What if Sun wrote the Solaris kernel as a drop-in replacement for the Linux kernel?

CK writes:

>Is it possible to shed some light on the reasons why this didnt happen
>earlier?

>I mean, what could have possibly gone wrong?

It is very difficult to explain how things happened as they happened;
but the most important thing at the time was lack of management buy in
and lack of interest in most of the organization.

>In a simplistic view, the company always wins (no I am no OSS FSF
>advocate). You can basically get other companies and people to do your
>development for you while you can still sell the same product, so to
>speak.

One can't be sure of that.

Casper
--
Expressed in this posting are my opinions. They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.

In article ,
CK writes:
> Words to the wise, Casper H.S. Dik wrote:
>
>>>I believe that the only problem Sun has is that they didn't foresee
>>>Linux coming and they wish they made Solaris open source 10 years ago
>>>and have the position Linux is holding right now. They are trying to get
>>>there now (a little late) and it's an uphill battle.
>>
>>And the worst of it is that some people inside Sun knew at the time
>>they needed to do this ....
>
> Is it possible to shed some light on the reasons why this didnt happen
> earlier?

I think there was a long period when the value of Solaris
was not understood outside of the Solaris engineering group.
It was believed customers bought the systems because they
liked sparc, or they liked purple machines in the computer
room. There probably was a time when this was true, but it
stopped being relevant some time before the management
realised. You can see this in the killing of Solaris x86 9;
after all, why would anyone want this as it wasn't sparc
and it didn't come with hardware with a purple front?
The reaction the management got to this decision was the
was their big wake-up call to the value of Solaris, and
that was the point where they started taking it seriously
and giving it the attention it deserved. That is why you
can see an enormous amount of work poured into Solaris
itself resulting in all the new features in Solaris 10
(which actually started during the latter Solaris 9
releases after x86 came back).

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]

Words to the wise, andrew-at-cucumber.demon.co.uk (Andrew Gabriel) wrote:

>In article ,
> CK writes:
>> Words to the wise, Casper H.S. Dik wrote:
>>
>>>>I believe that the only problem Sun has is that they didn't foresee
>>>>Linux coming and they wish they made Solaris open source 10 years ago
>>>>and have the position Linux is holding right now. They are trying to get
>>>>there now (a little late) and it's an uphill battle.
>>>
>>>And the worst of it is that some people inside Sun knew at the time
>>>they needed to do this ....
>>
>> Is it possible to shed some light on the reasons why this didnt happen
>> earlier?
>
>I think there was a long period when the value of Solaris
>was not understood outside of the Solaris engineering group.
>It was believed customers bought the systems because they
>liked sparc, or they liked purple machines in the computer
>room.

So, Solaris was seen as a "add-on" to selling the hardware?

>There probably was a time when this was true, but it
>stopped being relevant some time before the management
>realised. You can see this in the killing of Solaris x86 9;
>after all, why would anyone want this as it wasn't sparc
>and it didn't come with hardware with a purple front?

Yes, indeed, why would anyone?

>The reaction the management got to this decision was the
>was their big wake-up call to the value of Solaris, and
>that was the point where they started taking it seriously
>and giving it the attention it deserved.

So someone realized that it is possible to make money off supporting
an otherwise free OS. That about it?

> That is why you
>can see an enormous amount of work poured into Solaris
>itself resulting in all the new features in Solaris 10
>(which actually started during the latter Solaris 9
>releases after x86 came back).

Yes, I was wondering about that and figured that Sun would probably
move towards a more software-oriented company, though I doubt this is
really possible without killing of hardware altogether.
--
Claus Kick
Computers are stupid, but they create work

Words to the wise, Casper H.S. Dik wrote:


>It is very difficult to explain how things happened as they happened;
>but the most important thing at the time was lack of management buy in
>and lack of interest in most of the organization.

So, too strong a hardware focus?

>>In a simplistic view, the company always wins (no I am no OSS FSF
>>advocate). You can basically get other companies and people to do your
>>development for you while you can still sell the same product, so to
>>speak.
>
>One can't be sure of that.

That is why I said, in a simplistic view. Being neither business nor
management, the only view I can hold is a simplistic one.
--
Claus Kick
Computers are stupid, but they create work

"Floyd L. Davidson" wrote:
>
.... snip ...
>
> For example, I'd bet that Vista will be the last major
> "build a whole new OS" project based on a proprietary
> kernel from Microsoft. Another attempt would destroy them...
>
> As always, interesting times are just ahead!

Another? Read the URLs below:

--



cbfalconer at maineline dot net


--
Posted via a free Usenet account from http://www.teranews.com

In article <4678359e$0$644$5a6aecb4@news.aaisp.net.uk>,
Andrew Gabriel wrote:
>In article ,
> CK writes:
>> Words to the wise, Casper H.S. Dik wrote:
>>
>>>>I believe that the only problem Sun has is that they didn't foresee
>>>>Linux coming and they wish they made Solaris open source 10 years ago
>>>>and have the position Linux is holding right now. They are trying to get
>>>>there now (a little late) and it's an uphill battle.
>>>
>>>And the worst of it is that some people inside Sun knew at the time
>>>they needed to do this ....
>>
>> Is it possible to shed some light on the reasons why this didnt happen
>> earlier?
>
>I think there was a long period when the value of Solaris
>was not understood outside of the Solaris engineering group.
>It was believed customers bought the systems because they
>liked sparc, or they liked purple machines in the computer
>room. There probably was a time when this was true, but it
>stopped being relevant some time before the management
>realised. You can see this in the killing of Solaris x86 9;
>after all, why would anyone want this as it wasn't sparc
>and it didn't come with hardware with a purple front?

Sorry, I cannot concur.

Of course there was knowledge on the value of Solaris outside the
Solaris engineering group. There have been many people outside Sun (let us
call them the community) who did know this and even customers know this. Let
me name two important people: Jürgen Keil (Bonn) and Bruce Adler (Bay area).
Didn't you know that the "Oberfinanzdirection Hannover" (a German tax
authority) did run Solaris 8 on 12,000 PCs and more than 1000 noteboks? I am
sure you know about the "secret six" and what happened after Solaris 9 Beta
update was released and some days later declared dead....

You may not know that "Oberfinanzdirection Hannover" most likely switched to
Linux (this was the last press release) after the German Linuxverband did tell
them lies on Solaris and Sun did not respond.

You most likely do not know that Sun is currently doing everything to lose the
ability to sell to German (and probably all European) governmental sites for
the next 10 years because Sun is ignoring the OpenSource Initiave of the EU.
Other companies (including MS) did become members in the related e-Government
laboratories but Sun does not do anything :-(

My impression is that Sun did have no problem as long as there have been monthly
closed meetings with Scott, Andy and Bill. Then Andy left Sun and later Bill and
Scott did ignore the voice from reality and did do what all marketing people do
that have no contact to what really drives their company.

Another important issue is that Sun did choose the wrong chip manufacturer and
Sparc Workstations did no longer lead the market because the CPU chips did not
have sufficient single thread performance for a workstation. It is not
sufficient to have the best engineering samples if the manufacturer cannot
produce them in large quantities.

The last point is that it seems that Sun lawyers did take 6 years in order to
understand that the contract with Novell together with the AT&T SVr4 contract
did allow Sun to make Solaris OpenSource. At this point, your impression seems
to be correct if you limit it to _inside_ Sun. Sun Engineering did start the
Opensource project for Solaris around Y 2000 and a bit later marketing declared
Solaris x86 dead although Solaris x86 is the only way to keep Solaris in the
eyes of the people because Sun has no competive Workstations anymore.

Now everything looks nice as Sun did decide to make Solaris really open.
Well, the problem is that as long as Sun's internal burocracy prevents putbacks
from Sun external people, the goal is not yet reached....

Sun Marketing does still not understand what needs to be done in order to save
Sun and OpenSolaris.

Let me use this as my contribution to our 2 year OpenSolaris anniversary....


--
EMail:joerg-at-schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin
js-at-cs.tu-berlin.de (uni)
schilling-at-fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
URL: http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

In article <6jfg73pnu2n68jlsqtfg6bbn39i8upfjhr@4ax.com>,
CK writes:
> Words to the wise, andrew-at-cucumber.demon.co.uk (Andrew Gabriel) wrote:
>>I think there was a long period when the value of Solaris
>>was not understood outside of the Solaris engineering group.
>>It was believed customers bought the systems because they
>>liked sparc, or they liked purple machines in the computer
>>room.
>
> So, Solaris was seen as a "add-on" to selling the hardware?

Rather more something essential to enable selling the
hardware. Without an OS, no one would buy a computer.

>>There probably was a time when this was true, but it
>>stopped being relevant some time before the management
>>realised. You can see this in the killing of Solaris x86 9;
>>after all, why would anyone want this as it wasn't sparc
>>and it didn't come with hardware with a purple front?
>
> Yes, indeed, why would anyone?
>
>>The reaction the management got to this decision was the
>>was their big wake-up call to the value of Solaris, and
>>that was the point where they started taking it seriously
>>and giving it the attention it deserved.
>
> So someone realized that it is possible to make money off supporting
> an otherwise free OS. That about it?

No, I didn't mean financial value, I meant strategic value.
Solaris x86 was killed off because the management didn't see
this and thought people might as well use Linux.
Sun's customers responded by swiping the management with a
giant baseball bat to knock some sense into them. The message
was that Linux wasn't good enough and they bloody well wanted
fully commercial grade Solaris back. It was this response that
made the management suddenly realise they had something which
was far more valuable than Linux to many customers, and they
hadn't understood this before. The volumes of Solaris x86
licences subsequently shipped have proved this to be the case.

>> That is why you
>>can see an enormous amount of work poured into Solaris
>>itself resulting in all the new features in Solaris 10
>>(which actually started during the latter Solaris 9
>>releases after x86 came back).
>
> Yes, I was wondering about that and figured that Sun would probably
> move towards a more software-oriented company, though I doubt this is
> really possible without killing of hardware altogether.

I don't see that the two are connected. Indeed the introduction
of amd64 and Intel hardware product lines has very significantly
helped Solaris x86's cause within Sun. Although these are big
sellers for Sun, the majority of Solaris x86 is run on non-Sun
hardware. Furthermore, much as it might surprise you, a lot of
Sun's x86 hardware is actually sold to customers who want to
run Windows and find it amongst the best hardware for that too.
This indicates Sun is good at both hardware and software
products which stand up by themselves, as well as integrated,
and I could not see why Sun would want to drop either.

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]

In article <5dr1ulF35vsk0U1@mid.dfncis.de>,
js-at-cs.tu-berlin.de (Joerg Schilling) writes:
> In article <4678359e$0$644$5a6aecb4@news.aaisp.net.uk>,
> Andrew Gabriel wrote:
>>In article ,
>> CK writes:
>>> Words to the wise, Casper H.S. Dik wrote:
>>>
>>>>>I believe that the only problem Sun has is that they didn't foresee
>>>>>Linux coming and they wish they made Solaris open source 10 years ago
>>>>>and have the position Linux is holding right now. They are trying to get
>>>>>there now (a little late) and it's an uphill battle.
>>>>
>>>>And the worst of it is that some people inside Sun knew at the time
>>>>they needed to do this ....
>>>
>>> Is it possible to shed some light on the reasons why this didnt happen
>>> earlier?
>>
>>I think there was a long period when the value of Solaris
>>was not understood outside of the Solaris engineering group.
>>It was believed customers bought the systems because they
>>liked sparc, or they liked purple machines in the computer
>>room. There probably was a time when this was true, but it
>>stopped being relevant some time before the management
>>realised. You can see this in the killing of Solaris x86 9;
>>after all, why would anyone want this as it wasn't sparc
>>and it didn't come with hardware with a purple front?
>
> Sorry, I cannot concur.
>
> Of course there was knowledge on the value of Solaris outside the
> Solaris engineering group. There have been many people outside Sun (let us
> call them the community) who did know this and even customers know this. Let
> me name two important people: Jürgen Keil (Bonn) and Bruce Adler (Bay area).

I was referring to inside Sun here.
The value of Solaris outside Sun was clearly understood or the
subsequent events I described would never have happened.

> Didn't you know that the "Oberfinanzdirection Hannover" (a German tax
> authority) did run Solaris 8 on 12,000 PCs and more than 1000 noteboks? I am
> sure you know about the "secret six" and what happened after Solaris 9 Beta
> update was released and some days later declared dead....

Only too well.

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]

CK writes:

>Words to the wise, Casper H.S. Dik wrote:


>>It is very difficult to explain how things happened as they happened;
>>but the most important thing at the time was lack of management buy in
>>and lack of interest in most of the organization.

>So, too strong a hardware focus?

I would think so, yes.

That and a focus on high end, high margin systems.

Casper
--
Expressed in this posting are my opinions. They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.

CBFalconer wrote:
>"Floyd L. Davidson" wrote:
>> For example, I'd bet that Vista will be the last major
>> "build a whole new OS" project based on a proprietary
>> kernel from Microsoft. Another attempt would destroy them...
>>
>> As always, interesting times are just ahead!
>
>Another? Read the URLs below:
>
>--
>
>
>
> cbfalconer at maineline dot net

(Please do not put portions of the body of your article
into the signature!)

But, you do have a valid point. It is one that I did
not want to directly make only because it is of course
inflammatory, and what I had to say was already
enough... :-)

However, your first two cites do not directly support
that point, though they do provide background
information which indirectly can and should be
understood as the underlying cause of the problem.

The third one is more to the point, and says Microsoft
has already gone too far. I don't disagree entirely,
but I do not think it will destroy Microsoft as a
company if Vista is a catastrophe. What it does do is
position Microsoft in a "sudden death" playoff... make
the same mistake, or any other major mistake, on the
next go around and it *will* be curtains.

--
Floyd L. Davidson
Ukpeagvik (Barrow, Alaska) floyd-at-apaflo.com