Why not on a shared drive like dropbox?

Concurrent editing wouldn't work, though.

Beyond "not working", concurrent editing is likely to silently lose data, or corrupt data, or possibly just fill up your drive with lots of "conflicted copies" that you have to manually reconcile.

When I read about this project, it doesn’t seem like concurrency is much of a concern. I think it’s _supposed_ to be treated like sharing a document with someone else.

If a capsule is a recipe, like in one of the examples, that’s pretty cool. If a capsule is a recipe _manager_, then I think it would get messy.

Fun little experiment with building a PCB design agent - I didn't think the tech was here already but it definitely seems to be!

TIL about GPU databases. I did a bit if research but only found this one: https://github.com/bakks/virginian

What are the advantages of a GPU database?

TIL too, but probably what GPUs are good at: optimizing for throughput while hiding latency. Probably better suited for OLAP than OLTP.

GPUs win when a query can be decomposed into thousands of identical operations on independent data which can be run in parallel, which corresponds well to OLAP database workloads (but not OLTP ones), although in practice it really only corresponds to very large OLAP workloads with multiple chained compute-intensive operations (which make it niche in practice).

Analytical queries are dominated by scans, filters, aggregations, and joins over columnar data. Something like `SELECT region, SUM(revenue) FROM sales WHERE year = 2025 GROUP BY region` over a billion rows is very parallelisable: every row gets the same predicate and the same arithmetic. A GPU can throw tens of thousands of threads at that (memory bandwidth is the other half of the story - an H100 has ~3 TB/s of HBM bandwidth versus a few hundred GB/s for a CPU socket, and scans are bandwidth-bound workloads).

What provider are you using? I have it a shot through open router and saw some weird half formed words coming through occasionally, would love to switch over and give it a proper go

Direct API

Thank you!

What issues did you have with macinacloud? In was considering using that to build and sign some dmgs and test my app and would love to know what your experience was like

Ok, I actually tried MacinCloud again and had much better luck. Instead of the pay as you go plan which you need to spend $30 minimum on, I got the $60 dedicated server which allows me to install software, and in turn I was able to test my software without buying a Mac, yay!

Hah funny I ended up giving it a go and ended up having a lot of trouble - connection issues, display issues, etc. They tried rebooting the machine but no luck.

Just bough a used mac today!

Boxcryptor used to be amazing for this till they got acquired by Dropbox. Last I checked cryptomator couple years ago it still hand random file corruption risk, is that still true?

For now I just use hetzner storage share.

Put together a workflow for using AI to speed up product validation—research, branding, assets. Haven't tested it properly yet, writing it up so I don't forget.

The idea is I use AI generated assets to throw up a landing page and then see what actually performs well.

Are there other people doing the same thing or maybe even tools being built to do this all in one platform?

Give plasticity a shot - not FOSS but pretty decent CAD and works on linux

I know you are looking at this for education but I've actually ordered two to toss in my travel bag for when I visit factories to handle manufacturing bringup.

Great tool. Thank you.

This is amazing. I love that this exists. I'm surprised how negative some of the comments are.

I skipped to a random spot, and lasted about 5 seconds. Two drawings of dicks, 50% scribbles, and a couple frames of just the word PENIS. Yes, this is truly a noteworthy effort.

I’m as equally curious as you, just on the other side of the horseshoe.

Same here. If you skip to a random part of the movie, and move frame by frame, it becomes a game of "spot the penis". Just one or two can be "tolerated", but we're far from that here, so ... My opinion is that it does not really belong in HN

> My opinion is that it does not really belong in HN

It's a fascinating tech/social experiment. It belongs on HN because it was done and the outcome is what it is.

If you actually want to watch the Bee Movie then watch the original. I doubt generating a watchable alternative was behind anyone's motivation in this.

Because of a penis drawing? Are we in preschool? There have been those ever since caveman times.

I think the point is we're not in preschool, nor are we cave(wo)men, so we don't need the penis drawings.

A penis drawing isn't something that's needed, it's something that happens. Feel free to rail against them in a Kanute-like manner, but learning to cock an eyebrow and say "yeah, that figures" can be self-preserving.

For what it's worth I too prefer the plausible deniability of Georgia O'Keefe's outrageous lies. It's a more sensible chuckle.

No-one's railing against anything.

I was hoping to introduce my old “donging through the millennia” thesis work and ride that hobby horse around the room a little :(

Did the caveman Bee Movie have penises in it? Because then you’d have a really good point.

I'm sure this one had an absurdly short TTP (time to penis).

>I'm surprised how negative some of the comments are.

It's literally unwatchable - It's 85% scribbles and "jokes."

I can understand the negativity because this only appeals to specific sensibilities.

I admit that it’s low brow but it hits my fascination the exact right way. To me there is something special about so many people being organized around something so mundane.