> Is it? What was the W in SUNW?
"Workstation".
> What are the roots of Solaris? If we add
> Sun Ray Server, which is a Sun core technology relying on Solaris to the
> list, isn't Solaris a Server and a Desktop OS at the same time?
Solaris always was a server and a desktop, but that's not the point.
SunRay is a thin client, a modern day version of a dumb terminal, and it is
very easy to mistake it for a desktop solution; dumb terminals were connected
to servers, and SunRay is doing exactly that, connecting to a server. It has
nothing to do with the desktop, it's just a glue, a fancier display for one or
many
of that server's users.
(I took part in a SunRay pilot project at one Swiss institution, which
better remain unnamed.)
And even SunRay wasn't fully cooked, for instance, one could not connect a DVD
recorder to burn media, and audio was a nasty hack with a bunch of files being
dumped in /tmp.
Solaris used to be about quality first; nowdays, it's just about getting it out
of
the door as fast as possible, because someone else has already done it.
But if someone else has already done it, that means one is no longer a leader,
but a follower.
And obviously, Sun had to make a change from being SUNW to a server company,
because luckily, back then, they determined that there was "no bread" in
workstations
any more.
sgi in comparison stuck with the workstation paradigm, that's why very few
places
know that they made servers. And where are they now? Why, they're dead, eaten
up by rackable systems, for peanuts, a company selling PC-bucket servers.
There is a lesson to be learned from that.
> Of course not, nobody can be a master of all trades. That's why we rely
> on other projects like Gnome etc.
But contorting Solaris into Linux to make it more appealing to those GNOME guys
doesn't work; that alone is not appealing enough to them; on the other hand,
a better mousetrap like ZFS might be.
What that does is draw in all the "me toos", which don't have the technical
insight
to even understand the issues, when they make a request for Solaris to behave
more like GNU; those are NOT the people that will be writing any good or
quality code. Wrong crowd. And that too, is painfully obvious, at least to me
it is.
> AJAX? Not sure if I want that...
Well, maybe that's the problem: maybe you really aren't sure what you want -
that's why there are Steve Jobs and Larry Ellisons of this world to tell you
what
you want, and make tons of money doing it.
I know exactly what I want, or rather, what I need. I don't need uncle Steve or
uncle Larry to tell me what I want. And what I need is ZFS deduplication
in Solaris 10, because that's something that can be used NOW to deploy in a
critical
environment.
My point: ZFS dedup should have come out in Solaris 10 FIRST. Then after
PRIORITIES
have been taken care of, hack on it all-day-long in Solaris 11, or OpenSolaris,
or
"watchamacallit", play with it, build castles-in-the-sky or what have you,
because the
customers that actually pay for this stuff, and really DEPEND on this stuff
will be taken
care of.
They call that "good business sense", and also "good business practice".
> Where do you draw the line? Some say that ZFS is not in line with the
> UNIX(R) philosophy, still it's a great piece of software. And it was not
> a "Toyota Kaizen" style of development. It was about "alte Z?pfe
> abschneiden", throwing obsolete things out, thinking about new ways.
Sure it was, it was more kaizen than anything, an elegant solution.
A better mousetrap, built upon decades of experience and improvement.
An EXPERIMENT like IPS is not built upon decades of improvement
(hence it's written in Python, which is yet another
experiment-within-experiment),
it threw away things that are actually needed (pre- and post- install and
remove,
we already had this discussion), and so on.
Too much of an experiment, and not one single concept from anyone else taken
and implemented COMPLETELY. Not one!
But the really important thing is that someone can say
"look! It works just like Linux!"
Is that what's important?
> Solaris is a damn great Server OS, just like OpenSolaris will be
> someday.
Someday! But that day is not TODAY.
Meanwhile, I guess people like me should just "limp along" and be forced to
do their own engineering, instead of focusing on development on top of the
OS? Nice.
Again: what do I need Sun engineers for, then, if I have to do it myself?
> But OpenSolaris is also a fine Desktop OS for technical people.
> It can even be a great desktop for business users.
Desktop is dead, you read it from me. And companies like Google will prove it.
I don't care about shiny new gadgets and gizmos in OpenSolaris, when even
the most basic stuff like dhcpagent(1M) isn't working correctly, and when
I open a bug request, I'm told to provide more information, but the system
doesn't allow me to. Meanwhile, dhcpagent(1M) is still busted! But yeah, there's
a new version of IPS out, oh, OK! Never mind that basic network connectivity
doesn't work correctly.
So much about desktop.
> "Rome was not build in one day."
> --John Haywood (1497-1580)
OK, now if you could just explain to me, why it was decided to burn Rome as he
existed (SX:CE) and decide to start anew, instead of refining and improving
what was already there?
And how in the world can one company, or even a community, bet their very
existence
on an unfinished experiment?
> Things can never be done to satisfy everyone, and sometimes we have to
> agree to disagree.
No; Larry and Steve will not "agree to disagree" with you; they know what needs
to be
done, and how it needs to be done, and so they will do it. That's why they are
where they
are, and why others are not.
_________________________________________________________________
Windows Live: Friends get your Flickr, Yelp, and Digg updates when they e-mail
you.
http://www.microsoft.com/middleeast/windows/windowslive/see-it-in-action/social-network-basics.aspx?ocid=PID23461::T:WLMTAGL:ON:WL:en-xm:SI_SB_3:092010
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/ug-chosug/attachments/20091118/a46b11c4/attachment.html>
