[OpenIndiana-discuss] [oi-dev] Studio future in OI (Re: )

Nikola M. minikola at gmail.com
Sat Oct 19 08:43:53 UTC 2013


On 10/14/13 11:41 AM, Udo Grabowski (IMK) wrote:
> On 13/10/2013 11:57, Laurent Blume wrote:
>> On 11/10/2013 18:14, Udo Grabowski (IMK) wrote:
>>> This should not happen, this would kill us completely, we have 
>>> commercial
>>> applications compiled with Sun Studio, and a lot of software that can't
>>> even be compiled with gcc-fortran,
>>
>> I can't help thinking, what will happen to you when *Solaris* Studio 
>> stops
>> working on *OpenIndiana*? Aren't you worried about that? I don't 
>> exactly see
>> Oracle striving to maintain compatibility, and without source, it's 
>> bound to
>> happen...
>
> Openindiana must keep binary compatibility to Oracle Solaris, otherwise
> it will quickly be obsolete, as vendors of commercial software won't
> compile their software for OI, but only for Solaris 11, and FOSS packages
You can't expect to keep binary compatible with _Future_ releases of 
Closed source system,
you do not share kernel with.
You can expect to implement and exchange just specifications for changes 
and API descriptions or something, but just in case that cooperation 
between illumos and Oracle OS/Net exists.. (and I don't think it is).
It exists in other levels and user space code releases and I guess it is 
what we can use to cooperate.
There are many points where Openindiana and Solaris could continue to 
exchange code, in the userspace, porting applications specs, etc.

So you can only keep and freeze binary compatibility op to the certain 
level (like Solaris10 branded zone) or have freeze of features (to 
certain OI release) or make a new branded zone for your needs
or running applications inside KVM virtualized older OS.

Binary compatibility with a shooting target you can not even see is not 
realistic, even if you put a tons of money and manpower on it.
Binary compatibility with previous releases - then yes, but with 
freezing and virtualization.
> obtained from elsewhere in binary form would also break. This should 
> never
> happen.
I think this line of thinking got a bit to the points that are not 
realistic.

if you want to use open source project OS and software for your business,
you need to contribute to it's development with manpower and hardware 
and money resources.

That way you can steer it's path where you like it, but the future of 
every open source project is determined collectively, by actions of 
multiple contributors.
> If that ever happens, we would have to switch to Linux, where the Sun 
> Studio
> Compiler is also available (and binary compatible).
I don't think one can count on binary compatibility between kernel 
releases on Linux,
I think that was main feature of Solaris, not Linux.

I also think that Studio should continue to work on OI, BUT it is not OI 
that has power to maintain Studio.
You need to register as Oracle customer with paid support for Oracle 
Studio and then request that OI be supported with Studio releases, so 
you can compile software for OI in the future,
(*providing you contribute to OI to have /release in OI any time soon)

If we (users) need software for OI in the future, we need to compile it 
or pay someone compile it for us, and we need to support that process of 
sustainable and manageable distribution and applications, if we want to 
use it.
Even open source applications need to be patched, contributed upstream 
and compiled for every distribution release,
pre-releases needs to be tested and reported and fixed and all that 
require more involvement of the users in distribution construction 
process, but "if feature X don't work, then.. ." : right answer to that 
is: "... sure, you are free to make it happen".

if you have 3 closed-source entities (closed source Solaris, closed 
source Studio compiler and closed source applications from the vendor, 
support specific platforms) then you are trapped by them,
and not OI.
Illumos and OI gives you ability to actually do something about it and 
closed vendors don't have that in mind for you.

>
>>
>> The issue here is not monoculture, it's a choice of an ABI. As I 
>> pointed out:
>> you can use Studio and have the standard GCC ABI. So do have a try at 
>> those
>> compatibility options, and start assessing how you can make the 
>> switch. It
>> has *already* happen in Solaris' past. So surely, what our cave-dwelling
>> ancestors could do with their limited stone tools can be done more 
>> easily now
>> that we have actual technology.
>
> There's no way to urge vendors to compile for OI instead of Oracle
> Solaris 11, that's just unrealistic wishful thinking .... All of
> this ABI compatibility arguing is for software you compile yourself,
> and it won't work for existing binaries.
If you ask it from vendor, why not. It seems possible to recompile?

If you are forced to change platform because your vendor requires it, then
it is a problem of why being locked to that particular vendor and why 
the business model depends on vendor/s and why you don't control the 
source code of the applications you use for a business, yourself?

Those are all questions that inspired creation of illumos-based 
distributions like SmartOS , OmniOS, Nexenta, etc. They are all 
self-sustaining and exchange code through illumos or directly, to lower 
development costs.
Openindiana is also here to lower your development costs and productions 
costs, (you are spending money on development?)

So you might translate your urge for binary compatibility ,
to some realistic project within Openindiana, by allocating your 
development resources to Openindiana to make it happen, for your needs ,
while working with others that want to stabilize and advance distribution.
_It can shoot both targets but it needs your involvement_




More information about the OpenIndiana-discuss mailing list