One of the important steps in setting up a system for music production is selecting the software. This is of course no different when using Linux for music. And since almost all of the software for Linux is open source, there are some huge advantages, but they do come at a small cost. But if you know what you are looking for this is no big deal. So here are a number of things to take into account while selecting your software for Linux music production:
Before even looking at available software, it pays to define your needs. Don't just take your old software for granted and try to find a Linux replacement. Moments like this are great opportunities for thinking about what you want to accomplish with the software. Perhaps there is a better way to do it. Applications that are better suited to your objectives might well exist in the open source domain. Once you have sorted this out, you can start looking for the applications that will fit the bill. There is a list of Linux music software at Hitsquad. The list is extensive and well maintained, but it also sports many applications that are not supported anymore. While comparing alternatives, compare their functionality with your wish-list: does the software do what you want it to do? What are the functions that you can't live without and where are you willing to make concessions. Using Linux for music is still a relatively new experience, and while there are some rather mature applications in some fields like Ardour (recording) and Rosegarden (sequencing), other areas like music notation are still in their infancy so to speak.
Look & feel
Personally I find the way the software works, looks and feels also rather important. While this isn't really a "deal-breaker" for me, it is an aspect I take into account. For example, I would rather use a program with a few less features that I can work with in a way that is congenial to my style, than 'great software' that is hard to work with. So ask yourself: does the software work in the way you want it to work? Since I'm fairly new to using Linux for music, I'll use my Windows notation software as an example: when I first started out using computers for music notation (some 15 years ago), there where a number of applications available that I tried, but for some reason I found the way Finale worked to be quite natural for me. The way the software was conceived felt natural for me, even though many people complained about its steep learning curve. Later Sibelius came around, and because many people lauded its easy way of working I gave it a serious try. But while I found it a very good application for notating music, working with it never felt as good, as natural, so my end results suffered. Back to Finale it was.
In the proprietary software world it is quite hard to assess the situation of a company and what the future will look like for upstart companies. So there is a tendency to go with the established players, after all your not only investing your money in this software, but also your time and your projects. In the open source world, this is made a bit more transparent because in most cases, for most projects you can actually take a peek in the garage - so to speak. You can see who the developers are, how many there are, how often the software is updated, how bugs are handled, etc. You'd want to know if the software is well-maintained. It is a reality that not every open source project proves feasible, and many die in good intentions. So it pays to check if the software is being actively developed. But there are more than enough applications that are maintained very actively. And this automatically gets me to the next point:
You can become an active player in determining the success of an open source project by contributing some of your time and talents to the projects! Consider contributing to the projects of your choice. You don't have to be a developer to be of value. There are many options available. Providing feedback is of course easiest. You can help the developers tremendously with your ideas and experiences. And you can also spread the word about programs you use, thus helping to enlarge the user-base and bring other people on board. Some projects have a need for translators into different languages, so maybe you could help there. You can also contribute to the program's documentation if you have a bit of a writing skill. And of course if you are a software developer you could provide your services in programming, even if you don't have much time. Give this a bit of thought. Consider it a "thank you" to the people who invested their time and effort in creating the software for you to use.
Narrowing down the selection
Just as with the selection of a distribution, it pays to experiment with the applications of your choice. Don't invest too much time and effort yet, just experiment to see if the program does the things you want in the way you want it. Be open to alternatives, they might just provide what you need! After fiddling around a bit with a program you can make an informed decision whether you want to go ahead with it or not. Since almost all open source software is provided at no cost, you can easily try out an alternative. There is only the time investment to take into account. Of course at a certain moment you'll have to commit to a certain set of applications if you are to get any creative work done!
My personal choices
For now I've selected Denemo for music notation and Ardour for my recording needs. I have worked with Ardour before a few years ago, so this one is likely to stay. But since I'm still experiencing some hardware trouble, real testing has not yet begun. As for Denemo: While I like the Lilypond output, I'm not sure either Lilypond or Denemo can offer what I really need. I know there is a notation editor in Rosegarden, but I'm not convinced of that one either. Since I produce a lot of sheet music my demands on both work-flow and output are quite high. And I am not sure there is Linux software yet that can cater to my needs. I have not (yet) looked into the sequencing applications because I really have no need for it at this point. This might change in the course of this year though. Also, my personal choices are influenced by the fact that both Ardour and Denemo are GTK (Gnome) based, which is my preferred desktop environment. I'm still in the experimental phase and my time to "test-drive" is limited, so these choices might well change in the future. Anyway, I'll keep you informed of my experiences and choices as I go along. In the meantime I'd like to encourage you to share your experiences with Linux music production software in the comments.