[Pgi-wg] PGI decision-making process - seeking input

[Oxana Smirnova]

Hi all,

some suggestions on decision-making process in PGI were already made, both on this list and during today's meeting (trust me ;-) ). I was appointed by the group co-chair to prepare a proposal for simple internal guidelines (this ought to be in the minutes).

I therefore would like to solicit input from those of you who have opinion and did not express it yet. I'll try to outline the background:

1. PGI stands for "Production Grid Infrastructures", specifically:
1a. Stakeholders represent parties enabling production grid infrastructures:
  * administrators
  * middleware providers
  * users
2. (1) implies certain group specific, namely:
2a. Stakeholders enter with pre-existing deployed solutions, related interfaces  etc
2b. There are substantial numbers of customers behind every stakeholder
2c. There is a substantial, and ever growing, number of stakeholders

For reference, here's a list of distinct stakeholders represented in PGI which I deduced from currently active PGI participants, in no particular order:
UNICORE, DEISA/PRACE, Globus/IGE, EDGI, NDGF, ARC, EGI, gLite, GENESIS, TeraGrid, SAGA, NAREGI/RENKEI, NGS (and I probably missed some).
Several stakeholders are represented by more than 1 individual. Even if UNICORE, ARC and gLite will become a single EMI stakeholder, there will still be more than 10.

In general, the group should follow the OGF guidelines and aim to achieve consensus. However, due to the large number of stakeholders and high stakes, consensus is often practically impossible. In such cases, decisions must be made following a consistent *open* procedure, and not left to the judgement of a single person: this is how it is done in modern democratic societies.

A solution can be to:
a) Limit number of decision-makers by applying criteria such as affiliation (one voice per stakeholder) and/or attendance (75% attendance of all group meetings, or similar), or
b) Introduce voting procedures (quorum, majority, tie-breaking, veto), or
c) Combine (a) and (b).

Important decisions which involve complex documents and/or potentially affect functionality of the production Grid infrastructures, must be well prepared in advance:
  * proposal authors must provide sufficient material for the stakeholders such that they can use it to consult their user base if necessary
  * sufficient preparatory stage must be allowed, during which all the stakeholders will have the opportunity to study the proposals and consult their customers if necessary
  * materials must be made public on GridForge and advertised on the PGI mailing list, together with relevant deadlines
  * decisions achieved by consensus, and *especially* those achieved by voting, must be documented in respective public meeting notes for further reference


Please send your thoughts either to the list or directly to me by next Wednesday, November 17.


I anticipate irritated comments about overcomplicating the process and violating OGF rules. Believe me, there is nothing complicated, and this is a very simplified model of public decision-making process, successfully used in bodies from condominium boards to country parliaments. Ad-hoc procedures, absence of relevant documentation and reliance on a wise leader typically lead to stagnation and failures, even if there are individual success stories. I am confident OGF is not pursuing this latter track.


Cheers,
Oxana

-------------- next part --------------
A non-text attachment was scrubbed...
Name: oxana_smirnova.vcf
Type: text/x-vcard
Size: 270 bytes
Desc: not available
Url : http://www.ogf.org/pipermail/pgi-wg/attachments/20101111/b87a3906/attachment.vcf 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2357 bytes
Desc: S/MIME Cryptographic Signature
Url : http://www.ogf.org/pipermail/pgi-wg/attachments/20101111/b87a3906/attachment.bin