[oi-dev] Desktop Illumos Still Matters

[Nick Zivkovic]

I was just reading through the oi-dev archives (about the resignation
of the OI lead) where a former Sun engineer, claimed that the efforts
made by sun to port OpenSolaris to the desktop/laptop platforms (by
adding wifi support, by making a new audio system, etc), only hurt Sun
in the long run, and that it was all for nothing.

I agree that it probably hurt Sun in the long run. But I promise that
it was not for nothing.

I started using OpenSolaris some time in 2006 or 2007. I was 16 then,
and didn't know anything about programming. The availability of the
OpenSolaris distribution, is what gave me, a mere mortal, the ability
to store files in a 3-way ZFS mirror (a teen would not be able to
afford an enterprise RAID card). To play with DTrace and illuminate
the inner workings of the system. OpenSolaris is the platform that I
started writing my first serious code on. It is the platform that I
write code on to this day. It was and still is a joy to use DTrace
daily, and to transport data between machines as ZFS datasets, and not
as tar-balls. It is bliss to have inline compression and
deduplication. Fast and free snapshots.

Most Illumos devs take these things for granted. They are all
professional engineers, whom I respect deeply. But in 2006 I was a
mere teenager, who was merely enthusiastic about administering
unix-like systems. I guess switching between Linux distros, and being
different made me feel "elite".

But the truth is, I was a pretender. I was a mere hobbyist, looking
for a distraction. Something that would make life less boring and
humdrum. I meandered about the web, not knowing what kind of
distraction I was looking for.

Funny youtube videos? Yeah those were okay. But they got stale and
weren't as stimulating.

Video games? Stimulating, but tiring. Not fulfilling. Almost
masturbatory in nature.

Porn? Same as video games, though slightly more effective.

Going out / hanging out? Only in short bursts. Doing it constantly
made me tired and fatigued.

Setting up a Linux distro had its own kind of high. It lasted longer.
But once you've set up a distro, you can't really leave it alone. You
either have to try another distro, or upgrade the existing distro. I
can't explain how I longed for upgrades. Especially on Gentoo. That
was the high of _assembling_ something that worked.

Of making an otherwise blank machine come to life, and operate in a
customized manner. (As customized a manner as possible for a
non-coder).

But it wasn't until I tried OpenSolaris that I made a decision to
learn to code proficiently. For several reasons. The burst of
innovation that came with OpenSolaris (ZFS, DTrace, etc), made me
aware that software/computing had unsolved problems (or problems in
need of better solutions). The availability of DTrace, and the
prospect of being able to see what the machine is doing (and hence
what my own code is doing), was too tempting a technology to ignore. I
understood that the technologies in OpenSolaris would best facilitate
my new-found need to create, to build, to engineer solutions to
problems. Namely my problems, but problems nonetheless.

But if OpenSolaris were not usable on the desktop -- If it didn't have
support for things like audio and wifi -- I would never have
considered it. I would never have decided that writing code might be a
good idea. I would have never discovered one of the most exhilarating,
blissful, and strangest activities that we humans have come up with.
Coding has been the most profound and intense experience of my life. I
don't know exactly when it started changing me, I just know that it
did. And I am glad it did. If I hadn't chanced down the OpenSolaris
road, I probably would have ended up an uninspired burnout, a wasted
bag of cells.

If Illumos were to disappear from the desktop. Or even from the world.
I would be disappointed. But I would move on, because I already know
the bliss of writing code, and I'll do it elsewhere, even if the
environment and tools are not as agreeable. One wouldn't be able to
stop a writer from writing his heart out, by taking away his
typewriter and giving him a quill and parchment.

But when I think that there is some kid out there who is bored out of
his skull and wandering aimlessly, because, quite frankly, he is too
smart for average activities, who may never learn the joys of the art
of programming because some developers decided that they're content
letting Illumos's basic desktop capabilities atrophy over time as new
devices replace old devices... well, my heart breaks.

Because these kids might be those that never get inspired, and waste
their lives on stupidity.

I'm not saying that the Illumos engineers are doing the wrong thing.
In fact I think they are all doing the right things. I hope Illumos
retains the desktop essentials (usb1/2/3/N, wifi, audio, modern
browser); that's all that's needed to get someone started down the
path of writing code. Desktop Illumos isn't supposed to replace Ubuntu
and Mac. It is supposed to attract a certain kind of personality, that
will _create_ more than they consume. (believe me, I'm writing
fascinating code, that I'm not ready to reveal yet; well, I find it
fascinating).

All I am saying is that those seemingly useless projects (wifi, audio)
undertaken by Sun, were not useless to me, and not useless to Illumos.
And I suspect such projects will not be useless to future generations
of engineers, and future revisions of Illumos.

My apologies if this email is discombobulated or off-topic (it is very
late here, and I am out of coffee). But I needed to respond, and offer
opposition to the idea that desktop capabilities were without benefit.
Especially since I was one of the (presumably many) beneficiaries of
OpenSolaris. The benefit is happening at the edges of graph. And I
think that this benefit will ripple back to the center of the graph,
when the inspired amateurs, begin contributing code and seeking jobs
at companies that hire Illumos talent.

Thanks to everyone from Sun and the Illumos community.