Well, I've been expecting email like this. The NetBSD Project has really
stagnated. As a NetBSD user, I can agree with Charles in most cases.

> 1) There must be a strong leadership, and it is not the current one.
It should. But there also should be a productive community and core
team, which does the main things, but also helps and follows the
leader. Someone should be allways ready to take over the leader's role
if it is necessary.
Right now, even our core team is absolutly passive. In fact, I don't
see a real core, only a group of developers, who works more separate
than in team, and has no resoluteness to lead the all others.
It is really not for me to ciricize core, and I even don't know how
they work inside, but I can state how it looks from this side.
Even with concrete things - for example PF and ALTQ. We cannot integrate
this in structurically correct way, because we need an API. But there
is a need in community of this feature NOW. If there is no "strong
hand", no resolutenessof trying collaborate with other BSDs on this
work, then make a decision for temporary workaround. Three years
elapsed and not much was done...
We can also look at our sys/kern, sys/net, sys/uvm, etc. These parts is
mostly sleeping. As mentioned, no work with threading, SMP, where is
our AIO?.. Just periodically somone integrates few small things or
fixes, but there are no daily work on these parts. Of course, there is
developers who does great work on other parts. But at the current time,
probably only Yamamoto doing some larger work (with our UVM) in NetBSD's
heart. Well, of course we have quite new, good thing - Google Summer of
Code (in earlier times, we had KAME), which really helps us to integrate
features. But event if we look at FreeBSD's core part - they're doing
commits allways. Dillon in DragonFly does the commits daily. Theo in
OpenBSD too, even if their sys/kern is nothing better then ours
(sometimes I see fixes in their tree, which we done few years ago) and
they're just syncing it with us. But they have something we don't. It's
an idea and philosophy, which leads and inspires them to work. It
should not be only declared, but also felt. Even such things, like
"hackathons" gives not only technical benefit, but also emotional,
feeling that there is a team, which can work together productively. We
have an example near by - it is pkgsrcCon. So there is no such
conferences, meets or "hackathons" in NetBSD community?..
NetBSD, probably (I'm not sure), has the smallest community if we
compare with FreeBSD and OpenBSD. And I'm sure that it is not because
of technically things. Yes, I agree that we don't need a hype, and I
allways believed that quality is more important than quantity. But such
a small thing - there was a suggestions to replace a NetBSD.org website
with more "candy" (by the way, which looked nice even with lynx or
links), but nothing has done. Maybe it looks not important, but what
can user think, if he sees same, old, "dry" website many years?.. No,
not about stability. Yes, I like NetBSD because it is light and clean,
but there is some things, which people would not change only because of
some, probably, religious purposes, even it has no influence to
usability/quality/confortability/whatever. I believe that even
marketing has imporant influence to this (by the way, why we cannot
sell NetBSD CD's like OpenBSD does?).
And so on, and so on...

One of really essential point is direction of NetBSD. I see this system
in two directions - embedded systems and servers. In my subjective
opinion, we should concentrate more on these two things, rather than
others. Developing of desktop, multimedia and similar enviroment is
good for us - our developer and a small community. But with this small
count of developers and users, we cannot compete with other projects,
i.e. Linux distributions. We physically cannot provide some things
better, than huge i.e. Linux's communities.
And we shouldn't. We shouldn't work to wide. More we concentrate to our
direction, more we will be strong on that. It is simple that dispersed
energy is less stronger than concentrated one. These things is aimed
to NetBSD, but not to pkgsrc (by the way, I would like to have
something like pkgsrc-server branch, which would be updating ASAP)

At this point, I must turn one's attention to pkgsrc. We have a good
example, how our's (TNF) project lives quite good. Even many developers
work both the NetBSD both the pkgsrc project, but I think the latter is
not lacking of problems mentioned by Charles. Probably we can learn
from ourselves :)

Sorry, if my tone was unacceptable. That is just my few subjective
thoughts. 
I very hope, that this topic will influence NetBSD project.


-- 
Best regards,
Mindaugas