[indiana-discuss] About SXCE, SXDE, Indiana future

[Bryan Boone]

Hopefully, Ian (I'm not really on a first name basis with him;) )
Will model the release cycle as he did in Debian.

1) stable
2) testing
3) experimental

Users of Solaris would be able to glom onto whichever release they're 
comfortable with.

"Big Iron" servers, for instance, would be able to use a "sever" distro 
based on stable (Solaris 10...11...12...)

Nevada could be the next "Ubuntu" distro based on testing

Indiana could be the next "Sid" distro based on experimental.

Having these tracks could then allow different folks to tailor their 
distros (ala distrowatch distros) for different audiences (multi-media, 
development, cool themes, etc).  Last time I checked there were about 
60-70% distro based on Debian in distrowatch's top 100.  That's a lot of 
power/choices placed in the hands of distro developer and the user.

I don't think Sun would have to be all things to all people, just 
provide sufficient tooling. (ie, conistent release cycle and packing tools)

-Bryan


Brian Nitz wrote:
> Richard Elling wrote:
>> Andrew Watkins wrote:
>>
>>> I wonder is this the starting point of Sun having 2 supported Operating systems, a server version (Solaris) and a desktop version (Indiana/OpenSolaris). I find it very hard to believe that they would be able to convince big business that the new Indiana would run there big systems.  More in line what RedHat, Ubuntu and Microsoft are doing.
>>>
>>> I know we are talking what happens after Nevada (Solaris 11?), but what I see happening is:
>>>
>>> - Server Version (Solaris) with long term support, getting close to the Indiana i.e. packages system.
>>>
>>> - Desktop Version (Solaris Desktop) with a shorter term support level, but may be with more up to-date desktop packages.
>>>
>>> - OpenSolaris which will be the cutting edge, with the 6 months turn around (No support).
>>>
> A few of us have been advocating creating a separate desktop
> distribution Sun for a while.  There was even more reason for this when
> the old Solaris installer and package manager didn't allow the kind of
> package granularity that allowed the sysadmin to configure the system to
> balance the needs of a desktop user against the needs of a server.
> 
> The IPS package system and common sense install system allows much more
> granularity of install, patch and upgrade and if it is carefully
> designed, can make it possible to have one release cover the needs of
> most desktop users as well as traditional big iron Solaris users.
> 
> But if we encounter too many places where the needs of the desktop user
> conflict with the needs of the big iron server administrator, I would
> advocate steering Indiana towards the desktop user and leaving Nevada to
> handle the long term stability guarantees and slow upgrade cycle that
> large companies running big servers require.  It has been my experience
> that small and large companies are fed up with being forced into an ever
> more frantic upgrade cycle.  The only organizations who crave a rapid
> upgrade cycle are those with lots of money, those with critical bugs in
> their existing system, or those who profit from planned obsolescence.
> 
> Keep in mind that even Indiana can download fully QAed and stable new
> releases every day, companies must qualify their internal  and OEM
> software against this.  In some cases there is a legal or internal
> policy requirement for re qualification even if Sun (or another vendor)
> guarantees the a rock solid stable API/ABI.  Look at how reluctant Adobe
> is to rebuild their acrobat reader for Solaris X86 even though Solaris
> X86 has a rapidly growing user base and even though Sun stands behind
> the portability of applications  between Sparc and X86.  Imagine trying
> to convince Adobe to release a new acroread for our platform every 6
> months.  Now imagine thousands of organizations who don't have the
> software development resources of Adobe and who must do more than the
> "./configure && make && make install" and QA that Adobe would have to do
> to port acroread to Solaris 10 x86.
> 
> 
>>>
>> I disagree, that is a 1980s model.  Most of the server/desktop
>> separation in other companies
>> is because of the pricing premium expected to be extracted from
>> servers.  For Sun servers,
>> the server includes the appropriate Solaris license, so the pricing
>> model is very different
>> than software-only companies (a good thing).
>>
>> Also, with the increasing market for thin clients, that which is desktop
>> is also server.
>>
> So shakeout desktop bugs and stability issues in a fast upgrade cycle
> Indiana release and qualify a Sun Ray release against Indiana every year
> or so to provide a stable desktop platform.
> 
>> I do agree that the cutting edge will continue to be important, that is
>> where the action
>> is, not the slow-paced back-end.
>>  -- richard
>>
>> _______________________________________________
>> indiana-discuss mailing list
>> indiana-discuss at opensolaris.org
>> http://mail.opensolaris.org/mailman/listinfo/indiana-discuss
>>
> 
> _______________________________________________
> indiana-discuss mailing list
> indiana-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/indiana-discuss