Hacker News Reader

Comment by gavanm | Hacker News Reader

The high level issue here is around data ownership, and device ownership.

I think the nearest car analogy is ODBII ports and data access - ANT+ is a wireless communication protocol, mostly for reading statistics (I think it can also be used for issuing commands).

Hammerhead had a license to access the privately configured Shimano data - and then they were purchased by SRAM (who are Shimano Bike division's main competitor).

As a result, Shimano is (for now) limiting a competitors ability to see the data produced by Shimano components.

This feels very petty to me - most of the data is essentially going to be "which gear is the front/rear in", and "what shifting pattern do you want to use" - though it might also extend to preventing future interoperability (like preventing competing wireless shifting levers triggering the other manufacturers components) - which would be a loss for consumers.

View on HN

Well, if it feels petty, it's because it is.

The "data" You're talking about is basically just battery level and which gear are You in.

Just for context, SRAM (the second biggest component producer and the new owner of Hammerhead) publishes the same data in plain ANT+ format. It's basically an open standard over which sensors broadcast their data. Anyone who is interested can read it, no license agreements necessary.

AFAIK Shimano has decided to encrypt their packages and the license is for the encryption keys / algorithms.

so one of the future firmware updates will just be to rotate the encryption keys out of spite to make sure it doesn't work anymore.

What a time to be alive /s

That would be difficult given the very limited compute power on the sensor devices involved.

Technically you do have to agree to the ANT+ adopter license agreement in order to access the specifications. But it's free and available to everyone.

https://www.thisisant.com/business/go-ant/levels-and-benefit...

I have a Hammerhead 2 and Di2. The most commonly used functionality for me was using the buttons on the brake shifter hoods to control the screens on the HH. It meant you could control the device without moving your hands and I use it all the time.

Having access to shifter position and Di2 battery levels was also super useful. This move really pisses me off!

Guess I will not be updating the firmware of my HH2 for a while at least.

yeah, exactly what I was thinking when I saw the news. I've been specing a groupset update on my current bike just last week and although the primary factor has been component availability (the only Shimano groupsets I could find were on used market), I'm really glad I went for SRAM AXS (though their wireless protocol does raise some hair as well, need to look into it a bit more when I have a chance).

Lack of interoperability in any system is a bad thing for consumers. However in this case you could argue that Shimano are right. It's fine to allow interoperability without the data. A KMC chain will work with a Shimano derailleur. Shimano should have no problem with that. But Shimano have to assume that SRAM would now get all the aggregate data on the operation of thousands of their Di2 systems. So SRAM could see how much faster or slower the Di2 gear shifts are. How many times the user had to make the same shift twice maybe because of failure. If a user tends not to use the largest sprocket when putting in a lot of effort then maybe it means their gears aren't tuned correctly etc. So it's competitive intelligence that SRAM can use to out compete. When Shimano allowed Di2 interoperability their intention was that it would be for the benefit of the consumer, not their direct competitor.

yeah, that used to be true up to 11 speed groupsets. Things are changing however. Take a look into the 12 speeds now. Good luck finding a non-shimano 12 speed chain that works on Shimano derailleurs / chainrings.

As for the part about speed and quality info, sorry, but that's "not even wrong". There's absolutely no way You could get this kind of info from Di2 ANT messages.

All You have is things like "the gear has changed" and "battery level is now X", not "when" has the user pressed the button or how many microsteps has the derailleur done at which speed. That goes through their own wires (I think it's basically a version of CAN).

I think GP wasn't referring to use a groupset with mixed SRAM/Shimano components, though

It shouldn’t be up to Shimano at all. You buy a deceive and use it, it’s your device, your data, do whatever you want with it.

If a person attempted this kind of control over another person they would be called a manipulative psychopath honestly.

It's not that they want to control the person. They want to control what is sent to the competitor. There's a phrase "good fences make good neighbours". This is the history of industrial competition and cooperation. So we want to have cooperation and sharing which is what advances prosperity for everyone. But as soon as you share stuff people can abuse it. We see this with social network data. We see this with AWS abuse of open source. The practical way to have cooperation is to have sharing over which you have some control otherwise people will stop cooperating. So Shimano need some way of saying: We want to share this data with the user and with third party software companies. But we do not want to share this data with a competitor. At the moment there is no way to do this so they just do not share at all (or at least not with that "third party" software company). Shimano probably want to cooperate but they also want to be able to compete fairly. Right now it's not fair competition because SRAM are not sharing SRAM aggregate data with Shimano.

> It's not that they want to control the person. They want to control what is sent to the competitor.

Distinction without a difference. Shimano is using technical and legal means to prevent owners of their devices from using their property as they wish. And we have bought into the confusion that just because someone manufactured an item, they have some sort of legitimate claim to how it is used even after they sell it.

> There's a phrase "good fences make good neighbours".

I couldn't agree more. Which is why it angers me that Shimano is moving its fence, expanding into consumer's back yards. I don't want every company that sells me something to then try and control how I use that something.

They are attempting to control their customers.

When someone says you’re free to do whatever you wish so long as they allow it, we’ll then you aren’t free at all.

What’s next? Do you look forward to a future where a Mercedes will actively forbid you to drive to a Tesla shop?

> So Shimano need some way of saying: We want to share this data with the user and with third party software companies. But we do not want to share this data with a competitor.

The bike is mine, the components are mine- I bought them, why does shimano need any way of saying where I can put my data produced by my devices?

What's next, is a fridge company going go turn off the fridge if I pup non-approved food in it, will an oven refuse to heat an unauthorised meal? Is a smart lock going to refuse to open the door if I bring in more people than my rental contract allows?

We are on a high-way to a feudal kindom and tyranny.

I guess that SRAM can afford to buy a dozen of Shimano gear shifts, put them on a bench and measure any parameter they care even without accessing the software. Real world data could be interesting but are they really so useful?

"A KMC chain will work with a Shimano derailleur. Shimano should have no problem with that."

IMO if Shimano could DRM their chains and other consumables they would.

they already put a gazillion warnings everywhere only to use Shimano chains.

They could. They tried to make a metric chain (2cm spacing rather than 1 inch) but it didn't catch on. Campy has kind of done that with their weird pull ratios right? Or does that only make sense if Campy was second to some drivetrain size (I'm not sure who was first or second to each cassette size).

Reply on news.ycombinator.com