====================================================================== CFJ 1089 The Proposal submitted by Swann and entitled "The Great Officer's Veto Scam Proposal" has not been distributed. ====================================================================== Judge: Kolja A. Justices: Steve (S), Morendil (J), Michael (C) Judgement: TRUE Eligible: Antimatter, Blob, Chuck, Crito, elJefe, Harlequin, Jester, Kolja A., Michael, Morendil, Oerjan, Steve, Swann Not eligible: Caller: Murphy Barred: - Disqualified: Time Agent (defaulted) On hold: General Chaos, Sherlock ---------------------------------------------------------------------- Appeal eligibility: On hold: elJefe, General Chaos Originally ineligible: Sherlock, Murphy Judge: Kolja A. Default Justices: Steve, Morendil, Michael ====================================================================== History: Called by Murphy, 18 Mar 1998 01:50:52 -0800 Assigned to Kolja A., 27 Mar 1998 11:36:08 +0000 Judged TRUE, 27 Mar 1998 17:02:34 +0100 Published, Fri, 27 Mar 1998 18:40:38 +0000 Appealed by Blob, Sat, 28 Mar 1998 13:58:08 +1100 (EST) Appealed by General Chaos, Sat, 28 Mar 1998 07:55:09 -0500 Appealed by Steve, Wed, 1 Apr 1998 10:57:52 +1000 (EST) Appeals process begins, Thu, 2 Apr 1998 12:33:39 +0100 (BST) Morendil judges to sustain, Sat, 4 Apr 1998 18:16:50 +0100 Steve judges to sustain, Wed, 8 Apr 1998 13:16:06 +1000 Michael judges to sustain, Wed, 8 Apr 1998 08:39:11 +0100 Final decision published, Wed, 8 Apr 1998 08:44:30 +0100 ====================================================================== Appelate decisions ---------------------------------------------------------------------- Justiciar Morendil: The Judge of this Statement presented a clear and concise argument in favor of the decision. However, I find some of his reasoning flawed; in particular, I believe that Rule 1339, cited therein, is evidently not relevant, and I further believe that Rule 754 is not relevant either. Should we interpret Rule 754 such that the difference between the text submitted by Swann as "The Great Officer's Veto Scam Proposal" and that posted by Promotor Blob under that title is *not* "inconsequential", that would not necessarily imply that Swann's Proposal was *not* distributed. Such an interpretation would merely imply that Rule 754 is silent on the matter. Judge Kolja observes that the difference between the two texts is not one of "spelling, grammar, or dialect, or the substitution of a word or phrase by a synonym or abbreviation", and draws the conclusion that, since Rule 754 does not cover this particular kind of difference, that difference is significant and the truth of the Statement follows. However, I find this line of reasoning suspicious in the extreme, because it would lead us to regard as distributed a Proposal submitted with the text "A Rule shall be enacted reading thus :" but posted by the Promotor as "Yo man, let's create a Rule thingy like this here:", a discrepancy far greater than the minor difference between Swann's submission and what Promotor Blob posted. However, I believe that Rule 754 does not govern in such cases; it is my interpretation that it only concerns *original* statements by Players. For instance, Rule 754 would apply in the case of a Player posting a text labeled as an "Aplication to start a Gourp", which would be considered unambiguously equivalent to an Application to create a Group. In the case at hand, however, the Rules require the Promotor to distribute the text that was submitted to em, not a variation on it; therefore, it is not enough that (as Rule 754 has it) "the meaning or validity of such communication is not altered in any way by such discrepancies"; it must further be the case that such a text be "the same", in some reasonable sense, as the one which was submitted to em, for *that* text to have been distributed. In other words, the issue really hinges on the notion of identity. No two copies of the text of a given Proposal are, of course, 'identical' in the strictest sense; nevertheless we use an approximation of considering two bodies of text to be 'identical' when they differ only in minor particulars. We don't use a formal definition of how 'minor' the variations need to be, however, expect as far as Rule changes are concerned (Rule 1339); we rely on an intuitive notion of it, which, I suppose, could be expressed as 'two bodies of text are identical if a computer program such as diff would consider them identical.'. Because this notion is not formalized, except in Rule 1339, I believe the Rules are silent on the matter. I also think it is in the best interests of the game not to allow a risk of Players voting for a Proposal unless they are provided with the "exact" (in the above vague sense) text of that Proposal. I further think that the discrepancy between the text submitted by Swann and that distributed by Promotor Blob violates the amovementioned intuitive notion of identity. I thus find myself in agreement with the original Judgement of TRUE, although unable to support the reasoning behind it. I therefore choose to Sustain. ---------------------------------------------------------------------- Speaker Steve: I follow Justice Morendil a long way in his reasoning. I agree with him that Rule 754 is not really relevant here (although the question of the relevance of R1339 is one I will return to later), and that the issue turns on the question of the identity of texts, that is, on the circumstances under which we should be prepared to say that two texts are 'the same'. I agree with Morendil that on this issue the Rules are silent. I agree that our Judgements should therefore reflect commonsense and the best interests of the game, in accordance with R217. However, the argument he gives that the best interests of the game are served by applying a very strict notion of identity ('identity under diff' as we might call it) has been effectively rebutted by Kolja, who pointed out that R107 guarantees that whatever gets distributed is what Players are voting on. So the question of whether identity under diff really is the best notion of identity for us to use remains open, as far as the arguments that presented by Morendil are concerned. That is not to say that identity under diff should be rejected - far from it. It is one of a number of standards now awaiting a defence which are competing for our attention. Certainly identity under diff has advantages which Morendil does not explicitly mention. It is simple and easy to apply, yielding clear answers in most cases. Perhaps it is to be preferred for this reason. Yet there are arguments against it. Consider capitalization, for instance. Former Rulekeepor Wes arbitrarily changed the capitalization of many Rules in ways that would violate identity under diff, and yet we accepted those changes as not changing the identity of the Rules. So historically speaking, simple identity under diff does not seem to be the standard that we have used in judging whether two texts are the same. One could argue in addition that identity under diff is not the standard that we should use, on the grounds that it is too strict. If strict identity under diff is to be our standard, then no typo introduced into a Proposal by the Promotor, however miniscule, will ever be considered insignificant. I don't think that is a desirable outcome. (That's not to say that no typo - in particular, the one introduced into P3710 - should be considered significant.) In that case, the simplicity of identity under diff, supposedly its strength, may actually be a weakness: it may be too simple to meet our needs. The challenge then is to say in what ways identity under diff should be relaxed, or with what it should be replaced. I propose to do what I can to develop an alternative standard. To do that, we need to have a clear understanding of what our needs are. As I've already intimated above, I don't think it's desirable (ie in the best interests of the game) that every typo introduced by the Promotor into a Proposal should invalidate the distribution of the Proposal. I think we would be best served by a standard which enabled us to say, at least some of the time, "Well, sure, there's a typo, but it's still the same Proposal." So strict identity under diff is out. Working from the other end, equally clearly we don't want to allow a situation in which it would legal for the Promotor deliberately to make substantial changes to a Proposal, and then offer as a defence, "Well, sure, it's completely different, but it's still the same Proposal." So, as a proposed standard, 'anything goes' is also out. That knocks out the extreme positions, leaving a large territory in the middle to explore. One temptation is to begin by distinguishing between accidental and deliberate changes introduced into a Proposal, and to completely rule out the latter. But I think that temptation should be resisted. If we had case where a Promotor had corrected a spelling mistake before distributing a Proposal, I think our standard should allow us to say that the Proposal has been distributed. A better strategy returns us to R1339: if two collections of text have different effects when adopted, then those collections of text are not the same. (Note that this does not imply that if two collections of text have the same effects when adopted, they should be regarded as the same.) Nevertheless, if a change introduced by the Promotor, either accidentally or deliberately, results in a text different in its effects when adopted than the text submitted to the Promotor, then I think that this is sufficient to show that the texts are not identical. Since R1339 is very clear that the change introduced by Promotor Blob into Swann's Proposal results in a text with a different effect, I think we have no choice but to conclude that Promotor Blob did not distribute Swann's Proposal. I therefore sustain Kolja's Judgement. ---------------------------------------------------------------------- Clerk of the Courts Michael My fellow Justices have done all of original thinking on this question. This is appropriate because my own position is somewhat compromised by my position as Judge of the similar CFJ 1090. I agree with the conclusions reached by Steve and Morendil and my ultimate decision is to sustain Kolja's original reasoning. My own instinct as to the best way of deciding whether or not two texts are the same is to perform a "diff" modulo any explicit provisions granted by the rules. My fellow Justices seem averse to this idea, and have come up with alternative ways of demonstrating that the relevant texts in this case are different. Particularly compelling is the demonstration that the meanings of the two texts are different, in the sense that if they were to be adopted, their effects on the game would be different. (This argument, an extensional one, appeals to the fact that if f(x) is not equal to f(y) then x is not equal to y, where x and y are our texts and f is our function for constructing the meaning of our texts.) I am happy with the intensional argument to the effect that i) the two texts in this case are demonstrably different, and ii) the rules do not allow us to consider them the same. In any case, the original judgement is clearly correct. ====================================================================== Original Judgement Judgement: TRUE Reaons and arguments: Rule 1483/3 (Definition of Proposals) reads, in part, A Proposal is created whenever a Proposing Entity delivers some collection of text to the Promotor with the clear indication that that text is intended to become a Proposal. The collection of text thus delivered is a new Proposal, and the Proposing Entity which delivered it its Proposer. A collection of text is said to be Proposed when it becomes a Proposal. Based on this definition and the publically known facts, it is clear that the proposal ("collection of text, delivered to the promotor with the clear indication etc..") referred to in the statement for this judgement was created, and exists. This proposal was then given a number and placed in the proposal queue (even if its text was not rendered correctly in the inofficial part of the document in which the Promotor published the proposal queue at the time). For the number note that at the time in question R109 was still in the old form 109/2, which assigned proposal numbers at the time of submission, not distribution. For the text note that R1036 states that "The actual text of the Proposals in the Queue need not be published." Therefore, the proposal queue did contain the proposal. However, while the text of a proposal or its scrambled rendition in the publication of the Proposal Queue has no legal significance for the state of the proposal queue, R1770 (Distributing Proposals) requires that "Proposals" be distributed, i.e. the body of text that consitutes the proposal. In the case of P3710, this didn't happen, and the difference between the distributed body of text and the true proposal 3710 is not insignificant. The only rules that deny minor differences between texts any significance are Rule 754/3 (Spelling and Grammar Errors) and Rule 1339/6 (Precision in Rule Changes) which allow "Differences in spelling, grammar, or dialect, or the substitution of a word or phrase by a synonym or abbreviation [...] as long as there is no ambiguity in meaning" resp. "Variations in whitespace or capitalization in the quotation of text in an existing Rule to be removed or replaced". However, the differing parts of P3710 and the distributed text were "Proposal at time" and "Proposal