Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[kstuart]

So I upgraded to a Master license so I could give MC for Linux a try.

What is the simplest Linux version, for a 2009 PC that otherwise runs Windows, that also works well with MC21 ?

By "simplest", I mean "least time consuming".

(I used to develop GUI software for Unix, so I should be able to figure out any technical aspects - but I just don't have a whole lot of time to devote to this experiment, hence "simplest".)

Thanks!

[mwillems]

So I upgraded to a Master license so I could give MC for Linux a try.

What is the simplest Linux version, for a 2009 PC that otherwise runs Windows, that also works well with MC21 ?

By "simplest", I mean "least time consuming".

(I used to develop GUI software for Unix, so I should be able to figure out any technical aspects - but I just don't have a whole lot of time to devote to this experiment, hence "simplest".)

Thanks!

I would recommend Debian Stable (currently called "Jessie" or Debian 8 ) .  It's the distribution (distro) that MC is developed on, and it's the only "officially supported" distro, so is the most likely to work as expected out of the box. Debian Stable is also relatively easy to install, runs on anything, and is generally more, well, "stable" than other linux dsitributions.  What this means in practice is that they freeze software versions for 2 years at a time so things are unlikely to break within a release.  They provide security updates, but that's generally it.  This makes Debian a good environment for "appliances" and servers, but less great for a workstation.  The only "trick" is that you have to install MC from outside of the normal distro repositories, but that's true of MC for every Linux distribution.

The only pause I would have is if you have specific hardware needs - do you have a specific DAC in mind or need DSD to work?  If so, you might need a more modern kernel than Debian ships by default which could raise the "fiddle" factor (DSD support is fairly new on Linux and new DACs are getting kernel support all the time).  

A tip: during the install wizard, you might want to consider choosing xfce instead of the default desktop package, as xfce is (when compared to the default desktop environment) lighter-weight, more similar to windows, and is likely to run better on older hardware. 

[kstuart]

Thanks, that's very helpful !

Any reason to prefer 32-bit Jessie to 64-bit Jessie for a system only running MC21 ?

PS  Definitely do not need DSD support in the DAC, only in MC21 itself.

[Awesome Donkey]

Ubuntu or Mint might be a little easier than Debian Stable, IMO. Less steps involved.

[bob]

Thanks, that's very helpful !

Any reason to prefer 32-bit Jessie to 64-bit Jessie for a system only running MC21 ?

PS  Definitely do not need DSD support in the DAC, only in MC21 itself.

I don't think that 32 vs 64 makes any difference to MC. Perhaps if your PC is low on memory 32 would be more efficient.

[kstuart]

Ubuntu or Mint might be a little easier than Debian Stable, IMO. Less steps involved.

Less steps involved in the MC21 part of the install ?

[kstuart]

What about "Linux Mint Debian Edition 2" that is based on Debian Jessie ?  Perhaps that would be "best of both worlds"?

[Awesome Donkey]

Well, if you look at these tutorials;

Debian: http://yabb.jriver.com/interact/index.php?topic=99333.0
Ubuntu: http://yabb.jriver.com/interact/index.php?topic=99334.0

There's a pre-step with Debian - you have to add yourself to sudo and audio groups, so technically one less step.

Linux Mint Debian Edition may or may not work fine, I tried it in the past and it did work fine however I haven't tried it again in a good while now. Ultimately it doesn't matter between Debian and Ubuntu/Mint, it just depends if you want one less step and your preference. Overall ease of use wise (and coming from Windows especially), Mint would probably be my choice.

[kstuart]

After doing a little reading, for the purpose of just running MC21, I am leaning towards "Linux Lite" - their 3.0 Final (stable) has just been released.  It uses xfce, as suggested above, and is based on Ubuntu 16.04 LTS:

https://www.linuxliteos.com/forums/release-announcements/linux-lite-3-0-final-released/

Any problems installing MC21 on that ?

[mwillems]

After doing a little reading, for the purpose of just running MC21, I am leaning towards "Linux Lite" - their 3.0 Final (stable) has just been released.  It uses xfce, as suggested above, and is based on Ubuntu 16.04 LTS:

https://www.linuxliteos.com/forums/release-announcements/linux-lite-3-0-final-released/

Any problems installing MC21 on that ?

The only problem is that I'm not aware of anyone here running it (and in fact haven't heard of it before).  So if you run into issues you are potentially on your own because it's not actually a supported distro and none of us will know exactly how to help.  My experience has been that smaller "boutique" distros often make inexplicable changes to the functioning of various packages that are poorly documented (if documented at all).  And the communitites are tiny, so invariably folks wind up on the forums of the parent distro (in this case ubuntu) with problems that no one can reproduce because someone downstream tweaked a compile flag and didn't tell anyone (this happens all the time with Arch derivatives).

So I'd advise you stick with a "well travelled" distro like Debian, Mint or Ubuntu. The latter two are still debian-based so have the high likelihood of working easily with MC which only officially supports debian.  All three are well-maintained, and well represented on the forums here (so you stand a good a chance of getting help even with an unsupported distro).  My personal advice is still Debian, especailly if this machine will be an "appliance" for playing music as it's more minimal out of the box, and its much less likely that a stray update will bust everything down the road.  If you plan on using the machine as a general purpose computer (i.e. a workstation)  ubuntu or mint may be better than vanilla debian, but I'd advise installing Xubuntu (the xfce ubuntu flavor) or Linux Mint Mate edition (rather than the Cinnamon edition).  Linux Mint Debian Edition is not the main focus of the mint maintainers; if you want to stay closer to debian, just stick with debian.

My two cents

[BryanC]

After doing a little reading, for the purpose of just running MC21, I am leaning towards "Linux Lite" - their 3.0 Final (stable) has just been released.  It uses xfce, as suggested above, and is based on Ubuntu 16.04 LTS:

https://www.linuxliteos.com/forums/release-announcements/linux-lite-3-0-final-released/

Any problems installing MC21 on that ?

Any distro can be made 'lite' after the fact. The problem with the smaller distros is that they usually have correspondingly small package repositories, 'quirks' that make forum/wiki tech support more difficult, and a propensity to disappear completely after a few years.

The biggest differences in 'liteness' comes down to the DE used, not the distro. Thus you will want to use a lightweight DE like XFCE or LXQT/LXDE. Xubuntu would be a good choice if you like the look of Linux Lite since it uses the same DE (XFCE), and is built from a distro that is very similar to the one that MC for Linux is developed on.

[kstuart]

Thanks for the informative and thoughtful comments.

I did some searching on Linux Lite vs Mint Mate or Mint Xfce, and the overall sense I got is that Linux Lite was somewhat more stable and used less resources.  I did find this 2016 comment on Mate vs Xfce:

"I use both, they both have their pros and cons, but I honestly prefer XFCE, it's just far more stable then Mate."

On Linux Lite vs Xubuntu vs Mint Xfce, I found the following comments:

"For my non-technical response, it has to be the great support of the developer, Jerry, and others on the Linux Lite forum. There is no Xubuntu forum and the Xubuntu users get lost in the Ubuntu forum. With Mint you may have a similar problem because it is only one of the many DE it offers. I let others provide more technical reasons."

and

"Linux Lite is designed to minimize the 'learning curve' for users transitioning from Windows to Linux, while using less CPU/RAM resources than Ubuntu/Kubuntu/Xubuntu to do so."

and this post was persuasive:

"The entire premise behind Linux Lite, is to give a Window's user the best chance at discovering the world of Linux and to present Linux in the most anxiety free way possible. The creator of Linux Lite, Jerry Bezencon has made his premise known in several interviews for online Linux publications since the inception of this distribution.

Specifics? To mention a few,

1. The choice of the Xfce DE shouldn't be easily dismissed. It's layout is simple, intuitive, functional and as configured in the iso, reminiscent of XP, Vista and Windows 7.

2. The structure of the system menu and more importantly, it's layout and labeling. Far too often a new user can be utterly confused and turned off by the names of Linux applications and many times, those applications are redundantly provided by the distribution (More than one application provided for the same task), Mint is famous for this. It's been mentioned before in this thread but needs repeating, the system menu (Whisker) in Linux Lite is labeled with common usage names and not obscure Linux application names.

3. Have you looked at the "Linux Lite Control Center"? Everything from system configuration, desktop configuration to software sources can be found here and it's intuitively designed to fit within the premise of Linux Lite.

4. Forget about terminal and apt-get, do you remember trying to understand Synaptic Package Manager during your first go around with Linux? Updating apt and adding sources? Then trying to locate the application you want (with a weird name) and hoping you don't make a mistake and soft-brick the system? I do!... Take a look at "Lite Software" app located in the System sub-menu of Whisker. There's a plethora of "intuitively labeled" and commonly used applications listed for the new user to simply click and install. No terminal, no synaptic package manager and no manual adding of PPA's. It also keeps an easily accessible,  running tally upon what apps the user has added to their system and provides easy removal too.

5. Last but most importantly (imo) is the "Linux Lite Help Manual". A NEW USERS DREAM! Answers to everything can be found here. BEST of all, it's Precisely,Concisely and Specifically written for the Window's user in mind. I can't begin to say enough about this document and how extremely valuable it would've been for me, when I first started using Linux many years ago.

In the sometimes egocentric and even arrogant Linux universe, it's a wonderful to find a developer and distribution truly engaged in making Linux and open source... Easy, presentable and FUN to potential new users. "

and the similar:

"LL really cuts it for me over any other distro (including Xubuntu) or Windows OS, for the following
reasons in order of importance (most important first):

1. The  LL forum: a real community of ever-helpful members who are willing to share their time and
    knowledge, and very importantly,  to match your level of understanding.  Ubuntu, Mint & Zorin
    forums seem much larger and as a result you feel somewhat anonymous, and so not quite
    the same feel of community about them - a unique selling point of LL

2. LL is easy to use for ex-Windows-OS users, particularly those who are not power users, e.g the familiarity of
    desktop layout and its intuitive gui which, and very importantly, does away with a need for
    command line
  
3. Applications rather than the package names are used - this gives clarity particularly for noobs who
    are not familiar with linux package names

4. Lightweight & does a great job of  resurrecting older PCs."

=====

So, I'm probably going to give Linux Lite a try - I can easily do it in a way that does not interfere with my normal use, since I built the PC with a large spare partition for future dual boot use.

And, if it works with JRiver MC21, then that will give a data point for future Linux Lite users.

BUT, I'm still open to other input...

[kstuart]

Since Linux Lite 3.0 is based on Ubuntu 16.04 LTS, the following comment seems another reason to try it, especially until Mint 18 is available:

" Media Center requires libs of at least the same version found in Debian 8.0 Jessie. Ubuntu 14.04 LTS (and Linux Mint 17.2 which is based off 14.04) use older libs than the ones needed by Media Center. The fix uses a PPA which upgrades some of the libs, thus allowing Media Center to work again. "

In other words, one less thing to futz with.

I have an old 2003 Windows XP laptop which is sitting idle, so I am going to try Linux Lite on it first, since it seems an ideal pairing.

[Awesome Donkey]

" Media Center requires libs of at least the same version found in Debian 8.0 Jessie. Ubuntu 14.04 LTS (and Linux Mint 17.2 which is based off 14.04) use older libs than the ones needed by Media Center. The fix uses a PPA which upgrades some of the libs, thus allowing Media Center to work again. "

This isn't an issue anymore, actually.

[kstuart]

This isn't an issue anymore, actually.

MC was changed ?

[Awesome Donkey]

Yes, when I installed MC in a Mint VM a month ago, it worked without issue. Bob changed the libs and made it work again.

[kstuart]

Interesting:

http://www.ap-linux.com/

[JimH]

Why not start with a well known distro?  Even one we support?  It's the path of least resistance and it's hard to see what you might gain by wading into the bush.

[Awesome Donkey]

Snake oil, IMO.

[kstuart]

Snake oil, IMO.

Then that means you approve !

"Snake Oil" is equivalent to "Fish Oil" and it actually has many health benefits:

https://en.wikipedia.org/wiki/Omega-3_fatty_acid

100 years ago, "Snake Oil Salesmen" were guys who put vegetable oil into bottles, and put "Snake Oil" on the labels.

So, "Snake Oil Salesmen" were frauds and con men - not oil harvested from snakes.

[Awesome Donkey]

http://yabb.jriver.com/interact/index.php?topic=103769.0

[kstuart]

Thanks for the link - as one poster pointed out - since the distro is free, it even fails that necessary quality of "snake oil" (or more properly "snake oil salesmen").

(The concept of "trying to sell something that does not provide a benefit" does not need a special term, since these days, that applies to a huge portion of the economy. )

The thought process on "audiophile linux" was:

* I did install Linux Lite on my 2003 XP laptop and it works fine in that underpowered situation.

* Then I tried installing it on this laptop that I use for web browsing, email, etc.

* Install went great, it actually shrunk the one existing partition (Windows) on my SSD and installed dual boot without fiddly intervention.  The install only asks simple questions like language, location, etc.

* When I started using Linux Lite - which was much more similar to Windows 7 than Windows 8 - I realized that the main problem is not the OS, but the fact that over many years (decades), I've done countless small tweaks.  Some I could do without, others were a necessary part of my workflow.

* The bigger realization was that Linux Lite was almost exactly the same as Windows 7 - and so why not use Windows 7 (rather than spend many hours recreating all those tweaks)?

* The only advantages of Linux at this point are related to the commercial aspect of Windows - i.e. its cost and MS' failure to do important things that have no profit incentive (such as USB2.0 Audio support present in all other OSes).

* And then I realized that the real new 21st Century OS is Android.

* The only other aspect of Linux that is relevant to MC21 usage, is sound quality - a significant number of users report better sound quality than Windows.   This could be due to aspects of their Windows installation, or it could be a real improvement.

* Note that a repeated double blind test conducted using audio convention attendees, showed subtle improvements to sound quality simply by using a faster PC (both PCs having "bit perfect" playback). 

[mwillems]

* The bigger realization was that Linux Lite was almost exactly the same as Windows 7 - and so why not use Windows 7 (rather than spend many hours recreating all those tweaks)?

Because while it may seem the same as windows 7 in terms of the user interface, the user interface is only a small outward facing part of the OS.  Under the hood it does things very differently, sometimes with benefits, sometimes with detriments (see below).

* The only advantages of Linux at this point are related to the commercial aspect of Windows - i.e. its cost and MS' failure to do important things that have no profit incentive (such as USB2.0 Audio support present in all other OSes).

This is quite false.  Linux has many advantages and disadvantages as compared to windows, but many of them are not immediately visible.  For example, Linux systems can be (and in my experience are) more stable than windows systems.  I've had Linux systems go 6 months or more without rebooting, with no flakiness or degradation in performance.  With windows weekly (or daily) reboots are the norm because eventually things start acting strangely.  Additionally windows (even windows 7) phones home and sends telemetry and other data to Microsoft.  Linux doesn't do that.  Those are just two obvious examples, there are quite a lot of differences all the way around (network stability, better options for data integrity, etc.).  Assuming that the two are equivalent because a distro can be made to look like Windows 7 is missing quite a lot of what's going on "under the hood."

 
* And then I realized that the real new 21st Century OS is Android.

Not to be pedantic but this is a perfect illustration of what I noted above: android is actually Linux.  Normally when people refer to Linux in conversation they're actually referring to GNU/Linux which is a combination of the Linux kernel (he Linux part) and a userspace built on a foundation of GNU core utilities and the GNU C compiler (the GNU part).  Android is not GNU/Linux because it uses its own userspace, but android uses the linux kernel, so is just as much linux as GNU/Linux in that both use the linux kernel as their core OS.  It's just the userspace that's different (the part the user sees), which illustrates my point pretty well; most people aren't even aware that android is a linux OS.

* The only other aspect of Linux that is relevant to MC21 usage, is sound quality - a significant number of users report better sound quality than Windows.   This could be due to aspects of their Windows installation, or it could be a real improvement.

* Note that a repeated double blind test conducted using audio convention attendees, showed subtle improvements to sound quality simply by using a faster PC (both PCs having "bit perfect" playback).  

Whether or not there are gains to be had, you're unlikely to see these benefits in a tiny, questionably maintained Arch Fork with an RT kernel (which actually provably harms performance in any application that doesn't involve live audio recording).  Try a mainstream distro with a light-weight desktop environment and you'll see all the benefits that you're likely to see.
 
Something that may not be clear to you: in the open source community, for the most part, people volunteer in their free time to develop and maintain software.  Larger distros actually work better than smaller distros because: 1) they're a larger target for developers, so developers work harder to make sure things work on them, 2) there's more manpower available to actually do the necessary work of software development and maintenance, and 3) more users means the user created wiki documentation will be higher quality and the forums will be active enough to be able to answer questions.  

These things don't necessarily scale linearly, but very small distros tend to be both very low value-added and poor quality to the extent they try to "add value."  Small teams can do great things when its their actual job, and/or when their efforts are appropriately focused (say on one or two software programs), but it's a very different story when you're essentially signing up to use an entire OS that's essentially one person's hobby project, in which they might lose interest tomorrow, take ill, etc.  When I see a "boutique" distro like aplinux that makes several questionable design decisions (e.g. a rolling release model for an audio appliance?  old RT kernels?) and has fewer than 200 threads on their entire user forum, it does not inspire me with confidence.

Do as you like, of course, the joy of Linux is that there's almost infinite choice, but recognize that there are deeper waters just out of sight  

[kstuart]

I think we have a different meaning of "better" in this case.

For example, let's say that you had an automobile engine that had 10% more horsepower and 10% better mileage.  In an abstract way, that is better.

If I convinced a friend to replace the engine in his car with that one, his personal experience would likely be indistinguishable from before.   Only a yearly calculation of gas cost would show a difference.

Now imagine I had to convince the friend to buy a new car to get the 10% improvements.  His experience would be overall worse, because it would take him hours to learn the high beam switch, the radio controls, etc. etc.   Since we have already established that he would not notice the 10% differences, then his personal experience overall would be worse.

Lastly, if all auto manufacturers used the new engine, there would a society-wide improvement in gas usage, which would likely be statistically positive.

===

I arbitrarily chose the example, but it turns out to be a good one, because then the equivalent of "Tesla electric car" is "Android".   Linux and Windows share a lot of their disadvantages, the biggest one being a requirement for average users to do a lot of setup and a lot of maintenance.   They also share a lot of 1970s computing concepts.

( I was aware that Android is based on Linux, but that is irrelevant.  For example, the "Linux" version of MC21 cannot be installed on Android, so clearly JRiver uses the same distinction of meaning.)

[kstuart]

I should mention here that the input from members in this thread IS helpful, these posts are just discussion of issues raised and should not be taken personally.

As far as the concept "distros that have modifications can cause problems with MC21" - that is true with anything other than the Jessie distro used by JRiver devs.  It is impossible to predict what change will cause a problem, otherwise action would be taken in advance.

So, it seems to me that unsupported distros are all equivalent in that respect.

===

The smaller group vs larger group issue can also be examined in iTunes vs WMC vs MC21 vs small and medium freeware media players development groups.

Each size has development group has its particular advantages:

* Small groups headed by one person can be the most responsive to user requests, but only up to a point.  Prior to MC, I used a player with the name of an insect , the dev was quite outgoing and open to requests, but ultimately he only had enough time for major requests that were requested by large numbers of users (as is appropriate).

* Medium groups (the typical open source project) are actually the most likely to go aground and die.  Unlike the small group, medium group devs tend to be largely anonymous and so don't tend to get the personal feedback that most volunteers require as their real reason for participation.  So they are constantly leaving.

In a medium group, if you are just a user, your feedback and needs is just statistical.  You will only get improved MKV support if a significant number of other people happen to ask for it as well.  Generally, high profile stuff like 4K and 3D get more attention.

* Large groups (Microsoft, Apple, Samsung) are impossible to contact (even a large group, including some businesses, could not get MS to even notice the request to add USB2.0 Audio support).  What you get is generally what increases profitability.  Over time, these companies get less and less customer responsive (Google is constantly phasing out popular products and features that really don't cost them a significant amount to support, as one can read in their product Forums that have tens of thousands of pleas to not phase out some small aspect of Gmail.)

So, no particular size is automatically better.

Overall, human decisions are generally based on a mix of emotional reactions, personal self-interest and a tiny smattering of objective evaluation.

For example, having a couple of rude and ill-behaved guys use your product commercially, without asking your permission, can color your reaction to a whole group of hobbyists, the rest of whom may be very well mannered and considerate.

[mwillems]

I think we have a different meaning of "better" in this case.

For example, let's say that you had an automobile engine that had 10% more horsepower and 10% better mileage.  In an abstract way, that is better.

If I convinced a friend to replace the engine in his car with that one, his personal experience would likely be indistinguishable from before.   Only a yearly calculation of gas cost would show a difference.

Now imagine I had to convince the friend to buy a new car to get the 10% improvements.  His experience would be overall worse, because it would take him hours to learn the high beam switch, the radio controls, etc. etc.   Since we have already established that he would not notice the 10% differences, then his personal experience overall would be worse.

Lastly, if all auto manufacturers used the new engine, there would a society-wide improvement in gas usage, which would likely be statistically positive.

A better example is a car that breaks down 10% less often on average and has brakes that are 10% more powerful; your friend may not ever notice the advantage (statistics being what they are), but no one enjoys a car that breaks down and needs immediate repair, and everyone likes to be able to avoid accidents.  Stability, better privacy, and better data integrity are things that you don't notice by design; you only notice their absence  .  Many of the differences are about managing certain kinds of risk, which can be hard to quantify

I arbitrarily chose the example, but it turns out to be a good one, because then the equivalent of "Tesla electric car" is "Android".   Linux and Windows share a lot of their disadvantages, the biggest one being a requirement for average users to do a lot of setup and a lot of maintenance.   They also share a lot of 1970s computing concepts.

Something else that may not be immediately obvious; Linux does not have a unified user interface; there are dozens of different user interfaces available.  There are Linux desktops that have entirely abandoned the desktop metaphor and other "1970's computing concepts" (no task bars, no minimize or maximize buttons, no desktop icons, etc.).  There are linux desktops that look and act pretty similarly to android (but with more robust multi-tasking).  There are actually linux desktop environments that are similar to OSX too for that matter.  Consider that so far you're evaluating an operating system based on a conscious decision to find the version that's most similar to windows, and then concluding that the two were functionally identical.  Because Linux is about choice, it can be made to be similar to Windows, but it need not be.  If you're looking for something that looks and acts a bit more like android, try the Gnome desktop; it won't run particularly well on very old hardware, but is very, very different than Windows.

But I certainly won't argue your point that linux requires the user to do more much setup and maintenance than android; it certainly does.  It also gets security fixes in a timely manner and is a general purpose operating system (both unlike android).  But as you note, that's beside the point, you can't run JRiver on android, and that's the goal here, right?

As far as the concept "distros that have modifications can cause problems with MC21" - that is true with anything other than the Jessie distro used by JRiver devs.  It is impossible to predict what change will cause a problem, otherwise action would be taken in advance.

So, it seems to me that unsupported distros are all equivalent in that respect.

You're forgetting the community.  Lots of folks here run distros that aren't Jessie.  I run plain Arch Linux.  MC works great for me.  If you ran plain Arch, for example, I could provide some assistance/tips and tricks (and have for users running Arch here on several occasions).  Awesome Donkey maintains the excellent Ubuntu/Mint guides and helps folks out with those, etc.

By running a distro that's unsupported and that very few other people run, you're drastically reducing your chances of being able to get meaningful assistance.  So all unsupported distros are not equal in the sense that if you hit a quirk with, say, Ubuntu, there are dozens of people here who might have some idea how to help; not so with the smaller distros.

===

The smaller group vs larger group issue can also be examined in iTunes vs WMC vs MC21 vs small and medium freeware media players development groups.

Each size has development group has its particular advantages:

* Small groups headed by one person can be the most responsive to user requests, but only up to a point.  Prior to MC, I used a player with the name of an insect , the dev was quite outgoing and open to requests, but ultimately he only had enough time for major requests that were requested by large numbers of users (as is appropriate).

* Medium groups (the typical open source project) are actually the most likely to go aground and die.  Unlike the small group, medium group devs tend to be largely anonymous and so don't tend to get the personal feedback that most volunteers require as their real reason for participation.  So they are constantly leaving.

In a medium group, if you are just a user, your feedback and needs is just statistical.  You will only get improved MKV support if a significant number of other people happen to ask for it as well.  Generally, high profile stuff like 4K and 3D get more attention.

* Large groups (Microsoft, Apple, Samsung) are impossible to contact (even a large group, including some businesses, could not get MS to even notice the request to add USB2.0 Audio support).  What you get is generally what increases profitability.  Over time, these companies get less and less customer responsive (Google is constantly phasing out popular products and features that really don't cost them a significant amount to support, as one can read in their product Forums that have tens of thousands of pleas to not phase out some small aspect of Gmail.)

So, no particular size is automatically better.

I think you might've missed part of my observation: people behave very differently when something is their actual job than when it is a hobby.  One noted developer has called OSS development the "Cascade of Attention Deficit Teenagers" development model.  As I noted, small groups can be great when it's their job and/or they're focused on a manageable task.  A large task (whole OS) combined with a small volunteer force trying to do it in their spare time, is not even remotely comparable to a professional group of similar size performing a comparatively small task (media player).

In any case I'm only offering an observation as someone whose spent a few years following open source projects (which come in all sizes by the way); my observation is that small volunteer teams working large software scopes tend to either fail to add value or fail to do a good job, especially over time.   It's frustrating to have to relearn things at a time not of your choosing because a project you relied on has vanished or gone into desuetude.  

You can come to your own conclusions through use over time as I have, and you may have a different experience.  Nothing I've offered above is intended as gospel; it's just observations to help manage risk.  You likely have a different risk tolerance than I do and there's nothing wrong with that.

[terrym@tassie]

Responding to your original post, I converted a couple of old laptops (one 32 bit the other 64 bit) to Linux from Windows last year to use as MC clients. I originally installed Debian Jessie for maximum compatibility with JRiver MC but I found after testing a lot of other Linux releases that I preferred Linux Mint 17.3 XFCE.
I still retain Debian on both machines in case I hit an issue with MC that requires testing but as a long time Windows user I find Linux Mint XFCE the most "understandable" and easy to use and maintain. It also seems to perform very well on old "underpowered" hardware.
Purely my own subjective opinion FWIW.

Terry

[kstuart]

@Terry I think that Linux Mint Xfce is very similar to Linux Lite, the main difference being that Lite made a conscious decision to remove any differences from Windows that were "optional", for example, they renamed some things to match the Windows equivalent.

@mwillems I actually agree with all the points in your last post, in fact, at the same point I was thinking that - just for the purpose of running MC21, the two best choices would be to run Jessie, or to run whatever was the most commonly used by MC21 users - as opposed to how big the distro dev team is.

Perhaps someone could start a poll in this Forum (which of course limits one to "forum visiters") about which distro they are using for MC21 ?

[mwillems]

Perhaps someone could start a poll in this Forum (which of course limits one to "forum visiters") about which distro they are using for MC21 ?

You can get a rough idea by looking at the number of views on the various install guides at the top of the forum.  The Ubuntu/Mint guide is ahead by a mile (if you bracket out the build threads, which shouldn't count).  You can also see the same result over time by looking back at the MC 20 for Linux forum which had guides for a few more distros (Ubuntu/Mint also in first place).

That also dovetails with what you would expect based on general adoption of the various distros in the Linux community: it's hard to get a reliable figure, but the top three distros in terms of number of users/popularity are typically considered to be Mint, Ubuntu, and Debian in that order (followed by OpenSUSE and Fedora in 4th and 5th place). 

[kstuart]

quote: "I think you might've missed part of my observation: people behave very differently when something is their actual job than when it is a hobby.  One noted developer has called OSS development the "Cascade of Attention Deficit Teenagers" development model. As I noted, small groups can be great when it's their job and/or they're focused on a manageable task.  A large task (whole OS) combined with a small volunteer force trying to do it in their spare time, is not even remotely comparable to a professional group of similar size performing a comparatively small task (media player)."

I think you've just argued against Linux itself, in favor of Windows or Mac.   Debian and Mint are both volunteer projects. Linux itself was originally a volunteer project, but it has become so important to a number of companies that currently 75% of the code is written by paid developers.  Of course, Ubuntu is a project of one of those companies (Canonical).  So, overall Linux is a mix of volunteers and paid developers.  And...

quote: "Stability, better privacy, and better data integrity are things that you don't notice by design; you only notice their absence.  Many of the differences are about managing certain kinds of risk, which can be hard to quantify."

I found the following from Wikipedia interesting on that:

Kernel code quality

In an interview with German newspaper Zeit Online in November 2011, Linus Torvalds stated that Linux has become "too complex" and he was concerned that developers would not be able to find their way through the software anymore. He complained that even subsystems have become very complex and he told the publication that he is "afraid of the day" when there will be an error that "cannot be evaluated anymore."[12]

Andrew Morton, one of Linux kernel lead developers, explains that many bugs identified in Linux are never fixed:[13]

    Q: Is it your opinion that the quality of the kernel is in decline? Most developers seem to be pretty sanguine about the overall quality problem. Assuming there's a difference of opinion here, where do you think it comes from? How can we resolve it?

    A: I used to think [code quality] was in decline, and I think that I might think that it still is. I see so many regressions which we never fix.

Con Kolivas, a former Linux kernel developer, compared some OpenSolaris kernel code to Linux code and was surprised at the difference in code quality:[14]

    The summary of my impression [after reading the OpenSolaris code] was that I was... surprised....the [OpenSolaris] code, as I saw it, was neat. Real neat. Extremely neat. In fact, I found it painful to read after a while. It was so neatly laid out that I found myself admiring it. It seems to have been built like an aircraft. It has everything that opens and shuts, has code for just about everything I've ever seen considered on a scheduler, and it's all neatly laid out in clean code and even comments. It also appears to have been coded with an awful lot of effort to ensure it's robust and measurable, with checking and tracing elements at every corner. I started to feel a little embarrassed by what we have as our own [Linux] kernel. The more I looked at the [OpenSolaris] code, the more it felt like it pretty much did everything the Linux kernel has been trying to do for ages. Not only that, but it's built like an aircraft, whereas ours looks like a garage job with duct tape by comparison....[OpenSolaris] looks like an excellent design for a completely different purpose. It's built like a commercial design for commercial purposes that have very different requirements than what most of us use Linux for, but it does appear to have been done so very well. It looks like a ** Star Destroyer, and the Linux kernel (scheduler) suddenly looks like the Millennium Falcon. Real fast, but held together with duct tape, and ready to explode at any minute.

Theo de Raadt, founder of OpenBSD, compares OpenBSD development process to Linux:

    "Linux has never been about quality. There are so many parts of the system that are just these cheap little hacks, and it happens to run.” As for Linus Torvalds, who created Linux and oversees development, De Raadt says, “I don’t know what [Linus] focus is at all anymore, but it isn’t quality.”[15]

[mwillems]

quote: "I think you might've missed part of my observation: people behave very differently when something is their actual job than when it is a hobby.  One noted developer has called OSS development the "Cascade of Attention Deficit Teenagers" development model. As I noted, small groups can be great when it's their job and/or they're focused on a manageable task.  A large task (whole OS) combined with a small volunteer force trying to do it in their spare time, is not even remotely comparable to a professional group of similar size performing a comparatively small task (media player)."

I think you've just argued against Linux itself, in favor of Windows or Mac.  

There are literally over a thousand volunteer software developers working on Debian, and that's not counting the people who write documentation, do debugging, host servers, etc.  It's not a couple of guys in their garage.  You can overcome the inherent limitations of a volunteer-only effort by having a large number of volunteers.

Debian and Mint are both volunteer projects. Linux itself was originally a volunteer project, but it has become so important to a number of companies that currently 75% of the code is written by paid developers.  Of course, Ubuntu is a project of one of those companies (Canonical).  So, overall Linux is a mix of volunteers and paid developers.  And...

Yes that's exactly my point.  Some combination of paid staff and/or a large number of volunteers is necessary for a modern operating system.  The larger Linux distros have one or both.  I don't think that more people and/or more paid staff necessarily scale up indefinitely, but if you have neither paid staff nor a large number of volunteers working on a modern OS, expect trouble.  

quote: "Stability, better privacy, and better data integrity are things that you don't notice by design; you only notice their absence.  Many of the differences are about managing certain kinds of risk, which can be hard to quantify."

I found the following from Wikipedia interesting on that:

Kernel code quality

In an interview with German newspaper Zeit Online in November 2011, Linus Torvalds stated that Linux has become "too complex" and he was concerned that developers would not be able to find their way through the software anymore. He complained that even subsystems have become very complex and he told the publication that he is "afraid of the day" when there will be an error that "cannot be evaluated anymore."[12]

Andrew Morton, one of Linux kernel lead developers, explains that many bugs identified in Linux are never fixed:[13]

    Q: Is it your opinion that the quality of the kernel is in decline? Most developers seem to be pretty sanguine about the overall quality problem. Assuming there's a difference of opinion here, where do you think it comes from? How can we resolve it?

    A: I used to think [code quality] was in decline, and I think that I might think that it still is. I see so many regressions which we never fix.

Con Kolivas, a former Linux kernel developer, compared some OpenSolaris kernel code to Linux code and was surprised at the difference in code quality:[14]

    The summary of my impression [after reading the OpenSolaris code] was that I was... surprised....the [OpenSolaris] code, as I saw it, was neat. Real neat. Extremely neat. In fact, I found it painful to read after a while. It was so neatly laid out that I found myself admiring it. It seems to have been built like an aircraft. It has everything that opens and shuts, has code for just about everything I've ever seen considered on a scheduler, and it's all neatly laid out in clean code and even comments. It also appears to have been coded with an awful lot of effort to ensure it's robust and measurable, with checking and tracing elements at every corner. I started to feel a little embarrassed by what we have as our own [Linux] kernel. The more I looked at the [OpenSolaris] code, the more it felt like it pretty much did everything the Linux kernel has been trying to do for ages. Not only that, but it's built like an aircraft, whereas ours looks like a garage job with duct tape by comparison....[OpenSolaris] looks like an excellent design for a completely different purpose. It's built like a commercial design for commercial purposes that have very different requirements than what most of us use Linux for, but it does appear to have been done so very well. It looks like a ** Star Destroyer, and the Linux kernel (scheduler) suddenly looks like the Millennium Falcon. Real fast, but held together with duct tape, and ready to explode at any minute.

Theo de Raadt, founder of OpenBSD, compares OpenBSD development process to Linux:

    "Linux has never been about quality. There are so many parts of the system that are just these cheap little hacks, and it happens to run.” As for Linus Torvalds, who created Linux and oversees development, De Raadt says, “I don’t know what [Linus] focus is at all anymore, but it isn’t quality.”[15]

Those quotes don't really address the points I made. I said that Linux is (once configured) more stable, more respectful of a users privacy, and allows for better protection of data integrity than Windows.  I don't pretend it's perfect, or that Windows doesn't have it's own advantages.  Those advantages just aren't in system stability, user privacy, or data integrity.  One key difference between Linux and Windows is that people can look at the Linux code and make critical comments, which often produce improvements, or just fix things themselves (like Con Kolivas has done on several occasions).  No one who doesn't work at Microsoft can see the Windows kernel or the various software stacks, so there's no way of knowing if it's better designed except by it's outputs (which are objectively worse in some ways and better in others).

I think this has gone about as far as it's likely to go productively, and is quite off topic by now. I apologize for my part in de-railing the discussion, and I'll show myself out.

[kstuart]

I'll leave the tangential sub-thread alone - we've both made our points.

Meanwhile, I looked further at the Audiophile version of Arch Linux, and it does resemble your remarks about a undermanned project.  There has been no release for a year, and one Forum mod admitted to not having enough time to devote to the project.  From reading the Forum a little bit, it looks tricky to install, which seems unfortunate given the lack of Forum participation.  So, I am unlikely to mess with that.

In contrast, Linux Lite has significantly more participation, has a release a few days ago, and installed in a easy and seamless fashion (probably the easiest and quickest OS install I have done, which is remarkable considering I allowed it do its own automatic partition-shrinking.)

In the last month, it has gotten as many hits at DistroWatch than Arch Linux, and over a longer period, has significantly more than Xubuntu (Ubuntu with Xfce).  So, it seems to be one of the larger distributions after the top 3 of Ubuntu, Debian and Mint (perhaps we can call those "medium size projects").

Quote: "If you're looking for something that looks and acts a bit more like android, try the Gnome desktop; it won't run particularly well on very old hardware, but is very, very different than Windows. "

I'll give that a try using a Live CD run...

[BryanC]

I'll leave the tangential sub-thread alone - we've both made our points.

Meanwhile, I looked further at the Audiophile version of Arch Linux, and it does resemble your remarks about a undermanned project.  There has been no release for a year, and one Forum mod admitted to not having enough time to devote to the project.  From reading the Forum a little bit, it looks tricky to install, which seems unfortunate given the lack of Forum participation.  So, I am unlikely to mess with that.

In contrast, Linux Lite has significantly more participation, has a release a few days ago, and installed in a easy and seamless fashion (probably the easiest and quickest OS install I have done, which is remarkable considering I allowed it do its own automatic partition-shrinking.)

In the last month, it has gotten as many hits at DistroWatch than Arch Linux, and over a longer period, has significantly more than Xubuntu (Ubuntu with Xfce).  So, it seems to be one of the larger distributions after the top 3 of Ubuntu, Debian and Mint (perhaps we can call those "medium size projects").

The first rule of picking a linux distro is to ignore the distrowatch 'popularity' list. It's for advertising purposes, and the website owner has admitted as much. On top of that, people that have chosen a distro don't spend their time browsing distrowatch (do you really think Linux Mint is the most popular distro?--preposterous). What you should be looking at are the archives and activity level of the distro's forums, bug reporting system, longevity, backing organization, etc. For instance, the Ubuntu forums have 2.1 million threads compared to 3000 threads in the Linux Lite forums. There are 2 million members on the Ubuntu forums versus 1000 for Linux Lite.

The advice you've received in this thread thus far from numerous posters is spot on. Linux Lite is just another of a gazillion debian/ubuntu spinoffs that flies by night. Stick to one of the big distributions that is backed by a company or an organization or you will get hung out to dry. Not a matter of if, but when. Linux Lite is led by one guy. What happens when he gets sick, bored, loses his job, etc? Companies can survive that but not small projects like an Ubuntu spinoff. You can make your own decision but you did ask the forum so I feel obligated to warn you.

[aoqw76]

I would say xubuntu. It's pretty light on resources, and as far as i can recall, did not require any fiddling to get jriver mc installed and working.
My other choice would have been lubuntu but i couldnt get remote desktop to work (as in, i could not connect to it remotely from windows over vnc) so went for xubuntu instead and that just worked.
I have used mc for linux since a version existed, and while it's true the early versions were temperamental, it was not an os issue. Latest version is very stable and i leave it running 24/7.

[kstuart]

Stick to one of the big distributions that is backed by a company or an organization or you will get hung out to dry.

You've convinced me, your logic is perfect, and that means installing Windows 7 - and with Windows Media Player instead of MC21.

A Forum with 2 million members is essentially equivalent to the Internet.  I can jump up and down in such a Forum and be unnoticed for 10 years.  In a group of 2 million members, everyone is a stranger, and status competition becomes the order of the day.

1,000 members is plenty.

Not a matter of if, but when. Linux Lite is led by one guy. What happens when he gets sick, bored, loses his job, etc?

* Format the partition
* Download an iso of a different distribution
* Install it

More evidence that Linux has the same ridiculous issues as Windows:

http://www.omgubuntu.co.uk/2016/03/ubuntu-drops-amd-catalyst-fglrx-driver-16-04

[Awesome Donkey]

Ubuntu 16.04 depreciating fglrx is due to AMD not supporting X.Org Server 1.18 in the proprietary driver. They're working on the AMDGPU hybrid driver, which should be out at some point this year (with GCN 1.2+ support at first, GCN 1.0-1.1 at some point in the future). And it isn't just Ubuntu that's affected by this, either.

If you have a pre-GCN AMD card, your only choice is the open-source drivers.

[mwillems]

More evidence that Linux has the same ridiculous issues as Windows:

http://www.omgubuntu.co.uk/2016/03/ubuntu-drops-amd-catalyst-fglrx-driver-16-04

Actually this is another key difference between the platforms, but is one that cuts in Windows favor.  Discrete graphics cards work better on windows, and AMD in particular is reknowned for their miserable Linux support with their proprietary drivers.  Nvidia cards are generally well supported on linux with the proprietary driver, but the performance is still incrementally worse than on windows.  Intel has published actual opensource graphics drivers and their igpus perform about as well on Linux as Windows.

In any case losing the proprietary driver isn't the end of the world as there are community written open source drivers for amd and nvidia as well, but while they generally support a wider range of chips than the proprietary drivers (older chip support hangs around longer in the open source drivers), they've historically been much lower performance.  

So windows has a performance advantage; there are non-proprietary options on Linux, which is nice, but the best you can hope for is similar or slightly worse performance than windows (with Intel hardware, or nvidia hardware with the proprietary driver).  AMD has been (for years) the worst case for Linux graphics among the major manufacturers in terms of performance and official driver support.

[kstuart]

And it also shows that just getting rid of one giant corporation that does not act for the benefit of users - MS - still leaves you with several other giant corporations involved (in this case, AMD).

So, also on the subject of hardware drivers...

My inspiration for this project was two-fold -

One - to see the current state of the art in Unix desktops (at least in terms of Windows-like ones, I still haven't gotten around to the Debian/Gnome Live DVD).  I found Linux Lite 3.0 which provides an automated install easier than any Windows install (including shrinking disk partitions automatically), and provides virtually every capability found on Windows for almost all consumers.  I installed LL 3.0 on an old laptop with XP, and it works very nicely now, and I plan to leave it that way.

Two - to investigate the conventional wisdom that "Linux sounds better than Windows".

So, I installed MC21 on LL 3.0 - which installed exactly according to the 16.04 LTS instructions.  I spent some time tweaking the settings for best results.  I listened to my usual test tracks on MC21 Windows, then on Linux and then on Windows again.

While MC21 Linux sounded good, and did nothing apparent wrong, nevertheless it is pretty clear that my MC21 Windows setup has significantly more realistic instrumental tone and more low level detail.

I can only attribute this to hardware audio drivers.   While Schiit has been lobbying for MS to include USB2.0 Audio drivers, it may be that the drivers they developed with Cmedia are actually better than what MS or others might come up with.

This situation may be why Computer Audiophile still uses Windows - and JRiver MC for Windows - on their music servers.

[Awesome Donkey]

Drivers shouldn't make any difference with audio quality, actually. One could argue on Windows that the drivers bring ASIO and WASAPI support, thus bypassing DirectSound and the system mixer. But on Linux with MC, by default, ALSA is used which bypasses any system mixing as well (though Pulseaudio is also an option). On my PC with three operating systems (Windows, OS X and Arch Linux) all three sound the same in MC, which is a good thing.

Just make sure you're using the right ALSA device in MC and avoid using Pulseaudio.

I found Linux Lite 3.0 which provides an automated install easier than any Windows install (including shrinking disk partitions automatically)

Sounds like it's using the Ubuntu installer, which Mint uses as well. Yeah, it works pretty great.