Ok, the cooled down version. My apology for the whitenoise.

On 6/18/20 10:09 PM, Hermann Meyer wrote:

If you are still interested in the future of Linux Audio Session
Management you should really be happy that the Session Management is now
in the hands of linuxaudio.org

It's positive that people are still working on it indeed. It's also a good thing that people who prefer a QT gui, have a option to use Argodejo (new GUI for NSM). It's positive that nsmd and a gui (argodejo) will be easier accessible for regular Linux users and that it's chosen to be the session manager to focus on and to recommend for Linuxaudio. It's positive that they stick with the original NSM design, not reinventing the wheel.

I'm not positive about the renaming to new-session-manager though. Non-session-manager is around for many years now, renaming it, adds confusion, also in relation to the existent documentation around on the Internet. I would really reconsider whether this change of name is needed and a wise thing to do. This part is also the 'hijackish' part for me, I know, it's free software, but still, it's a bit too much claiming NSM now with a whole plan behind it.

It was Jonathan himself who suggested to put nsmd in it's own repository, so it would be easy to package Argodejo with nsmd for the variety of Linux distributions (the dependency on NTK and using waf made it difficult to get it in distros like Debian). It would be better to settle around with the original developer on how to organize this. But with a bit more patience, this drastic approach was not needed I think. If a fork turned out to be the only workable option, it might have been better to just fork nsmd and rename that, and don't touch the non-session-manager (GUI) at all. The version available on the new github repo is not a fair replacement for the original GUI at the moment anyway. It has problems with the graphics. It's also not the intention to develop it further, so why fork and rename it.

The most important change is the change of the maintainers/developers of course. All though I'm sure it has positive sides in some aspects of it, don't forget we have NSM because of the work of Jonathan. He has come up with a session manager that works better then previous attempts, all though technically less complex it seems. Quite a accomplishment. It might sound good to have it labeled under 'linuxaudio' now, and aim for a more community driven development, but being nice and easy to each other isn't always good for the software. In fact, one of the reasons why NSM works, is because the original developer didn't make all kinds of compromises, for the sake of being a nice and friendly community.

So, a new way of maintaining the software, isn't success guaranteed. If I compare Carla with the NON tools for instance, it's a totally different approach to software development. Where NSM is minimal and restricted, Carla has lots of features, even quite some experimental ones. I'm not judging about which approach is better, I'm happily using both of them, but I'm convinced that NSM is a success because it's minimal, restricted and without compromises. So for session management a minimal approach is surely a better one imho. Also building a GUI for NSM, is different coffee then actually judging about if and which features should be implemented in NSM and how that should be done.  So, I've to see how this different maintainer-ship will turn out, but the original developer has proven that he 'gets it' when it comes to session management, for the others it's something we'll have to wait for and see. But not having the original designer of NSM as maintainer for whatever reason, I see it as a loss, not as improvement.

Anyway, at the end of the day, the real important thing which improves user experience significantly at the moment is adding NSM support to applications. Hacking around applications without NSM support, without adding NSM support itself as I also see happening, is time spending on the wrong things and adds only to a more confusing, less stable and a worse user experience imho. I'm glad this Argodejo/NSM project doesn't seem to take this path and I hope it keeps that way.

It has been said that there are not enough applications adding NSM support. I think a list of +- 30 applications isn't that bad at all actually, with quite some important ones included. For myself I've just a short list of applications that would have been nice to support NSM: Radium (in next release), Seq64/66 (work in progress), Hydrogen (work in progress), Carla-single (in roadmap), Muse (possibly maybe), Rakarrack, Sooperlooper, (Luppp needs a bugfix).

In meantime, the original Non-Session-Manager with the NTK toolkit is still available, easy to build and as I prefer it above Argodejo myself, I'll still be heading to the NON project and more will do so most likely. So you can state that it will replace the Non-Session-Manager, but at the end the users decide I guess and for the user it is actually just a matter of which GUI you prefer for the Non-Session-Manager, with the advantage of Argodejo being easy to package for Linux distributions and having a more responsive and active developer.  I think this could have been offered to the user in a less radical way, without releasing 'New-Session-Manager' (which has a lot of characteristics, but being new is not one of them).


My 0.02$
_______________________________________________
Linux-audio-dev mailing list
Linux-audio-dev@lists.linuxaudio.org
https://lists.linuxaudio.org/listinfo/linux-audio-dev

Reply via email to