[OpenIndiana-discuss] OpenIndiana branding

[Lou Picciano]

Richard, Jesus: 


This conversation is helpful to me; I am wrestling with this right now, as I am modifying a build or two to be Oi-specific. 


What's the consensus? What's the best, most reliable 'tag' to use for indicating a given build is for Oi_148? 


Of course, the question suggests: 'What the best way to remain most compatible with the largest number of other packages and subsequent incarnations of Oi'? 


Lou Picciano 


----- Original Message ----- 
From: "Richard L. Hamilton" <rlhamil at smart.net> 
To: "Discussion list for OpenIndiana" <openindiana-discuss at openindiana.org> 
Sent: Saturday, January 15, 2011 10:47:52 AM 
Subject: Re: [OpenIndiana-discuss] OpenIndiana branding 


On Jan 15, 2011, at 3:33 AM, Jesus Cea wrote: 

> -----BEGIN PGP SIGNED MESSAGE----- 
> Hash: SHA1 
> 
> On 14/01/11 21:53, Alasdair Lumsden wrote: 
>> OpenIndiana is technically SunOS 5.11, and to maintain compatibility 
>> we'll be keeping this for the foreseeable future. Perhaps in the very 
>> distant future a version bump might occur, but we're talking 5+ years 
>> here I imagine. 
> 
> While I support "SunOS 5.11" naming, I am a bit afraid of compatibility 
> divergences with the coming Solaris 11, also known as "SunOS 5.11". So 
> we will have two different systems reporting "SunOS 5.11" that are not 
> actually the same system. 
> 
> I don't have any better proposal, though. Maybe considering any 
> incompatibility with Solaris 11 a "bug to be fixed" could be a good 
> advice. Time will tell, I guess. 

Together with keeping uname -v unique, so that any unavoidable 
or intentional differences that remain can be identified, IMO that's reasonable. 

Even those differences that aren't bugs or can't be fixed should IMO be _documented_, so that one could know that for a given oi_xxx, what the differences were to the most similar Solaris 11 build. They might well be mostly that some recent or proprietary feature wasn't available, I'd imagine. 

Hopefully proper configuration scripts that as much as possible probe for features rather than _assuming_ features given an OS version will largely avoid problems when building source. But for those that don't, and for binary compatibility, it would be IMO very desirable for the differences to be documented. Also, the exercise of documenting the differences should serve as an additional check on avoiding unintended differences. 


_______________________________________________ 
OpenIndiana-discuss mailing list 
OpenIndiana-discuss at openindiana.org 
http://openindiana.org/mailman/listinfo/openindiana-discuss