Solaris Coming to IBM's Power Architecture?

johnm writes "Jonathan Schwartz, Sun's pony-tailed number two, dropped this little snippit in his blog where he talks extensively about what he thinks 'open' means: 'For example, as we continue porting Solaris onto IBM's Power architecture (demo coming soon!)...' Does this mean you'll soon be able to ditch OS X and stick on Solaris 10 onto Macs?" While coming off as an ad for Java, Schwartz also raises some valid points about Unix and migration.

Comments

I like his definition of open.

[Anonymous Coward]

For those of you that didnt RTFA here is the best part. Jonathan writes that the definition of open: Only a customer can define the word open.

Re:I like his definition of open.

[Anonymous Coward]

Which of course should instead read something like:

Only a customer can judge whether a given product/implementation/whatever is sufficiently open.

Re:I like his definition of open.

[jayminer]

Yes, this is one of the important points we (so called zealots, including myself) do not want to believe when we come home and fire up Konqueror/Mozilla etc. and jump in.

At work, I'm sure that many other Slashdotters are in communication with customers about open technologies.

For me, "open" may mean that it's totally hackable, modifiable and should include "fun".

For Joe, "open" may mean that it's possible to code to make it able to talk with his new XML based ERP system.

For Jane, "open" may mean that it's possible to save in an spreadsheet of office package X at home and embed it in the word processor of office package Y at work.

And so on..

We do have "our" preferences for the meaning of "open", but in the real world, we must achieve the fact that what we call "wide" open, may be restrictive for another person. This is what, at first, we should respect. Then we may have a peaceful settlement to all "open" wars around here or there.

Re:I like his definition of open.

[leandrod]

> We do have "our" preferences for the meaning of "open"

And they are meaningless.

There are two application of the 'open' term in Informatics.

Open systems conform to open standards. Solaris is an open operating system.

Open source, well, you know, Solaris ain't an open source OS.

Re:I like his definition of open.

[rewt66]

Solaris is an open operating system.

Yes, it is - as long as you stick to the POSIX specification. As the article points out, as soon as you go past that, Solaris isn't open any more, and neither are any of the other UNIXes. It's not open by the parent's definition (which I like, BTW) because there isn't any open standard for the non-POSIX parts of Solaris.

Part of the point of the article is that there is a lot of stuff in the "non-POSIX" part of Solaris. And if you use it, you're stuck with Solaris, and so Solaris isn't open to you.

I agree with the parent's definition. But Solaris isn't as good a fit as he thinks.

BTW, it's possible for open source to be closed in terms of open standards. (Of course, you could always re-engineer a new standard from the source, but then you run into one of the usual problems with standards - there are so many of them to choose from.)

Re:I like his definition of open.

[leandrod]

>> Solaris is an open operating system.

> Yes, it is - as long as you stick to the POSIX specification. As the article points out, as soon as you go past that, Solaris isn't open any more

That is not quite what the article said, nor the reality.

There are lots of other open or de facto standards besides POSIX that an OS can conform to, and Solaris does conform to several.

For example, LDAP is an open standard, SMTP and TCP/IP are de facto ones.

Even if you define the MS W16 API as once a 'proprietary standard', Sun's WABI conformed to it.

True, Sun loves to add proprietary apps. But those apps don't necessarily change the OS itself.

Re:I like his definition of open.

[cygnus]

Open source, well, you know, Solaris ain't an open source OS.

is that like LAME? should it be called SAOSOS?

"SAOSOS ain't an open source OS"

Re:I like his definition of open.

[Anonymous Coward]

You are ignoring the Single Unix Specification published by the Open Group. It specifies a lot more Unix functionality than POSIX, and is supported by all current unix-like systems (including Linux, which admittedly has a few obscure non-conformities).

Incidentally, the Open Group uses the word open in yet another meaning: any company is allowed to join, by simply paying the membership fee and agreeing to follow the rules.

Re:I like his definition of open.

[Fefe]

No. The definition of "open" in IT is: "open as in ``open your wallet".

And the source code to Solaris is open. You can get it and look at it, and fix bugs in it, if you like. It does not conform to the Debian pipe dreams, but heck, what does? Now that Debian considers removing drivers from their kernel which contain binary-only firmware images...

And "open standard" is a euphemism for a protectionist practice where companies get together and try to protect their market against outsiders. These standards usually codify the sum of the proprietary extensions of the people in the committee, and each of them has an obfuscated "reference implementation" and/or a few patents covering their share. That's why you will almost never find an open standard where some company is in the committee that does not have any leverage in the things required by the standard.

What we need is a real open standard, where only stuff is permitted in that is not covered by patents, can be freely used by everyone without any licensing claim by anyone, and where there exists a free software reference implemention. Then we can talk. Until then, all this "standards" talk is just trying to confuse you into believing you actually have a choice.

Re:I like his definition of open.

[leandrod]

> The definition of "open" in IT is: "open as in ``open your wallet".

So you are unaware that open systems (as in Unix) are what has driven down the (formerly expensive) prices of proprietary systems (then IBM)?

Take open systems from us (as in, let each current POSIX system diverge enough) and you will see MS, VMS and IBM prices hiking even higher than currently.

> the source code to Solaris is open

So try modifying and redistributing it to see how Sun likes it.

> "open standard" is a euphemism for a protectionist practice where companies get together and try to protect their market against outsiders

You have it in reverse. Companies (and users, including bigcorps and govs) banded together to break open the market from single-company's 'de facto standards', such as IBM, Microsoft and Digital had and still have.

> What we need is a real open standard, where only stuff is permitted in that is not covered by patents, can be freely used by everyone without any licensing claim by anyone, and where there exists a free software reference implemention.

Do you mean like POSIX, the Internet and other small, irrelevant stuff?

Re:I like his definition of open.

[SlowMovingTarget]

A large part of the communication problem is that these terms are open to interpretation. (Puts on grammar hat) You see, "open" is not a complete adjective. For it to have any real meaning, you must modify it to say what something is open to.

Solaris is at once open and closed, but to what is it open, and to what is it closed? That's what really matters, and that's why it does indeed depend on the user's preferences of interpretation.

Re:I like his definition of open.

[leandrod]

> Solaris is at once open and closed, but to what is it open, and to what is it closed?

No preferences of interpretation here, it is closed source but an open system conforming to (some) open standards.

Why not?

[Yaa 101]

I run Linux even on old world (gray) Macs... It shouldn't matter what os can run on what hardware, yet I know... ...utopia.

Re:Why not?

[Psymunn]

All i want to know is when can i run OS X on my SPARC architecture

Re:Why not?

[Cobalt Jacket]

You already can. Only it's called OpenStep. And it's about ten years old.

Re:Why not?

[timeOday]

All i want to know is when can i run OS X on my SPARC architecture

About a month after Apple open-sources it, i.e. never.

Re:Why not?

[name773]

indeed. however, gnustep is free and can be run atop most unixes and even cygwin

Re:Why not?

[maeltor]

Darwin's source code is freely available, as is a derivative of that code, OpenDarwin.

Re:Why not?

[Halfbaked Plan]

Yes, but it's 'portability' is limited to Intel and certain PPC architectures. Don't try to get it running on older PPC architectures or anything else.

Not that it's important or even worthwhile to port Darwin further. It's already been done better elsewhere.

Re:Why not?

[andreyw]

Does anyone /yet/ KNOW why Darwin 6 (and OS X 10.2) don't work on the 604e?

This guy didn't figure it out: http://www.opendarwin.org/pipermail/hackers/2003-J anuary/003730.html

Re:Why not?

[Guy Harris]

Darwin's source code is freely available

They said "OS X", not "Darwin". There's more to OS X than Darwin, and most of that "more" isn't open-source.

Re:Why not?

[maeltor]

Right I know that, but "Darwin" is what basically runs OS X. I don't want to get into a trivial argument, but what I meant was a LARGE section of OS X is already open source...hence my statement.

Re:Why not?

[Guy Harris]

but "Darwin" is what basically runs OS X

...for properly-chosen values of "runs". Yes, it's the core OS underpinnings - but a lot of the OS X user interface is not part of Darwin, and is not open source, so the question that started this thread:

All i want to know is when can i run OS X on my SPARC architecture

if by "OS X" they really meant "OS X", complete with Aqua and the Finder and Quicktime and..., really isn't answered adequately by noting that Darwin is open source.

Re:Why not?

[Halfbaked Plan]

All I know is I have some nice pre-G3 Apple machines that I'd love to run something more current on. Not talking about ancient hardware. This is decent PCI bus stuff.

But since the goal of Apple's software team has always been to sell more Apple hardware, I'm probably out of luck.

not really

[Anonymous Coward]

porting an OS is more than adding support for a CPU architecture. hardware drivers, for example...

Re:not really

[Bobas]

Not much work there in case that driver stuff is written in a valid and portable C, apart from proprietary hardware specific to said architecture.

Re:not really

[Anonymous Coward]

solaris kernel driver interfaces are not compatible with the MacOSX ones. porting driver code is more than just recompiling.

Re:not really

[gl4ss]

* solaris kernel driver interfaces are not compatible with the MacOSX ones. porting driver code is more than just recompiling.
*

what he meant was that most of the drivers would be portable from other flavours of solaris were they not just for spesific hardware on ppc.

(like on linux & etc)

Re:not really

[Lando Griffin]

Very true. I wonder if they will support my good old RS/6000 7015-R50, with all its Microchannel adapters. It sure would be nice to be able to run a supported OS on that box, since AIX 4.3 is not supported anymore and IBM dropped support the MCA machines from 5.2. How about it, Sun?

Re:not really

[Anonymous Coward]

Actually 95% percent of Solaris is platform
independent, the only part that isn't is the
hardware specific stuff that plugs into the
HAT (Hardware Address Translation) layer.

- Andrew

Ditch OS X For Solaris?

[oscast]

Why would you want to ditch OS X for Solaris?

Re:Ditch OS X For Solaris?

[HungSquirrel]

Because having the power of Unix coupled with a pleasing interface and scores of usable desktop applications is a disgusting perversion of everything Unix stands for.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Because as far as server technology goes, Solaris is superior to OS-X and far beyond what Linux has to offer. Belief to the contrary simply shows that you are not aware of the full capabilities of Solaris.

Re:Ditch OS X For Solaris?

[Undertaker43017]

I was wondering the same thing. I can already run OS X and Linux on my Mac why would I need Solaris?

I would suspect that Sun's intent is to impact AIX on IBM PowerPC's platform and not Mac's.

Re:Ditch OS X For Solaris?

[oscast]

enlighten us then please... oh wise one.

Re:Ditch OS X For Solaris?

[mbaciarello]

Well I guess that's the opinion of someone who's actually working on Solaris. Proudly biased?

I'd rather say add Solaris to your Mac, next to OS X and Linux...

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Belief to the contrary simply shows that you are not aware of the full capabilities of Solaris.

No, it means you are not aware of the full capabilities of OS X.

End Of Topic.

Re:Ditch OS X For Solaris?

[danamania]

Choice. perhaps that's where all your training lies. perhaps you have a mac at home and you work with solaris in the daytime. perhaps you just have an opinion that solaris is better. Lots of reasons =)

However, a port of solaris to the POWER architecture doesn't necessarily mean an immediate version for PowerPC machines, or Macs.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

The Mach kernel's message-passing scheme for IPC has been known to be slow. Microkernels also typically have worse performance than monolithic kernels.

Solaris uses a monolothic kernel. Solaris' scalability has been proven for many years on hardware with many more than two processors.

For industrial grade iron, there is no reason to use MacOS-- it is too young and is not intended to be used on high-end server hardware.
For a desktop machine, there is no reason to use Solaris. The nicities that you get when you target an OS at a desktop machine, such as power management, snappy UI response, and high-priority threads controlling multimedia, are at best unnecessary for a server OS.

Re:Ditch OS X For Solaris?

[rainman_bc]

I dunno, I don't see too many solaris based web servers on Netcraft's list of longest uptimes. They're all running FreeBSD - and guess what OSX is based on?

Doesn't mean it's necessarily as capable as Solaris in the enterprise computing world, but it's probably more secure, and likely more stable.

Re:Ditch OS X For Solaris?

[macmastery]

I don't expect that working with the processor is the limit of what you need to make an OS work. How about all of the other custom hardware Apple uses?

Re:Ditch OS X For Solaris?

[Anonymous Coward]

"For industrial grade iron, there is no reason to use MacOS-- it is too young and is not intended to be used on high-end server hardware."

FreeBSD is too young? Its used on high-end server hardware all the time.

OSX is already being used on high end server hardware. Where have you been?

Re:Ditch OS X For Solaris?

[Anonymous Coward]

OSX is not based on FreeBSD. OSX uses FreeBSD's libc, and many of its tools.

Darwin's kernel is based off of XNU whish was derived from Mach. FreeBSD has nothing to do with that.

Re:Ditch OS X For Solaris?

[gonk]

The Solaris boxes are not exposed to the internet. They're core servers, doing real world in the background.

robert

Re:Ditch OS X For Solaris?

[Anonymous Coward]

I don't see any 16 processor machines that run OS X... In fact, I don't even see any four processor machines running OS X.

Re:Ditch OS X For Solaris?

[chez69]

OS X is not FreeBSD, they just use some of the userland bits. It is a completely different kernel

Re:Ditch OS X For Solaris?

[Anonymous Coward]

there is no server hardware with more than two processors running MacOSX.

where have you been?

furthermore, OSX is not based off of FreeBSD. OSX contains FreeBSD's libc, and many of its tools. Other than that, it has nothing to do with FreeBSD. OSX is based off of XNU which was derived from Mach.

As far as "high-end" server hardware, FreeBSD's SMP performance is abysmal, that is why if you check the current activity logs for FreeBSD, they are in the process of rewriting that part of the OS.

Re:Ditch OS X For Solaris?

[macmastery]

Yeah, AIX was added to Macs in 1995 (but only to certain models. See "Apple Network Server").

Re:Ditch OS X For Solaris?

[leandrod]

> OS X is not FreeBSD

They are both BSDs.

> they just use some of the userland bits. It is a completely different kernel

Not true. The Mac OS X is a monolithic, perverse mix of the BSD kernel and the Mach microkernel.

Re:Ditch OS X For Solaris?

[divisionbyzero]

I know you are kidding, but your joke points out somethng interesting. You said desktop applications, not enterprise applications. That's where Sun's time in the sun (haha) served it well. Many enterprise class applications were made to run on Solaris or ported to it. That's where it has MacOS X beat hands down. The other stuff that Solaris can do (e.g. scale to 128+ processors, etc) is important, but not crucial.

Anyhow, I don't think any of this has anything to do with Apple. It's clearly IBM that Sun is after. First they say they will 'buy Linux' (i.e. SuSe) which is IBM's Linux vendor of choice and now they are saying they will also invade IBM's hardware. Good luck to Sun. Competition is good for everyone, except the losers of course.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

MacOSX has a Mach-based kernel with a BSD4.4 interface. there is no BSD kernel in Darwin's kernel.

you might have wanted to say, "Not true, Mac OS X is a monolithic, perverse mix of BSD4.4 interfaces wrapped around a modified Mach microkernel"

Re:Ditch OS X For Solaris?

[jadenyk]

Personally, I would do this if I were still running my web server off of my Mac.

Other than some kind of server environment, it would be fun to dual boot between the two. I don't know that anything could make me want to ditch OS X though.

Re:Ditch OS X For Solaris?

[bhima]

Really, I find the ppc 970fx cpu very interesting, perhaps more interesting than most (we use a derivative at work) but still Solaris who gives a fuck? (I could be speaking for my company a large pharma sort of place) when I say "piss off" *BSD suits my needs and NetBSD is orgasmically groovy!

Oh and I use OS X

why would I even need SCO^H^HOLARIS?

Re:Ditch OS X For Solaris?

[the MaD HuNGaRIaN]

With Oracle now running on OS X, and the fact that masses of Enterprise Application vendors use Java, that argument is dwindling away--as any enterprise app written using J2EE will run on OS X just fine.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Its more about the scores of obnoxious OSX advocates that puts me off, personally.

Re:Ditch OS X For Solaris?

[fiftyvolts]

That's not exactly true. In Mac OS X terms when one refers to hte Kernel one refers to the Mach Microkernel as well a the BSD, I/O Kit, File system, and Networking Components.

The actualy kernel is the Mach microkernel , but it is surrounded by all kinds of crazy crap

Re:Ditch OS X For Solaris?

[gl4ss]

because the cluster you bought from sun wouldnt even run os x?

Re:Ditch OS X For Solaris?

[rainman_bc]

Not trying to start a flame war here. I mean, I respect Solaris and all, but it's not right to discount FreeBSD's capabilities either. It's a pretty swanky OS IMO. For stability and security, it's pretty awesome.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Oh, I don't know... Maybe because it is a REAL UNIX OS.

Re:Ditch OS X For Solaris?

[leandrod]

> The actualy kernel is the Mach microkernel

Not.

When one talks a microkernel, that's not a complete kernel. It is the basics of a kernel, one needs to add servers to that in order to get an OS kernel.

In Mac OS X, there is only one server: BSD. And it is mixed in a monolithic kernel with Mach.

Contrast that with the Hurd which has Mach (or L4) plus several servers, or the other BSDs that have no microkernels.

Re:Ditch OS X For Solaris?

[Elwood P Dowd]

The Mach kernel's message-passing scheme for IPC has been known to be slow. Microkernels also typically have worse performance than monolithic kernels.

That's why OS X doesn't use it. XNU, honey.

Solaris uses a monolothic kernel.

Apple has poked holes in their microkernel everywhere they thought needed performance. I see no reason to believe that the micro-vs-monolithic debate can be so easily applied to OS X.

Solaris' scalability has been proven for many years on hardware with many more than two processors.

You are implying that Apple's support for 4 processors or more is not mature. In fact it does not exist. Their support for two processors is just fine, and that's the hardware that y'all would be talking about swapping OS X out for Solaris on. If that made sense.

For industrial grade iron, there is no reason to use MacOS-- it is too young and is not intended to be used on high-end server hardware.

Assuming that your definition of "high-end server hardware" does not include Xserves (mine doesn't) then not only is it not intended, it's not possible. There's no comparison to be made.

Explain why you'd want Solaris rather than OS X on a Macintosh . That was the debate. (I know there are reasons. I don't care about this idiotic debate. But you're talking stupid.)

Re:Ditch OS X For Solaris?

[jschottm]

Because some of us don't like the way that Steve^WApple has chosen to dictate that a UI should be? Because some of us like to be able to reliably do everything we need to do to maintain a machine via SSH? Because some of us like to maintain similar operating environments between our servers and our desktops (I spend most of my time on OS X in the server version, which has some significant differences in configuration tools, which leaves me fumbling around when I do deal with the desktop version)? Because benchmarks have shown that Linux runs faster than OS X, so Solaris may as well? Dtrace?

Just because OS X happens to fit your hand doesn't mean it fits mine. Linux et al allows me to customize the tool to fit my hand. Steve wants me to change my hand to fit the tool.

Just bec

Re:Ditch OS X For Solaris?

[torstenvl]

You know, I'm getting really sick and tired of you ignorant idiots.

"The upgraded kernel, based on FreeBSD 5.x..."

Go argue 'til your heart's content with Apple about Apple stuff. Until then, STFU and RTFM.

Re:Ditch OS X For Solaris?

[kjs3]

Those enterprise applications would not nessesarily run on a hypotehetical Power Solaris. Just as there is a big gap between apps supported on x86 Solaris v. SPARC Solaris, someone would have to convince ISPs that porting and supporting is worth it. The answer seems to be "no" for x86.

Re:Ditch OS X For Solaris?

[joshmccormack]

With Oracle now running on OS X, and the fact that masses of Enterprise Application vendors use Java, that argument is dwindling away--as any enterprise app written using J2EE will run on OS X just fine.

But there are some big differences between Solaris and other big time commercial Unices and *BSDs and Linux on the one side and Mac OS X on the other.

Solaris on Sun hardware has some failover and maintenance jazz that Apple hardware doesn't.

Solaris and the others can be stripped down to bare bones to conserve resources and make the box more secure - I haven't seen this done with Mac OS X.

Mac OS X does things differently. From what I can tell, for example, the default shell is bash and typing vi will get you vim. Now I know the default shell in Linux is usually bash, as well, but many shell scripts that are written for ksh or c shell or something of that sort will behave a little differently in bash, and I'm sure there must be all sorts of nuances I haven't found yet.

Re:Ditch OS X For Solaris?

[Ytsejam-03]

First they say they will 'buy Linux' (i.e. SuSe)

This article has a more realistic perspective on things. If Sun were going to buy SuSe, they would have done it before Novell bought them. After all, Schwartz himself said that Novell's products are "far less intersting" than Suse. Why pay the extra money for a bunch of Novell products that they don't want?

Re:Ditch OS X For Solaris?

[metamatic]

So that you can help support SCO in their valiant fight.

OS X is BSD-based, whereas Solaris is SCO-IP-encumbered System V UNIX.

Re:Ditch OS X For Solaris?

[Undertaker43017]

Thanks! I forgot about those. "Dead Ends" are sometimes interesting to look back on, as long as you weren't one that got stuck with one, like the 50 pound paperweight Mac XL, I have in my attic. ;)

Re:Ditch OS X For Solaris?

[bhima]

perverse but usefull!!!

Re:Ditch OS X For Solaris?

[Anonymous Coward]

>>>
Explain why you'd want Solaris rather than OS X on a Macintosh . That was the debate. (I know there are reasons. I don't care about this idiotic debate. But you're talking stupid.)
>>
For a desktop machine, there is no reason to use Solaris. The nicities that you get when you target an OS at a desktop machine, such as power management, snappy UI response, and high-priority threads controlling multimedia, are at best unnecessary for a server OS.

Re:Ditch OS X For Solaris?

[hexghost]

Depending on whether its a new install or an upgrade, the default shell is either bash or tcsh. As for stripping the box down to bare bones, I'm not sure what you mean, but OSX starts with no services running, which is pretty bare bones. You can also disable the GUI environment if it so pleases you (not like it hogs that much sitting at the login screen anyway).

Re:Ditch OS X For Solaris?

[joshmccormack]

Netcraft's list is just reporting the servers furthest forward, aren't they? You can't tell what app servers or database servers are running, right? So all you're really telling is what Apache et al are running on.

And Mac OS X is not FreeBSD. Similar? sure. Loosely based on? I'll buy that. But there are some major differences. Take a look at this usenet post (http://groups.google.com/groups?hl=en&lr=&ie=UTF- 8&selm=3CF65A12.9020000%40coldmail.com.invalid ) or search out others.

Re:Ditch OS X For Solaris?

[MouseR]

The Mach microkernel of Mac OS X scales up to 64 processors.

So? IBM still can't spew em out fast enough. Macs currently (and barely) ships with two processors only.

128 processor support isn't much of a selling point on Macs. For now anyway.

Re:Ditch OS X For Solaris?

[joshmccormack]

I was wondering the same thing. I can already run OS X and Linux on my Mac why would I need Solaris?

I really can't picture Sun targetting the majority of Mac OS X users. Maybe the scientific community.

I would suspect that Sun's intent is to impact AIX on IBM PowerPC's platform and not Mac's

Wow... that would be a serious target to attack.

They could also just want to hedge their bets and offer alternatives. There seem to be a lot of things happening with powerpc.

Re:Ditch OS X For Solaris?

[Elwood P Dowd]

While I wouldn't consider any Macintosh to be a high-end server, they are also not all used as desktop machines.

So I didn't consider that statement to satisfy my request for explanation. If that's what you meant, though, then I should have. Sorry.

Anyway, there are only two ways this port could be useful: You want to run Solaris on non-Apple PPC hardware (a pServer), or you already have Solaris machines & Power Macs, and now you want the same platform everywhere.

Other posters have made both of these points. So there is (barely?) a reason to run Solaris on a Mac.

Re:Ditch OS X For Solaris?

[Elwood P Dowd]

Wow. That's the dumbest thing I've read on Slashdot in weeks.

No, that is not an OSX computer with 2200 processors.

That is at least 1100 different OSX computers, each with no more than 2 processors.

Re:Ditch OS X For Solaris?

[dAzED1]

the only thing they're after from IBM is for IBM to do a partnership, which would be of huge benefit to both companies. I have forever wished IBM would just buy Sun. Would make for some great mixing.

Re:Ditch OS X For Solaris?

[squiggleslash]

I think the submitter put the cart before the horse.

I believe the reason Solaris is being ported to PowerPC is because Sun wants to jump ship from SPARC to PowerPC. It's so you can change processor, not OS.

Re:Ditch OS X For Solaris?

[HungSquirrel]

Most Linux distros have vim and a symlink 'vi' that points to it. There are a few exceptions, like Slackware (which uses Elvis for some reason no one can comprehend).

Re:Ditch OS X For Solaris?

[dAzED1]

ummm...that's a beowulf. Its not a single processor space, single memory space, single platform, environment. Hell, one could make a 10,000 node windows98 beowulf cluster...that's doesn't mean win98 supports 10,000 CPU's...

Re:Ditch OS X For Solaris?

[leandrod]

> perverse but useful

While Mac OS X is useful - people do useful things with it -, how useful is a kernel which is slower but not more flexible or safe than a monolithic one? The whole idea of microkernels was flexibility and safety.

Re:Ditch OS X For Solaris?

[rainman_bc]

Fair enough...

Here's what Apple says.

Looks like they've used portions of the kernel and libc. I understand Darwin / OSX FreeBSD, but I'll argue it's quite comparable.

I stand corrected :)

Re:Ditch OS X For Solaris?

[IPFreely]

It's clearly IBM that Sun is after. First they say they will 'buy Linux' (i.e. SuSe) which is IBM's Linux vendor of choice and now they are saying they will also invade IBM's hardware. Good luck to Sun. Competition is good for everyone, except the losers of course.

Yeah, when all those companies start buying IBM hardware just to put Sun Linux or Solaris on it, IBM will be in a world of hurt..... I mean, big hardware sales and service without the cost and headaches of software support. Can it get any worse?

Re:Ditch OS X For Solaris?

[mrm677]

Scalable to > 100 processors out-of-the-box. I don't need some tricked out kernel build from the folks building special 512-processor Linux machines.

Compatibility with 64-bit apps written 10 years ago.

A decent threading model that has been in place for years. Last time I checked there were 2 competing proposals for a new Linux threading system

CC-NUMA memory allocation.

Hot-swappable CPUs and consolidation. I can dynamically split single Solaris instance, running on 128 processors, to N instances each running on 128/N processors.

Mature user/kernel profiling tools.

Stable device driver model. Drivers from Solaris 2.6 will work fine in Solaris 10. Meanwhile any Linux kernel patch that changes task_struct will require rebuilds of certain Linux device drivers. Yes...not a problem with all open-source drivers, but the world isn't all open-source (ask nVidia)

The kernel is more modular. I can swap in a different scheduler.

Trusted Solaris is available if needed

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Funny, for me it is the obnoxious Linux
advocates.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

DTrace and ZFS come to mind. Other then that not much since OS X/FreeBSD's Unix implementation is pretty good relatively speaking.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

I dunno, I don't see too many solaris based web servers on Netcraft's list of longest uptimes. They're all running FreeBSD - and guess what OSX is based on?

Yeah, kind of like a pig is based on mammals.

Doesn't mean it's necessarily as capable as Solaris in the enterprise computing world, but it's probably more secure, and likely more stable.

Likely eh? I like that. Basically you're talking out of your ass.

Re: Ditch OS X For Solaris?

[Anonymous Coward]

Merge them; use Solaris as the kernel instead of Mach, with the rest of the Mac OS as it is. Then, take an Enterprise class Sun box, migrate to PowerPC and support 100's or 1000's of users running Mac OS desktops on Sun Ray Stations (think modern X-stations), with all of the current Mac OS X apps. In fact, Sun and Apple should merge while they're at it!

A couple of years back, I was at a Sun facility somewhere in CA, they were all using these Sun Ray stations against E6500's, with 100's of users on each box. It was kick-ass, the performance was great (they did their network right), but I laughed at the receptionist trying to use Xmail.

Re:Ditch OS X For Solaris?

[_damnit_]

Who cares what the "uptime" on netcraft is? Why would anyone think that's important? Most serious shops have scheduled maintenance windows for upgrades, app and OS patching, replacing EMC with anything else, etc. Most critical apps are failover so there is no real downtime to 5 9's services. Plus, most vendors and DC managers are bonused/measured against unplanned downtime.
Taking netcraft uptime rankings as a measure of any OS or platform's reliability is laughable. It may be fun for kids to aim for the longest uptime, but in the real world who cares?

Re:Ditch OS X For Solaris?

[bhima]

I think, to some extent, slower has become meaningless. Monolithic vs. Micro has become beyond /.

like I said *.bsd is very, very useful,

Solaris whho needs it?

Re:Ditch OS X For Solaris?

[ToasterTester]

To use a real Enterprise class operating system. OS X is a nice desktop OS and okay for web server or servers in a farm. Also read the reviews of OS X server they aren't that good it still has a ways to go. OS X server is not as mature as OS X desktop. Apple's main focus has always been desktop boxes, servers were alway an after thought.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Bah... that was "insightful"? Check out Netcraft's FAQ. It states that the uptime of "HP-UX, Linux, NetApp NetCache, Solaris and recent releases of FreeBSD cycle back to zero after 497 days ... Thus it is not possible to see a HP-UX, Linux or Solaris system with an uptime measurement above 497 days". That's why the list is dominated by FreeBSD (by necessity old versions, as their uptimes are 3-5 years) and BSD/OS.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

That page refers to Tiger.

Interesting. So they're ditching the Mach kernel in favour of FreeBSD's huh? That's quite a change.

Re:Ditch OS X For Solaris?

[leandrod]

> to some extent, slower has become meaningless. Monolithic vs. Micro has become beyond /.

Yet one is bound to question why go for the extra complication of Mach if simple BSD is faster and no less flexible or safe than the Mach and BSD combination in Mac OS X.

And in several situations one really wants that extra performance, memory, robustness and simplicity a, well, faster, leaner, more robust and simpler design gets.

> bsd is very, very useful

No one ever argued to the contrary.

> Solaris whho needs it?

Anyone who needs more than two processors and can't go GNU/Linux.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

You said: Competition is good for everyone, except the losers of course.

To which I reply: Then most /. readers are probably really opposed to competition. That explains why there are so many lunix users here. OMG!!! ROTFL!!!!!111

Re:Ditch OS X For Solaris?

[clarkcox3]

Now I know the default shell in Linux is usually bash, as well, but many shell scripts that are written for ksh or c shell or something of that sort will behave a little differently in bash

Umm, if a script is written for a specific shell, then put:

#!/path/to/my/specific/shell

in the first line. Problem solved.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Custom hardware? You mean like PCI, AGP, PC2700 memory, IDE, ATA, SATA, USB, FireWire?

Re:Ditch OS X For Solaris?

[Anonymous Coward]

That's funny. For me it's the obnoxious Windows and BSD advocates. (I think we've covered all the important systems now)

Re:Ditch OS X For Solaris?

[name773]

they're both atrocious...
dsb* ot hctiws

Re:Ditch OS X For Solaris?

[jbolden]

What he means is tearing out huge numbers of libraries and system services. For example on Linux we used to do things like:

a) compile a custom kernel with far fewer services available
b) Change program locations and links so that random calls by path wouldn't work (i.e. something like ls would be /sdf/sajfs/ysfs/sj while cat would be /uwsius/usiufs/sc etc...
c) remove dozens of commands entirely

etc...

You probably could do this with Darwin, you couldn't think of doing it with OSX

Re:Ditch OS X For Solaris?

[name773]

exactly, it's a different kernel than freebsd

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Scolaris sounds like some kind of horrible disease. That probably because SCO and Solaris both suck the hairy ass of Steve Ball-me.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

See what I mean? Here's one of those obnoxious BSD fucktards now. He thinks he's oh-so-clever about how he "encrypted" his switch to BSD SIG. Stupid fucker. Why not try Linux on for size. It works, and it works amazingly well. I've got a Linux box that has an uptime of nearly 800 days as my internal application server at home. Can't beat that with your fucking worthless BSD crap.

Re:Ditch OS X For Solaris?

[tonyr60]

"Most serious shops have scheduled maintenance windows for upgrades, app and OS patching"

Yes, they sure do here. About every 3 years hardware gets replaced. Apart from that, the Sun boxes just keep on running. Patches don't require a reboot these days, and defective hardware like memory and disk gets replaced on the fly.

Re:Ditch OS X For Solaris?

[torstenvl]

I don't believe they're ditching anything. I think it simply means that, as far as Apple is concerned, both components are to be considered part of the kernel, as Mach is pretty useless without anything running on top of it.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Process partitioning; you can lock users to single CPU's or banks of memory.

Mature 64bit and Mature device mgmt is never to be looked over, highly valuable.

Awesome threading capabilities, into the guts of the kernel. This may be something linux has finally caught up with, but you can take a single task with threads and span it across all CPUs.

The list goes on.

Re:Ditch OS X For Solaris?

[jedidiah]

While that is true. It is also true that Sun
support of non-sparc cpu's is not reliable. Also,
Sun's 3rd party vendors have tended to treat
non-sparc versions as the proverbial "red headed
stepchild".

Sun & friends have a whole lot of negative karma to work off from their x86 port before a PPC one will make any sense.

Re:Ditch OS X For Solaris?

[Donny Smith]

It wouldn't make shit - it's a huge overlap in J2EE, high end servers, UNIX, storage and much more.

Re:Ditch OS X For Solaris?

[sp0rk173]

Sure you can. My OpenBSD box has been up for 12 years. Solid. Seriously. And it gets 5,000,000,070,000.342 hits a day, while still being my gateway router (i provide internet services to an alien planet via high-throughput molecular two-way microradiowaves behind the router - approximately 17 billion users with ~3 mechagigabits upstream and ~4 gajillobits downstream per user). All on a 90 mhz pentium machine. Can't beat that with your mediocre linux junk.

Re:Ditch OS X For Solaris?

[ninkendo84]

And pray tell, what is the percentage of the population of servers that actually *need* any of those features? How many admins want to swap in a different scheduler for their kernels? How many servers are (or how many admins actually want) >32 processors anyway? Isn't it much easier to cluster your servers together? I mean, it's easy... And isn't it much more likely that solaris' user base is entirely due to the average sysadmin thinking "Hmm, everyone else seems to use solaris, and it's been around for a long time! I'll use that!" Wouldn't most admins prefer usability over all those esoteric features?

And for that matter, won't "hot swappable cpus" require sun hardware? Last time I checked, you couldn't pull the CPU out of a multi processor xserve.

And keep in mind, this is the company that told us hardware should be free, and people should only pay for software. (not really on topic, but I just thought that was an exceptionally dumb statement.)

Re:Ditch OS X For Solaris?

[jedidiah]

> Scalable to > 100 processors out-of-the-box. I
> don't need some tricked out kernel build from
> the folks building special 512-processor Linux
> machines.

"Those people" are the same people that SOLD Sun it's current NUMA technology.

Re:Ditch OS X For Solaris?

[jedidiah]

Take Solaris off of Sparc and it ceases to be a "real Enterprise class operating system" due to lack of 3rd party support.

The spiffiest OS on the planet will do you no good if vendors won't keep the apps current and relatively bug free.

Solaris + x86 + 3rd party vendors is a combination quite capable of generating grudges that last a lifetime.

Re:Ditch OS X For Solaris?

[LostCauz]

"Patches don't require a reboot these days"

Yes they do. Don't lie.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

A decent threading model that has been in place for years.

Actually, Sun recently replaced their much-touted M:N thread library with a Linux-like 1:1 thread library. So much for the "M:N must be better because Solaris uses it" theory.

Last time I checked there were 2 competing proposals for a new Linux threading system.

That has been resolved; NGPT forfieted, leaving NPTL as the last thread library standing.

Re:Ditch OS X For Solaris?

[Elwood P Dowd]

Very few of those features apply to any machine that runs Mac OS X, which is what the dude was asking about.

Re:Ditch OS X For Solaris?

[DevilM]

Can it get any worse? Well... IBM could lose their license to AIX because of SCO at which point Solaris on PPC might look pretty good.

Wasn't it Sun who supported SCO?

Re:Ditch OS X For Solaris?

[Logicdisorder]

You do know that OS X is based of Unix? BSD was the bases that Solaris used as was OS X. And come on I have yet to see another GUI for Unix that is as easy to used as OS X. What I would like to see Sun to is to add to the Open Power Architecture that IBM is doing now. Sun has pretty much said they are going to be dropping there CPU so why not added the best parts to the IBM CPUs.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Why you worthless cumsopped fagroon BSD jockey! Why don't you spill some REAL data. Check this:

17:02:02 up 854 days, 14 min, 78 users, load average: 0.12, 0.11, 0.06

You can't touch dis!!! Linux is the power of a GNU generation!

Re:Ditch OS X For Solaris?

[divisionbyzero]

Profit margins are much higher on software than hardware, and that's not incudling services.

Re:Ditch OS X For Solaris?

[JonAnderson]

>Actually, Sun recently replaced their much-touted >M:N thread library with a Linux-like 1:1 thread >library. So much for the "M:N must be better >because Solaris uses it" theory. How is it Linux like? I don't see how you can qualify that statement. The Solaris kernel is fully multithreaded (and preemptible - and has been for a long time)

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Come on. The bazillion attempts at creating a usable, attractive UI on top of any *nix OS have resulted in half-decent results at best.

The reason Apple has chosen to dictate what a UI should be is to ensure a consistent look-and-feel to its end-users -- something which Linux has NEVER done well. Neither has Windows.

The reason Apple CAN dictate what a UI should be, is because they do it better than anyone else.

If messing around in vi editing config files, fighting to add decent looking, anti-aliased fonts, fixing messed up keyboard mappings, etc. are your idea of limitless customization, have fun.

I personally find that my hand fits quite nicely around Steve's tool...wait, that didn't come out right...

Re:Ditch OS X For Solaris?

[0racle]

The Right tool for the Right Job. OS X looks very capable, but only a fool thinks any OS can do everything well, besides when has choice ever been a problem?

Re:Ditch OS X For Solaris?

[Anonymous Coward]

i bet you get laid a ton.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Yeah... and Monkeys might Fly from Thine Exalted Butt.

Re:Ditch OS X For Solaris?

[larien]

OK, few points:

Compatibility with 64-bit apps written 10 years ago.

Bull. 64-bit Solaris started with Solaris 7 which must have been about 98/99; maximum age on a production 64-bit Solaris app is 6 years.

That said, Solaris has reasonably good binary compatibility with apps from SunOS 4 and any 32-bit app written to comply with the ABI specs of previous Solaris releases.

Hot-swappable CPUs

Hot-swappable if (a) you can find documentation to confirm that you can hot-swap system boards and (b) the system will let you DR the board out. The latter is a kicker, as I've seen 12Ks which wouldn't DR out a board because of a caged thread on one of the CPUs; I don't think we got an answer as to whether it was a kernel thread or part of Sun Cluster that was locking the board.

I can dynamically split single Solaris instance, running on 128 processors, to N instances each running on 128/N processors.

The smallest unit on a SunFire system with domains is a system board, i.e. 4 CPUs + all RAM on that board. Solaris 10 introduces zones which take that limitation away; time will tell how well that runs.

Stable device driver model. Drivers from Solaris 2.6 will work fine in Solaris 10

I'm pretty certain that isn't supported.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

ow many servers are (or how many admins actually want) >32 processors anyway? Isn't it much easier to cluster your servers together? I mean, it's easy...

No, it's cheaper, not easier. If I was an admin, I would much rather admin one 32-way box versus 16 2-way boxes.

Actually, I am an admin, and I do admin 32-way and 64-way boxes (HP superdomes) and labs full of 1 and 2 cpu workstations. Each dome takes up twice as much time to admin as a single workstation. Net benefit of being 16x more efficient for admin work.They are a whole lot more efficient at real work when you need, say 10GB of ram for a single job too.

PS - don't assume that "labs" are only for academia either. Few uni's, even well granted ones, couldn't afford the amount of hardware I have to work with.

Re:Ditch OS X For Solaris?

[sp0rk173]

router:sp0rk {77} uname -a
OpenBSD router.charterpipeline.net 3.5 GENERIC#34 i386

router:sp0rk {78} uptime
8:17AM up 4380 days, 19:11, 4562 users, load averages: 0.14, 0.12, 0.12

/home/sp0rk 3:09PM
brainguts: uname -a
FreeBSD brainguts 5.2.1-RELEASE-p7 FreeBSD 5.2.1-RELEASE-p7 #3: Tue May 25 22:49:47 PDT 2004 sp0rk@brainguts:/usr/obj/usr/src/sys/BRAINEL i386

/home/sp0rk 3:09PM
brainguts: uptime
3:09PM up 935 days, 23:22, 5 users, load averages: 1.00, 1.00, 1.0

And for good measure...
sp0rk@opterskank ~ $ uname -a
Linux opterskank 2.6.7-gentoo-r11 #3 Tue Jul 27 21:44:24 PDT 2004 x86_64 5 GNU/Linux

sp0rk@opterskank ~ $ uptime
15:10:57 up 1 day, 3:38, 4 users, load average: 0.00, 0.00, 0.00

Now the work machine...
/home/escott 3:11PM
soilsa120a: uname -a
FreeBSD soilsa120a 4.10-STABLE FreeBSD 4.10-STABLE #9: Fri Aug 6 14:50:23 PDT 2004 escott@soilsa120a:/usr/obj/usr/src/sys/GENERIC i386

/home/escott 3:11PM
soilsa120a: uptime
3:12PM up 395 days, 8 mins, 4 users, load averages: 0.12, 0.08, 0.03

So, there you have it.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

"I think the submitter put the cart before the horse."

Of course they did. Have you ever been behind one?

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Scalable to > 100 processors out-of-the-box. I don't need some tricked out kernel build from the folks building special 512-processor Linux machines.

Not quite right. You need a special build of Solaris to boot your home UE15K, that one that comes along with the standard disk set from Sun will not run on a > 100 processors box.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

I may be totally utterly wrong about this, but didn't Sun get the good bit of Cray, and SGI got all the rest that Sun didn't want? If so then it's not /really/ the same people.

Re:Ditch OS X For Solaris?

[name773]

router:sp0rk {77} uname -a
OpenBSD router.charterpipeline.net 3.5 GENERIC#34 i386

router:sp0rk {78} uptime
8:17AM up 4380 days, 19:11, 4562 users, load averages: 0.14, 0.12, 0.12

has 3.5 been out for 12 years?
to quote openbsd.org, "The current release is OpenBSD 3.5 which started shipping May 1, 2004."

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Actually, the 1:1 thread library has been available for at least 4 years in Solaris. It has only recently become the default.

Re:Ditch OS X For Solaris?

[sp0rk173]

Psh, it's called patching, man. But seriously, i'm glad everyone saw right past the point of my post - that you can easily skew data, and that guy's uptimes are no more valid than the ones i posted.

Re:Ditch OS X For Solaris?

[_damnit_]

While I'm sure McNealy loves comments from customers like you, let's not exagerate and plain mistake the truth. OS patches of any significance require reboots to take effect -- period.
On the replacement of memory, the new enterprise models do allow you to Dynamically Reconfigure (DR) a board out and replace cpu/mem on the fly. The older models (except for the E10000) did not allow for DR. The Ex500 models could in theory, but the practice was flaky at best and required a big performance trade-off of disabling memory interleaving across boards and enabling the kernel cage to restrict the kernel to the memory of (typically) board 0. This also did not work if you had nonpagable memory on other boards (Oracle or realtime processes). Such restrictions really hampered the use of DR on that generation of machines.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

The point is Solaris is friendly and pleasant to the SysAdmin - also uniform and stable. I have to look after 300 Solaris machines - wouldn't want to look after 300 RedHat/SuSE/other commercial distro machines.

Re:Ditch OS X For Solaris?

[Guy Harris]

OS X is not FreeBSD, they just use some of the userland bits. It is a completely different kernel

...which happens to have a significant amount of BSD code in it. See, for example, this description of the OS X kernel architecture, which says

Above the Mach layer, the BSD layer provides "OS personality" APIs and services. The BSD layer is based on the BSD kernel, primarily FreeBSD.

Re:Ditch OS X For Solaris?

[LizardKing]

BSD was the bases that Solaris used

The original SunOS was BSD based, but Solaris was built on top of SVR4. Extra features from the old BSD based SunOS were shoe-horned into Solaris, and various bits put into a compatability directory (ucb).

Re:Ditch OS X For Solaris?

[LizardKing]

How many servers are (or how many admins actually want) >32 processors anyway? Isn't it much easier to cluster your servers together?

I see this come up quite often on places like Slashdot, the old "why buy big iron when you can cluster?" argument. The reason people pay for big iron (be it from Sun, SGI, IBM or whoever) is because writing a multithreaded app that runs on one machine is far simpler than writing a distributed one. If it's well written, then it will outperform a distributed app unless you use a disproportionate amount of clustered hardware.

For many applications, a bunch of separate machines can be a good solution. Think of a server farm for a large website, where each machine handles requests on a round-robin basis. However, the moment they need access to a single resource (such as a database of user accounts), then you need a single heavyweight machine to handle this resource. Otherwise you're introducing a potential bottleneck, have to write a very elaborate distributed system or invest in some serious enterprise software (which probably runs on a large Sun box anyway).

And keep in mind, this is the company that told us hardware should be free, and people should only pay for software.

I think you misunderstand what Sun were trying to say. What they meant was that you buy a lease, which essentially means that the hardware is free, as you're actually paying for support. Of course, if you stop paying for support, then they take the hardware back ...

Re:Ditch OS X For Solaris?

[LizardKing]

I dunno, I don't see too many solaris based web servers on Netcraft's list of longest uptimes. They're all running FreeBSD

As I pointed out in another post, the FreeBSD boxes are probably just a bunch of basic machines serving the actual webpages on a round robin basis. The heavyweight work (data stores, user accounts, news feeds or whatever) will be done by machines that NetCraft doesn't see - and they are often big iron like Sun or IBM servers.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

making something "based" on some kernel does not mean that they have been using any of that kernel's source code. for all you know it could be marketing lingo analogous to saying that the Darwin kernel follows the BSD 4.4 spec with similar improvements that the FreeBSD people did recently.

I could write a kernel "based" on Solaris'. That doesn't say anything about the quality. Now, if I said that there are pieces of Solaris kernel in mine, then we're getting somewhere.

Besides, Apple would be making a mistake by basing their entire kernel off of FreeBSD's. FreeBSD 5.x STILL does not have good SMP support, so claiming it as a win is pointless.

http://www.freebsd.org/smp/
the FreeBSD 5.x kernel isn't even fully preemptive yet.

Darwin STILL uses Mach as a base, and Mach messages are STILL being sent around. Apple's improvements meet Apple's needs for a multimedia OS, not a server OS.

Re:Ditch OS X For Solaris?

[Anonymous Coward]

Nice. Well said. :)

Re:Ditch OS X For Solaris?

[burns210]

"You can also disable the GUI environment if it so pleases you (not like it hogs that much sitting at the login screen anyway)."

Beyond that, any processing power it DOES use is put on the video card, not the main processor.

Power ? PPC

[macmastery]

Isn't the power architecture a superset of the designs used in PowerPC-based Macs?

Re:Power ? PPC

[Guy Harris]

not exactly. read this Guy's comment

Power != PowerPC

[genericacct]

Correct me if I'm wrong, but I don't think PowerPC is code compatible with IBM POWER RISC. They are similar, but PowerPC was a joint project with Motorola.

Re:Power != PowerPC

[Anonymous Coward]

You are wrong.

Re:Power != PowerPC

[2nd Post!]

POWER == PowerPC, but PowerPC != POWER

POWER is a superset of PowerPC. See here.

Re:Power != PowerPC

[wtd]

Might it be more accurate to say that PowerPC is a subset of POWER? An imperfect one at that.

Re:Power != PowerPC

[dgatwood]

Technically, the grandparent is correct insofar as the original POWER had instructions that didn't work on PowerPC (anything dealing with the MQ register, for example) except for the 601 (which was the first POWER family chip to be used in Macs, BTW, and the only one used prior to the G5).

That said, I assume they're porting to POWER IV and V, which are user-instruction compatible with PowerPC, though the supervisor instructions differ significantly. Thus, a POWER series port would be a good start towards making it work on random PowerMac hardware, but initially, such a port would only work on the G5 (and even then, wouldn't support altivec and would probably require additional code to recognize the CPU version...). Additional code in various assembly files (start.s stuff and various VM system changes) would be needed to make such an OS work on older PowerPC CPUs.

Re:Power != PowerPC

[Anonymous Coward]

> > POWER == PowerPC, but PowerPC != POWER

Equality is commutative. (A = B) <=> (B=A). This is nonsense.

> > POWER is a superset of PowerPC.

Now that makes sense.

> Might it be more accurate to say that PowerPC is a subset of POWER?

That's the exact same statement as "POWER is a superset of PowerPC". Thanks for playing, though.

Re:Power != PowerPC

[2nd Post!]

What, so how do you express transitivity?

Is POWER >> PowerPC better?

Re:Power != PowerPC

[Anonymous Coward]

POWER == PowerPC, but PowerPC != POWER


Sleep through our abstract math classes did we? Someone needs to look up the definition of equality.

Re:Power != PowerPC

[Benanov]

I thought some macs used the 603 chips, or are those not POWER chips?

Re:Power != PowerPC

[claudius0425]

This is not exactly true. Neither is a subset of the other, however they do share a large common subset. That said, most modern POWER chips can emulate the extra PowerPC instructions in microcode. This same capability could in theory be included on a PowerPC chip (in reverse), but would be more difficult, as several extra registers (such as MQ) are needed for the POWER spec; it has never been done.

Cain

Re:Power != PowerPC

[vegetasaiyajin]

Equality is commutative. (A = B) (B=A). This is nonsense.

Except in asymptotical notations (Big-O, Big-Theta, Big-Omega).
n^2+n=O(n^2)
n^2+1=O(n^2)
But not viceversa.
Otherwise you could say br n^2+n=n^2+1

Re:Power != PowerPC

[dgatwood]

603 is a pure PowerPC, as were the 604 and 604e. 601 was, IIRC, basically a hybrid between Power2 and PowerPC, but it is commonly described as basically being a Power2, though its MMU may have been its own animal, not sure.

Re:Power != PowerPC

[macmastery]

That's like 6+ years ago now. Ancient history. Everything Apple ships now is G4 or G5 (PPC 7xxx or 9xx).

Re:Power != PowerPC

[Anonymous Coward]

> What, so how do you express transitivity?

I think you mean, how do you express that A is a subset (or superset) of B. "Transitivity" is a more abstract concept. Both of these operations are transitive, but so are many others. (Equality, inequality relations, etc.)

You can see some set theory notation here. There's an operator that looks sort of like an underlined capital C that says "<left thing> is a subset of <right thing>"

Re:Power != PowerPC

[Ianoo]

Equality is commutative. (A = B) (B=A). This is nonsense.

He was coding in "Metaphorical C" (a derivative of Objective C). It allows you to express clauses like "equal but not equal" using this syntax.

Re:Power != PowerPC

[Anonymous Coward]

Asymptotic notation is a confusing and annoying shorthand carried over from Euler. (Nothing against the man; he was absolutely brilliant, just not too concerned about formal soundness.) When writing for myself, I always write "(n^2+n) \in O(n^2)" and read "O(n^2)" as "the set of functions with finite limit f/(n^2)". It's just braindead to use the equal sign when the symmetric axiom for equality doesn't apply.

On the other hand, Knuth seems to enjoy it. Maybe I'm missing something.

Re:Power != PowerPC

[alexre1]

Technically not. O(n^2) is a set, of which n^2+n is a member.

n^2+n "belongs to" O(n^2)
n^2+1 "belongs to" O(n^2)

So even in asymptotic notation, equality is commutative.

Re:Power != PowerPC

[Anonymous Coward]

< When writing for myself, I always write "(n^2+n) \in O(n^2)" and read "O(n^2)" as "the set of functions with finite limit f/(n^2)".

Wow, your way makes much more sense.

Re:Power != PowerPC

[2nd Post!]

Man, everyone is calling me on this.

So read it aloud:
POWER is equal to PowerPC, but PowerPC is not equal to POWER.

Sorry if it was sloppy, but I figure the inclusion of the word 'but' breaks any inherent abstract math in the phrase.

As far as I understand English, an object A can be the equal to object B by being superior, but in the process object B is therefore not the equal of object A.

Re:Power != PowerPC

[javiercero]

No... no... no...

PowerPC is a subset of POWER, in the sense that PowerPC lacks some of the POWER instructions, not the other way around. And no instructions are not emulated using microcode since both are RISC machines and have thus no microcode!

IBM uses PowerPC on some AIX boxen, and are able to run POWER binaries by emulating some of the instructions via OS calls. Pretty simple stuff no need for microcode at all... just translate a POWER instruction into n equivalent PowerPC instructions. And voila...

Re:Power != PowerPC

[javiercero]

True, the 601 was a hybrid of sorts. It implemented most of the 32-bit PowerPC ISA, plus most of POWER. The main difference was that the 601 did not implement the POWER cache control instructions.

The POWER instructions which are not part of the PowerPC instruction set are:

cache control: clcs, clf, cli, dclst
arithmetic: abs, doz, dozi, nabs
MQ register: div, divs, mul, sleq, sliq, slliq, sllq, slq, sraiq, sraq, sreq, sriq, srilq,srlq
others: lscbx, maskg, maskir, mfsri, rac, rlmi, rrib, sle, sre

The POWER MQ register is used in 64 bit multiplication and division: these instructions write to two destination registers, and this feature is not supported in PowerPC. The 601 chip, however, supports all of the above instructions (except cache control) and it even provides the MQ register.

Re:Power != PowerPC

[Guy Harris]

PowerPC is a subset of POWER, in the sense that PowerPC lacks some of the POWER instructions

True.

not the other way around.

False. PowerPC has multiply and divide instructions that don't use the POWER-only MQ register; those instructions were added to PowerPC to compensate for the loss of the MQ-based multiply and divide instructions.

The IBM "POWER" processors might happen to implement the full PowerPC instruction set; if so, that'd make them PowerPC processors as well.

Re:Power != PowerPC

[tbjw]

The proper term is 'symmteric' and not 'commutative'. One uses 'commutative' as a rule for operations and 'symmetric' for relations.

Re:Power != PowerPC

[Chanc_Gorkon]

You sir are wrong. Compile a rpm package on Yellow Dog running on a Mac and move it to AIX (yes AIX can install RPM's) and it works...usually. Barring any library issues that is. Same thing with other PPC based machines. Now if you had coded it using X feature off of x chip, in assembler or if your complier had an option to take advantage of enhancements to the processor it's currently running on...then you may not be able to move it. The grandparent of this post is wrong. POWER=PowerPC and PowerPC=POWER...mostly. Of course if you take advantage of some of IBM's work on POWER and move it to a Motorola chip, you will have issues. But if you just compile for a PPC, it should work on either system.

Since this is Sun...

[FooAtWFU]

Since this is Sun we're talking about, will we end up with Power Architecture Java as a result?

Again

[Anonymous Coward]

This is not the first time Solaris was ported to PPC. Back when Apple, M$, IBM, Novell, Sun, NeXT, and MOT were all more friendly, Sun had ported Solaris to PPC and the ABI was then became the SYSV 32bit PPC ABI.

Even M$ had WinNT ported to PPC and IBM even had OS/2 ported too but those were the days.

Re:Again

[namilax]

And still none of these PowerPC-native operating systems ever worked on Macs, even Macs with Open Firmware! (as far as I know anyway, I only tried to install Windows NT once on my Power Mac)

Re:Again

[dbirchall]

WinNT wasn't so much "ported" to PPC as PPC was one of the architectures it originally supported. (Along with x86, of course, Alpha - the world's first 64-bit PC was in 1993, not 2003! - and, if my memory serves without looking at my NT4WKS CD, MIPS?)

Re:Again

[Anonymous Coward]

Ah! Your memory serves you well. Solaris 2.5.1 was ported to Power. It ran only on an IBM RS/6000 Model 43P, the smallest box IBM made. It actually ran pretty well. It wasn't exactly an academic exercise, but suffice it to say the product never really found a market. They finally took it out of the catalog a few years ago.

Makes you wonder what they're up to. Could this be a prelude to Sun trying to sell themselves to IBM while they're still worth something? Surely they've seen what has happened to SGI, DEC, and DG. Of those previous Unix Workstation Vendor Flamouts (tm), only DEC could be said to have had a decent burial.

Don't get me wrong: I'm not saying Sun is going to die tommorow... the revenue off of DoD maintenance contracts alone will keep them on life support for another decade. But this would give them a chance to get out at better-than-firesale prices.

Could also mean I get to see Solaris on a "fast" machine one last time.

Re:Again

[TomSawyer]

Motorola licensed NT from Microsoft and ported it to PPC. The arrangement was priced out of Motorola's interests when Intel came to fear how much better even NT ran on a superior processor architecture.

IBM talked about getting OS/2 on PPC but it never happened.

Re:Again

[upsidedown_duck]

Could this be a prelude to Sun trying to sell themselves to IBM while they're still worth something?

Wouldn't Sun+IBM be like wearing purple pants with a blue sport coat?

Re:Again

[flaming-opus]

Sort of. Windows NT origionally was written for the i860, which was abandoned. The first release (3.1) ran on x86, mips, and alpha. the 3.5 release ran on powerpc. FYI

Re:Again

[chez69]

it did happen, it was never marketed for general use.

Re:Again

[IPFreely]

I wouldm't say they were attempting to sell themselves.
As processor architecture and performance changes ofer time, it becomes more and more expensive to keep up. Many times before, we've seen companies switch processor and/or hardware because their old basis was not keeping up. Apple switched to PowerPC from 68k. DG switched from 88K (or something older?) to intel. NeXT switched was attempting to switch from 68K to 88K but jumped to intel at the last moment. HP is making the jump to Intel IA64.
Older processor families dissapeared because they couldn't keep up or were too expensive to keep up. Software moves on.

I bet Sun is seeing Sparc performance advantage fading away and a cost sink they can't keep up on. IBM is doing a lot of work to make POWER keep up, and they're doing a good job. Porting Solaris to POWER could be a precursor to Sun making POWER hardware themselves rather than just using IBM hardware. Or maybe they will cut back their hardware all together and go software only on IBM hardware. They still have enough software value to make a go of it.

In either case, it's not necessarily a dying gasp.

Re:Again

[SEE]

Actually, IBM did ship "OS/2 Warp Connect (PowerPC Edition) Version 1.0." To get it, you had to have both a significant commercial relationship with IBM and a computer it would actually run on. But a handful of copies did make it into the wild.

Re:Again

[elmegil]

In either case, it's not necessarily a dying gasp.

That won't stop The Street[tm] from punishing us for it anyway.

Re:Again

[Anonymous Coward]

posting this as a coward :)

NT3.51/NT4 were available for Alpha, x86, MIPS, and PPC.

I don't know about MIPS or PPC, but Windows 2000 was still being developed for Alpha until at least RC2....that's the latest version that I ever saw actually running.

It's too bad that they didn't release Win2k to the public supporting Alpha....it's always been fun to have Windows boxes that are immune to all the x86 viruses...hehehe.

wunderbar!

Re:Again

[Anonymous Coward]

Twitter, is that you?

Please come back, you have been slacking on your dick sucking as of late. And you know how I feel about anonymous dick suckers!

- Stallman

Re:Again

[Anonymous Coward]

> Windows NT origionally was written for the i860, which was abandoned.
> The first release (3.1) ran on x86, mips, and alpha.
> the 3.5 release ran on powerpc.

Rumour has it that the current release runs on PowerPC too (google for rumours about the XBox 2 development system)

Re:Again

[Empty Threats]

Solaris 2.5 and 2.5.1 were both ported to PowerPPC, complete with a Sun compiler suite.

They supported all the Power Series machines (including the PPC thinkpads!) and most of the 43P line. I don't know which 43P's don't work, actually, because the official Sun 2.5.1 PPC edition HCL is lost in the mists of time.

Re:Again

[dumbnose]

And.....what's your point?

Re:Again

[Anonymous Coward]

Yep, it was a 2.5 / 2.5.1 LAR (limited availability release). Additionally, at the time there were ports for Solaris for PA-RISC and MIPS as well.

Decisions within Sun to make hardware >> software basicly meant that these ports never appeared to the general public.

Re:Again

[rs79]

Wouldn't Sun+IBM be like wearing purple pants with a blue sport coat?

(looks in mirror)

And there's something wrong with that?

Re:Again

[Anonymous Coward]

I have an old IBM 43P 7248. Does anybody know if that was supported?

By the way, I think only 2.5.1 was ported, not 2.5.

Curiously, Macrovision still has the FLEXlm daemons and utilities for Solaris/PPC, donwloadable from:
http://www.macrovision.com/services/support/flexlm /lmgrd.shtml#unixdownload

Re:Again

[Kevin Burtch]

Yep, 2.5.1 is the only release that supported PPC... and I'd love to get my hands on a copy.

I have a Motorola PowerStack, an early PReP (pre-CHRP) based machine, that I'd love to see running Solaris.

Re:Again

[cygnusx]

Yup -- in fact, the i860 was codenamed the N-Ten, which is where NT got its "NT" moniker from initially. Marketing then called it "New Technology" and licensed the letters from Northern Telecom.

Open is open

[gtrubetskoy]

Only a customer can define the word "open." That's my view.

To me "open" simply means you can figure out what happens, "customer" has nothing to do with it. When I wrote mod_python I did not think of myself as a vendor and I don't think of mod_python users as "customers". You can't just think of everything in terms of "business", it's not like that at all.

Re:Open is open

[flaming-opus]

You might not want to think about everything in terms of business, but one can, and it's important that he does think about things that way.

Sun is in big trouble. They sell a bunch of decent servers that are not really unique from what the rest of the unix world is selling. They are obviously not able to keep ahead of the competition by making sparc the best processor around, so they have to come up with some other way sell something worth paying for. Solaris, for all its issues, is a reliable, scalable OS that runs a lot of applications. Solaris is a great asset to Sun; If they can leverage it on IBMs processor and make money doing so, it would really help the company.

Sun has moved beyond the "we can do everything in house" days, and is trying to figure out which battles are worth fighting. If they choose the wrong battles, they might go the way of dec, data general, and Sequent.

Re:Open is open

[LoveTheIRS]

Schwartz is arguing that being open means being substitutable for another product. This argument is definately not ignoble. However, I think the open source idea (a. la. the GPL kind of open source), is that there doesn't even have to be a competing code base. If a supporter of a (GPL) open source code base starts coducting business poorly, a entity using the code base does not even have to substitute out the current product they are using. The entity can instead take their support dollars to another corperation supporting the same code base.

Re:Open is open

[CJSpil]

You mean get bought by IBM? That at least is what happened to Sequent

Hardware compatiblity?

[peterprior]

Sun produces both the OS and the hardware for their machines. Apple produce the OS and hardware for their machines. Thats what makes things Just Work (tm). Plonking Solaris on a Mac isn't going to do much for hardware compatibility :|

Re:Hardware compatiblity?

[gl4ss]

..what this probably just means is that they will be starting to make machines with powerpc cpu and selling them with their os(or selling some ibm made hardware with their own os, knowing fully what hardware it will be running on).

Re:Hardware compatiblity?

[ThePuD]

bsd who?

Re:Hardware compatiblity?

[hackstraw]

Plonking Solaris on a Mac isn't going to do much for hardware compatibility

Much more so than plonking Solaris on a PeeCee. Solaris has been available for x86 for years.

Re:Hardware compatiblity?

[misleb]

The reason Apple and Sun can integrate the OS so well with the hardware isn't so much because they designed the hardware (although I'm sure that helps). It is because the set of hardware is very limited and easy to test/certify. You can say "This OS is certified to run on a SPARCStation 20" and it will most likely run well on that hardware. How many video cards, SCSI adapters, motherboards, CPUs, etc does Windows and Linux have to support? How many PC vendors are there each with their own quirks? The permutations are infinite.

Re:Hardware compatiblity?

[vuvewux]

Maybe Linus should start making hardware for Linux, so that it would finally Just Work.

Re:Hardware compatiblity?

[Amiga Trombone]

..what this probably just means is that they will be starting to make machines with powerpc cpu and selling them with their os(or selling some ibm made hardware with their own os, knowing fully what hardware it will be running on).

That could be. For many years, IBM was a Sun re-seller (don't know if they still are). There's no reason that deal couldn't work both ways.

Just because they hate each other, doesn't mean they can't have a mutually beneficial partnership.

Re:Why Solaris ?

[dgatwood]

I can only think of one: unified development and deployment environment. If you're deploying on some Sun box with a bunch of CPUs, it would be nice to be able to run the same OS on both platforms. Admittedly, you could do the same with Linux, though it may not scale as well to the multiple CPUs on the Sun. At least that used to be the case....

IBM's POWER != PowerPC

[DamienMcKenna]

PowerPC is (or at least was) a collaborative effort between the AIM innitiative: Apple, IBM and Motorola. It borrows a lot of concepts and (with the G5) technology from the POWER series but they are not AFAIK compatible. The POWER series is used for IBM's big iron servers and workstations.

Damien

Re:IBM's POWER != PowerPC

[Anonymous Coward]

Not really true, the old POWER series processors are no more and all the ISA on the newest POWERs machines (POWER 3, 4, and 5) are that of PowerPC.

Re:IBM's POWER != PowerPC

[kzinti]

Yup. PowerPC was derived from the POWER architecture; this page: http://www-106.ibm.com/developerworks/library/l-po whist/ gives all the details. (My favorite: the PowerPC can run in either big-endian or little-endian mode - although every use I've heard of runs it in big-endian mode.)

Re:IBM's POWER != PowerPC

[ZackSchil]

VirtualPC used little endian mode to speed up x86 emulation. That's why G5s can't run VPC until the next version comes out.

Re:IBM's POWER != PowerPC

[Guy Harris]

(My favorite: the PowerPC can run in either big-endian or little-endian mode - although every use I've heard of runs it in big-endian mode.)

Solaris ran it in little-endian mode. (Yes, "ran" - Solaris was ported to PowerPC ages ago, back when IBM was thinking of coming out with a line of PowerPC-based PC's, with various OSes available including Windows NT and Solaris.)

Sun == erratic

[hot_Karls_bad_cavern]

Wow. What are they doing over there? Let me preface this by saying i work with solaris daily, i like solaris (love/hate, you know what i mean if you use it), and well, the ultras i have in the house just will not die (not for my lack of trying though).

However, after all these "sorta" announcements from different heads of the crew, i'm getting uneasy about Sun. Java open/closed/free/not-free/for-the-love-of-pete-whi ch-jre-j2se-jrs94x-should i get? Solaris open/closed/free/sorta/java-desktop? Heh, okay just poking fun there, but seriously, do they not seem a little like their top guys don't talk all that much and just make random announcements at this con or that? Yesh.

And i KNOW the roof will raise over the suggestion of dropping osX in favor of Solaris on mac....er, wow, my mind is blown that one might consider doing that for anything other than fun...for a few minutes. Wow, Sun is just makin me uneasy these days - glad i'm not in charge of any huge shops (i assure you that you are glad for that too ;-)

Re:Sun == erratic

[flaming-opus]

The workstation market is proving less and less profitable for sun, and for all the other unix games in town. Since a linux-PC or Mac is so close to a workstation, fewer and fewer people are willing to pay a big premium for a workstation. Thus it's probably not worth sun's engineering time to continue developing workstations. They will continue to develop their higher-end products from the ground up, but use commodity parts at the low end.

Why they would do this on powerpc when they already have an opteron product line, I don't know. I imagine they will not productize any powerpc systems, and are just doing this to thumb their nose at AIX. Either that, or they are making contingency plans in case they decide to become a total software company. In any case, I bet you can keep a limited port going with 2-3 engineers, especially since Solaris already runs on sparc and x86.

Re:Sun == erratic

[SEE]

What you're watching is the flailing of a company that knows its old buisness is doomed.

The RISC performance crown is POWER. The price-performance crown is x86. SPARC is stuck in a market slice between these two, and is getting squeezed. And SPARC is unlikely to be able to invade the x86-and-PPC-dominated desktop market, which means its development will always have fewer resources behind it than the squeezers. There's life left in the SPARC platform, but the way the wind is blowing is clear.

So what to do? Well, Sun's trying lots of things, hoping one sticks. If SPARC is in trouble, maybe Solaris can become the universal high-end Unix, running on any machine (that is, x86 and POWER). Maybe the Java Desktop System can secure Sun a slice of the Linux pie, even if Linux (backed by IBM) improves until leaves no room for Solaris. Maybe Java can save the company. Maybe if Sun open-sources key products, it can get the benefits of open development and still be the company people turn to for commercial support of them. Maybe . . .

Who knows? Maybe something will work. It's worth a shot, at least.

Re:Sun == erratic

[kaiwai]

Regarding the SPARC comments, one must realise that they're eventually going to kill their UltraSPARC and replace it with SPARC64, the ultimate successor to SPARC64 will be ROCK, which is that "geeze wizz" chip with 8 cores, and running 4 threads simultaneously per core, giving a grand total of 32 threads per-cpu, meaning, a nice little 16 way machine will clock and impressive throughput.

With that being said, watch SUN's SPARC workstation shipments fall through the floor once their Opteron workstations start shipping along with software companies coming on line with their Solaris x86-64 versions. Companies will no longer have to justifying an arm and a leg for under performing, over priced workstations.

With that being said, they still don't get it. They assemble their machines in America, yet all their parts come from China/Taiwan. Am I the only one here who finds that stupid? why not assemble them in Taiwan or China, like how IBM has outsourced their worktation/desktop production, just as Apple no longer manufacturers their workstations.

Solaris on the Mac?

[cacepi]

Does this mean you'll soon be able to ditch OS X and stick on Solaris 10 onto Macs?

IBM Power != Mac, silly rabbit.

Re:Solaris on the Mac?

[Anonymous Coward]

The distinction is becoming almost none now adays (again).
There was a time when there was not going to be any....

Re:Solaris on the Mac?

[Guy Harris]

The distinction is becoming almost none now adays (again).

Cool! So I can install from a Panther CD onto a pServer or IntelliStation POWER workstation?

There's more to system architecture than just CPU instruction set architecture. Macs and RS/6000's/IntelliStation POWER workstations/pServers may share a CPU instruction set architecture, but they don't, as far as I know, have compatible firmware, and don't necessarily have the same support chips, I/O bus, peripherals, etc..

by sun's "open" definition

[adamshelley]

by suns open definition, to be able to change easily from the platform, windows is the most open operating system there is.

Linux makes it real easy to get rid of windows.

Re:by sun's "open" definition

[rewt66]

Only if you don't have to take any apps with you...

New world...

[Anonymous Coward]

I run Linux even on old world (gray) Macs

The gray Macs are new world - the gray aliens with the big black eyes that like to hover over Scotland and make circles in crops.

Of course if you are talking abould a Big Mac - they have always been kinda gray inside ...

beware vaporware

[linuxislandsucks]

beware vaporware announcements from a cto that has trouble reading a blance and income statement..

Re:beware vaporware

[hotspur_fan]

1) He's not CTO. He's COO and President. Greg Papadopoulos is CTO.
2) This is not some official PR vaporware announcement. It's some entry in his blog.
3) If anything, for things like this, I trust the execs that can read C code over the ones that can read a balance and income statement. Nothing worse than management who thinks "Things will just work" and underestimate the technical resources and time required to accomplish tasks.

Solaris and Gnome over OS X?

[Offwhite98]

I wonder if it would be worthwhile. I know that Sun had a close relationship with the Gnome community to help improve the usability of Gnome but I still feel that OS X is a much better total UI than Gnome.

I could be wrong, but Solaris and Gnome still have some rough edges which need smoothing out. My biggest critisms of of Solaris/Linux/Gnome is they move onto the never version and new features before the round out and polish the last version. That last 5% of effort to make the software shine is really what sets makes the average computer user feel it is 100% better.

Re:Solaris and Gnome over OS X?

[leandrod]

> I still feel that OS X is a much better total UI than Gnome.

Yet Gnome is rapidly approaching.

Re:Solaris and Gnome over OS X?

[nsayer]

Yet Gnome is rapidly approaching.

For suitable definitions of "rapidly".

Re:Solaris and Gnome over OS X?

[biostatman]

I was recently faced w/ this possibility when I found out that blastwave had ported GNOME 2.6 to Solaris. I actually prefer GNOME for various reasons, one thing in particular is that I like Evolution. I had previously ditched my Ultra 60 for a G3 tower running OSX.

So I wiped clean my Ultra, installed Solaris 9, installed GNOME 2.6. Unfortunately, GNOME on the sun box was DOG slow, and Evolution would crash as soon as I opened up a mail folder. Going back to the OSX machine (similar specs in terms of CPU speed and both have 512MB RAM, the Ultra has SCSI, G3 IDE) there was a quantam leap in speed and app stability.

Its too bad - the "last 5%" - app stability and performance was what kept me from migrating back.

I want to be clear that this isn't a slam at all on the great work that the blastwave folks have done - I am ever so grateful for their efforts.

Re:Solaris and Gnome over OS X?

[Anonymous Coward]

Why not just install gnome on your OSX by using fink (http://fink.sourceforge.net/)?

Re:Solaris and Gnome over OS X?

[Unregistered]

Most people choose a product based on features. They say they want polish, but then pay for features. That's why windows lacks polis, for example. MS knows all they need to do is add features and people will continue buying. Also the same reason why apple advertises all the new features in each version of osx and barely mentions that it's faster and more poished.

Re:Solaris and Gnome over OS X?

[fthomas64]

but I still feel that OS X is a much better total UI than Gnome.

Understatement of the millennium.

Re:Solaris and Gnome over OS X?

[Anonymous Coward]

My biggest critisms of of Solaris/Linux/Gnome is they move onto the never version and new features before the round out and polish the last version.

Yea, the newer verison often seems to be the never version. I pre-paid and I'm still waiting on Gnome Nukem Forever.

Easy decision

[tommasz]

No one in their right mind is going to ditch OS X on a desktop machine for Solaris. No one. It might have a chance as a server OS but given that you can already run Linux on the Power architecture, there's no compelling reason to consider Solaris unless you're already a Solaris shop and want to buy Power machines.

Re:Easy decision

[jschottm]

I'd consider it if Linux didn't offer the same capability. For some of us, the unconfigurable, different-just-to-be-different UI is horrible.

Re:Easy decision

[Anonymous Coward]

I'm in my right mind and I'd do it in a heartbeat.

I like Solaris and am comfortable with it and, more importantly, it lets me configure things more than Mac OS X.

THERE IS NO DOCK REQUIREMENT IN SOLARIS.

And no offense to Mac OS X, but Solaris crashes less often. Of course I wouldn't have the high quality Sun hardware if I ran on commodity Macs (no ECC RAM, for example) but oh well, it couldn't be the hardware oh no, not that.

Re:Easy decision

[burns210]

People talk like Linux and Solaris are on equal footing... They arn't. Linux is awesome, I am a big fan(though I still love my mac). However, Solaris(trusted solaris, in particular) is better refined, more secure, more stable, more tested, high government rating.

Solaris isn't the lowend solution to running a webserver, email server, or what not.. But then, if you have an Apple Xserve, why would you want Linux to do those tasks?

Solaris still has Linux beat hands down in the highend. Linux just isn't there yet... yet.

when do we get Real Stuff and not Sound Bites?

[MarkEst1973]

Schwartz's blog and just about every press announcement from Sun lately seems to be nothing but smoke-up-my-ass vaporware and/or hollow promises.

You can consider that sentence flamebait or you can take it is my open letter to Sun to "Put up or Shut up". I, for one, would like to see some more follow-through on many of these announcements, like an open source Java and Solaris.

Re:when do we get Real Stuff and not Sound Bites?

[Johnny Mnemonic]

I agree.

It's almost as if their strategy is "ignore the man behind the curtain! Look at all the shiny things we might do!"

But: will any of those things keep them from losing $1B every quarter? Unless you can answer that question in the affirmative, and it'll ship soon, and not cost more to develop than it brings in, it's just looks like an attempt to distract from their failing core business.

Re:when do we get Real Stuff and not Sound Bites?

[Jim_Maryland]

Please define exactly what your definition of "open source" is. As the blog indicated, open source potentially has a different meaning for people. If by open source you mean that you want access to Solaris, check out Solaris Source (although this really just states: "Source code for the Solaris Operating System is available for qualified educational institutions and partners; please contact your Sun sales team for details"). I don't know what disclosure agreements are necessary for the access, but a determined individual should be able to get access.

Forget Macs, P series!

[telemonster]

Who cares about running Solaris on the Mac G5, look at IBM's efforts to convert Solaris/Sun shops over to AIX/RS6k shops! If you browse IBM's page looking at the pSeries servers (the Power series) you will notice ads about migrating from Solaris to AIX. This is a big inititive at IBM.



From our standpoint, it's goes a bit like "ewww AIX" ... Solaris on the pSeries boxes would definitly be interesting. I believe IBM rebadges quite a bit of commoditiy hardware and marks the price up 900% (Older advanced 3d graphics cards for RS6000s were $30 s3 cards with different PCI identification tags and such)... so it might be easy to pick up support for quite a bit of the peripheral hardware from the Linux world.

I'm not sure I'd shove it into a production environment, and what if IBM starts to throw curveballs into the works to thwart the people running Solaris. Still totally funny if you ask my opinion. Talk about a comeback to IBM's marketing strategy, but at what cost to Sun's hardware sales.

Re:Forget Macs, P series!

[Anonymous Coward]

IBM gives away AIX with pSeries. If Sun wants to give away Solaris and earn IBM pSeries sales, IBM will probably not complain.

Re:Forget Macs, P series!

[tofu2go]

i don't think IBM would want to thwart people running Solaris on IBM hardware. IBM's software divisions make software for Solaris already. i don't think IBM makes money off the OS. they only need an OS like AIX to be able to provide a one-stop total solutions package. if people chose to run Solaris on IBM hardware that's fine, so long as IBM makes money on the hardware and software stacks.

i don't really think there's money to be made in Operating Systems unless you're planning to be like Microsoft, lock people in and charge exorbitant prices. with Unix platforms, most enterprise applications are Java-based, so lock-in is less of an issue at the OS level; lock in is more likely to occur at the Application Server level. money is made on middleware, not the OS.

vendors tend to think that if a customer uses their OS and their hardware, they'll use their middleware, but that's not the case here. at the application server level, IBM's WebSphere is doing better than Sun's SunONE, so what does IBM care about AIX vs Solaris? WebSphere runs on Solaris.

Re:Forget Macs, P series!

[Wesley Felter]

IBM does not give away AIX any more. A "Linux-ready" pSeries box with no OS costs less than a pSeries with AIX.

Re:Forget Macs, P series!

[grawk]

Anyone who says "ewww AIX" hasn't run AIX in a production environment. There's a reason why banks and insurance companies run AIX. It's rock solid. And now that IBM's hardware is the fastest in the world, there's no compelling reason to run anything else.

Re:Forget Macs, P series!

[Anonymous Coward]

Intresting. We priced out a pSeries p615 recently, both Linux-ready and w/AIX and they were the same price.

Re:Forget Macs, P series!

[Anonymous Coward]

Rock solid? Hrm, not always. AIX on IBM has been known to be flaky as hell. I've been in one company where a big HA cluster was less reliable than the previous non-HA environment they replaced. We've now had a p690 crash in my current job; not just an LPAR, but the entire frame.

In summary, don't say AIX is rock solid when it isn't.

Posted anonymously because I'd probably get my ass whipped by my current employer (a bank FWIW; preferred platforms are the AS/400s and HP, although there is a lot of IBM as well).

Re:Forget Macs, P series!

[memfrob]

There's a reason why banks and insurance companies run AIX. It's rock solid.

As the saying goes, I'll use AIX once they finish porting it to UNIX. :P

Re:Forget Macs, P series!

[Amiga Trombone]

I'm not sure I'd shove it into a production environment, and what if IBM starts to throw curveballs into the works to thwart the people running Solaris. Still totally funny if you ask my opinion. Talk about a comeback to IBM's marketing strategy, but at what cost to Sun's hardware sales.

This inclines me to believe that if Sun is planning on porting Solaris to Power, they plan on building their own servers based on it, or at least collaborating with a third party who will. If they were planning to support IBM's hardware, it certainly wouldn't be any fly in IBM's ointment, since it would just be one more option IBM could offer on a partitionable server like a P670/P690. The customer would be able to run Solaris, AIX and Linux (and OS/400 in the next iteration) in their own partitions on the same frame. This certainly doesn't do anything to hurt IBM, it just helps them migrate customers off of Sun's equipment on to theirs. The only way this makes sense is if Sun is planning to use Power as the basis for some of their own servers. Otherwise, what's the advantage to Sun? They make their money as a hardware vendor. The revenue from Solaris itself is negligable.

Re:Forget Macs, P series!

[oh]

I've been in one company where a big HA cluster was less reliable than the previous non-HA environment they replaced.

Not an uncommon story. High Availability clusters are great in theory, but they introduce complexity. If you have a system 5 components, and a failure of any one of them would stop your system from working, is that worse then a system of 100 components, of which any two failures would have the same impact?

This is an extreme example, but quite often you will have more failures caused by the clustering then are saved.

Of course when it all works it is beautiful. The feeling when you ring a user up and ask if they have just had any problems and having them say no, when you know full well that a server has just rebooted.

This targets AIX, not OSX

[grunt107]

Sun is attempting to hijack the UNIX server market served by IBM's AIX server line.

This could also signal the end of large $ spending on R&D by Sun (why innovate when others are creating the computers that the OS can run on?).

Re:This targets AIX, not OSX

[mihalis]

This could also signal the end of large $ spending on R&D by Sun (why innovate when others are creating the computers that the OS can run on?).

Or...This could also signal the end of large $ spending on R&D by IBM (why innovate when others are creating the OS that the computer runs?).

Re:This targets AIX, not OSX

[Bralkein]

Well... no, I don't really think so, if everyone started running Solaris on their IBM machines instead of AIX, I think that would piss IBM off a bit. They want IBM customers to run IBM software on IBM machines, so they'd put money into R&D to make AIX better, so that there would be no reason to switch over to Solaris at all.

Re:This targets AIX, not OSX

[Anonymous Coward]

No, IBM is sending AIX to the garbage heap where
it belongs. Linux is their (questionable)
future.

Re:This targets AIX, not OSX

[grunt107]

Combine the 2: IBM LAIX ("Lakes") OS

Re:This targets AIX, not OSX

[grawk]

Sun stopped spending big money on R&D ages ago. Certainly compared to other players in the market. That's why sun dropped off the top of the top 500 so dramatically. They've essentially abandonned the sparc architecture. Now they just have to hope that someone wants to run solaris when they could run Linux or AIX instead.

Re:This targets AIX, not OSX

[JonAnderson]

1) Where did you get this from? Can you please post a link to Sun's R&D budget relative to it's industry peers to support this statement. 2) sparc has not been abandoned. 2 processor lines were shelved because alternative designs(Rock and Niagara) were proving to be more competetive. Intel did the same thing but were branded visionary. At least Sun has a processor strategy (hp.com). 3) Yes, you could use Linux or AIX - but which would give lower TCO???

Open is...

[Offwhite98]

...leaving the case off because you had tried to install a new hard drive and you got frustrated with it so you just left the PC sit open for a week because you just did not want to deal with it anymore.

Re:Open is...

[avandesande]

Hah--- I have noticed that my computer's dependability depends on if the case covers are on... if it's open it works for months but as soon as I put it back on something breaks.

Re:Open is...

[confused one]

That's 'cause you need more neon and a few more fans.

Contradictory...

[RU_Areo]

Johnny boy states that "Only the customer can define open" but then proceeds, to define it, not to mention plug Sun's products. This seems contraditory

Re:Open is open, but to who?

[davecb]

I think he's preaching to the business community, for whom the ability to buy a different brand of computer for a new lab is a Real Big Thing.

Remember "vendor lock-in"? Used to happen with IBM mainframes, then Windows, and now, regrettably, with Unix variants.

The freedom to be able to chose a vendor is important to businesses and universities, and in principle to anyone who doesn't want to be locked to a particular vendor. Such as Sequent, who sorta doesn't exist any more...

I used to do a ton of porting for the purpose of unlocking stuff from vendor X or Y and making it run on "stock Unix", which is to say, pretty much anywhere. heck, I still do, on request (;-))

--dave

Where do Microsoft come in?

[polyp2000]

Anyone care to speculate on the possible benefits for Microsoft if this happens? Since Sun are now sleeping with the boys from Redmond, there must be an alterior motive here... must there not ?

Nick...

Re:Where do Microsoft come in?

[upsidedown_duck]

Since Sun are now sleeping with the boys from Redmond, there must be an alterior motive here... must there not ?

As far as I know, Sun is still not an OEM for Microsoft by any measure, nor does Sun sell any Microsoft software at all. If they made any deals, I would think that they were of the licensing/patents/IP/etc. nature. I would hope Sun is smart enough not go the way of companies like Intergraph or, almost, SGI. Sun is certainly no friend to MS, and anything more than lawyer-games between the two companies would be like Kerry signing up Bush for VP (or vice-versa, to be fair).

Licensing issue

[Anonymous Coward]

But you KNOW that SCO owns java's license and java's source was stolen from xenixware.

I am bill gates and I approve this message.

Mono Project

[ddkilzer]

Apparently, Jonathon doesn't know about the Mono Project, since moving a .NET application to another platform wouldn't require a rewrite.

for Solaris to truely work on apple's product line

[mrfibbi]

They would have to strike a deal with Broadcom. Right now the reason I had to wipe YDL from my powerbook was that there was no driver available for airport extreme (apple's 802.11b/g card), severly impairing the notebook's mobile functionality.

Broadcom refuses to diologue with the linux community. I wonder if they'll consent to do it with sun?

Can we say behind the times???

[Supp0rtLinux]

Sun just doesn't get it. We already have MacOS X and Yellow Dog for PPC. We have Sun and Linux the SPARC. We have M$ and Linux for the x86. Linux is the common denominator. Why the heck would we care about Solaris on PPC?

Sun is trying every last ditch effort they can to stay afloat. The company that believed the world revolved around Solaris and SPARC is now supporting X86 and AMD64 and talking about PPC. They're offering Linux solutions. Everyone else sees the sinking ship that is Sun, but Sun themselves. Unfortunately, I can't help but think the old adage of "a day late and a dollar short" is going to apply to Sun very shortly, if not already.

Re:Can we say behind the times???

[Anonymous Coward]

The PPC != POWER architecture. This is about IBM's big iron, not Apple's overpriced brushed aluminum jokes.

It has nothing to do with OS/X, everything to do with AIX and Linux. The article poster just happens to be as ignorant of technology as you are.

Re:Can we say behind the times???

[FlashBIOS]

And OS/X != OS X

You should make sure you know what you are doing when you call someone else ignorant.

Re:Can we say behind the times???

[upsidedown_duck]

Linux is the common denominator. Why the heck would we care about Solaris on PPC?

Maybe Sun could get OpenGL working well without requiring the user to recompile the kernel and install new modules for their particular version of XFree86, for example.

Re:Can we say behind the times???

[Anonymous Coward]

Linux is common only in name.

Solaris beats Linux hands-down. Always has.
Always will. And it's far cheaper! Solaris 10
will make this even more evident.

I could care less if current Solaris ran on
Macs. What I want is Apple & Sun to merge and
put Apple's Aqua & applications on Solaris.

Re:Can we say behind the times???

[duffbeer703]

Brushed aluminum has been proven to increase the performance of just about anything.

The brushed aluminum wing on my Corolla has made it a speed machine on par with any Porsche

Re:Can we say behind the times???

[Anonymous Coward]

We already have MacOS X and Yellow Dog for PPC. We have Sun and Linux the SPARC. We have M$ and Linux for the x86. Linux is the common denominator.

Hmmmm. NetBSD runs on all of those Platforms.

Not that your nickname here....Supp0rtLinux would admit to.

Re:Can we say behind the times???

[Brandybuck]

Is that your Corolla? I say that on the street and I just about drove off the road I was laughing so hard.

Re:Can we say behind the times???

[Anonymous Coward]

Because they're not porting to PowerPC! They're porting to POWER! PPC !~ POWER

Re:Can we say behind the times???

[Anonymous Coward]

His was a spelling/usage issue, he was responding to a gigantic not-getting-it-at-all issue. They hardly compare.

Re:Can we say behind the times???

[confused one]

Actually, it's you who may not get it...

Sure, Linux and BSD run on PPC, Sparc, x86 & AMD64 hardware. Some companies don't want to run Linux; they'd rather run a fully supported & certified, scaleable Unix. If Sun can get Solaris to run on all of these platforms (including Power4 & Power5), then they'll be the only Unix vendor in this position.

Re:Can we say behind the times???

[Cyno]

Its because, if you ask Sun, Solaris is better than Linux and will always be. It would be open source, too, if they were allowed to release it under an open source license. But unfortunately no matter how many times they pay for extra UNIX protection from SCO for their customers SCO keeps telling them they can't open source it. So I think if they port it to every platform available and pay SCO a few more million to continue its fight against Linux Sun will come out the victor. What do you think? :)

Honestly, though, its like our current Administration. They cannot admit that Iraq was a mistake or it will show up as a sign of weakness and they will be attacked and killed. Its a psychelogical perception thing, like most of capitalism. Once the competing technologies solidify there will be two factions: the believers and the thinkers.

Re:Can we say behind the times???

[Anonymous Coward]

Some companies don't want to run Linux; they'd rather run a fully supported & certified, scaleable Unix.

You really are trolling. Linux is fully supported, certified, and scalable.

Re:Can we say behind the times???

[FlashBIOS]

No. He was responding to an "opinion" that he disagreed with by calling it ignorant. Disagree != ignorant.

On the other hand, not knowing the name of one the most popular operating systems in the world, when he is a geek himself, is ignorance.

Re:Can we say behind the times???

[Meowing]

The PPC != POWER architecture. This is about IBM's big iron, not Apple's overpriced brushed aluminum jokes.

Of course, AIX runs on both POWER and PowerPC, and can use the same binaries. That is because there is little difference between the instructoin sets, and especially because IBM use PowerPC in their own server products.

Re:Can we say behind the times???

[Anonymous Coward]

tell me something...how many dicks in your life you suck?

You know

[shfted!]

I've used Sun workstations a lot, but they sure felt sluggish. I guess they really could use a little more Power. Heh.

Open Standards vs. Open Source

[booch]

His main premise is that Open Standards are more important than Open Source. On this point, I completely agree. Conforming to an open standard, which anyone is allowed to implement, is a great thing for customers. As long as they depend only upon the standard, they can choose whichever vendor they want. This is effectively a commoditization of the market.

What he fails to realize (or admit) is that Open Source has other advantages that build upon Open Standards. Even if an Open Source program doesn't conform to any well-recognized standard, the availability of the source can provide the same advantages. If you don't like the way Ximian is building their free Evolution mail reader, you can find another vendor who will take the existing mail reader and build you a custom version, fully compatible with the old. Also, Open Source programs typically embrace Open Standards with a passion. Look at Mozilla for a good example.

In addition, Open Source provides new advantages that Open Standards do not. The main advantage is control. If the company goes out of business, and you want to stick with their product, you can do that. If the vendor doesn't want to implement a feature that you want, you can do that. You get the advantages of commoditization, plus the ability to customize and modify things to fit your own needs.

Re:Open Standards vs. Open Source

[aristotle-dude]

You are forgetting a couple things:
1. Some companies don't have or want to waste IT resources on development of products they use.
2. Companies that use software built on open standards/formats can easily move to another vendor.
3. If a vendor goes out of business, their product will not stop working unless incompatibilities arise with the OS.

Re:Open Standards vs. Open Source

[booch]

1. Something like 90-95% of all development is for in-house projects. A lot of companies do spend a lot of resources developing software. Even if they don't want to spend their own resources, they have the option to, or to get someone else to help them with it. The Freedom that Free Software provides is more control of their own destiny.

2. I don't disagree with that at all. I thought I had stated as much in my initial post, or at least implied it by agreeing with Sun's statements.

3. There are a lot of things that can go wrong when a vendor goes out of business. New bugs and vulnerabilities may be found in the product. The company may want to use it in new ways. They may change the systems that the product interfaces with, and find incompatibilities. Patches, updates, or upgrades of the OS or underlying libraries/infrastructure/dependencies may cause the software to break. (XP SP2 anyone?)

Why Solaris on POWER?

[BinxBolling]

The Mac hardware thing is mostly a red herring, I'm guessing.

Here's my guess: Sun is considering the idea of dumping SPARC in favor of POWER. As things stand, they're way back in the raw performance game. Why continue investing R&D money into their own line of chips, if this is what it buys them?

Note that I'm not suggesting that they would become a pure software company -- my guess would be that they still design and build their own systems, just not their own chips.

Re:Why Solaris on POWER?

[upsidedown_duck]

Sun is considering the idea of dumping SPARC in favor of POWER.

This is pretty unlikely, as I would see Sun adopting Opteron before anything from IBM. More likely is they are either trying to become as platform-neutral as Linux or they are trying to be a thorn in IBM's side, somehow.

Re:Why Solaris on POWER?

[Anonymous Coward]

Not a chance. With Fujitsu they will be
competitive with POWER. Niagra & Rock
will blow the doors off Intel.

Putting Solaris on Power & Itanic is just a
way to show that Solaris can do anything
Linux can do - better. And a way to kill
AIX & HP-UX for good.

Re:Why Solaris on POWER?

[Wesley Felter]

Actually, Fujitsu's SPARC64 V has pretty decent performance.

Re:Why Solaris on POWER?

[_damnit_]

Sun just signed a big deal with Fujitsu to use their Sparc chips in their next Big Iron box which they'll jointly develop. After that generation, Sun plans on having their next "throughput computing" chips running their servers. While you may ponder whether it's at that point they'll abandon Sparc, I doubt it. That's quite a few years off and it doesn't take that long to port an OS if you really concentrate your personel.

Don't worry, Sparc will be around for a while longer.

Re:Why Solaris on POWER?

[Wiz]

Pretty good, but not compared to to the other high end chips.

Itanium 2 & POWER 5 walk all over it. I suspect the Opteron will too quite easily, and it costs a lot less.

Re:Why Solaris on POWER?

[Anonymous Coward]

Yeah, and the AMD/Microsoft partnerships they kept making sounded so ghetto that they needed a cooler sounding announcement.

Re:Why Solaris on POWER?

[misleb]

Why would Sun choose Opteron, with all of its x86 cruft, when they could get the more pure and open POWER architecture?

Re:Why Solaris on POWER?

[Amiga Trombone]

This is pretty unlikely, as I would see Sun adopting Opteron before anything from IBM. More likely is they are either trying to become as platform-neutral as Linux or they are trying to be a thorn in IBM's side, somehow.

Power has the advantage of being an open architecture, as opposed to Opteron. If Sun were planning to do customized implementations of a processor for their hardware, Power would be the logical choice. They don't have the option of extending Optreron.

Re:Why Solaris on POWER?

[chegosaurus]

I don't agree. I think Sun are heading towards their own backplanes, I/O etc but with Opterons instead of SPARCs. Seems like the sensible thing to do. They know how to build a good computer, but they have to see the SPARC consortium can't hope to compete with intel and AMD.

Solaris is so damn sweet. Really. If you're really into Linux you owe it to yourself to try Solaris 10. And give it a really good try, not just "Install is ghey. CDE is teh suX, back to gentoo".

I *really* want a lovely Mac laptop, but I pretty much need Solaris on my portable. mmmmm, Sol 10/OS X dual boot.... I'd be surprised if this ever surfaced, but I so hope it does.

Re:Why Solaris on POWER?

[upsidedown_duck]

Why would Sun choose Opteron, with all of its x86 cruft, when they could get the more pure and open POWER architecture?

Well, I was just thinking that Sun hates AMD less than IBM and Intel, and that Sun's newest servers are based on Opteron. Opteron is less crufty than Xeon, at least, and, actually, Opteron's interconnect architecture isn't very different from the UltraSPARC IIIi. I wonder if a motherboard could be designed to work with either processor (given same pin-out or some adaptor).

Re:Why Solaris on POWER?

[Anonymous Coward]

Could it be that re-hiring Andy Bechtolsheim and aquiring AMD/Opteron based Kealia had something to do with their direction? Do people even bother to read the news anymore? Pay attention.

Java. Open for webapps but not for linux apps

[mpcooke3]

Jonathan Schwatz is right with regard to webapps, the attempt at making apps portable across JVM's and application servers is very useful and does make java seem more open than say .NET

However due to the JVM license restrictions and lack of a good OSI compliant license for the JVM and class library source, Java is not open enough to become a primary linux development language.

It is a critical time for SUN if they ever want Java to become a primary development language for Linux.

To keep up with the competition linux development is going to have to move to higher level languages for the development of core desktop applications, such as the Gnome suite. As far as I can tell there are only a few realistic options, Mono, Java, Python or something new.

With the M$/patent concerns over mono the immaturity of python and the lack of any other options now would be a damned good time to properly open source Java.

That is if they care about being a leader in the inevitable linux desktop takeover :)

Re:Java. Open for webapps but not for linux apps

[ThisNukes4u]

There is never going to be a new primary developement language for Linux except for the one that is already being used, C/C++. As long as the kernel is written in C, and Glibc is still the primary way of doing programming in linux, C is here to stay. Hell, you can't even write a true assembly language program in linux without using Glibc unless you know every kernel call to do what you want to do.

Re:Java. Open for webapps but not for linux apps

[mpcooke3]

Whether Mono, Java or python is layered ontop of Glibc or not is largely irrelevant. If you think there won't be a migration away from C/C++ to modern languages that provide protection from buffer overruns, pointer errors and with decent garbage collection then you either think linux is just "dead" or you are very nieve.

If what you say is true it would indicate linux will never catch up with windows/osX and I'm not convinced that is true now that there is so much interest in linux and so much money is being poured in by the likes of IBM.

Right

[bayerwerke]

Yeh, right after they buy Novell.

Solaris does not excite me

[mi]

Good kernel, but awful user-space. Outdated utilities (starting with the /bin/sh itself), awfully user-unfriendly "out of the box" install. Linux may not follow the standards either, but, at least, the command-line editing works out of the box... FreeBSD remains my choice, of course...

Re:Solaris does not excite me

[Anonymous Coward]

Outdated probably, but /bin/sh isn't the place to start I can't recall the number of times I've been screwed by some Linux programmer making a script that only works only in bash and has "#!/bin/sh" written on top. Or POSIX complient shell scripts working everywhere else but Linux (however this is becoming very rare).

I've seen terrible mess where someone replaced all the basic system utilities (most the stuff in /bin) with the gnu equivalent...

Re:Solaris does not excite me

[Anonymous Coward]

Outdated utilities (starting with the /bin/sh itself)

Help me out here; are you saying that /bin/sh itself should be changed? If so, I disagree -- one of Solaris' major strengths is its compatibility with what went before, and replacing /bin/sh with, say, bash would seriously jeopardize this. /bin/sh should remain the Bourne shell, so that scripts continue to Just Work.

Now, if what you're saying is that Solaris should include more modern shells, I agree -- and so does Sun, apparently, since Solaris has included bash, tcsh, ksh (including dtksh) and zsh in the standard install for many years now. The key part is that they've done this without breaking what came before, and that's worth a lot in most if not all environments where real work needs to get done.

Change for the sake of change, and even change in the name of progress, can be a bad thing. As a decades-long vi user, I find vim to be obviously superior -- but still annoying when things I expect to work based on my vi experience don't.

Re:Solaris does not excite me

[mi]

All sh-scripts I saw on Solaris -- starting with Sun's own -- use /bin/ksh.

I'm not advocating change for the sake of change, but for the sake of bug-fixing and feature adding. A vi, that can't handle my window width is broken.

An awk, that can't handle, what nawk can, need not exist. And so on.

Why can't I edit a command line in the default shell? Why bother shipping csh at all, when tcsh is the fully compatible and (much) newer version?

Why does not find(1) accept ``-print0'' (and xargs -- ``-0'')?.. It is 21st century, files with blanks and other non-printables in their names have been around for years...

These are just random picks. All commercial Unixes, I saw, suffer from these problems.

teh spoke??!1~`

[Anonymous Coward]

OMFG I <3 YUO

More /. misinformation...

[Anonymous Coward]

Statements like, -"Does this mean you'll soon be able to ditch OS X and stick on Solaris 10 onto Macs?"- shows how little the author understands the technology (s)he is reporting. The bigger question is with a real, solid OS like Solaris 10 (which can run native linux apps) running on Sparc/x86/Power platforms, why would a corporate customer even hassle with linux. And even if Sun were porting Solaris to Apple's PPC architecture like the authour inappropriately suggests, why would you want to downgrade your desktop to CDE or gnome?

Silly rabbit, Solaris is for Servers!

[nsayer]

Does this mean you'll soon be able to ditch OS X and stick on [sic] Solaris 10 onto Macs?

Ugh, why would you want to?

Now, Solaris on an XServe... That makes sense... Server class hardware that doesn't suck, yet doesn't cost an arm and a leg, running perhaps the best multiprocessor Unix ever... Mmmmm.

The ironic thing in my view is that this is sort of what CHiRP was supposed to be - a happy universe where you could buy an RS6000 and run MacOS on it, or a Mac and run Solaris on it, or whatever. But then His Steveness decided that the clones had to go...

Re:Silly rabbit, Solaris is for Servers!

[Anonymous Coward]

"Now, Solaris on an XServe [apple.com]... That makes sense... Server class hardware that doesn't suck [intel.com], yet doesn't cost an arm and a leg [sun.com], running perhaps the best multiprocessor Unix ever [sun.com]... Mmmmm."

Didn't you mean:

Now, Solaris on an XServe [apple.com]... That makes NO sense... Server class hardware that doesn't suck [AMD/Sun serrvers]], yet doesn't cost an arm and a leg [apple.com], running perhaps the best multiprocessor Unix ever [BSD]... Mmmmm.

I'd take BSD running on an Ultrasparc any day over Solaris running on a xserv.

Re:Silly rabbit, Solaris is for Servers!

[nsayer]

Didn't you mean...

No. If I had meant that, I probably wouldn't have signed my name either.

Only a customer can define the word "open."...

[omeomi]

I like how he starts out with, "Only a customer can define the word 'open.' That's my view.", and then procedes to define the word 'open'.

Re:Only a customer can define the word "open."...

[Anonymous Coward]

Maybe that is the way customers define it to
him.

but, open source makes it easier to migrate

[Anonymous Coward]

I agree with his notion that being able to easily migrate from one system to another is important. However, this is not entirely unrelated to open source, because open source software makes it easy to migrate between platforms. If there are no binaries available for your platform, you can recompile your application. If the platform itself has bugs, you can fix them. And probably most importantly at all, if your application and the platform aren't interoperating properly, you can read both sets of source code and figure out why.

The author seems to suggest that if the platform is standards-complaint, this won't be an issue, and in theory, that's true. But in practice, standards are complex and sometimes unclear, and so compatibility is rarely 100%.

Re:but, open source makes it easier to migrate

[fitten]

I agree with his notion that being able to easily migrate from one system to another is important. However, this is not entirely unrelated to open source, because open source software makes it easy to migrate between platforms. If there are no binaries available for your platform, you can recompile your application. If the platform itself has bugs, you can fix them. And probably most importantly at all, if your application and the platform aren't interoperating properly, you can read both sets of source code and figure out why.

Yes, but the relationship in your statement is backwards, in my opinion.

Ease of migration isn't related to Open Source. Open Source is related to ease of migration. It may sound nit-picky but Open Source did not bring about ease of migration. To some degree, the desire for ease of migration caused Open Source to come about.

That being said... the majority of the ease of migration provided for by OSS is between different ports of Linux and not necessarily between Linux and other operating systems. The Linux Zealots' goal isn't necessarily about choice. Their version of the world isn't so different from Microsoft's in that there is only one OS... the only difference is which OS it is.

The power of the Schwartz

[mesmartyoudumb]

I see your schwartz is as big as mine!

which defintion?

[grocer]

First line of article:

"Only a customer can define the word 'open.' That's my view"

Conclusion:

"Open as in door, is different than open as in source. Unix, linux, Windows - none are open, I'd argue. There is no agreed upon specification, no neutral test to determine validity, and no guarantee made by vendors other than rhetoric."

Apparently, Schwartz wants a gatekeeper to insure that all libaries and ancillary programs are standard between Websphere, BEA, and JES. In short, he's complaining that IBM keeps adding features outside of the TCK/AVK "standard" (apparently defined by Sun), pushing Sun out of the market.

Geesh, here's a novel idea -- innovate! Out-feature IBM, open source the environment and libraries, package support with a linux distribution, and then sell, sell, sell!

Re:which defintion?

[Anonymous Coward]

Geeze, when MS did this to sun it was all "Embrace and Extend" used as a whip on them. When IBM (now somehow a /. "friend") does it... nothing, "Sun, you just need to out do them!"

Well you are right of course, but why was that not the mantra with MS? MS did a better version of Java than Sun, why? Becauase Sun had completely dropped the ball on Java and left it in a very low state of useability with virtually zero development tools.

If you leave your work up to others to do... they will do it for you IN THEIR OWN WAY!

Seems like Sun is still being Sun, and the /. hipocrisy is as strong as ever.

Sun put Solaris on PowerPC a LOOONG time ago...

[timepilot]

It's Deja Vu all over again.

Solaris 2.5.1 had a "PowerPC edition"

http://docs.sun.com/db/doc/802-4127

This was back in the day when you could also get NeXTSTEP for Motorola 68K, Intel x86, SPARC, and HPPA.

Competition is Great

[WindBourne]

Several years ago, Solaris ran on Sparc and x86. Of course, the solaris X86 was the bastard child.

Likewise, mainstream Windows ran on X86 compatable only (yeah, NT ran on alpha, but that was a decade ago; And yes I saw NT on PA-RISC, but it was never released).

In addition, Windows will have a a 3'rd world distro that will cost but a fraction of their current stuff, but have 99% of what they currently offer. Historically, Bill Gates encourages theft of Windows as a way to check growth in other areas. That happened to Borland, Sybase, etc. These days MS claims that linux growth in 3'rd world country is so that it can be replaced by Windows. If so, then why do they think that a low cost version will be bought by end customers, when they can have it for free?

Linux and BSD run on many arch. and the 2 of them are making huge inroads into older OSs. Suddenly Windows and Solaris want to port to everything. Solaris on multiple platforms and low-cost to free windows is simply an attempt to stop Linux from wiping out sales

What people fail to see...

[l4m3z0r]

Alot of people seem to be saying how awful/useless this is and how sun is dying. Those who don't are talking about the various tech aspects that make solaris a viable thing to use. Both sides are failing to see the main reason why Sun would want to do this IMO. In order to be able to say to potential customers, look you have these architecture machines, our OS can run on them. I know that I would be much more interested in hearing some company pitch a solution if all my various architectures could be supported by one OS. My Power arch machines, my x86 machines, and of course the fancy new sparc stations they are trying to sell me could now all run solaris. I would love to hear that, I'd be say great, I can cut IT costs by having the support team worry about only one OS and the migration would be cheaper because I could reuse the expensive Power machines that we bought a few years back.

I abhor diversity when it comes to computers its just a pain in the ass. Any chance I can get to have all my equipment running the same software I'd jump at. Jon's arguments apply mostly to the business end, he isnt trying to pitch superior tech, just a superior business/IT plan.

Re:What people fail to see...

[Anonymous Coward]

That would be really cool... One OS on all the different hardware platforms. The not so cool part is that all of the applications are still specific to the hardware they are running on. Running Solaris on IBM hardware will still require you to run hardware compiled specifically for your HARDWARE/OS combination.

Re:What people fail to see...

[Anonymous Coward]

to bad sun doesn't make any type of platform agnostic language/toolkit. If you could somehow come up with a compiler that generated intermediate byte-code to run on all that different hardware, you would really be able to pitch a solution.

oh well, I guess I will go get a cup of coffee and ponder what could have been.

And it wasn't ported to 64 bit according to this

[xyote]

blog entry.

All I gota say is KILLER!

[Thaidog]

KILLER! There I said it.

advantages

[sad_]

seems to me sun is working at making solaris look more and more like linux all the time.

you can run linux apps on solaris x86 and it should/will be LSB compliant, solaris will almost run on every platform linux will run on. gnome will replace CDE, more and more gnu/linux tools get introduced into solaris. last but not least, they talk (but where is the action??) about making solaris open source!

can anybody see a trend here?

There Is No 2200 CPU OSX Machine

[cmholm]

I don't see any 16 processor machines that run OS X

"Haven't you heard of this one running OSX on 2200 processors"

The Virginia Tech cluster isn't a machine, it's a pile of PCs communicating via MPI, like any other Beowulf cluster. What the previous poster meant was OS support for SMP... CPUs in one box handled by one instance of the OS. I'd be more than happy to see a 4 or 16 CPU Apple, but there ain't one. Anyway, as others have said, I think this Solaris ploy is aimed at IBM RS/6000 boxes, not Macs.

Re:for Solaris to truely work on apple's product l

[adam872]

They already have. Broadcom NIC's already exist in Sun servers (I have one, a v240).

I think the poster completely misses the point..

[jamesdood]

I don't think Sun gives two shakes about Apple and MacOSX. Let's face it, the SPARC architecture is dead, the most compelling CPU in the RISC market is the IBM POWER5, this chip is not the braindead G5 in the Mac.. (sorry apple fanboys) but is a true enterprise class processor. So it might behoove Sun to port Solaris to run on High-End IBM hardware. If you could get Oracle to run on a big IBM P series box with Solaris you could do some serious data-warehousing. (Without DB2 and AIX!) Maybe oracle is helping fund this effort??

Sun Porting to Different Platforms?

[bigirondawg]

Hmm... sounds like they're getting desperate for business and cash. Probably only a matter of time before they decide that Linux copied some Solaris code and decide to sue...

Re:I think the poster completely misses the point.

[grawk]

Why would you want to run Solaris instead of AIX on ibm hardware? AIX is more reliable, easier to manage, and works woonderfully with oracle. That's what most of the financial world runs on, after all. I don't see what the killer app for solaris on pSeries hardware is. Except that it'd be faster than Solaris on Sun hardware...

Power != Macintosh

[csoto]

Sure, Macintosh is my favorite PowerPC platform (writing this on one), but it's not the ONLY platform. IBM's eServer pSeries is a far better match for Solaris. I love our Xserve G5s (fastest damn servers we've seen in a while), but they already run a great OS. For a shop like us, if we could get the performance, but without AIX (we're Solaris friendly), that would be fine.

Of course Solaris on Opteron is also starting to look better, given Suns' recent attention to that line...

Yeah, but why? -- ANTI -WINTEL!

[Anonymous Coward]

I think Sun's just sick of the WinTel image they got from kissing ass to Microsoft and AMD

Who cares? Java source is available anyways.

[Anonymous Coward]

As a matter of fact, you CAN look at the code for Java. Have you ever? I'm a java developer and I sure haven't. As a matter of fact, I don't give a flying fuck how they implemented java.util.LinkedList. Java's free anyways, and Solaris is too for 1 processor (and I think they have a right to sell it for server use, they built it). Why do you zealots whine so much about 'open source' if you can do everything you'd want with it anyways, and will probably never look at the code.

Re:I think the poster completely misses the point.

[stu_coates]

Maybe oracle is helping fund this effort??

Maybe, just maybe, IBM also has a hand in it...

Consider this...
SCO are arguing that IBM has illegally used some of its IP in AIX. If (and it's a big IF) this is true, and SCO win the court case, IBM is left without a high-end Unix (like) offering. In steps Sun with Solaris and IBM can now continue to sell its high-end boxes with Solaris on them. Sun is happy as it's shifting Solaris licenses, IBM is happy as it doesn't have to maintain AIX, and now has 2 Unix (like) offerings: Linux on the low to mid range machines, and Solaris on the high-end.

Just a thought!

Business model?

[Anonymous Coward]

Uh, so if they open their operating system, and allow it to run on hardware not supplied by Sun... then they are not making money on the hardware, and they are not making money on the software... so what is their business model? Face it, Sun is doomed no matter what they do.

Not that I'm saying you're wrong, but...

[Svartalf]

Do you have anything other than opinion backing up the claim you just made? It's been modded "informative" and unless it's backed up with some sort of proof (it's my understanding that Darwin/MacOS X is a BSD personality layered on top of Mach...) I can't see the comment as anything other than conjecture and thereby not "informative".

Re:Not that I'm saying you're wrong, but...

[leandrod]

> Do you have anything other than opinion backing up the claim you just made?

Yep. Just Google around for the Mac OS X kernel and its creator, plus perhaps Mach, BSD, kernel, microkernel. I could give you the guy's name and the name of the merged entity of BSD plus Mach, but I'm too lazy now.

> it's my understanding that Darwin/MacOS X is a BSD personality layered on top of Mach...

I thought that too. But the 'layering' is actually a commingling.

Re:Not that I'm saying you're wrong, but...

[julesh]

But the 'layering' is actually a commingling.

I've gotta ask... what the hell is the point of that?

Benefits of a microkernel -- good software engineering practice, small enough to understand.

Benefits of a monolithic kernel -- faster performance.

Benefits of mixing the two -- as far as I can see, none of the above. You might get to keep the performance, but then you may as well stick with a monolithic kernel.

Re:Not that I'm saying you're wrong, but...

[leandrod]

> what the hell is the point of that?

Exactly my question!

> You might get to keep the performance

Not even that, the commingling of BSD and Mach is obviously fatter and slower than simple BSD.

Make up your minds, geeks

[Anonymous Coward]

I thought all you uber-geeks were crying for Apple to move the Mac OS to x86 so you can use it cheaper. Now you want to run other OSes on Mac hardware?? Make up your mind!

Confusing

[Anonymous Coward]

To me, companies like Sun and IBM exist to provide services and integration.

In 1996, we considered moving the biz from a homegrown system (oldSCO and windows) to a fully integrated solution, from IBM or Sun. To us, it was what one does to get out from under the expense of running a homegrown system.

We were stunned. The salespeople wanted to do end-to-end replacement, at about 10X the cost of our internal solution.

Sp when people tell me that Sun and IBM compete to sell people hardware and/or software, I'm always like "eh?" these companies make their money selling complete solutions, through their VAR channels, etc.

So why this odd strategy? IBMs strategy makes sense--leverage the free platform of Linux to underpin your solution portfolio. This seems like an obvious and worthwhile strategy, except that, in all honesty, IBM does not appear (in the foreground) to be giving a lot back, especially on the usability/desktop front.

Sun's strategy--offer Solaris as an an alternative to AIX on IBMs hardware. Eh? WTF? IBM is already replacing AIX with Linux on their hardware...why Solaris? Why would anyone buy hardware from IBM to run Sun software?

Did Darl McBride come up with this strategy? "When SCO wins, IBM won't be able to run AIX or Linux on their machines--you'll own the world, along with us! Bwahahahaha!"

This is a strange strategy. To me, Sun would be better off to leverage their existing chip expertise to use AMD64 in superior ways to the commodity chipsets...things like bandwidth to perihperals, etc...places where enterprise use runs into bottlenecks compared to desktops.

Sun would also drop Solaris and use their software expertise to radically improve the usability of Linux, along the lines of OS-X. They would create application sets like garageband as an add-in.

Sun should buy multimedia software companies and bundle their products as add-ins. They would sign agreements with NVIDIA, etc.

The end result: A shrinkwrap linux that is as usable as OS-X and ass Staroffice (capable of running with Exchange) capability to to Single-sign-on with the DOD Common Access Card out of the box, full java development (of course), NVIDIA 3D support without downloads, as well as the aforementioned media applications. In other words, willing to distribute binaries under a license, which RedHat is not willing to do.

Sun is missing the boat. I'm guessing that a few years of work on their part would grant them a viable, sustainable market share of the desktop.

Hell, they could start a downloadable music service.

Sun is in worse trouble than I thought.

[Anonymous Coward]

The more this guy talks, the more totally clueless he seems.

As the CTO, he's supposed to be setting the vision for Sun? Hoo boy - I'm glad I dumped my Sun stock long ago. My personal impression is that they really don't seem to have a prayer of a chance.

It's sadly amusing - Sun reminds me a lot of DEC in the last years of decline. They could change I suppose; but I wouldn't bet on it. The only question now is when some PC maker is going to buy them up.

Which serves them right, after Sun decided to support SCO against Linux

Not likely

[jbellis]

read some of Schwartz's blog entries. Sun hates IBM. Also, Sun already offers servers running Solaris on AMD processors; they wouldn't gain anything by porting to another commodity CPU.

Re:Not likely

[he who meows]

>>they wouldn't gain anything by porting to another commodity CPU.

They wouldn't gain anything at the moment. I think the idea is to try and hook people who already paid for POWER machines into Solaris. That way, when they need to buy another machine, it will be an easier sell for Sun's own hardware. It could probably be argued thats one of the reasons for Solaris X86 as well; Do you think they would give solaris x86 out for free without a vested interest?

Re:Not likely

[Anonymous Coward]

But why on earth would anyone be buying a POWER machine if they did not buy it with either AIX or in Apple's case, OSX?

That's precisely why this move dosen't compute. Sun hates IBM. Sun, as far as I can remember hasn't been involved with Apple, and current Apple machines simply make no sense for a Solaris port.

What is the potential market for Solaris on POWER, unless Sun completely ditches SPARC? The scientific community? In most respects, they'd be better served by clusters, and most X11 programs will run under OSX with little coaxing. POWER Solaris makes no sense for them. Government? Space?

The whole thing makes no sense. In almost every way AIX could handle whatever reason someone wanted to make of a huge multi-cpu single image computer, and undoubtedly just as well or better than Solaris could. The only way porting Solaris to POWER makes sense is if there is some mysterious application that ABSOLUTELY POSITIVELY ***NEEDS*** Solaris, and it's either worth bookoo buckos to make it run on POWER (a scenario that I just don't see), or that our government wants to upgrade it's bazillions of SPARC machines (you know, the giant network under Cheyenne Mountian recording and parsing every phone call) to POWER, but retain Solaris... In other words, some crazy echelon machine...

But even that's pretty far out there. Personally, I see NO market for Solaris on POWER.

Re:Not likely

[Anonymous Coward]

Okay. The gears have been turning and turning, and I have come up with EXACTLY one scenario that makes sense with Solaris on ***PPC*** Not POWER, mind you. PPC.

Portables. Good Solaris (RISC) portables (yes, I also know you can put x86 Solaris on a laptop). Yeah, I know there have been a few in the past, but they were all based on relatively un-modern SPARCs, and were huge. I can see how this might be in some way doable.... I've seen service guys use portable Solaris machines (x86) to service certian types of hardware (a ScotchPrint 2000 printer, which runs x86 Solaris)... A Powerbook G5 would be ideal in that scenario, but as I said, it's a stretch.

But other than that, I've got nothing.

Re:Not likely

[he who meows]

There is no direct market. Like I said, the whole point is the possibility of swaying people who already bought POWER machines to *eventually* buy Sparc machines. Why would people switch? I don't know. Maybe a company wants to avoid vendor lock-in. Maybe some previous CTO of X company decided to buy a bunch of POWER machines a few years ago. Now, the company wants to explore their options, or port their software to solaris without buiying a hundred new workstations, or ultimately jump to Sparc hardware. This allows people using POWER machines to be able to test Solaris without a lot less loss. Also, I really doubt that they are trying to target Apple users. There is a whole world of POWER and PowerPC machines outside of Apple.

Re:Not likely

[he who meows]

Check out: http://www.tadpolecomputer.com/ Tadpole still continues to make 64-bit Sparc laptops (along with X86 laptops), and has for a long time. But you're right about it being a stretch. The number of Solaris users on laptops is probably rather small, and mostly for tech support stuff.

Re:Not likely

[Amiga Trombone]

read some of Schwartz's blog entries. Sun hates IBM.

Which means what? Until recently, they hated Microsoft even worse.

Desperation makes for strange bedfellows.

Re:Not likely

[andreyw]

And even a slick PowerPC laptop costs less...

"Yay"... I think...

Re:Not likely

[squiggleslash]

So what if Sun hates IBM? Apple hated IBM too, remember?

SPARC is coming to an end, it's not keeping up with the other technologies out there. The CPUs, while technically excellent for their time, are failing to scale and the resources simply aren't available in the same way as they are for ix86 and PowerPC. Sun, realistically, has three choices:

• Stick with SPARC, and be left behind
• Ditch proprietary hardware and jump straight into high-end commodity PCs - they're sure to make as much of a success of it as, say, Commodore
• Switch to the ONLY other commercially viable and supported mass-market computer platform, and the only one other than SPARC that's pretty much open - PowerPC

If you were in Sun's shoes, what would you do? You wouldn't stick with SPARC. You certainly wouldn't want to try competing with Dell. Meanwhile PowerPCs are mass-market cheap, they're under active, competitive, development, thanks to IBM, Nintendo, and Apple, and reference platforms themselves aren't that far removed from those Sun produces right now - PCI (and successors), OpenFirmware, etc, etc.

Whether Sun hates IBM or not has nothing to do with anything. Right now Sun has to do the right thing for Sun. If that means out-of-court deals with Microsoft, or buying CPUs from a combination of manufacturers that includes IBM, then so be it.

Ooh, I know this one!

[jpmorgan]

That would be Mach, right?

I See No Ships

[Vagary]

Are you sure they don't recognise they're sinking and this is the best response they can come up with? After all, it's pretty obvious: they're a hardware-OS company whose two key product lines (SPARC and Solaris) cannot compete with the free/commodity competition. What else are they supposed to do but flail around looking for a market?

At the very least, announcing something to do with IBM will keep their stock afloat since the average stockholders probably knows even less about Solaris' market position than the average /.er.

Schwartz is a spin doctor

[Anonymous Coward]

He writes:

"To make matters worse, if you're running your Solaris app on industry standard x86 servers, and you want to move to AIX - well, you can't, because IBM doesn't make its operating system available on even its own x86 servers. Of the Unix suppliers, only Sun makes Solaris available on x86. Even IBM's x86 servers."

This paragraph is a cute little straw man. If it's my app, and I wrote it, then I have the source code and I can recompile it to another architecture. Some code may have to be rewritten, but that's also true even if you target another OS on the same architecture. There is also a hidden "pro" argument in here for the open sourcing of 3rd party applications, even though he wanted to skip that issue.

I've seen this kind of spindoctoring come out of him before. Spindoctoring isn't just what you say, but also what you omit.

IBM Workstations

[lcde]

I thought IBM was going to come out with a G5 based workstation. So..

I figured Sun would start using these instead of the Sparc systems. Both are 64bit and this is a way that they can move on from a (perhaps?) soon to be dead arch. the x86 solaris/linux machines are testing the waters to see how stuck on Sparc people are.

Just my 2 cents.

snake oil and rhetoric

[Anonymous Coward]

what a crock...

Sun has a monopoly on Java. Sun is the only place you can license Java and the TCK.

So, to quote Jonathon ...

"If you love a product, but the vendor providing it triples its prices, how easily can you move from that product to a competitive product? "

and just to make it more obvious ...

"If you love Java, but Sun triples its license prices, how easily can you move from that product to .NET ? "

Since all licenses come from Sun, they control the final price and they can screw every licensee and every customer. By his own definition Java is not open as it is controlled by Sun.

As they get more desparate for money, what do you think they will do ?

Re:I think the poster completely misses the point.

[Derkec]

Excellent point. This has a whole lot to do with having an arguably superior OS to compete on IBM hardware and not as much to do with running on former OSX machines.

That said, Sun and Apple have a pretty cozy relationship on the Java side of things, so perhaps Sun is trying to highlight the Apple hardware as enterprise ready?

Ha-ha that is so funny!

[Donny Smith]

Obviously your thinking is one dimensional.

It will get real bad for IBM.
It's existing IBM customers moving from AIX and WebSphere to Sun's stuff at a fraction of IBM's maintenance fee.

You're laughing now?
(www.whysanity.net/monos/ggr2.html)

Re:Ha-ha that is so funny!

[IPFreely]

One dimension has two directions. And you don't seem to know much about corporate decision making.

Nothing's going to happen to existing IBM customers. It's more like Sun customers who already have and use sun hardware and software now buying IBM hardware to run their Sun software on. No loss for IBM there.

Customers don't but software based on their hardware choice, they buy hardware based on their software choice. If the software decision has been made ("we're using Sun software"), then it is only a matter of what hardware to buy to run it on, IBM or Sun.

If you think software cost is a consideration, then that cost decision would have been made long before the choice of hardware was even considered. If customers were going to move from IBM software to Sun software, there's nothing about hardware that will hold them back, no matter what Sun software runs on. They'll junk their IBMs and go buy new Sparcs in a minute.

Sun is moving away from hardware sales.

[crovira]

By spreading their 'industrial strength' OS to every platform and trading on their reputation, they are hoping to survive the shift.

Actually its a smart move.

Hardware has been commiditized into oblivion...

Re:I think the poster completely misses the point.

[jamesdood]

Having run both Solaris and AIX, (On E10k's P690's) I can say straight up that Solaris is one hell of a lot easier to work with than AIX. Especially if you have to do any porting of open source software. And Oracle runs MUCH better on Solaris than AIX, plus most Oracle DBA's are more comfortable using solaris than AIX. Perhaps this would serve as a stop-gap until Sun comes up with some serious hardware.

Were you drunk?

[Donny Smith]

>a) compile a custom kernel with far fewer services available

WTF?
# chkconfig --del

>b) Change program locations and links so that random calls by path wouldn't work (i.e. something like ls would be /sdf/sajfs/ysfs/sj while cat would be /uwsius/usiufs/sc etc...

What?????

>c) remove dozens of commands entirely

Haha! Why'd you do that?
# chmod 400 /path/to/command

Re:Were you drunk?

[jbolden]

You miss the point.

1) You wouldn't have chmod or chkconfig
2) Even if you did you wouldn't know what they were called or where they were.

The result is you couldn't find your way around.

Re:Were you drunk?

[asdfghjklqwertyuiop]

a) compile a custom kernel with far fewer services available

WTF?
# chkconfig --del

WTF exactly. chkconfig doesn't have the slightest thing to do with a custom kernel.

Re:Were you drunk?

[Donny Smith]

>WTF exactly. chkconfig doesn't have the slightest thing to do with a custom kernel.

It has in the sense why would you compile a kernel with "fewer services" (this is a WTF in itself - services have nothing to do with the kernel but I wasn't the one who said that) when you can simply disable services you don't want to run?

Re:Were you drunk?

[Donny Smith]

(I didn't mean to execute chmod as hacker, but as sysadmin - instead of doing things they do it'd be sufficient to chmod files you don't want executable and disable services you don't want to run. Not to mention iptables, Bastille, intrusion detection and other tools that can be used to prevent unauthorized access).

Would such screwed up OS work at all?
Its own programs wouldn't find their way around and the sysadmin would screw up the system sooner or later.

On the second question - what if one tries manually executing several files (auto-complete with the TAB key is always helpful) until they've found five-six commands that suffice for a script?

1. find all files
2. check each file's attributes
2a. If the file is an executable, run 'man executable$i > executable$i.txt' to get man pages or just ./executable$i > executable$i.txt to collect error message strings.
3. Do a loop on the text output - parse headers of collected man pages to find the actual name of each command (the first word in the first line until the left parenthesis - for example from fdisk man page:
# awk -F"(" '{ print $1 }' fdisk.txt | head -1
FDISK
Now this needs to be converted to lowercase and one can automatically create Bash aliases for those commands (or for the important few, such as fdisk, for example).

A real hacker would do this with one-line script... Once someone's gained root access, none of these little tricks can stop him.

Re:I think the poster completely misses the point.

[jamesdood]

That could be a reason to be considered.. I think that it wouldn't make too much sense for IBM to do this, unless it would help them push more tin out the door (I hear margins on P series machines are quite nice) so nothing is beyond the realm of possibility.

Sun is wising up. This is a good thing.

[amper]

It's funny, I was just discussing this with a friend the other day. I really think this is the smartest possible move for Sun. It has been becoming increasingly obvious that Sun is seriously lagging behind in processor development. A move to Power and/or PPC would enable Sun to stop sinking money into the pit that is SPARC.

Although it has been pointed out several times here that POWER!=PPC (or Apple), I think Sun would be well served to make certain that any port they do runs on at least the Power Macintosh G5 platform (and any later Apple hardware). This would give Sun access to the many, many existing Apple workstations out there so as to provide Solaris with exposure to the Mac community.

Let's face it, although Mac OS X is a great OS, Apple doesn't really seem to be doing much to chase after the enterprise market, even though they now have what could be an enterprise-class OS (with some better documentation, anyway). The XServe is a fine machine, but it's hardly what I would consider "enterprise", with the possible exception of high-density clustering apps.

Solaris is a very good OS with a huge amount of support in the community, and good installed base at the higher levels. If Sun could get Solaris running on Macs and IBM RS/6K (or whatever they're calling them these days???), it could open up many more doors for them, while still enabling them to possibly design their own brand workstations and desktops on the POWER/PPC platform to compete with both IBM and Apple. That could also mean Mac OS X support on a Sun box.

I can't help thinking that this may be a precursor to shopping Sun out to one of the aforementioned competitors. Apple could use Sun, and vice-versa. An IBM+Sun pairing would probably mean the death of Sun.

Re:How come...

[Anonymous Coward]

Vote for Bush! He will protect us!

Re:Open is open, but to who?

[Jahf]

Used to happen with IBM mainframes, then Windows, and now, regrettably, with Unix variants.

Oh come now ... vendor lock-in was happening with Unix LONG before Windows was a viable competitor.

Is it a purchase ploy?

[gamgee5273]

The first thing I thought about Solaris on Power is: "Gee. How hard would it be to integrate Solaris apps into OS X Server or AIX?"

Not that difficult at all, I assume. I'm not a coder, so I'm not all that certain. But, if Solaris is running on the Power architecture, the PPC is only a few steps away.

Could Sun be doing this to make itself an attractive acquisition for IBM or Apple (if Apple is truly serious about expanding its place in the enterprise?)?

Re:Open is open, but to who?

[Anonymous Coward]

Remember "vendor lock-in"? Used to happen with IBM mainframes, then Windows, and now, regrettably, with Unix variants.


Now?? Maybe you're young but vendor lock in tactics with UNIX vendors started in the 80s with the whole SysV-isms and the graphics system wars. This is long before Windows was even an issue. In fact, the whole senseless bickering amongst UNIX vendors is one of the reasons that Windows NT adoption shot up in the mid 90s. So many UNIX variants were adding features simply to better the competition, many were poorly designed or not well thought out. The Windows API looked like a pillar of stability by comparison, vendor lock-in or not.

From the perspective of many decision makers the stability of one large company supporting a nonstandard looked better than a bunch of companies in stupid consortiums (Open Group, X) backstabbing each other for customers.

Yeah, vendor lock-in bites, but it is something the UNIX community has been dealing with for a long time before Windows was around.

Who would run Solaris on pSeries?

[swamp boy]

I believe that Sun is really talking about IBM's pSeries (formerly known as RS/6000). My question is who would run Solaris on a pSeries server? By the same token, I can't imagine anyone running AIX on a Sun server either if it were available. I really doubt that Sun has *any* interest in Apple PPC hardware.

Re:I think the poster completely misses the point.

[grawk]

I can't imagine how you would think solaris is easier to work with than AIX, unless you didn't put any effort at all into learning how to use the IBM. As to porting open source code, with the linux affinity libraries, my experience is that most open source code just works. Just autoconfig, then bam.

Why use Java?

[Sirwar]

On the topic of Java...I want to know why anyone would make a program for Java, as opposed to C and such?
I honestly don't know much about developing, only experience using the software, and this is what I've noticed:
Java is slow, and often times is used through a (bleh) IE window.
Java in unstable. I've never worked someplace or used a Java program that didn't just memory leak itself to death, or just plain freeze up all the time.
Java needs much more RAM and CPU to run.
Given these points, why would it be considered more cost effective, or overall a better descision to make a Java application? I understand the portability, and thats great for script, but isn't it ultimately a bad choice for a professional, mission critical application?

Re:Why use Java?

[Frobean]

As someone whose job includes writing a lot of code, much of it in java, I can answer that.
First of all, java browser applets generally are more trouble than they are worth (IMHO). Java applications, on the other hand, can be quite nice. Second of all, as with many languages, there are a lot of bad java programmers out there writing a lot of bad java code so don't automatically blame the language for the ineptitude of the programmer. :-)
Java is not inherently unstable. I have several apps that run under heavy loads for extended periods on everything from PCs to some largish Sun iron without a single issue.
Speed has not been an issue for a while now. I use the 1.4 JVMs and the speed of those VMs has *never* been an issue.
I prefer java as my dev platform for a couple of reasons. Most of my code is intended to eventually run on a Sun server. Java allows me to develop and do the preleminary testing on my windows workstation which is a very efficient way to things for me. Also, I can usually bang out a program in java in less than half the time it would take me to code the same app in C or C++. Another personal preference for me is the garbage collection in java. I have to try pretty damn hard to write java code that leaks memory (it *can* be done tho).

Re:Open is open, but to who?

[davecb]

Hmmn: My recollection is that people were shifting off IBM series 3/3X minis first, then mainframes, after seeing the possibility (and relative simplicity) of switching amoung Unixes.

After the BSD/Bell split, there was both differentiation (which could be used for lockin) and standardization (for making the marketplace look big, to attract ports). VMS was the big incompatable vendor in those days. The Unix folks were't big and powerfull enough to lock down much.

Then Windows 3.0 came along and started a new era of lock-in.

--dave

Not Macintosh but IBM's Power5 stuff....

[FatherOfONe]

I don't know why everyone is saying Macintosh, no where in the article did he mention Apple. I believe he means running Solaris on IBM's mid range PowerPC systems. The ones that are running AIX now. It is my belief that Sun wants to convince customers that they could standardize on Solaris instead of Linux.

The meat of the article was that he feels that open means no vendor lock in. His point is that if you use Java and don't use any proprietary junk you could move your code with little effort. I agree in principle, but if I write the stuff in Java, then I am locked in to Java. Not that this is bad, but it would make sense for Sun to change the VM (perhaps open source it and go to the standards body and get it approved as a standard) and then get other languages to run on it. In a way it would be somewhat like .Net, but be open.

Re:Open is open, but to who?

[davecb]

In my view, that was more bad management than intentional lock-in. I usually assume stupidity if it suffices, rather than malice (;-)).

This follows from folks actually trying to write standards and work to them. I was on the standards-and-stability team for a major vendor for some years, trying to live the motto of "write once, run forever".

Alas, vendors don't care much these days, so I tend to think that strong language standardization (ie, Java and "write once, debug everywhere") is the best current push in that space...

--dave

not likely to boot

[sentientbrendan]

Others have pointed out that the POWER architecture is very similar to the powerpc architecture in terms of instructions sets. Be that as it may, I doubt that an operating system designed for POWER systems will actually run on apple's hardware any more than osx will run on an IBM POWER system which, to be clear, is to say that is will *not* run. The processor after all, is not the only thing that differentiates different platforms.

That said, maybe some open source group will write some kind of workaround similar to maconlinux.

Also, POWER (at least in the sense that the article uses it) does *not* mean G5. The G5 is not just a POWER4 popped out of an IBM workstation and popped in an apple tower.

it doesn't matter whether you call it "open"

[dekeji]

What matters is what the license actually says. And the Java license (probably the Solaris license, too) is too restrictive: Sun gets legal control over the evolution of the platform, they get legal control over who implements it, they get legal control over many third-party contributions, and they even get legal control over the future work of people who look at their "open" source code. That's just a bad deal for both users and contributors.

If we accept Schwartz's argument that Sun's licenses are "open", then we just have to make an additional distinction: "good-for-users open licenses" and "bad-for-users open licenses"; many of Sun's licenses then fall in the "bad-for-users open license" category.

And why would you...?

[Nom du Keyboard]

Does this mean you'll soon be able to ditch OS X and stick on Solaris 10 onto Macs?

And why would you do that? Can you run Safari under OS-X? Photoshop? Illustrator? iTunes? Trade discs with your other Mac friends?

I really think people buy Macs because they want to run the Mac operating system along with Mac applications.

Re:pfffffft

[JonAnderson]

Wow, you have seen strange pussies. Are you sure you haven't been fucking yaks?

Now we get to see how Sun fucks up POWER FB's

[edgedmurasame]

Ever since Sun played the EOL game with their framebuffers where they would leave in an old, obsolete cg6 when they undo support for perfectly good framebuffers such as the ZX and the double decker version of the Elite3d M6 in only 2 versions (removed in solaris 7 and 10 respectively). It's bad enough that Sun does it to their own. McNealy: Keep your hands off POWER. You do not need to ruin another architecture.

Re:for Solaris to truely work on apple's product l

[larien]

Isn't that the NIC that caused a product recall?

A customer's definition is not always right

[Anthony]

Only a customer can define the word "open." That's my view.

I was with Amdahl many years ago and we gave a presentation on UTS and "Open Systems" and one customer stated that they already had open systems because of Amdahl. They weren't locked into one vendor for hardware and that was enough leverage for him. [Amdahl had 80% market share among Govt departments in our city at the time]. He did miss the point that we were talking about operating systems and portable application APIs. Of course this utopic vision was clouded by the "Unix Wars" at the time where USL and OSF were about to expend inordinate amounts of energy fighting each other, giving another OS vendor an opportunity to gain a foothold in markets where it didn't really fit. But that is another sad story.

Who cares?

[Anonymous Coward]

Who the f*** cares what Sun does?

They're dead.

Get it? Dead.

Re:Who cares?

[Anonymous Coward]

> They're dead.

I'm not dead yet...I'm feeling better...I think I'll go for a walk now....

(you get the picture)

Re:Open is open, but to who?

[Jahf]

Doesn't seem like a fair comparison to me for 2 reasons:

1) The "Unix Wars" of the late 80's / early 90's were primarily due to various vendors trying to lock people in. See this article as one example. In fact it was this very action that many analysts say opened the door for Microsoft to succeed in moving into the market with Windows NT.

2) [hinted at above] Microsoft was not competing in the same market as Unix or mainframes with Windows 3.0 ... Windows 3.x was almost purely a desktop with some small chunk of the workgroup server space. This didn't change until Windows NT (yes, NT shared the 3.1 version but was extremely different). And Windows NT didn't start eating into the SMB server market significantly until NT 4.0, which was LONG after the Unix Wars has ended.

Re:Open is open, but to who?

[LinuxHam]

Such as Sequent, who sorta doesn't exist any more...

Ouch, I'll pass that on to all my co-workers with /Beaverton/IBM in their Notes email addresses! We IBMers are very grateful to Sequent for their NUMA contributions. It really took us to the next level in flexibility and scalability, and helped secure our lead in server shipments well into the future. Paired up with our newest developments in Power5, and our "run Linux on everything" philosophy, I'd say we have a pretty good server family.

Bite me

[Anonymous Coward]

faketoo: uname -a
FreeBSD faketoo 9.8.7-RELEASE-p69 FreeBSD 9.8.7-RELEASE-p69 #3: Tue May 25 22:49:47 PDT 2024

faketoo: uptime
3:09PM up 9462884935 days, 23:22, 5 users, load averages: 1.00, 1.00, 1.0

Re:Bite me

[sp0rk173]

I didn't change the release number, c'mon it's got to be somewhat believable.

From a former ISSC employee, Sun kick IBM's ass !

[Anonymous Coward]

From a former ISSC employee, I say to Sun go Kick IBM's ass !

Agreed -- check the pricing on AIX vs Linux

[msobkow]

I think you'll find that IBM has priced the PowerPC hardware running AIX or ready for Linux (SuSE or RedHat enterprise licensing) such that you pay roughly the same price either way. Certainly no business is going to quibble over the nickel and dime OS pricing difference when they're facing far greater costs on licensing all the products they still need (e.g. RDBMS.)

The hardware sale, the maintenance, the support -- those are still all IBM's. All they lose in the deal is the OS licensing fees, which hasn't been the major revenue generator for a while. Don't forget how cut-throat the competition between Sun, IBM, and HP has been for so many years -- they learned not to lean on an operating system as a revenue generator a long, long time ago.

Think of the vendor OS not as a particular package, but as your assurance that the vendor has taken responsibility for testing and maintaining the OS. Not a purchased product, but an on-going service to keep your systems protected from the various bugs and oversights that always hide in the code. It's that "chain of responsiblity" that OSS zealots forget about when they bleat that their favourite tool is just as good as the commercial product.

Business doesn't pay just for the functionality -- they pay for the assurance and to identify a responsible party if there are any legal/insurance issues to be dealt with. Without that big backing vendor, big business will not deploy OSS. It's too risky from a business perspective, regardless of the technology.

Re:I think the poster completely misses the point.

[sedawkgrep]

Why is it that everybody who thinks they have the insight and right to malign AIX is a developer?

Jeesh dudes - it's by far the best UNIX on the planet for a systems admin. Systems admins/engineers and managers are the ones who make the decisions on whether or not to buy this stuff - perhaps it is the best choice for what they want to do?

now for some real data...

[sp0rk173]

check this out. FreeBSD, up over 1700 days. Kinda makes your 854 days look puny =\. But seriously, no system should be up that long without reboots. Security flaws come and go, and kernels need to be patched and recompiled. Any real sysadmin knows that effective uptime (that is, is the system up when it needs to be up?) is far more important than actual uptime. Long stretches of uptime just means you have an old, vulnerable box sitting there, waiting to be cracked, wormed, or expoited. Plus...if your linux box has been up for over 800 days you're missing out on a LOT of neat new technology. Er...unless you're a debian user.

Re:now for some real data...

[Anonymous Coward]

You fucking dead OS using assclown!!! My box has been up for 800 days solid and it works fine as a file server/dhcp server/web server/dns server for internal purposes. It's not internet exposed, so there are no "neat new techologies" that I am missing out on unless there's been some HUGE paradigm shift in DNS, DHCP, Samba or HTTPD. You are an EXTREME motherfucking idiot!!!! Give it up, you lost. I posted real data and you posted your idiotic drivel. Typical posturing from a BSD luser. Face it, your OS is dead. There is little or no development taking place to bring the interesting stuff to your platform. BSD FAILS IT. You FAIL IT! Give it up motherfucker.

Re:now for some real data...

[sp0rk173]

Awww you got mad. That's so cute!

Re:now for some real data...

[Anonymous Coward]

Naw. I didn't get mad. I just gave you the beating you deserved. All you could do was repond with the typical loser stance of trying to belittle me by pretending that you weren't at all affected by what I wrote. In truth there is a part of you that is cowering in fear because you know that Linux is beating BSD in every way shape and form. So again... fuck off loser. You aren't worth the shit to smear you with.

Re:now for some real data...

[sp0rk173]

hah. The funny thing is, i'm not a BSD zealot. Infact, i'm using linux right now. It's my main desktop.

Real blog

[Anonymous Coward]

Is that his real blog, or does it go through 10 - 15 PR reps before the actual text gets posted.

one platform for unix

[pelorus]

IT'd ne nice to see more on Power variants. Just simply from the point of view of choice.

Vendor-lock-in is a curse. Cutting things down so that all big iron UNIX ran on Power variants would be a nice separation from the current model where the UNIX variants all run on their own processors and Linux gains ground because of its ability to run on commodity x86 hardware.

And yeah, I'm going to get flamed for this, I don't believe Linux wins on merit - other UNIX variants do it better in almost every category - Linux is cheap, it's available and it's open. It's the same marketing ploy that put Windows everywhere. And I damn well hope it succeeds A world with 30% of the market being UNIX and 70% of the market being Linux would be a nice place to live.

Re:I think the poster completely misses the point.

[stu_coates]

It really wouldn't cost IBM much to make this happen. They'd have to supply Sun with all the low-level hardware details of the p-series tin and maybe a couple of people to answer questions and Sun do all the rest. It'd make sense as a backup plan should the SCO situation turn nasty.... remember, Solaris is "SCO-friendly" since sun paid them a bunch of cash earlier in the year.

Sun wants to beat IBM at IBM's Game

[Anonymous Coward]

IBM has worked hard at getting Linux at all levels of IBM hardware. An IBM salesman can sell you a Linux, DB2, websphere, etc. package and sell you an Intel x86 box, a POWER box, even a Mainframe.

Sun could argurably do the same with Solaris. There are several advantages of running Solaris over Linux. Solaris is the gold standard by which my large faceless corporation measures all other OS. We like Solaris because Sun owns the codebase and can provide us support. Solaris scales like crazy.