Re: The future of Solaris?

From: Greg Smith <gsmith(at)gregsmith(dot)com>
To: Liraz Siri <liraz(at)turnkeylinux(dot)org>
Cc: pgsql-general(at)postgresql(dot)org
Subject: Re: The future of Solaris?
Date: 2008-12-11 03:54:34
Message-ID: Pine.GSO.4.64.0812101944000.28476@westnet.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On Wed, 10 Dec 2008, Liraz Siri wrote:

> Linux may still be behind Solaris in a few areas but I'll wager Linux
> will catch up and make Solaris completely, utterly obsolete in the not
> too distant future.

Great, free money is even better than free code; how much would you like
to loo...er, wager on that? Could use a new sure thing now that there's
no more money for me to make shorting SCO stock.

I've been hearing this particular refrain constantly since 1996, when I
first switched to working full-time mainly on Linux systems. Every year
seems like it's finally the year for Linux, even on the desktop, yet
Solaris is still here. In fact, it's better than ever. Solaris has been
aggressively closing the gap with Linux the last few years in the things
it was most behind on, while keeping a lead in some areas.

As for things it's ahead on, ZFS is still way better than any Linux
filesystem, with potential challenger Btrfs so far off from being
enterprise quality that there's a ton of redunant work going into ext4
just as a stopgap measure. (You could easily argue that new features like
"Time Slider" suggest Linux has actually been falling even further behind
this year). DTrace is doing wonders for people every day, while potential
Linux competitor Systemtap seems to have completely missed the point that
the idea is to make it easy to instrument things safely.

And the biggest thing that used to keep me away from Solaris, how painful
the packaging made it to get a functional system with the usual GNU tools
all installed, has been getting better fast lately, and looks almost
completely cleaned up as of last month's OpenSolaris 2008.11. The thing I
think a lot of people miss is that most of the value of a Linux
distribution is not from Linux itself. Remember: Linux is just a kernel.
Combine a Solaris kernel with the rest of the usual GNU and other tools
you see on Linux distributions, and most people won't even notice the
swap. There will be less supported hardware, and it will be a bit slower
at some things, but at least the kernel will be stable moving forward.

Which brings me to...the primary thing that really bugs me lately is that
Linux kernel development is increasingly not focused on stable releases,
it's all about rapid innovation at any cost. And the anti-business
politics of some key contributors is really getting in the way of
pragmatic adoption. Check out this great rant from Theodore Tso about how
badly things are broken in that area:
http://thread.gmane.org/gmane.linux.file-systems/26246/focus=26492

The part I like is "sometimes people have suspected that some changes made
had benefits that were so marginal that it seemed that the main
justification was to screw over externally maintained
drivers/filesystems". That's sure how it feels to me. I have some
closed-source bits and some things that compile outside of the core kernel
that I rely on, and every new kernel point release I get breaks one of
them--often for trivial improvements and with *zero* regard even for the
transition periods promised by the kernel team itself. Here's the last
one I got personally burned by: http://lkml.org/lkml/2008/2/17/319
Planned deprecation period? "sadly it was untenable". And people wonder
why I stay as far back from the current kernel as feasible.

> Besides Sun Microsystems hasn't been a financially healthy organization
> for quite a few years, as evidenced by its rather dismal stock
> performance:
> http://finance.google.com/finance?q=java

Bah, Google Finance makes it hard to refer anybody to a specific chart.
Stupid AJAX. How about we stare at these three for a minute:

http://finance.yahoo.com/echarts?s=JAVA#chart1:symbol=java;range=5y
http://finance.yahoo.com/echarts?s=rht#chart1:symbol=rht;range=5y
http://finance.yahoo.com/echarts?s=novl#chart1:symbol=novl;range=5y

Wait, which one of those was the weak one likely to fail? They all look
pretty poor to me. Despite its recent fall, Sun still has the largest
market cap of the three. You say it's been going badly for "quite a few
years", but the only serious divergence from its competitors was only this
year. Sun would be doing better right now had they not decided to light
$1B on fire back in January, that's where their stock really accelerated
its dive downward.

RedHat is actually by far in the best financial shape of the three, at
least they make more money than they spend.

--
* Greg Smith gsmith(at)gregsmith(dot)com http://www.gregsmith.com Baltimore, MD

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Greg Smith 2008-12-11 04:20:21 Re: Startup process thrashing
Previous Message Joshua D. Drake 2008-12-11 03:43:03 Re: Data Replication