Submitted by “There are dozens of articles like this one on the net. Over and over people suggested solutions like this for different reasons and although I know that such thing probably won’t happen any time soon, from my point of view now it is the best moment ever in the history of both operating systems to merge in a one powerful alliance. And the hell has already frozen over, hasn’t it? First I will give short description of both OSes, so we can see the strong and the weak sides of them and see if the combination should eliminate the shortcomings and make the good points even better.” Update: Sun is giving out free Solaris 10 DVD sets.
quote: “Is it just a dream?”
yes it is!
the leadership of both apple and sun are egomaniacs believing that their product is the best and second to none. For this reason alone this will never happen.
Let’s keep the number of sensational but useless comments down, OK?
the leadership of both apple and sun are egomaniacs believing that their product is the best and second to none. For this reason alone this will never happen.
This is true. They’re both literally calling their OS “the world’s most advanced OS”:
http://www.google.com/search?hl=en&lr=&q=most+advanced+os&btnG=Sear…
Apple only need cherry-pick the bits they like, which they’re already doing; ZFS, DTrace.
Apple wants everything it does to be kept as locked down as possible (as can also be seen in the recent iPhone news). It wants to control everything and fights against consumer freedom with teeth and claws.
Sun, on the other hand, has become quite OK with giving its users some freedom.
This discrepancy alone would make Sun and Apple unlikely bedfellows.
bah MSFT is doing the same thing with their multimedia everywhere campaign – Vista Media Center + XBox360 + Zune player.
I still haven’t figured out what Solaris is for other than overkill for servers. But part of that is because I’ve never been able to get it installed right due to its failure to use USB keyboards and mice.
OpenStep ran atop of Solaris (as well as MS Windows NT) and OpenStep essentially was the equivalent of the Mac OS X GUI and development parts (Aqua, Quartz, ObjectiveC etc.) and was originally (to some extend) co-developed with Sun Microsystems.
Sun jumped the fleet and instead started focusing on its Java technology for platform independence instead of NeXT Computer’s Yellow Box APIs and Fat Binaries for multiple platform support.
Personally I’d love to see it happen, Mac OS X userland on Solaris, but I highly doubt it’ll be that way…
The article mentions DTrace and ZFS as reasons, but since they are already going to be part of Leopard (DTrace is in the new Xcode tools, support for ZFS is already partially present in the current Leopard builds), they actually support Apple’s current approach: porting interesting/useful technologies.
As for one of the other reasons, “new buzz OS for geeks” – yeah, it is a dream, just not what most people (read:customers) dream about
While I would dearly love to see the Mac OSX gui grafted on top of Solaris, the article does not present any actual compelling reason why this should happen.
The article asserts that Solaris’ failing is its userland utilities. That is a debatable point. A large number of Solaris commands exist in order to ensure that the OS complies with various standards. If you don’t want that standards certification, don’t complain about the OS…. find another that suits your needs better, or start to make use of SunFreeware or Blastwave.
And Sparc isn’t going away in a hurry, so asserting that it should be deprecated is just plain stupid. I don’t think the article author fully understands just how important the Niagara (and Rock) architectures are.
I’m not saying that SPARC will go away, but neither Rock, nor Niagara are workstation processors, they are both aimed at server. So my suggestion is to port only server relevant parts of MacOSX to SPARC.
merging would be silly. thoguh i htink if solaris wanted to make things prettier ni user land they should either really help contribute to the gnome projust or buy out trolltech and go with a nice QT solution.
Don’t like this article doesn’t make sense. Why would Apple with want to buy into the Solaris Kernel. Yes, they may spend a couple of hundred million on mac os x every year. But apple like’s to control the direction it’s moving in, and not jumping on someone else’s band wagon.
Didn’t like the way the author described sun’s workstations as archaic. Sun just needs a better exterior hardware designer to make their workstations look funky! Give me a reason as a geek to buy a sun workstation over a macpro!
> Sun just needs a better exterior hardware designer to
> make their workstations look funky!
And why would they want to do a thing like that? Remember, Sun doesn’t market to the consumer market. They market to the serious business market.
Imagine that you’re the head of the IT department and two of your employees are making recommendations on hardware platforms that will run the software that you will host 1 billion dollars worth of transactions per day.
One of your engineers presents you with a boring gray box with a boringly safe uptime and a boring operating system that doesn’t do anything unexpected and doesn’t follow the fads and doesn’t upgrade for years at a time (Solaris 2.8 support for new software and updates is still going strong even though it is 7 years old).
The next engineer presents you with *funky* hardware with fad-like designs and fad-like GUI features that have little to do with financial transactions and won’t be seen by any engineer very often because it will be in the back room somewhere and fad-like continuous upgrade cycles that give you new features that have nothing to do with financial transactions.
Which system would you chose? Apple and Sun are in two different markets. Merging both companies would be a disaster.
Edited 2007-01-15 03:40
“Which system would you chose?”
Sadly a lot of IT managers would chose the “funky” hardware.
i totally agree.
we sell systems that run on both osx ans solaris and solaris is my preferred system to sell and maintain.
I’m not necessarily disagreeing with you here, but in terms of funky cool looking professional workstations, one can’t forget nifty pieces like this:
http://personal.cfw.com/~tkprit/images/sgi.jpg
and this:
http://www.mteege.de/pic/tezro_3.jpg
and other assorted machines from said company in the past.
This is ridiculous.
I swear a very similar article was posted to OSAlert about two or three years ago. The article argued that Sun should buy Apple. Neither article is very convincing. Apple’s target market and SUN’s target market do not overlap very much at all. Both are geared towards their respective markets and both are excellent products within their markets (technologically at least). Sure both products may benefit from some cross contamination but I don’t see a huge benefit from completely uniting Solaris and OSX especially considering the amount of time and money it would take. Most desktop users would hardly notice a different kernel and most server admins couldn’t care less about a pretty desktop. Besides GNOME is good enough at this point for SUN that it would be a a waste of resources to attempt to graft Aqua on top of the Solaris kernel. Apple may benefit more from a combination of technologies but not enough to significantly increase their marketshare. This is all just a geek’s dream.
What i really want is a rock solid cleanly developed OS kernel and a very nice OSX style GUI.
I wish SUN develops a OSX level usable GUI on top of solaris.
Wow that would be like a dream come true.
Help KDE|Gnome|Enligthenment out and run that on your rock-hard/flexible OS of choice.
Sun = somewhat open source
Apple = more and more blackbox
It could and maybe it should happen to Mac OS X Server version.
There’s no visible performance penalty on the desktop version coming from the fact of using Mach3, at least considering the majority of the applications.
On the other hand, Mac OS X Server is seriously compromised by its Mach3 kernel as benchmarks show. For the majority of the usual server tasks it is heavily surpassed by *BSD and Linux.
But then, Sun would gain nothing with such happening.
The level of performance really depends on the load. On moderate workstation loads, its very easy to get an OS X machine to the point where it takes seconds for user input to process. And workstation tasks like compiling are notably slower in Darwin than in competing systems.
Moving to Solaris would help many tasks (I/O bound stuff), but honestly as a desktop kernel, even Solaris is second to Linux (or maybe third to Linux and FreeBSD). Solaris has a server-tuned scheduler, and it shows. In terms of “teh snappy”, using GNOME on Solaris reminds one very strongly of using GNOME on Linux 2.2. The sheer volume of complaints about interactive lag in earlier Linux kernels seems to have motivated so much work on the CPU and I/O schedulers in Linux that modern 2.6.x kernels get close to Windows at low levels of load, and blow Windows away at high loads.
Of course, moving to Linux or Solaris or FreeBSD is probably a day-dream. Apple could get 90% of what they need with a fraction of the effort by improving the various subsystems in XNU. Sure, the result will never compare to Linux or Solaris, but Apple doesn’t need it to.
well, i just had to order that free media kit from sun. yes i know, it is quite sad that that is all i got out of this article.
I ordered as well. I was also quite sad when I took time out of my day to submit the bit about the free DVDs and then saw it grafted onto what is essentially a nonsensical blog post.
I find it quite nice of Sun to start delivering free Solaris DVDs again; the last time they did it was in 2004. These are nice little touches that win people people over, just like Ubuntu’s ShipIt did.
i sure hope this does not happen, apple would ruin solaris. the opensourcing of solaris was a great step, and since apple would never go along with opensourcing their stuff(not that i think its worth it), it would be quite a step back.
I wager all these “Apple should do $X” articles are implicitly saying “I don’t like the direction Apple is taking”.
The next step (no pun intended) could be people realizing there’s other viable products|solutions out there and contributing to them (via usage, promotion, testing, development).
Solaris will be re-released under GPL v3. Desktop implementations of the operating system will go mainstream and we will all be using a hacked up Solaris with GPL v3 applications in 10-15 years.
Either that or the GNU/HURD project will get finished.
Hmm, I can see this scenario as well. But always in the move the future is. *IF* this would happen, what would happen to the Linux kernel project? Would Torvalds and friends eventually switch to GPL3 as well? Surely the technical merits of both kernels are excellent. Interesting times.
GNU/Solaris + Debian with ZFS would be a very nice distro indeed…
while debian gnu/solaris would not be my choice, gnu userland on solaris is definetly something i could see myself running.
i have tried solaris(granted not much), and stuff i really cant get used to is their userspace.
so nexenta(gnu/solaris) is actually quite interresting for me.
http://www.gnusolaris.org/
“Either that or the GNU/HURD project will get finished.”
Oh man, that made my day, too bad OSAlert doesn’t allow modding up for humour
I won’t comment about MacOS since I haven’t used it seriously since 1998. But some of the comments by kloty about Solaris are way off the mark.
“Outdated” CDE (Or Common Desktop Environment) is not only shipped with Solaris, it is also shipped with AIX and HP-UX. This is so US Goverment applications written to support the Common Operating Environment (COE) can work, and is required for an OS to be certified for Government use. RedHat had to include OpenMotif in RHEL in order to sell to the US Government, so while CDE might be outdated, it fits a very specific requirement from a major customer. Additionally for those of us who are “old school” or prefer something faster, CDE is great when all you need is a few term windows.
Sun’s decisions as far as the Java Desktop System is to provide a quality, stable corporate desktop, not a “geek” desktop. You don’t do that by using every piece of bleeding edge software. And for that matter I am sure that IBM doesn’t ship the latest and greatest versions of Gnome and KDE (IBM is just as “backward” as Sun is, if not more so about stability and compatibility). I see this as and old and tired argument that only means a lot to Linux users who are accostomed to new features every week, I prefer the stability. The version of JDS that ships with Solaris 10 11/06 works just fine for me.
And what difference does it make what processor Sun uses in a workstation? Obviously kloty hasn’t checked Sun Store recently, because I found the following there:
http://www.sun.com/desktop/workstation/ultra25/
http://www.sun.com/desktop/workstation/ultra45/
I think Sun’s commitment to SPARC workstations is pretty clear.
What software has not been ported to x86, are we talking about commercial applications or F/OSS software? Examples would be nice.
The “Linux supports more hardware” argument is also very old and tired. I think Sun’s efforts have gone a long way to draw more people into using Solaris x86, now it is up to hardware vendors to start supporting Solaris x86 (the problem is simply not limited to Sun). Sun never positioned Solaris x86 to be a desktop OS, so I would not expect it to work with a variety of hardware like Linux does. I have used Solaris x86 since 1999 (Solaris 7) and it works extermely well as long as you use hardware specified on the HCL (just like any other OS).
As other posters have said, there is nothing in this for Sun, only Apple would benefit.
What software has not been ported to x86, are we talking about commercial applications or F/OSS software? Examples would be nice.
Easy; Adobe Framemaker and Adobe Acrobat Reader – both are available on Solaris SPARC but not x86 – closer to home, still waiting on their Sun Ray software to come out of beta and on Solaris 10 x86.
Regarding F/OSS – how about making Solaris work so that everything that I try to compile doesn’t break because of some stupid badly setup; how about setting up the paths CORRECTLY when Solaris is installed rather than expecting the user to do that – that is time wasted spent fixing something that Sun should be setting up.
Edited 2007-01-15 05:14
Shouldn’t you address the lack of a Solaris x86 version of Adobe software with Adobe? Sun has no control over what Adobe does or does not port to a particular architecture. Sun might be able to influence a vendor, but I think users have a lot more power than a vendor does.
What are you compiling that breaks because of the “setup”? The PATH used by Solaris is minimal and can be easily extended by the user, it’s been that way for years, and not just for Solaris (AIX and HP-UX are pretty much the same way). My experience with RedHat Enterprise Linux shows that their default PATH isn’t worth a plugged nickel either, why do I need /usr/kerberos/bin in my PATH? I always update the PATH manually when I use RHEL, so I don’t really see much of an improvement there either.
If it is a linking issue I use crle to update the linking environment so that libraries the application might need are found.
I don’t see setting a PATH and a linking environment that works for me as a waste of time. And I do not expect Sun to “do it for me”, I am capable of doing it myself and creating a .profile that works for me gives me the flexibility I want. Isn’t that the reason why people like UNIX, is because you can tailor the environment to meet your needs as opposed to some deafult setting?
Shouldn’t you address the lack of a Solaris x86 version of Adobe software with Adobe? Sun has no control over what Adobe does or does not port to a particular architecture. Sun might be able to influence a vendor, but I think users have a lot more power than a vendor does.
Bullcrap; Sun has money, it is a matter of setting up a metting with Adobe, and paying Adobe to port their products to Solaris x86 – Adobe won’t care, as long as it is not going to cost them a bean, thats their bottom line.
The problem with Sun, its filled to the rafters with cheapskates and penny pinchers who are unwilling to part with a bit of cash to improve their product, they’d rather it all occur ‘naturally’ – so whilst Microsoft and Linux are taking them to the cleaners, Sun is expecting a little ‘free love’ to solve their problems.
Oh, and regards to paths, how about the difference locations Sun place their opensource components, off my head I can think of atleast 3-4 different directories sprawled around the drive; and worse still, when you compile things, the said source code can’t find its dependences.
Sorry, the day I can grab a *NIX source code off the net, and go ./configure && make && make install without the whole thing going tits up, then I’ll know Solaris has made progress, until then, all I see over there is a perpetual buck passing and blaming for things that should have been corrected over 10 years ago.
And why should Sun pay to have Adobe port their products to Solaris x86. If the demand is there Adobe will do it without any expenditure from Sun. And as far as being cheap goes, how is this different from Adobe porting Acrobat Reader to Linux, did RedHat or Novell give Adobe money to port it, or did a bunch of Linux users get together and complained to Adobe until they ported Acrobat to Linux. My money is on the latter, because I don’t see either RedHat or Novell giving Adobe a dime.
So because you don’t know how Sun lays out their filesystems and default locations for software, there is a problem with Solaris. Here is how it works, for the bundled F/OSS stuff like gcc the location is /usr/sfw. For optional software (like Veritas NetBackup)it is /opt (including software downloaded from Blastwave.org (/opt/csw)). For software downloaded from SunFreeware or compiled with defaults, it is /usr/local. Now why is it I can compile OpenSSL, apache, Nagios, OpenLDAP, Samba and MIT Kerberos without any problems, and you can’t. Because I took the time to set up an environment that works using ksh, .profile and crle to add /usr/sfw/lib /usr/local/openssl/lib and /usr/local/lib to the linking environment.
‘Sorry, the day I can grab a *NIX source code off the net, and go ./configure && make && make install without the whole thing going tits up, then I’ll know Solaris has made progress, until then, all I see over there is a perpetual buck passing and blaming for things that should have been corrected over 10 years ago.’
eh? what OS actually does that? no OS ever builds straight up because you always need glib this, libexpat that.. blah blah. Solaris, OS X, Linux, BeOS, QNX, *BSD, AIX all the same.
Closest I’ve seen to working is BSD ports, NetBSD pkgsrc and Fink… oh and ruby gems
still waiting on their Sun Ray software to come out of beta and on Solaris 10 x86.
Waiting for what? Sun Ray for Solaris x86 was released 9 months ago. You can download the final version from http://www.sun.com/software/sunray/index.jsp
this reminds me of that seinfeld episode where George comes up with the perfect combination of watching tv, eating pastrami and something else (that we don’t want to imagine george doing)…all at the same time. Sometimes, really good things don’t make any sense together.
Solaris is a better kernel than what BSD offers (multi cpu’s anyone?) and apple provides the user interface.
I still wouldn’t use either though…
Not that I really care what you run but just so you know all 4 BSDs support multiple processors. FreeBSD even defaults to the SMP kernel as of version 6.1.
but does it do it well? i dont have experience with smp on bsd myself, but i know many that have, and they tell me that in terms of smp, linux is much much better at it than freebsd.
Solaris has the potential to become a desktop killer by itself, given its decision to have GNOME as the default desktop – and given the huge progress so far, it’ll move from strength to strength in the future, thus benefiting all those who rely on GNOME.
The problem is, whether Sun Microsystems has the drive, determination and passion to turn their decrepit operating system into something that resembles a modern desktop operating system – low latency audio system with a extensible and modern API, hardware support that doesn’t royally suck to the point of not even supporting the most basic devices that are shipped with todays computers.
The problem with Sun, they fired, fired, fired, and fired programmers by the dozen, and now they have to spread what limited numbers thinly across multiple projects – I think someone needs to remind McNealy and Schwartz that you have to spend money to make money – perpetually cutting back those who actually create the products (aka programmers) infavour of having incompetent marketers (pointing the finger at the current marketing manager) and half baked sales folk who neither can communicate the companies products to customers let alone evangelise existing cutting edge features available in their said products.
And just what is decrepit about Solaris? That because Sun doesn’t use bleeding edge features that appeal to a limited audience, it is decrepit? If you don’t like the feature set of Solaris, use the Linux distros that give you the features you desire.
You are comparing apples and oranges (no pun intended). Sun’s plan for JDS and CDE does not include the home desktop, where things like “a low latency audio system” are important. And saying Solaris is decrepit because it doesn’t support advanced sound cards is ridiculous.
Wall Street has been screaming for years that Sun has retained too much of its staff. And while I do agree that Sun’s marketing at times leaves something to be desired, they are a long way from Microsoft’s marketing department. From what I see in terms of the Solaris Betas I test every few months, the e-mail from OpenSolaris about the progress of various projects, I think Sun has more than enough programmers to handle the load.
And just what is decrepit about Solaris? That because Sun doesn’t use bleeding edge features that appeal to a limited audience, it is decrepit? If you don’t like the feature set of Solaris, use the Linux distros that give you the features you desire.
Excuse me, why are they selling workstations loaded with Solaris, and promoting the idea of running Solaris on laptops – really bad miscommunication about Solaris’s direction or the fact you’re making excuses for Sun’s lack of drive to fix deficiencies?
You are comparing apples and oranges (no pun intended). Sun’s plan for JDS and CDE does not include the home desktop, where things like “a low latency audio system” are important. And saying Solaris is decrepit because it doesn’t support advanced sound cards is ridiculous.
And you know babe, sound cards are used by more than just ‘stupid pathetic home users’, what about those in those ‘fortune 500 companies’ which Sun marketers have wet dreams over – you know, they use their computers for more than just writing documents and spitting data onto a spreadsheet.
The sound issue could be easily corrected by purchasing the infrastructure they need – buy out OpenSound and merge it into Solaris, problem solved.
Wall Street has been screaming for years that Sun has retained too much of its staff. And while I do agree that Sun’s marketing at times leaves something to be desired, they are a long way from Microsoft’s marketing department. From what I see in terms of the Solaris Betas I test every few months, the e-mail from OpenSolaris about the progress of various projects, I think Sun has more than enough programmers to handle the load.
Wall street bitches and moans about every damn thing; heck, if Sun had one employee, it would be whining about ’employee bloat’ because these numbskulls on the trading floor think you can make products without employing anyone, and if any employee asks to be paid a fair reward for the honest days work – they’re obviously some sort of dirty communist who wishes to take down the system.
The problem isn’t bloat, but poorly utilising their employees; the number of employee’s they had, they should have already corrected all the deficiencies I mentioned and more – right now, I should be able to put Solaris onto any old machine, and for all the hardware to be supported. The issue isn’t about employee’s but bad management of human resources, and ultimately, that responsibility lands squarely on the shoulders of those managers in Sun who seemed to have locked themselves into the ivory tower a little too long.
It seems to be that management at Sun are suffering from the some sort of reality distotion field because none of the managers as far as I see can actually see any deficiencies in their own products.
If they think that there is nothing seriously wrong with Solaris or their middleware, if I was in the position of CEO, it gives me a clear message that they’ve lost touch with reality and their employment at Sun should cease immediately.
Edited 2007-01-15 20:50
First, Sun doesn’t even sell laptops anymore. You cannot find the Ultra 3 on store.sun.com, so it appears that Sun could not compete with General Dynamics (who bought Tadpole, a company who makes SPARC laptops). And where do you see Sun promoting the use of Solaris on laptops. Provide us a few links because I don’t see it and I get a lot of email from Sun both at work and at home.
And why should Sun buy OpenSound, Sun makes servers and workstations and they provide the necessary device support for those machines. Sun isn’t trying to compete in the desktop market, so why should they buy companies and products that don’t fit their product line or target market, to make it easy for you? Your ranting sounds less insightful and more like a child whining about not getting what he wants!
No, I think Sun managers get it and they are providing what their customers want. They are not providing what you want, so you piss and moan about how Solaris sucks because you can’t use Adobe Reader because it hasn’t been ported to x86, can’t compile software, and cannot use your sound card. Have you ever thought of purchasing OpenSound, or is this another “I expect somebody else to pay for it because I don’t want to”.
And I seriously doubt that your insights into management would improve the outlook of Sun Microsystems considering what you have said here today.
Robert,
can you please tell me a single reason, why should I buy a workstation from Sun? It seems to me that you agree, that Solaris is a server OS. I always thought that Solaris x86 is aimed mainly for workstations, so if there are on reasons to buy a Sun workstation, why bother with x86 Solaris at all? My article was about how to make Solaris more attractive for workstation users. From all the comments (including yours) I see that nobody would miss Sun in the workstation market, they should just stick with servers.
You wanted to have examples for non-ported software from SPARC to x86 Solaris. Take a look at the big three from the EDA industry: Mentor, Cadence and Synopsis. There is not a single big software package available from these companies for x86 Solaris (I know that Cadence offers only a GDS-Viewer for x86 Solaris). And as far as I know there are no plans to port their software to x86 Solaris. Linux is good enough.
So, if Sun won’t offer better solutions on desktop I see Solaris becoming a server only OS, like zOS.
Anton
If I am not mistaken this list of applications seems familiar (a pretty narrow list if you ask me), so because these applications don’t run on Solaris x86, Sun should get out of the workstation market? While I am not going to search for a list of applications that have been ported to Solaris x86, I believe that Sun should continue in the workstation market, and as time and popularity of the Solaris x86 platform increases, so will the amount of applications. As I have stated here before, Sun dropping Solaris x86 cost them and it will take time before some people come back, if at all. I would also not necessarily be so quick to put words into the mouths of vendors. What they don’t do today is not necessarily what they will not do tomorrow.
Do you compile your code on your server or your workstation? For me I use both based on what I need at the time. For testing software and OS configurations there is nothing like a workstation, so I think you have misread me. The biggest issue I see here is the confusion between workstation and PC, Solaris while running on x86 hardware is not your typical PC OS, just as Linux at one time was in a similar situation before its increase in popularity.
A PC is what you use for a variety of purposes, including entertainment. Kaiwai’s comments about poor sound support fall into this category. A workstation is what you use to compile code, run test databases on, etc. Sun’s workstation lineup (both x86 and SPARC) have sound capabilities to meet corporate requirements, not those of people looking to play games and use the machine more as an entertainment device. While work is being done by some people toward that end, that is not what I use Solaris x86 (or SPARC) for. But this is what I use it for, what someone else expects is entirely different, thus the difference in opinions.
A PC is what you use for a variety of purposes, including entertainment. Kaiwai’s comments about poor sound support fall into this category. A workstation is what you use to compile code, run test databases on, etc. Sun’s workstation lineup (both x86 and SPARC) have sound capabilities to meet corporate requirements, not those of people looking to play games and use the machine more as an entertainment device.
Maybe you should read what you wrote and take onboard the statment of not putting words in peoples/vendors mouths.
I never said I wanted to have a sound API that was great for gaming; what I want is simply the ability to be able to load it on my Toshiba laptop, have my sound card supported out of the box (realtek) and my wireless card actually supported with minimum fuss and bother.
I’m quite happy to knuckle down and pipe bits together, but what I am not going to do is install an operating system that doesn’t even support the most basic of hardware.
Oh, and soundcards in business – what about video conferenceing and the likes; like I keep stressing, a business uses their computer for more than just writing letters; most IT people seem to be oblivious to that fact.
I love it when Sun is timely, here is a list of applications that work on a Solaris x86-64 workstation or server:
http://www.sun.com/bigadmin/apps/data/views/all_applications_x64.pa…
I guess there is a reason after all for Sun to make workstations!
Thats nice, but one problem – they’re server applications, not workstaiton or desktop.
First, Sun doesn’t even sell laptops anymore. You cannot find the Ultra 3 on store.sun.com, so it appears that Sun could not compete with General Dynamics (who bought Tadpole, a company who makes SPARC laptops). And where do you see Sun promoting the use of Solaris on laptops. Provide us a few links because I don’t see it and I get a lot of email from Sun both at work and at home.
Obviously you have some major comprehension problems because I never said they sold laptops – also, open your eyes and see that Sun has consistantly said publicly that they’re willing to support hardware outside their own product line up; both McNealy before he stepped down, and Schwartz have said this.
Add that to the number of programmers who seem to be going hell for leather to improve the Solaris support on the Acer Ferrari and many other laptops, its telling me that Sun not only want’s Solaris on workstations, accessed via thin clients (which they sell, thus making Solaris a desktop operating system by proxy) but also to get technical people running Solaris on their laptops.
And why should Sun buy OpenSound, Sun makes servers and workstations and they provide the necessary device support for those machines. Sun isn’t trying to compete in the desktop market, so why should they buy companies and products that don’t fit their product line or target market, to make it easy for you? Your ranting sounds less insightful and more like a child whining about not getting what he wants!
People use workstations for more than just number crunching.
No, I think Sun managers get it and they are providing what their customers want. They are not providing what you want, so you piss and moan about how Solaris sucks because you can’t use Adobe Reader because it hasn’t been ported to x86, can’t compile software, and cannot use your sound card. Have you ever thought of purchasing OpenSound, or is this another “I expect somebody else to pay for it because I don’t want to”.
If they were providing what customers wanted, they would have made a profit by now, which they haven’t – their profit is a benchmark on whether they’re delivering what the customer wants, and so far, its been saying ‘no’ given the outlook. The changing of the guard isn’t going to change that situation.
And I seriously doubt that your insights into management would improve the outlook of Sun Microsystems considering what you have said here today.
I seriously doubt the Sun management are mature given their constant tirade of bashing Microsoft when all their failings have been due to their own incompetancy and nothing to do with the perceived evils of Microsoft’s, what I consider, hardnose business tactics.
If Sun wishes to sit in the corner playing with itself, then so be it, but they shouldn’t then come out blaming all and sundry for the lack of drive to take advantage of situations when they arise.
When I make a mistake in terms of decision making, I take responsibility for it; I don’t do a Scott McNealy, fly half way accross the world to Aussie to simply suck down some turps and whinge about how I’m is a victim of the greater Microsoft conspiracy of taking over the world.
Like I said on the Solaris mailing list 6months ago; every manager in Sun should be given a copy of Solaris, told to go home and install it on their PC – maybe a first hand experience with their product will wake them up, and realise how deficient it truely is for the markets they’re trying to aim at – the technical worker either on a workstation and laptop (as well as their traditional midframe/server markets).
What, exactly, is Apple’s commitment to the server? I have heard about DTrace and ZFS, is there anything else? Will their servers be expected to be a big part of their plans? I was a bit surprised that nothing much was said of Leopard at MacWorld, nor is anything said about the “server” aspect of Leopard on their website. In fact, the name change reflects all the gadgets and gizmos than actual computers. So, this article states the many benefits of the Solaris kernel that would help Apple in the server department. However, I have not seen much that would indicate that Apple is making a big enterprise push. Does anybody else have some more concrete ideas about Apple and their servers.
I Bet MS will be first to base it's vista successor on Solaris.
Rather than gripe about the downfalls of X86 Solaris why try it on the real thing. You can readily find inexpensive used Sun machines on ebay (even in Australia).
I thought they are moving step by step towards more like FreeBSD architecture, and just handpick those Solaris bit they want like ZFS and Dtrace.
But any kernel changes to Mac OS will properly come after 10.5 or 10.6
I also think the Star Wars and the Harry Potter universes should be merged. That would be just awesome.
Solaris is designed as a Server OS, it could be used as a desktop but it is designed as a server. OS X is designed as a desktop and there is a server version but it is no way close to Solaris for a server.
Solaris as a desktop system would be slow and choppy. Solaris is not designed to handle all the graphics and media, as well the algorithms are not for handling one user with a heavy task. Vs. as a server where there is less screen IO and designed for many users with small tasks. These require a major OS redesign to make Solaris work as well as OS X for a desktop. As well OS X will need a lot of work to match Solaris as a server.
> *IF* this would happen, what would happen to the Linux kernel project?
I’d bet my money on that Linux development will continue like it has always been. GPL3 Zealots will migrate to GNU/Solaris and the rest of the world will remain with Linux, probably because they don’t care about the difference between GPL2 and GPL3.
>Would Torvalds and friends eventually switch to GPL3 as well?
I’d bet my money on Linux not switching to GPL3. If it does switch, I’d bet that it is because of a fork. Also, Torvalds would disregard GPL3 Linux.
… apple should open her desktop enviroment, then everybody will do what he want.
Hell couldn’t be *that* frozen. Maybe if Steve Jobs get visited by three ghosts on christmas eve…
Or 6, because, i would really like to mod some of you up one more notch.
Solaris might be different, strange or even slow from a desktop jockey perspective, but that’s totally missing the point of Solaris (or any high-availability OS) here and this includes the original Blog author.
Both Apple and Sun has its strengths and has different focuses.
Apple is no longer a “computer” company. Sun is still in the highly reliable high performance computing area. Adding more spaz to Sun stations are not going to do a lot of good (for the effort), and neither does adding enterprise level expertise to a consumer OS (Considering that almost all Macs sold are of the desktop/laptop variety).
The comments made by the article is as absurd as GM and Ford and Chrysler should merge with each other in the automobile market. Not going to happen. Just a dream.