[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

RE: agreements vs. Hobbes

• To: "'Simon St.Laurent'" <simonstl@s...>, xml-dev@l...
• Subject: RE: agreements vs. Hobbes
• From: "Bullard, Claude L (Len)" <clbullar@i...>
• Date: Thu, 8 Aug 2002 09:54:48 -0500

A Hobbesian world doesn't admit to the overlapping 
boundaries in domains.

The need for formal aggreements doesn't 
extend to all levels and every case.   A smart 
manager isolates out the mission critical, agreement 
sensitive aspects and works toward formal agreements 
on those aspects, careful to fix values that must 
be fixed and to leave dynamic that which must be 
dynamic.  In short, constants and variables.

The nasty brutish wrestling match starts over 
seeking advantages in the classification of 
members of these.  Rhetoric is used to persuade 
the crowd that something should be left variable 
or made constant when the decision is in an 
underdefined boundary case.  Daring to do less 
is an exercise in declaring large overlapping 
boundaries and leaving the members of these 
to choose sides.

len

-----Original Message-----
From: Simon St.Laurent [mailto:simonstl@s...]

At the end of Walter Perry's presentation last night, C. Michael 
Sperberg-McQueen suggested that Walter was describing a "Hobbesian world of 
processes in competition...each process for itself", while for himself, 
"call me a corporatist", he preferred to work with processes based on prior 
agreement.

After too many years of working with web browsers, which share a common 
agreement to some extent but which still have dark corners, I know that I'm 
inclined to doubt the prospects of large-scale distributed projects run by 
competitors proving genuinely willing to abide by the terms of the 
contract.  Contracts in the United States often start with the best of 
intentions, but but sometimes turn into battlegrounds, specifying the terms 
of engagement in a more bellicose style than was originally intended.

I'm curious at this point how the "XML project" of agreement-building is 
proceeding.  In most of my own work, I find that either I don't bother with 
contracts (my own rules files, which others have been able to adapt to 
their own needs) or the contracts sort of partially work (HTML, DocBook at 
O'Reilly).

Are most people working on building agreements across communities?  Are 
they working on the I-publish-you-discover approach common to smaller 
efforts and formalized by things like WSDL and (to a lesser extent) RDDL?

Personally, I'm happy to support prior agreements when all parties agree 
about the nature of the agreement and are willing to continue supporting 
that agreement over time, but I have some deep suspicions about the nature 
of agreement that leave me suspecting that technologists and technologies 
often live in a Hobbesian world.  The results are less difficult than those 
Hobbes predicted: we don't all seem to be living the "nasty, brutish, and 
short" life, nor do I see much need for us to throw ourselves before the 
mercy of an all-powerful tyrant, which I believe was Hobbes' solution [1] 
to the ugliness of this world.

• Prev by Date: process vs. "it is"
• Next by Date: RE: matching text with regexp in xpath
• Previous by thread: process vs. "it is"
• Next by thread: =?utf-8?B?UkU6IFt4bWwtZGV2XSBtYXRjaGluZyB0ZXh0IHdpdGggcmVnZXhw?==?utf-8?B?IGluIHhwYXRo?=
• Index(es):
  • Date
  • Thread