paulfitz

That Dropbox like system is literally all these worms need to propagate.

Thanks! Fair about styling :). You can bring your own stylesheet https://support.getgrist.com/self-managed/#how-do-i-customiz...

The enterprise features are self-hostable. Look at "your servers" on the pricing page for Grist. Individuals (and orgs with < $1 million in annual income) quality for free activation keys btw.

[grist employee here] Grist forms are open source and were used to keep the toilets clean at FOSDEM just a few days ago https://fosstodon.org/@grist/116001932837956733

Everything you see in our standard docker image is open source. Yes, you can enable and pay for enterprise features too.

Grist forms support uploads since 2025 https://github.com/gristlabs/grist-core/pull/1655

Since it is relevant here: support for uploads was code written by a French contributor, and reviewed by a developer working for the French gov (ANCT/DINUM) and a developer working for Grist Labs. Grist Labs has since maintained and improved on it. The forms feature itself was inspired by an integration built by Camille Legeron at ANCT.

I work at Grist, the "tableur collaboratif" (collaborative spreadsheet) listed on the La Suite homepage. We're in the interesting situation of being both a NYC-based company, and open source software the French gov has adopted and is helping to develop. Grist is mostly a node backend. So it is a complicated story. The key is having code the gov can review and trust and run it on sovereign infrastructure.

Grist https://www.getgrist.com/

A write-up of how the French gov uses it https://interoperable-europe.ec.europa.eu/collection/open-so...

I'm quietly adding "pull requests" for data to Grist https://github.com/gristlabs/grist-core/issues/1829 - been wanting to do this for a long time.

I tried doing this years ago as a stand-alone project and it was too much. I wrote a data diff/patch/merge tool called "daff" that worked okay. But I've always wanted to add this to a proper spreadsheet tool like Grist.

I really want people working on data projects to be able to work more like coders, with pull requests and reviews. Not all data projects are as curated as that, sometimes your data is just a big soup, but when it is curated, there should be a better workflow.

This is excellent, thank you!

Escaping is needed no matter what separators are used, but if a character from the astral plane is always present (like U+1F4A9 PILE OF POO) then you can be pretty sure the software is handling unicode well and isn't corrupting cells without you noticing.

So true about RFC4180. Admittedly this post kind of got out a little early, support for the format was slated for the first of next month...

The main problem is they are still in the basic multilingual plane, so U+1F4A9 PILE OF POO still has an edge there for tickling problems quickly.

Meant to link to this! Thanks, will update.

The overlap if you work 9 to 5 hours in that timezone has proven a bit of a drag for us.

Thanks for playing! :)

Grist Labs | Systems Engineer | Full-time | NYC OR REMOTE +/- 3hrs | https://getgrist.com

We're looking for someone to make our modern spreadsheet software run everywhere. To apply, there's a puzzle. Just do: docker run -it gristlabs/grist-twist and poke around. If the words battery correct horse staple mean something to you, you might have an advantage.

The heart of the software you'll be working with: https://github.com/gristlabs/grist-core/ More information about job and company: https://www.getgrist.com/job-systems-engineer/

The essential requirement for the job is comfort working with operating systems and containers, and having some back-end programming experience. People with a dev-ops background will do well. If you have strong open source experience, even better.

Have you seen people install your popular software in weird and wonderful places, and seen all the weird and wonderful problems that crop up? Ever wanted to really go all in on making an app or library that can go anywhere? If so, you're the kind of person we want to talk to. Our software can already be used everywhere: as a commercial SaaS, or as part of a government office suite, running on servers owned by enterprises and citizen self-hosters, desktops, air-gapped installations, compiled to pure in-browser javascript, rippling like a dream through the etheric plane (well, not this last one yet). But it isn't always easy, and that's where you come in!

A surprising number of people submitting "bread pickles tomato cheese" as the answer. No! That was just an example! (I was hungry)

The motivation for calling Grist a spreadsheet is that it has formulas, and cell values get updated automatically when something they depend on gets updated. Agree there is scope for misunderstanding here, maybe there's a better word? [Grist employee]

In Grist, reference columns let you do a lot of what you can do with a join https://support.getgrist.com/col-refs/ while still having spreadsheet-style immediate updates when underlying data changes.

Also, if you just want to do queries and don't care about instant updates, you can do any SQL you like including joins with a SQL endpoint or a custom widget https://twitter.com/getgrist/status/1710018836836077967

[Grist employee]

Snapshots are there, but you need to configure an s3-like store to enable them, e.g. MinIO (or S3 itself). [Grist employee]

In the few days since we wrote this post, Code Llama came out, including a flavor fine-tuned on 100B tokens of Python code - could be great for Grist, I look forward to re-running the benchmarks. This is an unexpected side benefit of using a common, popular language for Grist's formulas...

The exact LLM used in the experiment mentioned in this post was upstage-llama-2-70b-instruct-v2.ggmlv3.q2_K. Grist was configured to use it via llama-cpp-python and https://github.com/gristlabs/grist-core#ai-formula-assistant...

Neal Stephenson calls them earthsuits in Termination Shock.

Agree, I started paying attention to duckdb when it started allowing FROM before SELECT https://github.com/duckdb/duckdb/pull/5076