beat the closed one on price?", but you need to consider that the closed models are completely hobbled when trying to do anything security-related. For my use-case, I can't risk getting pwned because I'm using a model that refuses to secure my app while there is now an open model that obliges to obliterate any app that isn't protected.
I'd say this is more of an interesting take on page transitions. I was expecting mouse wheel scroll to zoom, so I instinctively scrolled expecting some kind of zooming effect.
I remembered there was a website featured here on HN that had an interactive tour of the scale of the universe ranging from the very microscopic world (if I remember correctly I think it even went down to Planck length) all the way to the macroscopic (black holes, galaxies). I'd be interested in such a zooming library that achieves something like that.
Thanks for bringing to my attention. I haven't tried but I can spot some differences in philosophy:
- Musel Cloud doesn't embed Google Docs or Sheets, it has its own text engine; the word processor is built-in. There's no reliance on Google for docx, xlsx editing. Musel's built-in, natively on canvas. File formats such as docx, xlsx are compatible with any other software.
- Musel renders everything using a canvas (and WebGL), it doesn't use HTML at all. It is truly a whiteboard through and through. Everybody else has to rely on HTML for their rich text layout. Musel Cloud doesn't rely on third-party apps or popup a separate modal. Because everything is entirely natively rendered using the canvas, layers work really well, zooming works really fast, and performance is great. Items on Musel are more like Photoshop objects; they're all raster.
- has a built-in drive that syncs (but only supports Windows right now). No reliance on Google Drive or Dropbox.
Musel delivers on the promise of seamlessness. Works on files you already have on your devices.
Hi HN, I soft launched Musel Cloud, a collaborative whiteboard where you can drag and drop files (docx, xlsx, pdf) on the board and edit them in-place. These are all natively rendered on the canvas so they blend in seamlessly with other items. It has a built-in file explorer that syncs with your devices. Is it a whiteboard with a Drive or a Drive with a whiteboard? I'm not sure, but it greatly reduces context switching.
Speaking of context, I made an interesting discovery while working on Musel Cloud. Those who work with Claude Code, Codex, etc... will quickly realize that you'll spend the majority of the time wrestling with context management. With proper context, the LLM can be consistently reliable whereas prompting alone can be hit-or-miss. It turns out whiteboards are really good at managing context and interfacing with LLMs.
Whiteboards are a really crowded category; I didn't even intentionally set out to develop one. I was building a RAG app and needed a way to gather and lay out documents. To my surprise, there weren't any existing product where I can drag and drop forms, documents, contracts, books, etc.. onto a board and then edit them.
I'm working on brand new type of collaborative whiteboard that allows anyone (or team) to drag-n-drop items from their devices onto the board.
The problem I'm solving: On a team, people and their files are scattered everywhere.
Solution: A canvas that attempts to open (and edit) as many file types as possible (images, xlsx, pdf, docx, cad). This means you can have people and files on the same page.
It's the only whiteboard that can natively render docx and pdf so far; these can also be edited directly on the board without having to use dedicated software.
It has a built-in Drive where you can store/backup files that syncs across your devices.
There's a few widgets such as Kanban, sticky notes, cards.
And of course, there's agentic LLM (Gemini 3 Pro) that can take actions such as viewing the board, reading documents on the board, and editing items on the board. For example, you can tell it to read a pdf, then write a spec sheet (in docx), or create tickets on a kanban.
I'm launching a private beta next month if anyone is interested in testing it out and giving feedback.
I've just finished creating a Magic the Gathering rules engine, and now I'm currently training an LLM agent to play games against itself through reinforcement learning.
I struggled learning about Mamba's architecture but realized it's because I had some gaps in knowledge. In no particular order, they were:
- a refresher on differential equations
- legendre polynomials
- state spaced models; you need to grok the essence of
x' = Ax + Bu
y = Cx
- discretization of S4
- HiPPO matrix
- GPU architecture (SRAM, HBM)
Basically, transformers is an architecture that uses attention. Mamba is the same architecture that replaces attention with S4 - but this S4 is modified to overcome its shortcomings, allowing it to act like a CNN during training and an RNN during inference.
You're right, I completely forgot about Etherpad. But I don't consider Etherpad the same as Quilljs or ProseMirror. Etherpad is something you host (an app itself) while the other text editors are something you can build upon/integrate into your app. I didn't want to host an instance of Etherpad, I needed a library.
I wanted an editing experience similar to Medium.com text editor: consistent, well-formed html output and formatting. At the time, Quilljs was one of the best open-source editor that produced consistent output regardless of the browser.
Also I didn't want to store html, I wanted to store operations in the database.
I have an app similar to Figma[1] (basically real-time Sketch) and I needed just the real-time text editing feature.
I was able to get real-time collaborative text editing to work on one of my side projects by combining quill.js[1], ot.js[2], and firebase[3]. This was a few years ago when there weren't any fully open-source options. I'm in the process of switching out firebase for an elixir/phoenix backend[4]. I'm really glad for this article because it captures the exact feelings that I have.
I would love a good in-depth tutorial on operational transformation but so far haven't found one that is satisfactory (for a complete beginner). Maybe I've been spoiled with all the great javascript/react tutorials out there that I'm hoping someone could do the same for OT. A tutorial or series about OT with detailed instructions from concept to implementation is something I would pay for to learn.
I'm currently building a browser-based WYSIWYG to make book covers. I've tried SVG and Canvas but it's terrible for implementing multi-line text. Trying to add kerning, tracking was almost impossible.
Not sure if anyone's heard of Readymag[0], but I'm really impressed with their editor - and it has excellent typography tools and UI.
Anyway, what I've noticed is that now many people are abandoning their own websites & blogs in favor of a centralized service such as Medium, Instagram, etc.. I remember a few years ago there were alot of fashion bloggers and now they're all on Instagram, updating daily. People have lost interest in designing, building, maintaining their own sites because it's too much work for the average person. Not to mention traffic going to the individual sites are neglible compared to social media. When posting a photo on your own site gets 10 visitors, but that same photo garners hundreds of likes on somebody else's platform, then you're going to be spending time on that platform.