[lpi-discuss] RE: Linux Professional Institute changes Recertification Policy -- the "2 issues, " plus LPI's direction ...

Anselm Lingnau anselm.lingnau at linupfront.de
Mon Dec 4 17:28:14 EST 2006


Bryan J. Smith wrote:

> At the same time, while I wasn't in Germany, if this _indeed_ came up
> in a TAC or other committee -- that's "good enough" for me because
> it's at these TAC/committees where _real_people_ who have a "stake"
> in the process show up.

Just for the record, I was at the TAC session in Germany and the issue most 
definitely did *not* come up. OTOH, there were all kinds of fun and games 
regarding an LPIC-3 »super-guru certification« comprising six speciality 
exams. Boy, would I ever hate to redo *that* every two years!

> [Government of LPI]
> As long as they are not made "behind the scenes," the Meritocracy+Executive
> approach is _very_efficient_.

A lesson from recent political history: Unlike with other European countries, 
the German government wisely decided to keep the vote on whether to join the 
Euro (common EU currency) within the parliament rather than put it to a 
referendum, because at the time it was fairly obvious that the general 
populace would have given them a resounding »Hell, no!« So, if chances are 
that the community will give you an answer that you don't like, better not 
ask the community in the first place.

As far as the »meritocracy plus executive« approach is concerned, having a 
meritocracy is well and good if you're running a free software project, such 
as the Linux kernel. This implies that the meritocracy does get to make the 
actual decisions. In a real meritocracy like the Linux kernel, you ought to 
be able to increase your influence to the decision-making process according 
to what you contribute. This extends all the way up to Linus's »trusted 
lieutenants« and Linus himself, who have proven their capabilities to the 
Linux community over and over again. Sure, not all their decisions are spot 
on the first time around, but we do trust them to get things right in the 
long run, and they generally do.

In this model there is no room for an appointed »executive board« with veto 
power, since if the people on the board had the interest and know-how to 
fulfil their positions based on their standing in the community, they would 
be part of the meritocracy already, and no special »board« would be required! 
Hence, BY DEFINITION, the people on the board are stupid about what they have 
veto power over. Thus their apparent function must be to hamper the work that 
is being done, for the benefit of their sponsoring organisations. I think we 
can all agree that converting the Linux kernel project over to »meritocracy 
plus executive« would be the single worst idea in the history of Linux 
development.

There seem to be very few projects around where »meritocracy plus executive« 
really works. Most of the bigger ones, with the possible exemption of Debian, 
are pure-meritocracy, often with a »benevolent dictator for life« at the top. 
Look at Python and Guido van Rossum, Perl and Larry Wall, X11, KDE or GNOME.  
Fedora may be the most obvious example of M+E, but even this is losing 
mindshare to Ubuntu (which is pure-executive) right, left, and center. If M+E 
is so »very efficient«, then why does this happen, and how do the other 
projects manage?

Now, LPI is not a free software project, and we should not expect it to be 
managed like one. We've recently seen a number of interesting people take 
over positions within LPI who do not have a history with the free-software 
community, but come with impeccable credentials from the business world. This 
is potentially a good thing, because it gives us a fresh handle on many 
things. On the other hand, there is a significant danger that too much of 
a »business outlook« will hurt relations to »the community«, which is where 
LPI originally came from, and the current LPI executive should keep this 
firmly in mind. After all, LPI purports to be a certification »of the 
community, by the community, for the community«. It would do well to align 
its processes closely with what the community expects and is used to in order 
to uphold this philosophy. Think »bazaar«, not »cathedral«.

If the public perception of LPI changes to »yet another certificate« 
(regardless of whether we, who are close enough to the process to at least 
get some insight, and who care, subscribe to this POV or not), all the credit 
that it started out with as the »community-based certification« will be lost, 
and this will be very difficult to re-gain. We -- and that means not just the 
LPI executive, but the sum total of all people involved in LPI work, 
including folks like us who go out to actually *prepare* candidates, or teach 
general Linux classes where we try to nudge the participants towards 
considering LPIC -- have to work at making LPI the #1 Linux certification, 
and as we all know this is not an easy task. It requires cooperation at all 
levels. In recent years we have made significant inroads towards convincing 
the community that a non-proprietary Linux certification is actually a Good 
Thing To Have Around -- when there used to be significant antagonism to the 
idea, and still is in places. *Please* do not jeopardise all this work by 
sacrificing so much of what we used to use as »ammunition« when »spreading 
the word« among the heath^Wuncertified, just to make LPIC more appealing 
to »business« minds. Your candidates still come from the community.

Anselm Lingnau

(This is my personal opinion and not that of Linup Front GmbH.)
-- 
Anselm Lingnau ... Linup Front GmbH ... Linux-, Open-Source- & Netz-Schulungen
Linup Front GmbH, Postfach 100121, 64201 Darmstadt, Germany
anselm.lingnau at linupfront.de, +49(0)6151-9067-103, Fax -299, www.linupfront.de


More information about the lpi-discuss mailing list