On Mon, 21 May 2007, Ian Murdock wrote:
On 5/17/07, Alan DuBoff <[EMAIL PROTECTED]> wrote:
On Thu, 17 May 2007, Ian Murdock wrote:
> Sure, Fedora is a great QA vehicle for Red Hat, but what about the hordes
> of people who are putting Fedora into production (and, yes, there are a
> lot of them--this is a different group of people than the group Solaris
> targets today).
But this is the same group of people that OpenSolaris would target. There
is not much difference. Most enterprise will pony up with RHES, just like
they would want to run Solaris.
Actually, I see them as two very different groups. The difference is
primarily the entry point.
I agree for the most part, the difference certainly is the entry point.
The problem for existing customers is that most are slow to upgrade. They
won't use nevada until it's been released and proven in many cases.
People are changing, and newcomers are more willing to use
opensolaris/sx/nevada as it is, but this doesn't say anything for the
large institutions, corporate 500s, and/or government affiliates that use
Solaris as it's been known.
These are the customers I see continuing to use Sun's Solaris.
We have a much larger opportunity to grow our community, and projects like
Indiana can only be good for that, IMO, and OpenSolaris is the piece that
will let us do that.
The latter market is what we should be targeting with OpenSolaris--can we
make OpenSolaris the OS those developers will reach for?
Yes, I think we can. We can do it by standing our ground with much of the
foundation as it's been done. I mean this in regards to letting the world
know about our goodness, and some of the problems Solaris/OpenSolaris have
been able to solve today. Sun has solved some of the most difficult
problems already. While many know and see things that need to be changed,
or what some call "linux-like", much of that is smaller problems. I would
much rather have the world's smaller problems on the plate than trying to
tackle implementing some of the larger ones...like a 128-bit filesystem,
for instance.
But how do we capture the network effect value? Fedora has no network
effect value for RH--where's the upgrade path to RHEL, or the big
company to call when you need support? This is where can do better with
OpenSolaris..
I don't see it so cut and dried. I see customers that must have support,
or need assistance when they have problems (i.e., the world doesn't all
know Linux in all cases either), to pony up and purchase RHES. For those
that need the support, it's there. No, it shouldn't be free, and that is
what Fedora is for, those that don't need the support, or acknowledgement
that the packages have been kinda tested together, don't need to pony up
their $$$s. I say it that way, as I am not completely sure how RH does
testing on RHES, but know that many corporate feels it's enterprise ready.
When Sun states that Solaris is enterprise ready, it's been put through
it's paces, up, down, and sideways. Still chances that a bug could crop
up, engineers are only human afterall, but I have a hard time beleiving
that RH puts RHES through it's paces like Solaris. Solaris/OpenSolaris
rides on this tradition.
I can already hear it: "Oh, that's just WRONG! You should NEVER put an
UNSUPPORTED OS in production!" Perhaps. But that's the way the world has
evolved..
I would say that some of the world has evolved to that stage, but
certainly not all.
All situations are different and each one needs to be looked at on a
case-by-case basis, IMO. Some will most likely be able to fit in the
model, and others won't. Hard to put a blanket over it all, in my
experience working in the high tech industry.
No CIO in the world said, "I've gotta get me some Linux!", or at least
in the late 1990s when Linux was taking off. He woke up one day and
realized Linux was already everywhere. Do you fight the trend or figure
out how to take advantage of it? The answer seems perfectly logical to
me.
No, but many of them woke up one day and asked, "who's running that
mail/web server and where is the machine?", only to find out it was setup
and running in the backroom, and performed much of the same function that
commercial solutions, such as Solaris, offered.
I have been there and done that, it certainly does/did happen. This is now
an opportunity for Solaris Express and OpenSolaris.
--
Alan DuBoff - Solaris x86 IHV/OEM Group
_______________________________________________
opensolaris-discuss mailing list
[email protected]
