chrisdalke

Another option I've used, self-hosts vector tiles from an MBTiles database: https://github.com/akhenakh/kvtiles

Do you have a particular investment thesis/focus? I think many early-stage startups are hesitant to take investment vs. bootstrapping unless that investment also comes with some significant domain experience, scaling advice, connections, etc. and is from an investor who's well-aligned with their vision.

For HN specifically: maybe helpful to write up a high-level summary: Who are you, what's your focus, what kind of startups are you interested in helping? In this forum the anonymity-factor is higher (vs. twitter or other venues) so that information is probably more important here.

Linking to the primary source, Joe Barnard’s YouTube channel, BPSspace: https://m.youtube.com/watch?v=SH3lR2GLgT0. He’s got many videos on this. Would recommend, it’s been very cool to follow his progress over the years.

Spotify won’t even let you view an artist’s songs in order, which frustrates me because I like to listen to entire discographies in order while working. I spent one weekend building a free tool to generate chronological playlists for an artist - https://www.timelineify.com.

A decent number of people use this, which I think indicates some clear gaps in Spotify’s basic display features. Luckily the Spotify API is very generous with the data they expose about artists.

Thanks! A little more detail at https://www.chrisdalke.com/projects/mini-pinball-machine/, and the Raylib-based code is on GitHub: https://github.com/chrisdalke/mini-pinball-machine

Related, check out raylib, the rendering/game engine made by the same author!

I’ve used Raylib and raygui on a few RPi based projects and recommend it. It’s a really simple, intuitive way to get an OpenGL-based UI running. Good alternative to web-based UIs because it has similar simplicity but runs on devices with lower specs. I built this pinball machine with raylib: https://youtu.be/iiBn7FVzlcc

I've been following Derek closely as he's been working for the past >1 year to make MockRocket a reality. Congratulations on the launch!

I saw a discussion on HN earlier today criticizing SPAs for offloading data processing to the user's machine, benefiting Facebook/Google/big tech at the expense of the user experience. I think that conversation sometimes misses the positive side of React/client-based apps.

MockRocket and other web-based creative tools demonstrate the strength of the web as an application platform, making complex tools more accessible and easily distributed than before. Derek's built a pretty slick editing & rendering experience that allows more people access to video editing, without having a powerful machine with a GPU. I'm incredibly impressed with the rendering fidelity he's been able to build into MockRocket and excited to play around with it more.

Another great application is https://www.photopea.com/ -- a pretty powerful photoshop competitor that also works exclusively in the browser. Looking forward to seeing more people build and distribute creative/technical tools that more people have access to.

I’ve been fascinated with wooden boat-building for most of my software career, and that interest has driven me towards the marine space. It’s cool to read of other people who have a cross of software/computer skills and woodworking.

I feel that there is a deep similarity in the kind of problem-solving and craftsmanship that fine woodwork and software development both require. I’m convinced some of my (minimal) skills I developed working with my father in woodworking and metal work translates to a better ability to visualize complex systems.

Boat building is a great example of the phenomenon the OP talks about. I agree it’s particularly challenging, mostly because nothing is planar. You could spend a lifetime exploring the nuances of stitch-and-glue plywood construction, or bead and cove cedar strip, vacuum-bagging fiberglass, etc. In wooden boats, projects are measured in decades…

Big fan of using DigitalOcean droplets instead of AWS for hosting simple backend APIs. A few perks:

- Fixed pricing, very predictable billing

- No CPU throttling vs. EC2 instances

- Bandwidth included in the cost of the droplet

I've built a few projects with Netlify hosting a React frontend and DigitalOcean running an API server for the backend/database. A single DO droplet can scale far beyond what I've used it for, especially when combined with a SPA that offloads much of the processing to the users' browser.

It's already a reality -- See https://www.saildrone.com/technology/vehicles. A few companies building serious fleets for ocean data collection.

I clicked on the site just to check it out and ended up spending 15 minutes on your demo puzzle... well played.

Awesome job on all the user interactions. The dragging controls are pretty intuitive, it always did what I expected, and it's very smooth/seamless. The little touches like the autorotation are very helpful. I can imagine that having really great interactions differentiates this from many other similar sites so good job there.

As far as I know, you can seamlessly join Timescale tables with normal Postgres tables in a query. Timescale is activated on a per-table basis.

I think you are both agreeing -- He's saying that what's "bogus" is the idea that the market is perfectly efficient and there are no opportunities.

Okay, so I've looked more into the stats based on their marketing material, and I just don't see how the math works out...

They're claiming a 200kWh battery which seems a bit high for the size of the boat.

An equivalent battery system for marine environments that exists today is the Deep Blue BMW i3 (https://www.torqeedo.com/us/en-us/technology-and-environment...) which is a Torqeedo-BMW collaboration. That battery is 40kWh and 278kg.

Now, obviously there is a BMS and other parts that don't need to be replicated 5x, but scaling that up to a 200kWh pack gives 1390kg. That is absolutely ridiculous. Candela's entire C-7 boat weighs 1300kg, and it's only able to maintain a good range by utilizing hydroplaning extensively.

I'm not sure how a conventional planing hull that might weigh well over 2000kg with interior components, passengers, etc will get any reasonable range, and I suspect they don't show average range or loaded weight in the stats because it just doesn't look great compared to their competitors.

Absolutely agree, couldn't have said it better myself. It's a systems engineering problem and requires rethinking of nearly every component, from the propeller design to the architecture of motors and controls.

Torqeedo is an interesting example, they've been very successful with low-power electric outboards because in contrast with other companies they've invested heavily in rethinking propeller designs to be suitable for electric motors. The result looks nothing like a gas motor: They have huge, slow-rotating propellers that peak at <1200rpm and a high-speed brushless motor that gets geared down so it operates at the top of its RPM and efficiency range.

The power required to plane in a boat looks a bit like a normal curve -- Power consumption scales upwards exponentially with speed until you either 1) Get going fast enough to hydroplane and reach a high efficiency regime 2) Hit the peak power of the drivetrain and fail to accelerate further

A capacitor bank wouldn't really help here, since that power has to come from somewhere. The issue isn't that the batteries aren't capable of outputting that peak power -- Even small battery systems are more than capable of outputting the 50kw+ it might take to get on plane. The real issue is that doing that repeatedly requires you to scale up your battery capacity and weight, which is a positive feedback loop.

Boats are not just getting on/off plane once on a tank of gas, they're doing it 10-20 times depending on driving conditions, surroundings, no wake zones. This might happen 100s of times for watersports like wakeboarding. So a big part of the equation is finding a way to save weight elsewhere that can then be used to insert more batteries.

Unfortunately that's probably not going to happen anytime soon -- These electric boat startups are the real deal, but they all tend to rely on a lightweight hull to get ahead of the power density issues.

Most composite boat hulls are way too heavy to stick an electric motor in and expect reasonable performance.

Not a huge amount of information to comment on (They don't even show the fully-loaded weight of the boat which is probably the most important factor), but the focus on a lightweight hull seems like the right path and is similar to other startups in this field.

Electrification of boats is uniquely challenging because it's a delicately balanced optimization problem -- You need to keep the boat up on plane to get a reasonable range and compete in the same class as other recreational motorboats. Getting on plane requires a high peak power output, but that in turn requires a heavier battery pack, making it harder to get up to speed...

Battery energy density is not quite at the point where you can just stick a battery and electric motor in an existing boat hull and expect it to perform, so the theme of all these companies is using a custom lightweight hull or hydrofoil to reduce drag/weight enough to avoid the energy density problem. Another interesting startup is Candela (https://candela.com/) who is taking on the problem from a similar angle; they use hydrofoils and a complex controls system to keep the hull foiling which is very low drag.

That said, I'm not sure Arc has done their research on the market fit for aluminum hulls. Based on their concepts it will take a lot of work to convince the recreational boating scene to buy a $300k aluminum boat with an inboard motor; typically aluminum hulls are used for fishing and shallow waters, and in those use cases an inboard doesn't make sense -- it might be better equipped in that case with a jet drive but would hit power issues there.

For sure miniature gensets exist! They're often used on gas RC / autonomous airplanes to generate supplemental electricity. Here's one: https://www.youtube.com/watch?v=1yB49G756OI

I think the answer is that the miniaturization reduces efficiency, adds a ton of tiny, expensive parts/maintenance, and doesn't generate a lot of power compared to what you can store in a battery. So while the idea is cool, it's really only applicable in a few specialized use-cases like long-duration drone flight.

You've got to also compare it to other available power generation schemes -- A single 250W solar panel is far cheaper, simpler, and eco-friendly.

This is great! Love seeing hardware hacks like this on HN :)

Beyond just being a cool project, the detail in the README on Github is really amazing and I've bookmarked this as a great reference for releasing open-source hardware as a git repository. CAD files, a bill of materials, detailed usage instructions, etc.

One nice use case of COBS is to clearly delimit packets being sent over a serial stream. Framing packets with zero bytes is particularly convenient for small projects because it’s easy to implement and debug.

I’ve used COBS and MessagePack combined for some embedded sensor projects. My implementation for Arduino (also includes some other functionality on top to encode data with MessagePack): https://github.com/telemetryjet/telemetryjet-arduino-sdk

Most recently I built https://www.timelineify.com/ - Spotify messes with the order of albums & singles, making it difficult to hear an artist's works in the order they were released. Timelineify generates a chronological playlist of an artist's entire discography with a few clicks!

I was interested in playing around with the Spotify API and this was a narrowly-scoped problem I personally had and could solve in a single weekend. I'm happy with the results, about 500 users have logged in and created a playlist.

I reach for Spring Boot as my tool of choice for most APIs, but I feel your pain. There’s always a particular corner, the dreaded “configuration” folder, filled with a collection of random annotations, single-purpose beans, filter chain setup, etc.

I find that most of my business logic ends up being really compact and powerful, but the tradeoff is that one chunk of the project is really dense.

Definitely agree -- It's on par with the Smithsonian Air and Space Museum, and very fun to explore since it's divided between four hangers. Definitely worth a visit for anyone in that area.

This is exciting, the best part in my opinion would be the downsampling function:

> SELECT lttb_downsample(time, value, num_buckets=>500) FROM data;

I cannot tell you how many times I've reimplemented downsampling for time-series charts. This would be incredibly useful!

Cool! You should record a demo video and post it here so we can see it in action.

By the way I'd love to chat about what kind of data collection you're doing on Cygnus-X1. I've been building a platform for data collection, visualization, and sharing from experimental hardware (https://www.telemetryjet.com/).

Do you mind if I send you an email? Based on Cygnus and the Baja telemetry tool you have public on your Github I'd be really interested in getting your feedback about the project. It might be something that you'd find useful on your projects :)

Nice little program! Pretty fun playing around with the PID gains and other settings.

Saw you have an option to add torque noise. I tried adding some accelerometer noise (Just added some randomness to the input angle to doPID) and got some nice oscillations. Maybe this might affect your PID loop on a real rocket depending on how fast you're running the PID loop vs. polling the accelerometer/gyro?

uPlot is a cool library, I use it on a few projects :)

This is really exciting, congrats on the launch! Exciting to see engineers coming from college engineering teams like FSAE -- I was a member of a solar boat racing team, which is a very similar problem space and also relies on "stacking efficiencies" in your drivetrain system.

If you don't mind, I have a few questions about how you collect data and evaluate the performance of your motor quantitatively. Do you have any particular software stack for data collection? I see the motor communicates over CAN -- Is that just to send control signals, or do you also expose sensor readings like temperature, power input/output, etc? Lastly, did you need to write any custom software to collect/visualize motor performance and what does that look like?

The reason I ask is because I'm in the early stages of building a data collection & analysis platform for experimental hardware (https://www.telemetryjet.com/), targeted at small engineering teams or individuals. I don't want to focus too much on what I'm building (this is your thread!) so I'll just say in general I'm really interested in learning more about how small engineering teams like yours collect & utilize data in the engineering process.

Congrats again on the launch.

Hi HN!

I often listen to an artist's entire discography in order when coding. Spotify displays singles and albums separately, and often jumps to an artist's radio instead of playing through in order. I got annoyed, and built a tiny tool that fixes this!

Timelineify lets you search for any Spotify artist, and produces a playlist of their work in chronological order. Search for an artist, filter by albums and/or singles, and sort by chronological or reverse chronological order. Timelineify will add a playlist to your account, for example:

Artist Timeline: Gorillaz // Timelineify

Hope this helps someone!

Demo video if you don't want to login to Spotify: https://youtu.be/G603-ruJ3YQ