If Di2 and mechanical were the same price..

[Smokin Joe]

Software is no different to any other manufactured product - ultimately there's a tension between cost, reliability and complexity. To say all software is CRAP is simplistic and misleading. Compared to almost any physical product it's certainly complex and likely to contain flaws, but those flaws (a) will often not be critical to the function and (b) can usually be corrected. If the flaws are critical to the function then that's a failure of testing, which is no different to supplying a bike with a crap seatpost clamp (Canyon) or dodgy hubs (Zipp).

The code to respond to a small set of bluetooth messages and move a pair of stepper motors isn't especially complex when compared with most software systems, although realtime "things" code has its own challenges. I'd suggest that the more likely causes of failure of an electronic system would be the hardware; in order of most to least likely to break, the battery, stepper motors and then changer switches.

So, taking my complaint earlier about brifters and using anecdote as evidence , I have two broken example brifters in my parts bin - Ultegra and Dura Ace. The Ultegra seems to have some snapped internal plastic and the Dura-Ace I think has lost a spring. I also have a "repaired" Ultegra where a return spring had popped out but I was able to reseat after a good deal of swearing. Bear in mind these are regarded as not repairable by Shimano. I'd recommend looking inside a brifter if you get the chance. Alongside some mechanical complexity, a lot of the plastic is no better than decent toy grade. Even at the top end, they're built to a price. They might be more reliable than a pair of microswitches, but then again they might not.

So the answer to that is to drop the indexing and go back to downtube shifters, which is just fine and no one is stopping you. You'll be making a choice to swap convenience (indexing) and I'd argue safety (keeping both hands on the bars while shifting) for simplicity and probably cost. Or go fixed. Even cheaper, more reliable and less complexity, but possibly less functionality, especially if you've got "built to a price" knees like mine.

My long-winded argument here is that electronic groupsets aren't a standalone cycling revolution/threat to humanity, but are just another point on the cost vs function spectrum of cycling. I think they get criticised unfairly and without evidence on reliability, where mechanical groupsets get a free pass in comparison. I think they have functional benefits over mechanical systems. I completely accept that they're expensive and the benefits may not justify the price to many people.

Damn good post, that man.

 

[I like Skol]

Damn good post, that man.

If only it was that simple.....

Trouble is, the product and software developers get cocky and start building in things like auto learning functions and trim adjustment. Then they think it might be good to bias shifting according to a riders current shifting patterns and predicting future gear selections. Then to top it all off they put in some diagnostics, battery monitoring and a sprinkling of limp mode options and suddenly counts cell B28 gets corrupted and the system stops talking when you are halfway up Wrynose Pass on a Sunday morning and the nearest Halfords that is open doesn't know how to work on Di2, but they can sell you a Clarke's gear cable for £1.99.......

For all day, everyday leisure cycling mechanical shifting just seems to make more sense. I fix machines for a living.

 

[Mike_P]

As a half way why does no one make a gear sensor for mechanicals or at least a rear gear sensor so mechanical users can have a Garmin or whatever give a read out.

 

[Alex321]

As a software engineer and system administrator in a previous life, it's a matter of principle for me. Bicycles don't need software, so I won't have ANY of that CRAP on one of mine. It's ok for speedometers, and yes, I have a 'smart' phone, and appreciate modern medicine, etc.

I don't care who else uses software on their bicycles, btw. I just won't, myself, because I know a lot about........software.

That "previous life" must have been an awful long time ago.

Very little modern software in such devices is remotely "CRAP".

Faults in Di2 are not going to be in the software, as it really isn't complex at all. Faults will be in the physical components.

 

[steveindenmark]

I have a lot of mechanical geared bikes. I do nothing but keep them clean and lubricated and they have worked faultlessly for thousands of kms. My least used bike is my full carbon Koga Solicio with Di2. Di2 is great. But it is only a luxury. Its great to whizz up and down the gears hearing the whirring sound changing the gears faultlessly. But if it goes wrong there is more chance you will need a visit to a bike mechanic. It a lovely thing to have. But certainly not a necessity. I would go mechanical if I had to choose between the two

 

[Fredo76]

Di2 may have better software than the Mars Rover, it's just that I won't be dealing with it, that's all.

The Venn diagram with the bubbles "Software Issues" and "Fred's Bicycles" will simply always have the empty set in their intersection.

My opinion of the software industry and modern life in general need not hurt anyone's butt. I didn't mean anybody's bicycle was crap, for the record.

 

[Bollo]

As a half way why does no one make a gear sensor for mechanicals or at least a rear gear sensor so mechanical users can have a Garmin or whatever give a read out.

There were a couple of editions of Shimamo brifters (early-mid 2000?) where you could attach their cycle computer via wire and it would tell you the selected gear. I bought one because as I’ve said I like the cycle tech. It was rubbish.

Since then I can’t say I’ve seen anything similar but for the right price it could find a market.

 

[ianrauk]

There were a couple of editions of Shimamo brifters (early-mid 2000?)

Called Flightdeck iirc

 

[Mike_P]

There were a couple of editions of Shimamo brifters (early-mid 2000?) where you could attach their cycle computer via wire and it would tell you the selected gear. I bought one because as I’ve said I like the cycle tech. It was rubbish.

Since then I can’t say I’ve seen anything similar but for the right price it could find a market.

I would have thought today a sensor on the rear derailleur and a magnet, say on the hanger, would suffice: the sensor detecting how far from the magnet it is and hence equating to what gear that derailleur is in.

 

[Bollo]

If only it was that simple.....

Trouble is, the product and software developers get cocky and start building in things like auto learning functions and trim adjustment. Then they think it might be good to bias shifting according to a riders current shifting patterns and predicting future gear selections. Then to top it all off they put in some diagnostics, battery monitoring and a sprinkling of limp mode options and suddenly counts cell B28 gets corrupted and the system stops talking when you are halfway up Wrynose Pass on a Sunday morning and the nearest Halfords that is open doesn't know how to work on Di2, but they can sell you a Clarke's gear cable for £1.99.......

For all day, everyday leisure cycling mechanical shifting just seems to make more sense. I fix machines for a living.

My bike was hacked by North Korea and the rear derailleur is now mining bitcoin!

Hate to break it to you, but the trim adjustment and battery diagnostics are already in there. On the SRAM etap I’ve also set my front derailleur to semi-auto, so I just click up and down and the system worries about the gear combo. I wouldn’t use this setting in something like a crit, but for daily riding or time trialling it‘s great.

Seeing I’m probably coming across as the thread electronic group set fanboy, I’ll share a major gripe, which is around sustainability and open standards. As far as I’m aware there’s no agreed standard for communication between electric cycle gear devices, which means that the manufacturers can change protocols, rendering older devices obsolete and prevents third parties from developing within the ecosystem. In fact, it looks like Shimamo are going the other way and have recently forced Hammerhead to remove Di2 functionality from their Karoo computer for commercial rather than technical reasons.

 

[wafter]

Faster, more precise gear shifts are important to those who ride competitively, whether that's with a number on their back, in a sportive or just out for a fast ride with their mates.

You don't have to have it, mechanical will still be around long after all here have gone. Those who buy it are not doing so because they are gullible or have more money than sense, but because it suits their requirements.

True, however to me it seems like a disproportionate tradeoff between a minute improvement in performance against significant cost, complexity and potential reduced reliability, serviceability and longevity. I can see the benefits, I just don't think they justify the drawbacks.

I'm not against what I consider to be real progress; and indeed my gravel bike has many relatively modern appointments - thru-axles, flat-mount hydro brakes, the now ubiquitous STI shifters.. however these seem like tangible, future-proof, value-for-money improvements.

Plus as I've said before, IMO adding electronics to an otherwise wholly mechanical system seems like madness tbh

 

[I like Skol]

I was thinking about this as I rode my filth encrusted commuter bike home from work in the rain last night.
It gets bare minimum maintenance to keep it running sweet. Gear cables replaced maybe every 4-5,000 miles then left alone.
The 10spd Tiagra components do a sterling job of shifting reliably despite the neglect and I was listening to the instant, slick, near silent gear changes in response to each press of the shifter. Indeed, the shifts are sometimes so slick and quiet that I sometimes couldn't be sure the change had happened after the click of the shifter so would click down again just to make sure.

 

[freiston]

I was thinking about this as I rode my filth encrusted commuter bike home from work in the rain last night.
It gets bare minimum maintenance to keep it running sweet. Gear cables replaced maybe every 4-5,000 miles then left alone.
The 10spd Tiagra components do a sterling job of shifting reliably despite the neglect and I was listening to the instant, slick, near silent gear changes in response to each press of the shifter. Indeed, the shifts are sometimes so slick and quiet that I sometimes couldn't be sure the change had happened after the click of the shifter so would click down again just to make sure.

I sometimes get this with my 3x9 bar-end friction shifting too (though at other times I can make it the opposite of instant and slick).

 

[freiston]

Without considering cost of Di2 (or any other electronic system), I'm not sure that what is available would meet my requirements - in a similar way that I struggle to find cable shifting systems that meet my requirements.

I have never used brifters and currently have a 3x9 system with bar-end friction shifting. I'm a bit strapped for cash at the moment but there is an outside chance that in the next year or so, I could have the resources to build my "dream bike" (with some budgetary constraints).

I have been thinking in terms of a 3x10 but have also been exploring 2x set-ups through https://www.gear-calculator.com. Whilst doing this, I have wondered if an electronic shifting system could handle front and rear mechs together to let me change through the whole gear range from lowest to highest in order, as opposed to having to change front and rear independently. Can they do this? Either way, I'm not convinced that it would "solve a problem" (my thinking was that I might be able to forego the middle ring and use the cross-over of ratios between small & big chainrings to replace it). I've never seen an electronic triple front shifter on the 'net or elsewhere.

Looking at the ranges I want, 20T or 22T at the front (46/34/24), 22T to 25T at the back (11-36), I can't see anything in the GRX range, electronic or cable, that is rated to handle that. A (MTB) Deore XT RD is rated to handle it.

But a Deore XT 10 speed will not work with any drop bar brifter or other drop bar indexed shifter - so, Di2 aside, even with current cable systems, Shimano don't provide, as an indexed system, the set-up that I want.

So, without ruling out Di2 due to cost, not only am I ruling it out due to specs, I'm also ruling out the indexed cable systems due to specs and instead, I'm looking at friction shifting options to get what I want.

I don't know much about how Di2 (or other brands) can be configured but I have seen Di2 systems specced as either 11 or 12 speed - can any Di2 system be configured to work as either, for example, 9, 10, 11 or 12 speed, or do you need either an 11 speed Di2 system or a 12 speed system etc.? Can they be used with just one shifter to go through the whole range of a 2x system or do you have to have both front and rear shifters?

From what I gather, Shimano don't make MTB Di2 systems - so there's no point in asking if Shimano Di2 road and MTB systems are compatible with each other. Presumably Shimano, introduced GRX as a third "discipline" to allow wide ratios to be used with drop-bars, but this seems to have created even more incompatibility between components and disciplines where I would have thought that a "programmable" electronic shifting system would be able to vanquish such compatibility issues.

 

[HLaB]

Although I have the dexterity back in my fingers during chemo I would have definitely tried Di2. I do like the ease of maintenance of cables on the occasional time things have went wrong but that's maybe justifying to my self why not to spend £££ more on Di2.