[lsb-discuss] Don't blame LSB and standards, please: was: Re: Fedora Plasma Product, feedback please

Robert Schweikert rjschwei at suse.com
Tue Apr 1 15:39:13 UTC 2014

On 03/31/2014 05:57 PM, "Jóhann B. Guðmundsson" wrote:
> On 03/31/2014 08:53 PM, Theodore Ts'o wrote:
>> I think it might be helpful if you were really explicit about*which*
>> problem you are trying to solve here.
> The problem I'm trying to solve is unification across distributions and
> applications which I though was clear and I thought this standard was
> supposed to be solving

Well there are at least 2 planes of "unification" one plane is, 
everything is in the same place and at the same version in every 
distribution. Lets just say that unless we start living in a parallel 
universe that's just not going to happen. The second plane of 
"unification" is the interface level for developers to use. This is 
where the LSB has helped to keep distributions honest such that for the 
most part developers can treat Linux as one platform despite having 100+ 

For most developers it is immaterial if executable A is found in /bin 
/usr/bin or /sbin or ... Or if python modules are located in 
/usr/lib.... or /usr/lib64.... as long as "import A_MODULE" works across 

Therefore, stating the goal as unification is still not very specific, 
as Ted had pointed out.

> but given that you ask me how to convince
> application developers of the benefits of standardization for a standard
> that has existed what close to 13+ years and you yourself has been a
> part of all these years as well as the point I was trying to make by
> moving away from "discussion" which is an word that shares its root
> meaning with "percussion" and "concussion," both of which involve
> breaking things up eventually leading up to having one's opinion prevail
> over another which can easily be settled here and now by simply saying
> you are right which should satisfy what the thought of self seeks which
> is the driving force behind discussion, to a conversation in a form of
> dialog I realize this is not the venue for such effort to take place and
> we probably put different meaning in that standard and what said
> standard is supposed to solve as well.
> Anyway I proposed a method for the "LSB people" to restore faith in the
> standard

Hmm, another bold statement, who says, other than you, that people have 
no "faith" in the standard? I agree that in today's environment the 
value proposition of the LSB is being questions as distributions are 
probably closer to each other than ever before, but that is a different 
topic than "faith in the LSB."

> which is the first step as in getting people to follow said
> standards ( no matter how good standards are, they are useless of no one
> is following them).

There are 7 certified distribution, thus those follow the standard by 
pretty much any definition. That however does not mean that there are 
not many more distributions that are LSB compliant. Today certification 
costs money, although I think we are more than willing to waive the fee 
for community distributions, there are still few incentives for a 
community distribution to certify. Almost non of the commercial ISVs 
support running their apps on a community distribution and open source 
applications generally get built and compiled by the distribution 
vendors. Therefore, the cross distribution idea is somewhat moot, as 
already pointed out by Ted as the upstream application project does not 
have to worry about distribution portability, they get that for free 
from the distribution vendors who send patches back.

> Whether you choose to follow my advice's and approaches or disregard it

Well, from my point of view your "advice" is mostly constituted of bold 
statements that at times do not necessarily reflect the reality of the 
world that surrounds us. That said, at the recent f2f meeting the 
working group has decided to focus on a new direction after LSB 5.0 is 
out the door and work on setting things up to support that new direction 
has begun. There's a Github project with pending pull requests, thus 
everyone has a chance to review and comment.

More fingers on the keyboard that contribute are always welcome. As with 
any open source project the direction is influenced by those that 
contribute more so than by those that send messages to a mailing list.


Robert Schweikert                           MAY THE SOURCE BE WITH YOU
SUSE-IBM Software Integration Center                   LINUX
Tech Lead
Public Cloud Architect
rjschwei at suse.com
rschweik at ca.ibm.com

More information about the lsb-discuss mailing list