My biggest worry is your supply chain. If you're mostly using AI, would love to see you build your own library of functions and drop as many direct and transient dependencies as possible.
Nice work!
Not to bag on Nextcloud but I do think that this is _much_ easier to understand and maintain, it's standard Go/React/Sqlite, nothing esoteric so if you did need to maintain it I'm sure anyone could pick it up. Of course I don't really know Php so that def flavors my opinions here.
I have an example of a one-shot TODO app: https://github.com/tinycld/todo just to demonstrate what's possible.
This does look easier than Nextcloud however I'm not sure thats the tech-stack. Generally the new project is easier/simpler because it doesn't have so many years of scope-creep, work-arounds, half-imolemented-then-abandond features.
We support postmark right now (it has spam filtering), mailgun in the near future. Plus probably a pure smtp server for those who _really_ want to do it all themselves.
How is the Microsoft Office compatibility managed by these tools? There is a popular SDK providing the compatibility? I can't imagine everyone reimplementing the full compatibility layer
Text uses https://github.com/ZeroHawkeye/wordZero to read/write the docx, parse it into a y-prosemirror and served with yjs to clients using https://github.com/skyterra/y-crdt
And Calc uses https://github.com/xuri/excelize using the same techniques but uses plain Yjs Y.Map/Y.Array using a sparsely populated table
So how is MS Office compatibility manged by these tools? It depends on who you ask.
I may be mis-understanding what you experienced though, If you'd like to provide more details and/or some screenshots please file a issue https://github.com/tinycld/calc/issues
Asking for a friend, who also enjoys building projects with LLMs, but publishing and supporting them not so much.
I'd love for people to write whatever crazy packages they can think of for it so it has a rich ecosystem. It has the ability for admins to add a git or NPM reference into the packages list and install packages on the fly like wordpress supports.
As for hosting as a service, maybe someday. I also own the tinycld.com so who knows.
Personally I have several small (tiny really) companies where employees wear a few hats in them, this setup allows me to add people to org (and email domain) A & B while not getting C.
It is _not_ setup at all for billing or anything like that though, but that would be a easy enough feature to add
One question, I could not understand if a mail provider is needed or it handles on its own?
We also have https://tinycld.org/llms.txt for writing packages and periodically test it's usefulness by one-shoting a simple TODO app: https://github.com/tinycld/todo
I was wondering what your reason is of developing it instead of just use nextcloud for your company or for mail, calendar, drive, docs and sheets? I mean at least with nextcloud you'd be up and running and have a battle tested solution.
I totally get the do it yourself approach and it's reason enough. And combining all the apps in one web interface is really solved well. Kudos. But I just wonder what your technical and maybe other reasons were.
* It's UI felt dated - which I agree is not an entirely good reason to not use it, but certain stakeholders in the org that will be using it felt it wouldn't be acceptable.
* Collabera Office, which underpins their doc solution is kinda weird how it runs, everything renders on the server and then is displayed on the clients. Not mobile friendly at all.
I did really like the ecosystem though and did consider just writing the plugins I needed for it. But then that went down a rabbit hole in itself with Php which I don't have fond memories of.Ultimately it was a case of "be the change you want to see in the world". This is my opinionated answer to what an online suite of apps should work like for better or worse. I'm in a good spot where I have a small group of users who I'm close to that can stress test it out and yell at me when stuff breaks.
I wrote TinyCld because Google in their infinite wisdom decided to cancel my 20 year old free Google Apps suite for "commercial usage". Which, to be fair it totaly was, and I obviously shouldn't complain about 20+ years of free usage.
But _still_, that wasn't the terms I signed up under, and nothing is more irratating than a good old fashioned rug pull. How hard could it be, right?
TinyCld is two things, a self contained workspace like the one we're all familiar with (mail/cal/contacts/drive/text/calc), as well as an easy to extend platform that brings all the batteries you'd need to write realtime web+native apps.
And yes, I used AI to write a lot of this. (really, ~200k loc by one author is a pretty good tell) Some may hate on that and thats fine, I get it. But I did look at each commit and there is a lot of bugfixing and tweaking features back and forth that went into it. If nothing else, I've got some good stories to tell here and have learned a lot.
Stack: Expo (React Native + web) on the frontend, PocketBase + Go on the backend. Standard protocols (IMAP/SMTP/CalDAV/CardDAV/WebDAV) so native clients work as well.
I'm currently using it myself with a few friends but has not been widely used beyond that.
Demo (no signup): https://tinycld.org/ and click the big Demo button.
One-line Docker install: https://tinycld.org/docs/installation Build a package in 10 min: https://tinycld.org/docs/creating-a-package iOS app: https://apps.apple.com/app/tinycld/id6762420971 Repo: https://github.com/tinycld
I'm very interested in any feedback anyone can offer. I've got a looong list of add-on packages I'm considering, suggestions welcomed!