Rockbox future strategy

[ethanay]

Sort of a continuation of:
http://forums.rockbox.org/index.php/topic,6751.675.html

This first post provides context for the second post.  Consider it part 1 and part 2.

First off, I love Rockbox.  I've been a silent fan, installing it on friends' and family members' devices for years for various reasons (sound quality, simple interface, skinnable, accessibility, etc).  I don't want to see Rockbox go away. I think it allows us to unlock and explore the full potential of the hardware, and provides real benefit to the end users.  My mom can actually operate and use her PMP.  It works really well.  Very thoughtfully-designed.  Great tools and features for all sorts of different listeners.

So, first and foremost, this post asks the question, "What can we do to keep Rockbox alive?"  Second to this is the question, "Is there a market demand for what Rockbox offers?"  and implied in that, "What does Rockbox truly offer?"  I don't mean a "features list."  I mean real, felt value for the end user.  Because at the end of the day, even (especially!) for a non-profit open source project, if there's no demand, there's no interest.  and if there's really no interest...

So I think first we need to get to know Rockbox and its potential in a new light.  We need to know the value it has created.  Why did it explode all those years ago?  Why did people start developing it?  And why did so many of us flock to using it?

I see two strategies:  continue piggybacking on existing (non-open) hardware stacks and even turn it into an app (which I suspect violates some of the value it created in existing on dedicated reliable PMP hardware), and maybe find/partner with some sort of open hardware stack project.  Or start a PMP-specific open hardware stack.

Although this discussion is focused on demand, I discourage thinking about competition with big manufacturers who follow trends, e.g, of "thinner, smaller" -- current players are ridiculously small and thin.  I can carry what literally used to be a wall of records, tapes or cds in my hand and replay them on demand at ridiculously high fidelity and playback reliability.

So I think we need to identify specific market niches that relate to the intersection of both the philosophy and the specific value of rockbox.  Open source.  Education.  Accessibility (e.g., blind users).  So that's the context for the second post...

[ethanay]

Part 2:  a specific strategy to access some targeted market values

Rockbox Open Hardware
-Develop a modular open hardware platform for Rockbox
   +Combo form factor:  user-serviceable (AAA standard?) battery, built-in screen (no screen out), stereo-mini out only, integrated microphone and antenna; integrated A-B repeat button.  Comfortable to hold (form factor, balance).  USAGE SCENARIO:  portable universal-media player; student study tool.
   +Other: integrated controls, SD card-based storage system (internal memory only for OS?); focus on future-proofing (e.g., standardized components), performance (sound quality and battery life), durability and accessibility for blind users.
-Assess market demand for rockbox open hardware platform
   +Institutional markets:  Universities (CompSci, EE, music), government (e.g., military)?
   +Assess consumer market: FLOSS forums, existing rockbox hardware user forums
   +Crowd-funding, with matching funds
-Create sustainable operating model and business ecosystem for Rockbox
   +An existing manufacturer (e.g., Sandisk) may supply open-source hardware based on verifiable demand.
   +Apple model:  Hardware sales support software development
   +Service model:  Software and community support remain free, allowing developers to provide software customization and target support services.

Both computer science (software) and electrical engineering (hardware) departments of universities will have interest in the platform from the software and hardware side of things.  Music programs can offer finished units to students as part of their tuition package, as a study aid (the importance of the integrated microphone, A-B repeat, multi-codec support, and variable speed and pitch controls).  Institutions have connections to other institutions. 

Sales cost max target is $100 per unit, and production costs require $150,000 up-front for setup.  Units cost $75/piece delivered (parts, labor and shipping).  6000 unit sales break even (crowd-funding and institutional pre-orders/matching support).  24 institutional contracts at 250 units per institution per year for one-year break-even.  I pulled these numbers out of my ass, FYI.  But it's a starting point at least.

Institutions can help develop prototype and proof of concept via collaboration between Rockbox, EE and CompSci departments (student projects), music programs as testers.  Reduces development risks and helps guarantee market nice.  Crowd-funding campaign with prototype completed and institutional support can reduce break-even period and increase access to individual consumer markets.  Approach RaspberryPi Foundation early to evaluate and avoid potential niche duplication and help with institutional introductions and distribution?

The software may even be flexible enough where, with the right hardware connections, RB can offer design and consulting on custom niche media player solutions, sort of like how http://64studio.com/ evolved from an open source GNU/linux software multimedia production platform to custom solutions for hardware makers and other niche markets.

[wodz]

There are 2 things which are killing rockbox:
1) smartphones as the way people listen music nowadays.
2) smartphones as suckers of development force.

Creating open hardware DAP is simply not a viable option. While it is quite easy to create electronic part of the DAP from on-the-shelf components, mechanical part is really huge investment (not taken into account effort to port rockbox to new platform). Please read how divergent are expectation for new DAP in AGPtek thread.

[Chibisteven]

Rockbox will likely be a product of it's time.  Newer technologies eventually displace older ones.  At some point smartphones will eat up what's left of the portable digital audio player market completely.

[gomezz]

Is there scope for a Rockbox app to on people's smartphones, tablets etc given that non of the available player apps hold a candle to it?

[wodz]

Is there scope for a Rockbox app to on people's smartphones, tablets etc given that non of the available player apps hold a candle to it?

Maybe. Android port proved that the only technically feasible option would be to turn part of rockbox into library and then build player app using native UI widgets and rockboxlib as backend. There are a few problems however. First is where to draw the line what should be in lib and whats not. Second is the lack of skilled developers who would like to work on this. Third is the existence of other playback libs which one may use to write decent player app.

[ethanay]

Thanks for the discussion.  I think this is worth digging deeper into:

There are 2 things which are killing rockbox:
1) smartphones as the way people listen music nowadays.
2) smartphones as suckers of development force.

Creating open hardware DAP is simply not a viable option. While it is quite easy to create electronic part of the DAP from on-the-shelf components, mechanical part is really huge investment (not taken into account effort to port rockbox to new platform). Please read how divergent are expectation for new DAP in AGPtek thread.

In design process we get it hammered into us to clearly separate observation vs interpretation (this is really helpful in every aspect of life, esp. relationships...and also really freaking difficult).  In Wodz' statement (and Chibisteven's followup) I can see the following observations:

Convergence is occurring
Rockbox and the PMP hardware niche in general is struggling
It is currently popular for people to use their smart phones as their PMP
[I would also add that the whole "spotify" model is challenging the concept of a personal music library upon which pmp's have been traditionally based]
The market niche for purpose-built PMPs is contracting and restructuring

I also observe the following interpretations:
The market is nearly-completely saturated and tapped
The existence of smart phones and convergence dooms purpose-built hardware

Here's another general trend that is related to all this: Purpose-built vs generalist hardware. 

Purpose-built hardware and product design will almost always outperform generalist hardware any day, but only for specific niches.  Case in point:  Leatherman will never put Vaughan tools out of business with a hammer multi-tool design, but only because Vaughan knows and has a well-developed niche. An android phone is a leatherman.  Most people will opt for a leatherman.  But there's always going to be a need for a hammer.  A purpose-built PMP is, likewise, a specialized tool like a hammer.  I think we can assume there's always going to be a need for it.

Likewise a smart phone does not kill off the tablet does not kill off the laptop does not kill off the desktop.  They are simply getting "refined" for their special market niches.  A desktop is now a more specialized market niche than it used to be.  And there are always untapped and emerging markets.
 
I want to push back on that interpretation that smart phones doom rockbox to extinction. We've seen that argument over and over again.  What I've observed happens is one of three things in the context of convergence, emerging technologies and contracting/shifting market niches: 
1. An industry or model "sticks to its guns" but refocuses its niche and strategic partnerships to take advantage of the new market structure.  Example:  Identify sources of unmet or primary demand for purpose-built PMP hardware and target them directly and mainly, even in strategic partnerships to better meet their (unmet?) needs.
2.  An industry or model shifts to follow the market trends and try to keep its current market share.  Example:  Port rockbox to android app.
3. An industry or model fails to adapt and it collapses and maybe disappears.

So the question becomes:  Does rockbox want to continue?  And if so, what strategy do we want to apply?

[pamaury]

Hi,
I think you missed a very (maybe the most) important point in wodz's post. If you read this post:
http://forums.rockbox.org/index.php/topic,51071.0.html
and observe the current trend, you will see that basically:

• the number of people that want something that is not a smartphone is small
• each of these people seems to want a different DAP

Thus it may be the case that for every one, there exists a purpose-built DAP than outperms smartphone, but this particular device might be worse that smartphone for all the others. My point is that you can't possibly build a different device for everyone (at least not currently) because it is too expensive. This explains why at the moment we are mostly left with smartphone and expensive DAPs.

Now that being said, I am ready to make a wild guess: headphones are going to kill DAP. Consider the question: why would you buy a DAP? If you really care about audio quality, smartphone have you covered unless you use a very good DAP and a good amplifier. The latter implies carrying a heavy box along your DAP. On the long run, I think DACs will move from DAPs to amplifiers that are either connected wirelessly or by usb/thunderbolt/coaxial (ie digitally). Then it does not matter what sound quality your device provides and smartphone is better/cheaper. Apple is betting on the end of the jack, so this is completely unlikely. On the other hand, if you don't care about audio quality, buy a smartphone: it's cheaper and does more stuff. What about people that matter about something else (physical buttons, small size) ? Well that's even less that the number of people that buy DAPs nowadays, so same problem as before: not worth the cost unless you are ready to pay a lot for your purpose-built device.

I think it's fair to say that Rockbox on smartphones is mostly a failure so far, no one has really done the work of creating an interface from scratch. And on smartphone you can use huge/fully-featured audio libraries that probably do most of what rockbox does anyway. Surely there is room for improvement but I don't think using rockbox as a library is the sane option.

Personally I am trying porting Rockbox to whatever devices seem to be common/successful and these days it's mostly audiophile DAPs. I don't believe we can influence manufacturers so we can only port to what they create.

[ethanay]

Hi,
I think you missed a very (maybe the most) important point in wodz's post. If you read this post:
http://forums.rockbox.org/index.php/topic,51071.0.html
and observe the current trend, you will see that basically:

• the number of people that want something that is not a smartphone is small
• each of these people seems to want a different DAP

I appreciate you adding these points to the discussion, I did not perceive them in the discussion up to this point.  Just to emphasize, it is a world of difference between *individual* consumers and *institutional* consumers (who represent many individuals under a single preference or need).  Up to now, Rockbox has targeted the individual consumer market, but this focus is another potential area where Rockbox is struggling and seems unaddressed so far in the discussion.  Institutional consumers (e.g., universities) still have need for education devices (e.g., for music students) and are fairly standardized in preference and needs (e.g., the needs of one music program for a student study device will closely mirror the needs for another program).  I think the continued singular focus on individual consumers as a core market anymore is folly, and wonder about the institutional focus as an untapped market that could provide a core sustaining focus for Rockbox.

But on an individual level, it all seems like conjecture:  how many discussions show why so many people have embraced and then abandoned DAPs generally or something like Sansa specifically?    What we know is that PMPs are not being manufactured much anymore, which implies a decrease in demand, probably for a variety of reasons.  And still, worldwide, even a small niche demand easily represents millions of people.  So there may actually be more opportunity for Rockbox after the market contracts and collapses even more and the last fragments of random, closed-hardware and proprietary firmware competition go with it, leaving only a few niche OEMs who may see high value in leveraging FLOSS for their PMP OS. Also, the demand for different hardware aside, the demand for different type of software indicates a modular software infrastructure and interface, which Rockbox could address if demand can consolidate again.

Even if it doesn't change anything, this is an interesting discussion and I hope it helps anyone who reads it think through the lifecycle of Rockbox or other similar FLOSS phenomena.

[wodz]

People switched to smartphones because a) they have it and are carrying it already b) smartphones are good enough for this purpose. Maybe c) access to online streaming services but AFAIK that strongly depends on particular market. It is simply more convenient for majority of consumers.

[ethanay]

I wonder how much using the smart phone camera vs a dedicated camera parallels this?  I think the photography world is a good parallel, as it has a similar set of hardware chain challenges (e.g,. post-processing and printing quality are like DAC, amp and speaker quality).  They are really only different in their focus on consumption vs production activities, and that might be superficial for this discussion.

Most people use their smart phones to take pictures, but there appears to still be a strong market for dedicated cameras both among consumers (mostly point and shoot) and professionals (mostly DSLR or mirrorless).  Why?  Some professionals are using their smart phones to take professional photos, but smart phones are mostly replacing only low-end point and shoot cameras, not mirrorless or DSLR:
http://www.pcworld.com/article/2466500/how-the-smartphone-defeated-the-point-and-shoot-digital-camera.html
http://www.pcmag.com/article2/0,2817,2398032,00.asp

The second article only mentions serious competition between smart phones and the low end point and shoot market niche.  The niche point and shoots (e.g., like a Canon PowerShot G series) are still chugging on because they already focused on their market niche, and its based on hardware performance not on convenience.

This is the point that the first article makes at the end:  "Sometimes, it’s better to be convenient than good."  That is the smart phone:  Convenient, but not really Good.  Rockbox will find its niche in a focused market that is already not concerned so much about convenience as it is about actual hardware and software performance, which are enthusiast (vs casual) consumers and professionals.  And this niche will always exist regardless of convergence and miniaturization trends (which have probably maxed out anyway, Apple even had to scale up its Nano because they were getting too small for comfort!):

"More expensive cameras appear to be bucking this trend, however. The research firm noted that unit shipments and revenue for detachable lens cameras have each grown by better than 10 percent year-over-year, with such cameras boasting an average price tag of $863. Meanwhile, higher-end point-and-shoots, defined as having optical zooms of 10x or greater, had grown by 16 percent in terms of units shipped and by 10 percent in terms of revenue through November. Such products carried an average price of $257, NPD said."

So one of the niches for Rockbox to continue in the future isn't with low-end "point and shoot" DAP devices where it got started, but the DAP equivalent of the DSLR or mirrorless professional (or high-grade consumer) hardware and quality, or even high performance consumer grade stuff like the Canon PowerShot G series.  Professionals and enthusiasts, both.  Musicians and students.  Recording, audio and sound engineers.  Producers.  Audiophiles.

And if you add Open Hardware into it, then you can capture electrical engineer programs and students.  And Rockbox could already be partnering with universities on the software engineering and design side of things.  But the hardware partnership would strengthen that a lot.

If Rockbox is trying to stay focused on convenience and low-end hardware, it will probably go away.  If it shifts its focus, it will stay around.  And there will always be SOME market for dedicated low-end DAP, and if Rockbox can partner proactively with those few(er) SOC OEMs, it can help ensure compatibility without all the hard work reverse engineering, which will ensure that decent affordable players still remain available.

Another point is whether the people currently involved in Rockbox still actually want to stay involved (or if they are shifting their focus or life priorities or are feeling burned out, etc).

I'm learning a lot in this discussion, thanks!

[saratoga]

I think the photography world is a good parallel, as it has a similar set of hardware chain challenges (e.g,. post-processing and printing quality are like DAC, amp and speaker quality).

This is a better metaphor than you realize.  The camera market is in freefall and most vendors are desperate to establish alternative products that don't compete with smartphones. 

[ethanay]

Except for...

"More expensive cameras appear to be bucking this trend, however. The research firm noted that unit shipments and revenue for detachable lens cameras have each grown by better than 10 percent year-over-year, with such cameras boasting an average price tag of $863. Meanwhile, higher-end point-and-shoots, defined as having optical zooms of 10x or greater, had grown by 16 percent in terms of units shipped and by 10 percent in terms of revenue through November. Such products carried an average price of $257, NPD said."

Which is my main point.  Only low-end consumer point and shoot cameras are in a freefall.

http://www.imaging-resource.com/news/2016/08/24/from-instagram-to-a-dslr-smartphone-photography-driving-dslr-sales
http://www.hindustantimes.com/tech/smartphone-photographers-driving-dslr-sales/story-z8Bh0Rfim6rZRHgPmp8B9N.html
http://www.diyphotography.net/smartphone-photographers-driving-dslr-sales/

So there seems to be big distinction between low-end consumer-grade hardware, enthusiast hardware, and professional hardware.  Vendors are slimming down their low-end point and shoot lines and refocusing, or going out of business.  The same is true of DAPs.  Rockbox has previously focused on the market segment that is slimming down and going away.  It will also go away unless it refocuses.

"Canon India saw its DSLR sales increase 26% in the first half of 2016. But while DSLR sales, at least in India, are increasing as a result of a greater interest in photography generated by smartphones, the compact camera market is suffering. Nikon India's senior vice-president of corporate sales and strategy Sajjan Kumar has said that smartphone cameras have adversely impacted the sales of beginner point and shoot cameras."

[saratoga]

Except for...

"More expensive cameras appear to be bucking this trend, however. The research firm noted that unit shipments and revenue for detachable lens cameras have each grown by better than 10 percent year-over-year, with such cameras boasting an average price tag of $863. Meanwhile, higher-end point-and-shoots, defined as having optical zooms of 10x or greater, had grown by 16 percent in terms of units shipped and by 10 percent in terms of revenue through November. Such products carried an average price of $257, NPD said."

Which is my main point.  Only low-end consumer point and shoot cameras are in a freefall.

The companies that make them are also in freefall.  Which is my point. 

"Canon India saw its DSLR sales increase 26% in the first half of 2016.

India is a pretty small market for high end cameras, and not enough to save anyone's business.  Canon (the global company) is in major trouble and trying to refocus its business away from cameras. 

[ethanay]

Away from low-end point and shoots, not cameras generally.  Canon will not abandon successful product lines just because unrelated product lines aren't working out.  They will just cut their losses and explore new opportunities.  Which is the same crossroads for Rockbox, it seems.

Otherwise, it's like saying people aren't listening to digital music anymore just because they aren't listening to it on a low-end DAP.  It's just not true.  There's probably a market unfolding for a high-value DAP that delivers exceptional performance and reliability and is built to last (upgradeable capacities; internal and external modularity and serviceability).  But it's not going to look (or cost) anything like a Sansa.  Most people won't buy it.  But that doesn't matter.  It's one of the opportunities for the DAP that isn't in a freefall.

Also, the stuff about smart phones driving DSLR is interesting.  India is one of the world's largest and fastest growing consumer markets, and like in China, they often reflect and follow (or sometimes lead) US/European consumer preferences.