[advocacy-discuss] "What is OpenSolaris" page

Octave Orgeron unixconsole at yahoo.com
Tue May 6 22:44:15 PDT 2008


Hi Everyone,

I think we all have some strong opinions about the release of Indiana as OpenSolaris. It is good that as users, developers, sysadmins, etc, we have something to say. I also see that there is a divide where our participation as a community may be in jeopardy. I see the two sides have the following points:

Traditionalist:
1. Need for UNIX/POSIX Compliance
2. Support for SPARC, Jumpstart, etc.
3. Users should be able to learn the OS and not be hidden from it.
4. Anger with the Indiana=OpenSolaris issue

Modernist:
1. Need for GNU look/feel
2. Focus on x86 before SPARC
3. Users should not have to learn the OS
4. Emulate the Linux way of doing things

There are definitely arguments on both sides that are valid depending on how we look at things. 

For example, the focus on capturing Linux developers on x86. On the one hand, you have developers who are familiar with Linux, but not Solaris or SPARC. On the other, you have experienced developers who are use to porting their apps between platforms. Realistically, very little of the open source software available today only compiles on Linux/x86. Projects such as pkgsrc and BSD ports should demonstrate this fairly well. So which developers are we after? Are we after the ones that build desktop applications or commercial vendors who sell things like databases and app servers? If it's the desktop applications, we definitely missed the mark here. There are more MacOS X desktops deployed than Linux desktops. And if the vast majority of the open source applications work on Solaris today, then where is the issue? Are we trying to make it easier for end users to compile software? If that's the case, perhaps we need a pkgsrc->IPS pkg depo? Would that not be
 the easier solution? In which case, we would have thousands of free applications that can be easily installed. If we are trying to capture commercial developers, we definitely have a great story with our development tools (Sun Studio, Java Studio, etc.). Perhaps making those components easier to install with the OS makes more sense?

It's nice to attempt to attract the Linux user base. But realistically, it's an uphill battle against users, developers, and sysadmins that are familiar with the Linux way of doing things. We can try to emulate the installation or GNU user-space, but that's not enough for Linux folks. They are passionate about their OS and concepts such as the GPL license. I'll put it another way. There are plenty of Solaris users that still consider SPARC to be the desired platform for their desktops. Try as you may, they won't dump their beloved SPARC gear. And even as Solaris and OpenSolaris have improved x86 support by leaps and bounds.. SPARC gear sells very well on EBay to feed a demand that's very real. So does it make sense to only focus on x86 when a huge majority of the install base(especially in the commercial sector) is SPARC? Not to mention how SPARC sales lead over x86 sales when it comes to hardware at Sun. If we don't keep that SPARC install base, doesn't
 that make the goal of increasing the user base that much harder? In the same way, we probably won't capture a large segment of the Linux user-base if that's all we're focused on.

So this brings up lots of questions and fun debate points that the non-Sun and Sun developers can chew over for months. But ultimately, I would pose the following big questions:

1. Considering that Apple has captured more of the UNIX desktop market from both Sun and Linux, are we focused on the right end-user experience?
2. How are we really going to stand out from Linux/MacOS/BSD/etc and attract users, developers, sysadmins, and businesses?
3. What has been successful in the past with Solaris and what has not?
4. What are the core values that make OpenSolaris different? (Compatibility,  UNIX/POSIX compliance,  Stability, Quality, etc.)
5. How can we capture enough developers to participate in taking OpenSolaris to the next step? Or do we accept that most of the development will come from Sun?

Ultimately, both the "traditionalist" and the "modernist" have to come together and face these big questions. Otherwise, we'll alienate each other and go our separate ways. I don't think that's what any of us want. We're all here to participate in the development of OpenSolaris. So I urge both sides to take the bull by the horns and get things back on track. Together we are much stronger and can accomplish more than when we divide and fight amongst ourselves.

Octave

 *-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Octave J. Orgeron
Solaris Systems Engineer
http://www.opensolaris.org/os/community/sysadmin/
http://unixconsole.blogspot.com
unixconsole at yahoo.com
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*


      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ


More information about the advocacy-discuss mailing list