I got significant feedback from people wanting third See Sailing compatibility with third party logs so have spent the last month (or so) with test data from B&G Hercules, B&G Triton and Expedition to make sure it all works.

Seen here using a Hercules log, and you can click here for the interactive version (go on, give it a go!).

Aside from that there's now a management page for uploads, subscriptions and customisable stuff (username, names of boats)...

A fullscreen mode and the 'share sheet' now captures the time, location of the yacht and camera position. This is ideal for sharing insights into manoeuvres with crew, coaches or followers in general. Check out this gnarly gybe found in some expedition data!

So right now I'm using it and squashing bugs as they come up, and getting round to writing some documentation. It will be live and for sale really, really soon! The initial focus will be on getting the imported data to market, with hardware to follow. It just seems easier that way.

Progress is good!

A fun time was had persuading a renderer that objects larger than Wellington harbour actually exist. Our man who sent me some B&G data sailed out of Cowes, round the needles and off to Le Havre ... being not exactly a little day sail. None the less, the renderer is happy about large areas now and I've found the time an energy to display the core numbers about sailing in a convenient 'overlaid on reality' manner. Like this...

So the yellow line is the wind line (ie 90 degrees to true) and green is showing our VMG. White is boat speed and angle relative to true wind and the grey arrow shows the reported angle across the deck. This particular dataset also has rudder data, shown here as a blue streak and the current rudder angle (five degrees). There's also a comically large windex on top of the imaginary mast that you can watch flapping around in the breeze.

In the wider view you can see the angles in context such as...

From (approximately) https://app.see-sailing.com/b/incantare/2024-4-6-12-32-41?t=1712363372

So now I have a dilemma on whether to get a Triton version out of the door first or the hardware version. Either which way I still have to write manuals and stuff. But ... it's looking pretty healthy.

A correspondent casually mentioned how See Sailing might be a good solution for him, except that he already had a B&G Triton and (what with it being really expensive and all) would really like to use the data from that. Sometimes the hints arrive quietly and sometimes they speak out loud, so last week I put a Triton importer together and set about having a look at this data.

But there's a hitch. I've been rendering an actually relatively small harbour and the data in question was for an offshore race round a peninsula. My "first try" renderer was none too impressed....

Far too much detail in all the wrong places. So this week, at least, I'm going to need to implement some description of prioritisation and level of detail. As a VFX developer I spent quite some time with these things so know about about them. Unfortunately the #1 thing I know about them is they can become massive time sinks for those who are frankly too interested in the problem.

But! Data came in from a B&G Triton. This is good!

After some investigative and preparatory work in 2023, and lots of turning that into an actual product in 2024, "see sailing" is nearly at the point where it can go into production. A brief summary of where we're at:

• The electronics are finalised and have been running on a simulated network for quite some time. Firmware has been similarly tested.
• There will be a change to the physical housing for the electronics since feedback was very much in favour of a socket rather than a pigtail.
• Mobile phone application is running on both iOS and Android. There are some fiddly bits I want to change, but they're not as important as actually getting the thing to market so they'll have to wait.
• Backend authentication, users, interfacing with Stripe (who manage the subscriptions) are done. Some user facing facilities are still missing but, again, getting it to market is the push for now.
• Front end is working just fine, still. The representation for wind has been removed at the moment because more than one person mistook it for a rendering error. This, obviously, is high on the list.
• Oh, and a basic website and marketing infrastructure are up.

The big question now is: what to do about rudders?

For the investigative work rudder angle was assessed using an accelerometer and some 'harder than I thought' mathematics. I am, for now, unwilling to take the accelerometer work to market because they're (a) more electronics and (b) highly dependent on calibration. I had been assuming that "just" using an NMEA rudder angle sensor would be the easy way to go, but they are proving surprisingly difficult to obtain.

The other option is to include an analogue rudder sensor "in the box" but that would raise the initial purchase price considerably and require further electronics development - both things I could live without right now.

But, to a basic level, it looks like "June 2024" is achievable and should be good to go. I need to go order some components...

-Dave