Sailing Exodus
The argument, both sides

Open source
vs closed source.

What each one actually gives you, side by side. We run both on this boat, deliberately, and this is the reasoning.

What they are

In one line each

Open source

You can read the code, change it, run it, and pass it on. Not free as in nothing to pay — though it usually is — but free as in nobody can take it away from you.

Closed / proprietary

A sealed product you buy and use. Somebody else owns the internals, supports it, and decides its future. You are paying for it to work without you understanding it.

What each one is genuinely good at

Read down either column

Open source wins on

  • It cannot be discontinued. Marine electronics get end-of-lifed constantly — a five-year-old plotter stops getting chart updates, a cloud service shuts down, and working hardware becomes a brick because someone in an office decided. Open source has nobody who can make that call.
  • It's yours to fix. Doesn't work the way your boat needs? Change it. Try that with a sealed MFD.
  • It runs on whatever you have. A hundred-dollar Pi instead of a three-thousand-dollar display — and if the Pi dies you buy another and restore a backup.
  • Your data stays yours. No account, no subscription, no cloud. It lives on the boat.
  • Everything talks to everything. Open formats are how a Victron charger, a Chinese BMS, a Raymarine radar and a homemade sensor end up on one screen.
  • You can read how it works — so you can learn from it, and judge whether it's any good.

Closed source wins on

  • It works. Every time. First try. Power on and it is a chart plotter. No config file, no plugin to restart, no SD card that filled up. On a dark night in a rising sea that is worth real money.
  • Built for the environment. Sealed, sunlight readable, vibration and salt tested, designed to be worked with a wet glove on a pitching deck. A tablet in a case is not the same thing.
  • One system, tested as one system. Radar, plotter and autopilot were engineered together and somebody was paid to make sure they work. Our open stack is held together by us.
  • Charts and support. Properly licensed, properly updated, with somebody to telephone. Nobody is on call for an open source project at three in the morning.
  • It'll be here in ten years. Resale value, spares, and an installer in most harbours who has seen one before.
  • No time cost. You bought an answer, not a project. Sometimes that is exactly correct.

And what each one costs you

The honest other half

Open source costs you

  • Time. A great deal of it, especially early.
  • You are the support desk. When it breaks at 2am, that's you.
  • It needs maintaining. Updates, backups, the occasional thing that stops working for no clear reason.
  • Nothing is certified. No type approval, no insurer who has heard of it.
  • You have to care. If you don't enjoy this, it is a chore that never ends.

Closed source costs you

  • Money, obviously. And again for charts, and again for the upgrade.
  • The formats are secret. Charging for hardware is fair; keeping your own depth reading away from you is not.
  • It only shows their corner. One manufacturer's gear, one manufacturer's view of your boat.
  • You can't extend it. If it doesn't do the thing you need, it never will.
  • Somebody else decides when it dies.
Why it costs what it costs — the fair part. Marine is a tiny market next to phones, so development spreads over very few units. Everything must survive salt, vibration, lightning and years of UV. Certification is expensive. Stocking spares for a fifteen-year-old product is expensive. Somebody has to answer the phone. The part that isn't fair is keeping the data formats secret.

In practice: what open source buys you on one boat

The translator in the middle

A translator in the middle, that speaks all three and belongs to nobody.

On this boat that translator is Signal K, running on a Raspberry Pi. It reads 0183. It reads NMEA 2000. It listens on the ethernet. And it turns all of it into one common set of names that any program can ask for.

What that means in practice:

None of this requires throwing away your marine electronics. We run Raymarine gear and we are glad of it. The open-source layer sits alongside, reads the same wires, and removes the part where one company decides what your own data is allowed to do.

Signal K OpenPlotter How the networks work Who built all this →

So how we actually split it

Both, on purpose

The thing that keeps you off the rocks should be the most boring, most reliable, least interesting device aboard. Our open-source stack is better in almost every way we care about — and it is also the thing we are constantly changing, which is exactly what you don't want from the system you are navigating on.

JobRuns onWhy
Primary navigation Commercial MFD Must work with no thought, every time, wet and tired
Collision avoidance Commercial radar and AIS Same. Not a place to be experimenting.
Everything else Open source on a Pi Monitoring, logging, analysis, alarms, and every question the MFD can't answer

Note the "for now". Every month the open side does more. The radar already draws on our own screen. Charts, routing and the log are already better on our side. At some point the proprietary box becomes the backup rather than the primary — but that happens when it has earned it over a lot of sea miles, not because we got excited about a project.

The honest summary: run both. Let the commercial gear do the job it is genuinely excellent at, and use open source for everything it was never going to do. Anyone telling you to throw out your plotter is selling something, and so is anyone telling you a Pi can't be trusted with anything.

The data this boat gives away

The part more sailors should do, because it costs nothing

The boat is already measuring all of this. You are only choosing whether anybody else gets to see it.

Where it goesWhat we sendWhy it helps
MarineTraffic Our AIS receiver's feed — every vessel we can hear Coverage is built entirely from volunteer receivers. There is no satellite watching your creek. If nobody near you is feeding, that water is simply blank on everyone's screen.
AISHub The same AIS feed A sharing co-operative: feed it and you get the whole network's data back. Our own plotter sees targets far beyond our antenna because other boats are doing the same thing.
Windy Masthead wind speed and direction, pressure, temperature Models are only as good as what they're corrected against, and over water there is almost nothing to correct against. A real observation in the middle of a bay is worth far more than one more inland airport.
Weather Underground The boat's own weather station Same argument, different network — and it feeds forecasts a lot of people actually look at.
Navionics SonarChart Live Depth soundings along everywhere we go Builds detailed bottom contours from real tracks. Official survey data on a creek can be decades old; this is from this season.
Open depth and bathymetry projects The same soundings, to the open side So the data isn't only inside one company's product.

Why the sonar logs matter more than they sound

Depth is the one kind of chart information that only a boat can collect.

Satellites can map a coastline. Nothing from orbit can tell you there's three feet at the mouth of that creek at low water. The only way that gets known is a boat going over it with a working sounder.

Official hydrographic surveys are expensive, slow, and concentrated on commercial shipping channels. Away from those, a lot of charted depths are extremely old and the bottom has moved since. Anywhere shoaling and silting — which is most of the Chesapeake — the chart and the water drift apart every year.

Crowd-sourced bathymetry closes that gap. Thousands of boats each contributing their own track, averaged together, builds a picture of the bottom nobody could afford to survey. The track you sailed last Sunday becomes a contour line somebody else trusts next spring.

It costs you nothing. The sounder is running anyway, the plotter is recording anyway, and the upload happens when you get back to WiFi. It is the single easiest genuinely useful thing a cruising boat can do, and almost nobody switches it on.

A fair caveat. Crowd-sourced depth is a guide, not a survey. Soundings depend on the transducer offset being right, the tide correction being right, and the boat going where the track says. Use it to find the deep water worth investigating — not to decide you'll clear a bar. We contribute to it and we still navigate on the official chart.

One more point worth making: contributing to the commercial networks is not a betrayal of the open-source argument. Navionics gets our soundings and so does the open side. The data exists in both places, which is better than it existing in neither. Withholding it to make a point helps nobody standing off a shoaling entrance in a falling tide.

How we're contributing

And what you could do, without writing any code

What we put back

  • The radar and fishfinder protocol notes — what we worked out, including the parts we're unsure about.
  • Source code for the displays, the BMS readers, the power grid page and the ship's log, going up on GitHub as each write-up is finished.
  • Boat observations fed back to the public weather, AIS and depth networks.
  • These write-ups, with the mistakes left in.

The data we give away →

What you could do

  • Write down what you worked out. A forum post about a fault you fixed is worth more than you think. Someone will find it at 2am.
  • Report bugs properly — what you did, what happened, what you expected.
  • Fix the documentation. It was written by someone who already knew how it worked.
  • Send real data. A developer who doesn't own your instrument can't support it. A capture file is gold.
  • Sponsor somebody. Here's where to start.
  • Tell people it exists. Most sailors have no idea any of this is possible.

All the write-ups → How the networks work Every tool we use