Overall, yes. But then, any sufficiently large demographic seems confused to me
because... well, people can be really, really different from each other.
If you're talking about the "I hate MS and everything associated with it"
brigade, then yeah, those folks are really annoying and useless. Also, the
Linux "unpaid salesmen" (typically referred to as "fanbois") are really
annoying, because they're more interested in selling you on their particular
pet distro than actually being honest and helpful. They also almost always
belong to the aforementioned brigade, which makes them doubly useless.
The annoyance of too many Ubuntu fanbois helped convince me to switch to
Debian. I just found Debian to a more serious distro where you can find good
help without having to endure fanbois as much. I was able to put together a
Debian system to do exactly what I wanted easier than with Ubuntu.
Overall yes. But see my comment above about large demographics.
Yeah, that really needs to be addressed. I read an article about a guy who
proposed to have an API added to Linux to aid in the installation of software.
He got a bunch of developers of various package managers together, and got some
feedback from them. The conclusion seemed to be that most of them were planning
to simply keep doing things their own way, weren't all that interested in a
standard Linux API for installation purposes, and didn't really have much of an
interest in working on a more common solution. C'est la vie. So what we have is
a bunch of package systems, and app developers who don't support them all
because it's too much of a pain in the ass.
When I package my software, I make a deb, and that's it. It's not that I have
anything against other package managers, but I happen to use apt-based distros,
and I just don't have the time nor inclination to bother with the package
managers of other distros. If those folks don't want to make it easier for me,
then I'm not interested in them.
Because nearly all of them have their flaws, being designed by people who have
some good ideas, but unfortunately, do not seem to choose to have those ideas
peer reviewed by appropriate people (other sound developers, musicians, etc)
before coding starts. This is a really big problem with lots of open source
development. People start coding without taking the extra step of first writing
up some sort of detailed documentation about how the API will work, and getting
that peer-reviewed by other folks who could give good feedback (and maybe
propose some changes that will head off a lot of future dissatisfaction with
the finished code, and forestall a reason for those other dissatisfied folks to
start their own sound system. For some reason, those other dissatisfied folks
repeat the same mistake of not doing what the first programmer should have
done).
I always write up the docs for my software before I even start coding. I
*never* use any sort of utility that creates docs from comments in the source
code, and I think that stuff should be outlawed because it encourages people to
fail to do enough "flow charting" and proper peer review before they start
coding.
Probably for the same reasons as above.
I'm sort of with you here. But actually I want more real choice. What I'm
getting with Linux is a bunch of choices that are, to me, pretty much all the
same thing. For example, I don't want a choice between a GUI API that runs atop
of X (and inherents its inherent limitations and design flaws) such as Gnome,
and another GUI API that runs atop of X too such as KDE. I want people to
support more real choices, like instead of dividing up programming attention
between those two, maybe support something that really is much more different
(such as directfb). I don't want a choice between Pulse Audio (ick) or ALSA
(better than Pulse Audio, but still with its problems -- problems that Pulse
Audio inherents because it rides on top of ALSA). I want a choice between sound
systems that really do have a radically different approach. Etc.
There are way, way too many Linux "choices" that are far too similiar, have the
same basic set of features and limitations, and seem to exist simply because
people aren't collaborating better. I'd rather see more collaboration among
people who are doing pretty much the same thing, and encourage "choice" where
it really offers something significantly different. The problem Linux has is
that people aren't collaborating enough when they should, nor are they doing
something significantly different when they should. We're getting choices that
aren't enough of a choice.