Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, ais523 wrote: On Thu, 2008-10-02 at 07:22 -0700, Kerim Aydin wrote: On Thu, 2 Oct 2008, Roger Hicks wrote: So we are saying that SELL votes aren't valid unless the VM is published during the voting period on which they are cast? That is somewhat ridiculous, isn't it? Saying that you can't use obvious, well-known or straightforward contextual information outside the voting period to interpret information published within the period is indeed ridiculous, unless we take the position that no conditional votes are valid without publishing a dictionary as well. IIRC someone (maybe Murphy) submitted a proposal to fix the situation via legislation, because it does indeed seem ridiculous. My point is not that it's true now and needs a fix (though a clarification is always useful) my point is that it's ridiculous to interpret the *current* rule as excluding readily-available information (as long as it's *referenced* at least indirectly by the publication in question). -Goethe
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, Oct 2, 2008 at 12:15 PM, Kerim Aydin [EMAIL PROTECTED] wrote: My point is not that it's true now and needs a fix (though a clarification is always useful) my point is that it's ridiculous to interpret the *current* rule as excluding readily-available information (as long as it's *referenced* at least indirectly by the publication in question). -Goethe published during the voting period seems pretty unambiguous to me, as stupid a criterion as it is. If Rule 478 didn't define what it means to publish something I could be persuaded that readily-available and referenced counts as published.
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2008-10-02 at 12:30 -0400, Geoffrey Spear wrote: On Thu, Oct 2, 2008 at 12:15 PM, Kerim Aydin [EMAIL PROTECTED] wrote: My point is not that it's true now and needs a fix (though a clarification is always useful) my point is that it's ridiculous to interpret the *current* rule as excluding readily-available information (as long as it's *referenced* at least indirectly by the publication in question). -Goethe published during the voting period seems pretty unambiguous to me, as stupid a criterion as it is. If Rule 478 didn't define what it means to publish something I could be persuaded that readily-available and referenced counts as published. I don't think it was a stupid criterion for the intended use of the rule. The idea was presumably to allow simple conditional votes such as FOR if proposal 5400 passed or whatever, which are indeed just based on published information. Basing it on contracts doesn't seem to have been an intended use-case of that rule, and IIRC the Vote Market was the first contract to try to define a new complex vote. -- ais523
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
Wooble wrote: On Thu, Oct 2, 2008 at 12:15 PM, Kerim Aydin [EMAIL PROTECTED] wrote: My point is not that it's true now and needs a fix (though a clarification is always useful) my point is that it's ridiculous to interpret the *current* rule as excluding readily-available information (as long as it's *referenced* at least indirectly by the publication in question). -Goethe published during the voting period seems pretty unambiguous to me, as stupid a criterion as it is. If Rule 478 didn't define what it means to publish something I could be persuaded that readily-available and referenced counts as published. It's only stupid when taking contract-defined shorthand into account, which is a non-trivial stretch from the original idea of conditional votes with an explicitly-stated condition. Even then, one could reasonably argue that knowledge of the contract-defined shorthand is implicitly allowed, in the same way that knowledge of standard English is implicitly allowed.
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2008-10-02 at 09:42 -0700, Ed Murphy wrote: Wooble wrote: On Thu, Oct 2, 2008 at 12:15 PM, Kerim Aydin [EMAIL PROTECTED] wrote: My point is not that it's true now and needs a fix (though a clarification is always useful) my point is that it's ridiculous to interpret the *current* rule as excluding readily-available information (as long as it's *referenced* at least indirectly by the publication in question). -Goethe published during the voting period seems pretty unambiguous to me, as stupid a criterion as it is. If Rule 478 didn't define what it means to publish something I could be persuaded that readily-available and referenced counts as published. It's only stupid when taking contract-defined shorthand into account, which is a non-trivial stretch from the original idea of conditional votes with an explicitly-stated condition. Even then, one could reasonably argue that knowledge of the contract-defined shorthand is implicitly allowed, in the same way that knowledge of standard English is implicitly allowed. Rule 754 explicitly allows knowledge of standard English, and of the rules. It doesn't allow knowledge of contract-defined terms. By the same an explicit MAY implies MAY NOT in all other cases that we have in the rules (via the definition of regulation), I can only conclude that there is no implicit allowance in voting conditions. -- ais523
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, ais523 wrote: On Thu, 2008-10-02 at 12:30 -0400, Geoffrey Spear wrote: On Thu, Oct 2, 2008 at 12:15 PM, Kerim Aydin [EMAIL PROTECTED] wrote: My point is not that it's true now and needs a fix (though a clarification is always useful) my point is that it's ridiculous to interpret the *current* rule as excluding readily-available information (as long as it's *referenced* at least indirectly by the publication in question). -Goethe published during the voting period seems pretty unambiguous to me, as stupid a criterion as it is. If Rule 478 didn't define what it means to publish something I could be persuaded that readily-available and referenced counts as published. You missed the critical first part of that rules sentence. Information. INFORMATION. Information is *not* merely the words in the message, it is something that informs. If you publish (during the voting period) a clear and adequate reference to something that may be outside that period, but is reasonably available to the other players during the voting period, you are publishing information during the voting period which clearly allows the result to be resolved. If, as you claim, you don't allow *any* references to outside material, you'd have to publish a dictionary every voting period. And a grammar guide. And maybe a kindergarten curriculum. Clearly absurd even under the *current* Rule. (Note: I'm not taking a position on whether in this particular instance the information is adequate and available, I'm just pointing out the folly of taking information to be only the character strings in the message with no context or reference). -Goethe (Number 2).
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, comex wrote: On Thu, Oct 2, 2008 at 12:49 PM, Kerim Aydin [EMAIL PROTECTED] wrote: Information is *not* merely the words in the message, it is something that informs. If you publish (during the voting period) a clear and adequate reference to something that may be outside that period, but is reasonably available to the other players during the voting period, you are publishing information during the voting period which clearly allows the result to be resolved. If I send a message to all players as well as a-d saying that Y = 4, and then conditionally vote only if Y=4, what happens? Whether Y equals 4 can be reasonably determined by all players from information published during the voting period. But I don't think the intent of R2127 was to allow that sort of thing. All relevant information should have to be published to a-b or a-o and stored in the a-b or a-o archive. That's why you put all those reasonably and adequately words into judicial standards or the clarifying legislation so that on a case-by-case basis so each case can define expectations. For example, I'd say that the above wouldn't work if Y=4 was buried in an obscure corner of an old a-b or a-o post, but would work if the vote provided a link (in the a-b vote) to an a-d post where it was clearly written. Case-by-case. Indeed, what if the Agoran decision is private, and I publicly announce that I vote FOR conditionally if Goethe privately voted FOR? With this interpretation, the truth or falsity of the condition can be reasonably determined /by the vote collector/ from information published during the voting period. That's entirely new ground so probably best to wait for private votes to start doing these case-by-case. But for part of this, one could argue was that a vote isn't truly public unless the information to evaluate it is equally available to all voters, who all have an interest in counting votes. To this end, up above, I mentioned that a minimal standard might be that it was information adequately available to *players*, not just the vote collector (I thought about that while writing it). Also btw, this private/public case was partially covered in the AGAINT case. -Goethe
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, ais523 wrote: On Thu, 2008-10-02 at 09:49 -0700, Kerim Aydin wrote: If, as you claim, you don't allow *any* references to outside material, you'd have to publish a dictionary every voting period. And a grammar guide. And maybe a kindergarten curriculum. Clearly absurd even under the *current* Rule. No you wouldn't, because of rule 754. See my appeal argument to CFJ 2203. Sorry, that wasn't published during the voting period. -Goethe
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, ais523 wrote: Rule 754 explicitly allows knowledge of standard English, and of the rules. It doesn't allow knowledge of contract-defined terms. By the same an explicit MAY implies MAY NOT in all other cases that we have in the rules (via the definition of regulation), I can only conclude that there is no implicit allowance in voting conditions. And was the SLR published within every voting period? Otherwise by your rules you can't refer to it. -Goethe
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2008-10-02 at 10:12 -0700, Kerim Aydin wrote: On Thu, 2 Oct 2008, ais523 wrote: Rule 754 explicitly allows knowledge of standard English, and of the rules. It doesn't allow knowledge of contract-defined terms. By the same an explicit MAY implies MAY NOT in all other cases that we have in the rules (via the definition of regulation), I can only conclude that there is no implicit allowance in voting conditions. And was the SLR published within every voting period? Otherwise by your rules you can't refer to it. -Goethe Rule 754 is more powerful than rule 2127, and they contradict each other. Rule 754 wins. -- ais523
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, Oct 2, 2008 at 1:12 PM, Kerim Aydin [EMAIL PROTECTED] wrote: And was the SLR published within every voting period? Otherwise by your rules you can't refer to it. -Goethe Note that the Rulekeepor's obligation to post the SLR weekly would be satisfied if e published, for example, on Monday and then the next Friday, allowing a voting period starting on Wednesday not to contain it. -- hopefully minor evil
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, Oct 2, 2008 at 1:15 PM, ais523 [EMAIL PROTECTED] wrote: Rule 754 is more powerful than rule 2127, and they contradict each other. Rule 754 wins. hmm.. Rule 683 is more powerful than rule 2127, and they contradict each other (because R683 requires that the voter clearly identify which option e selects). Rule 683 wins, all conditional votes are impossible. I'd submit a CFJ on it, but I think there's some precedent about that. -- hopefully minor evil
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, ais523 wrote: On Thu, 2008-10-02 at 10:12 -0700, Kerim Aydin wrote: On Thu, 2 Oct 2008, ais523 wrote: Rule 754 explicitly allows knowledge of standard English, and of the rules. It doesn't allow knowledge of contract-defined terms. By the same an explicit MAY implies MAY NOT in all other cases that we have in the rules (via the definition of regulation), I can only conclude that there is no implicit allowance in voting conditions. And was the SLR published within every voting period? Otherwise by your rules you can't refer to it. -Goethe Rule 754 is more powerful than rule 2127, and they contradict each other. Rule 754 wins. Where's the conflict? R754 allows abbreviations, dialect etc. and outside references (dictionaries, etc.) R2127 allows that information be published, which may include R754 abbreviations, dialect, etc. The standard is lack of ambiguity. If you were arguing that a vote was unclear because it was unclear or ambiguous in the way it used an abbreviation, all well and good. But I'm not going to support the idea that an abbreviation is automatically forbidden because an aspect of its definition was published outside the voting period. -Goethe
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, Oct 2, 2008 at 11:22 AM, comex [EMAIL PROTECTED] wrote: On Thu, Oct 2, 2008 at 1:15 PM, ais523 [EMAIL PROTECTED] wrote: Rule 754 is more powerful than rule 2127, and they contradict each other. Rule 754 wins. hmm.. Rule 683 is more powerful than rule 2127, and they contradict each other (because R683 requires that the voter clearly identify which option e selects). In what situation would R2127 allow a vote without a clearly identified option? -root
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2008-10-02 at 10:26 -0700, Kerim Aydin wrote: If you were arguing that a vote was unclear because it was unclear or ambiguous in the way it used an abbreviation, all well and good. But I'm not going to support the idea that an abbreviation is automatically forbidden because an aspect of its definition was published outside the voting period. Yes, I think I agree with you here. I've come to the same conclusion as you. I think our reasoning may be different, though. (I'm thinking that clear abbreviations must be allowed, because the rules and custom support that, and unclear abbreviations must not be allowed, because the rules and custom don't allow those.) -- ais523
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, comex wrote: On Thu, Oct 2, 2008 at 1:15 PM, ais523 [EMAIL PROTECTED] wrote: Rule 754 is more powerful than rule 2127, and they contradict each other. Rule 754 wins. hmm.. Rule 683 is more powerful than rule 2127, and they contradict each other (because R683 requires that the voter clearly identify which option e selects). Rule 683 wins, all conditional votes are impossible. R2127 gets around it by specifically defining what clearly identify means in conditionals. It's one of those definitional workarounds. -Goethe
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, ais523 wrote: On Thu, 2008-10-02 at 10:26 -0700, Kerim Aydin wrote: If you were arguing that a vote was unclear because it was unclear or ambiguous in the way it used an abbreviation, all well and good. But I'm not going to support the idea that an abbreviation is automatically forbidden because an aspect of its definition was published outside the voting period. Yes, I think I agree with you here. I've come to the same conclusion as you. I think our reasoning may be different, though. (I'm thinking that clear abbreviations must be allowed, because the rules and custom support that, and unclear abbreviations must not be allowed, because the rules and custom don't allow those.) Well, we're agreed then... I think unclear *anything* (dialect, abbreviation, synonym, etc.) isn't generally allowed, and all the real issues with this vote are because here it's unclear what unclear means (as opposed of course to times when things are clearly unclear). -G.
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2008-10-02 at 10:47 -0700, Kerim Aydin wrote:
On Thu, 2 Oct 2008, ais523 wrote:
On Thu, 2008-10-02 at 10:26 -0700, Kerim Aydin wrote:
If you were arguing that a vote was unclear because it was unclear or
ambiguous in the way it used an abbreviation, all well and good. But I'm
not
going to support the idea that an abbreviation is automatically forbidden
because an aspect of its definition was published outside the voting
period.
Yes, I think I agree with you here. I've come to the same conclusion as
you. I think our reasoning may be different, though. (I'm thinking that
clear abbreviations must be allowed, because the rules and custom
support that, and unclear abbreviations must not be allowed, because the
rules and custom don't allow those.)
Well, we're agreed then... I think unclear *anything* (dialect, abbreviation,
synonym, etc.) isn't generally allowed, and all the real issues with this
vote are because here it's unclear what unclear means (as opposed of course
to times when things are clearly unclear). -G.
This reminds me of an edit war on Wikipedia over whether a particular
project page should be tagged with {{disputedtag}} or not
({{disputedtag}} is placed on a page to state that there's a dispute
over whether the page is a policy or guideline). In other words, there
was an argument (which got really quite heated) as to whether there was
an argument or not.
http://en.wikipedia.org/wiki/Wikipedia:LAME#Wikipedia:Spoiler for more
information.
--
ais523
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, Kerim Aydin wrote: On Thu, 2 Oct 2008, ais523 wrote: On Thu, 2008-10-02 at 10:26 -0700, Kerim Aydin wrote: If you were arguing that a vote was unclear because it was unclear or ambiguous in the way it used an abbreviation, all well and good. But I'm not going to support the idea that an abbreviation is automatically forbidden because an aspect of its definition was published outside the voting period. Yes, I think I agree with you here. I've come to the same conclusion as you. I think our reasoning may be different, though. (I'm thinking that clear abbreviations must be allowed, because the rules and custom support that, and unclear abbreviations must not be allowed, because the rules and custom don't allow those.) Just a followup ais523, would you agree with the following statement? For the purposes of R2127, if information published in the same message as a conditional vote and/or directly associated with a conditional vote contains a clear abbreviation that is generally understood by most players or a clear and direct reference to secondary material that is generally easily available to players during the voting period, that secondary material may be used to clearly resolve the conditional vote, regardless of whether the secondary material was published during the voting period, as the information published within the voting period clearly refers to the secondary material and makes it available, thereby making it a substantive part of the published information. -Goethe
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2008-10-02 at 11:15 -0700, Kerim Aydin wrote: Just a followup ais523, would you agree with the following statement? For the purposes of R2127, if information published in the same message as a conditional vote and/or directly associated with a conditional vote contains a clear abbreviation that is generally understood by most players or a clear and direct reference to secondary material that is generally easily available to players during the voting period, that secondary material may be used to clearly resolve the conditional vote, regardless of whether the secondary material was published during the voting period, as the information published within the voting period clearly refers to the secondary material and makes it available, thereby making it a substantive part of the published information. I wouldn't agree with it unconditionally, although it's right most of the time. I think it would depend on how generally understood, or how clear and direct. For instance, I vote OCTAHEDRON, where OCTAHEDRON is defined at http://example.com/foo; would be a pretty clear and direct reference to generally easily available material, and I think most players would allow that. However, if the content of the website in question varied during the voting period, it would get a lot more murky. (I am reminded of the flash animation which changed its mind about OPPOSE/SUPPORT); this sort of reasoning makes it hard to figure out where the line was drawn. Does anyone know why rule 2127 was created in the first place? I'm wondering if the bar was intentionally set high to discourage that sort of scam. -- ais523
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, Oct 2, 2008 at 2:35 PM, ais523 [EMAIL PROTECTED] wrote: where the line was drawn. Does anyone know why rule 2127 was created in the first place? I'm wondering if the bar was intentionally set high to discourage that sort of scam. The archives show that Goethe originally proposed it requiring the information be published within 30 days prior to the end of the voting period, then retracted that when Michael suggested that 30 days was too long, and proposed the current form (minus the endorsing stuff which was added recently).
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, ais523 wrote: Does anyone know why rule 2127 was created in the first place? I'm wondering if the bar was intentionally set high to discourage that sort of scam. I wrote it, because I thought it would be fun to allow just the sort of activity that's now going on (sell tickets, endorsements, etc., deals, betrayals, etc. etc.) And of course some scams. And I think it worked, and most uses have been very legitimate, enjoyable and sometimes unexpected. It's meant to be reasonably strict to prevent paradoxes and limit scams but mainly to prevent excessive work on the part of the Assessor (e.g. not require massive searching of archives for the Y=3 that was posted three years ago, and prohibiting things like if BobTHJ has cast an odd number of votes in the last 3 years[strictly speaking available to anyone], FALSE, otherwise TRUE). So it wasn't meant to prevent direct and obvious references to a contract (or even an official report) just because the report was published a week before the voting started. A good legislative clarification that would get the original intent better would be replacing information published during... with something like information referenced in the voting message and readily available and resolvable by any reasonable player without requiring unreasonable effort, between the end of the voting period and the resolution of the decision... That would leave the courts to define squishy terms like readily available and reasonable and unreasonable of course. -Goethe
Re: DIS: Re: BUS: Another reason those SELL votes might not have counted
On Thu, 2 Oct 2008, Geoffrey Spear wrote: On Thu, Oct 2, 2008 at 2:35 PM, ais523 [EMAIL PROTECTED] wrote: where the line was drawn. Does anyone know why rule 2127 was created in the first place? I'm wondering if the bar was intentionally set high to discourage that sort of scam. The archives show that Goethe originally proposed it requiring the information be published within 30 days prior to the end of the voting period, then retracted that when Michael suggested that 30 days was too long, and proposed the current form (minus the endorsing stuff which was added recently). Oh yeah, I knew I had included a provision for at least some past information like recent reports (30 days was so that monthly reports were generally included). So what's preferable, a hard (but longer) limit like 30 days, or my just-now proto with squishy reasonably available and clearly identified? -G.
Re: DIS: RE: BUS: Another reason those SELL votes might not have counted
On Wed, Oct 1, 2008 at 3:47 PM, Alexander Smith [EMAIL PROTECTED] wrote: Yep, I got the timing wrong, and I've already admitted my mistake. (That's during or close to the period of time during which emails to [EMAIL PROTECTED] got held up for random lengths of time and arrived in random order.) It's still arguable, though, that you can make a conditional vote and only define what it means later; that's what I'm trying to establish with my TETRAHEDRON experiment. Don't be absurd. The meaning of the conditions were not defined after the fact. That text has been in the contract for months. -root
Re: DIS: RE: BUS: Another reason those SELL votes might not have counted
ais523 wrote: Yep, I got the timing wrong, and I've already admitted my mistake. (That's during or close to the period of time during which emails to [EMAIL PROTECTED] got held up for random lengths of time and arrived in random order.) And that's what happens on my end when (as you probably suspect by now) I respond to earlier e-mails before reading the later ones. It's still arguable, though, that you can make a conditional vote and only define what it means later; that's what I'm trying to establish with my TETRAHEDRON experiment. I don't see why not. It seems functionally equivalent to saying I intend to vote on this later, and then later voting normally (including with a defined-at-the-same-time condition).
RE: DIS: RE: BUS: Another reason those SELL votes might not have counted
Murphy wrote: I don't see why not. It seems functionally equivalent to saying I intend to vote on this later, and then later voting normally (including with a defined-at-the-same-time condition). Just wait until you see the definition of TETRAHEDRON, then you might change your mind. -- ais523 winmail.dat
Re: DIS: RE: BUS: Another reason those SELL votes might not have counted
On Wed, Oct 1, 2008 at 6:09 PM, Ed Murphy [EMAIL PROTECTED] wrote: I don't see why not. It seems functionally equivalent to saying I intend to vote on this later, and then later voting normally (including with a defined-at-the-same-time condition). Not if the definition comes after the end of the voting period. -- hopefully minor evil
Re: DIS: RE: BUS: Another reason those SELL votes might not have counted
On Wed, Oct 1, 2008 at 4:11 PM, Alexander Smith [EMAIL PROTECTED] wrote: Murphy wrote: I don't see why not. It seems functionally equivalent to saying I intend to vote on this later, and then later voting normally (including with a defined-at-the-same-time condition). Just wait until you see the definition of TETRAHEDRON, then you might change your mind. If you're trying to do some sort of combined vote/action like the SELL votes, I expect the vote would be successful (provided the condition is able to be determined one way or another), but the action would not be, for lack of definition. -root
Re: DIS: RE: BUS: Another reason those SELL votes might not have counted
comex wrote: On Wed, Oct 1, 2008 at 6:09 PM, Ed Murphy [EMAIL PROTECTED] wrote: I don't see why not. It seems functionally equivalent to saying I intend to vote on this later, and then later voting normally (including with a defined-at-the-same-time condition). Not if the definition comes after the end of the voting period. Then it would fail per R2127 (paragraph 3).
