This is an odd feature request, because it’s an anti-feature request. I want Roon to stop supporting running the Core on so many different configurations. It’s a much more mature product on Nucleus or ROCK. Windows and Mac and Linux - though very widely used - are disproportionate consumers of time and resources that could be devoted to better velocity. But it doesn’t matter which platforms end up being the chosen ones, just fewer would be better. I know the team has made lots of simplifications that allow it to move faster through .NET and a bunch of other abstractions. But I fear they are making trade-offs that do not make sense given the maturity and scale of the project.
This is non-trivial in nature. There are lots of folks who would be on unsupported hardware. This is not a short-term thing. It would be massively controversial. And I want to make clear that I’m not just suggesting this idea because I’m on ROCK - I’m suggesting it because I think fewer is better. If it was determined that Windows was better, I’d run out and get a Windows license and run it on my NUC. If Mac was the winner, I’d figure it out.
In short, it’s just too complicated. Let’s cut off some configurations, and increase velocity. I’m not sure whether the complexity is in Core install varieties, or remotes, or home network. But I think the Roon user base collectively could get more out of the Roon team if we got rid of some currently supported or “supported” approaches to running the software. Just because it works, doesn’t mean it should be allowed to work. The shadow cost / opportunity cost of all that time working on all the permutations is less time on feature / functionality.
Yep, developing for a single platform would clearly free up resources, no argument there.
Three questions.
First, what proportion of the current user base have a Windows, Mac or Linux core? I suspect it’s a non-trivial amount.
Second, how many of those would jump ship if their OS of choice was dropped? Again, I suspect it’s a non-trivial amount.
Third, how many people would be put off from trying Roon because it also requires the purchase of additional hardware? You can see where I’m going with this, so I won’t repeat myself again.
I understand what you’re saying, and it makes sense if you don’t happen to be a Windows, Mac or Linux user, but Roon is marketed as ‘The Ultimate Music Player for Music Fanatics’ … note the word ‘player’.
Can’t agree with this at all. You need the wider support to pull in users. When pulled many choose a more dedicated approach which is where Rock, nucleus and other Linux based servers come in. Without Windows or Mac Roons user base would be almost non existent. Just with windows and Mac you have no option to make it as maintenance free as possible.
More the merrier from my perspective. I run Roon acorss the whole gamut for core, remotes and endpoints. It’s one of the reasons I went with it.
I’d be happy if my currently chosen platforms were not among the ones selected as “go-forward” platforms, if it resulted in a better experience with fewer degrees of complexity. If Roon said “pack up your ROCK, we are going full X”, where X was a Commodore 64 or NeXT machine, and it promised to be flawless, I’d figure it out. If it were Mac, I’d deal. To me this is sort of like Rawlsian justice- I’d prefer to live in that world, no matter which starting position I’m in now. It’d just be a better works overall if there were few things going on to support.
Maybe the Roon team will say “actually, we are good with all the cores, it’s all the remotes that drive us crazy, we wish we could do native development on platforms X and Y, and skip all the rest”, and I’ll say “bring it on, even if I can’t use my phone”.
Maybe the point of my post was to ask the Roon team to consider making some harder trade-offs and piss off a few people (even including me potentially) in the name of making a faster, more consistent, more rapidly developing product. Maybe core, maybe remote, maybe elsewhere.
If anything I think the core itself is destined for the dustbin future wise , going by the changes that are afoot. Obviously upsampling would not be possible without one, but the rest we will see.
Yeah, I’d be in shock if they gave up in the near term on local libraries. I’m not going to suggest that - though I would understand it commercially if, say, 99.6% of their new members were “streaming only”. When I say “make harder trade-offs”, I do NOT mean “cut off your leg, because life would be simpler with just one”.
If I couldn’t run Roon Server on a Linux distro of my choice I would not be a Roon customer.
Also, I feel a lot of problems I see in the support section are either kitten-weak Wi-Fi related or result from attempting to use a 750k track library on a Dell laptop that was stolen from the office in 2004. It doesn’t matter what platform you’re on if your hardware sucks and that really isn’t Roon’s problem.
Anyone who has followed the entire path from the purchase by Meridian to the strategically sensible Roon spin-off will recognize that steps other than the commitment to one manufacturer and one system made this great development possible in the first place.
Roon has visibly bigger ideas than cutting off the market and the future (the leg) for itself.
5 Likes
SukieInTheGraveyard
(This one time, at Bandcamp... I spent too much money!)
12
Sorry but it’s a total “no” from me!
I have run Roon for over a year without any problems whatsoever on a fanless Windows PC that I also use for office work, browsing etc. I certainly don’t want, nor currently need, a dedicated computer for music streaming. I would be very surprised if Roon would desire, or could afford, to lose Windows and Mac customers.
I think everyone should wear the same colour shoes and shirt. It’s just to much trouble keeping track of the variety of colours n the world. Perhaps even that same size.
The difference is that when we all wear different shirts, the speed of making new shirts isn’t lowered. But in software development if you support all kinds of different environments and use cases, your velocity (speed of new feature deployment) and quality (frequency with which stuff breaks for some people when new stuff comes out) both often suffer. There are some ways to cope with this “shadow cost of complexity” but they often require significant scale. Not the kind of scale a Roon with <1m users can achieve.
I’m not anti-choice, just pro-velocity and pro-quality. And I like funky shirts.
I think anyone who suggests this should volunteer to deal with the forum threads talking about inconsistent databases. It will be good practice for dealing with the forum threads that result from this change.