Checking in Before Sasquan Thursday Business Meeting

The coffee and baked goods setup adds a highly civilized touch to meeting room 300 CC and there’s plenty for those here since the SRO audience Kevin Standlee worried about has not materialized. At least not today. More later, with actual news content…

117 thoughts on “Checking in Before Sasquan Thursday Business Meeting

  1. Mike, they generated a demo already. It’s using dummy data, to keep it really easy to follow. There are links here.

  2. @Cassy B: Several attempts to derail EPH were voted down.

    @rcade: Please note that Ron Oakes had just become aware of B.2.4 a week ago when he posted that comment to Kevin’s blog. He may not have yet been aware that it’s an extremely mildly worded proposal that would merely express the WSFS Business Meeting’s recommendation of open-source licensing to Worldcon committees, and has no mandatory effect whatsoever.

    Proposers Kate Secor and Ben Wolfe might have had an easier time with B.2.4 if they’d written it to recommend viewable/auditable source code rather than open source, as the latter’s definitional requirement to permit any party to fork covered code doesn’t sit well with many programmers for a variety of reasons. (I speak as someone who works with open source software for a living, and helps the Open Source Initiative evaluate licences.)

  3. Mike — they’ve already developed the code, and run tests with the 1984 data. They’ve got a demo — a whole PowerPoint they’ve prepared to demonstrate how it works, and would have NO trouble showing how it would work with this year’s nomination data if given access to it — and it won’t take them 19 hours to do it.

    EPH has been independently tested by several people, and if you really want the blow by blow account, may I recommend reading the EPH threads at Making Light. This was worked on by specialists handling voting/election data.

    And I see P. J. Evans has beat me to the punch.

  4. Hi all –

    Just a couple of comments to add to what you’ve seen here. First, my plane doesn’t leave until 3:00, so as long as the meeting ends on time, I’m fine. It’s 25 minutes to the airport, so I could push it if need be. Plenty of others from our group are here if need be, though.

    Yes, we need to be clear that the workload for the admins is not increased — or even changed. They just substitute one nomination counting program for another. The algorithm is easy to code if they don’t want to use what I’ve done (and I’ve been making the source code available for months now, so I’ve got no problem releasing it). We’d have a year to fine tune anything people thought necessary. My version also generates a round-by-round history file specifically for audit purposes. But we do need to make that case.

    The only other argument against that I heard was that we don’t have enough data to decide if there’s really a problem. Data or not, the hole in the rules is still there and I think we can all agree that it can be exploited anytime anyone wants to. I think it’s pretty clear there’s a problem.

    And speaking of data, unless they release ballots (and I don’t know that they will), I don’t see any chance of generating a statistically meaningful ballot set based on this year’s nominations in just a few hours. Jameson might be able to do that, but I just don’t think there will be time. I haven’t talked to the rest of the group about it yet. But I think the results from the Hugos on Saturday will highlight the problem, so I think that’s what some people are referring to. If they release the ballots, I can definitely run then under EPH, though. I’ve got the code with me, so it’s trivial to do. We’ll see, though.

    Kilo

  5. Examination of the source code should suffice. EPH would be complex to carry out by hand, but the algorithm itself isnā€™t so weird that it would be difficult to verify that a given program implements it properly.

    Assuming, of course, that the executable run is the same as that derived from the source code, an assumption too many people are willing to make without due consideration in my experience.

    Not that I’m ever going to admit to having exploited this flaw in human nature for fun or profit, because that would be illegal.

  6. Bloom seems to really want to postpone everything. Is he someone I should know are just a fan with an opinion?

  7. we donā€™t have enough data to decide if thereā€™s really a problem

    The last couple of years aren’t enough for someone? Or any of the things that previous years have noticed? (I do hope that’s someone who simply doesn’t pay attention.)

  8. Meredith! Glad to read you again, and I hope your absence was not due to harrowing life events.

  9. I wish someone had/would made/make the point that this talk about gathering data is not valid. You decide what to do based on collected data only in cases where you do not know the underlying system. In this case, where the underlying system is fully understood mathematically, the existence of a problem can be determined analytically; actual sample data just isn’t required.

    It’s like writing your email password in large letters on a public wall where everyone can see it; you don’t have to wait until several people log into your email and send nasty emails to your boss, or even one. You can determine through analysis that there’s a hole in the system and patch it up.

  10. Any idea if the attempts at derailing EPH were pup-driven? Or is it just the normal fannish cat herding process?

  11. I think it would be a nice proof-of-concept to be able to show how EPH would have treated this year’s nomination ballots. I know that may not be possible for one reason or another, and if it doesn’t happen I won’t be mad. But it would be great to show “see? Here a slate locks everyone out of the ballot–and with EPH you can see the slate gets this one slot, leaving four slots (or three or whatever) for the honest favorites of the other 85% of the nominators” with recent nominating data in which a slate was known to be operating.

    I think it would put the finishing touch on making the argument convincing. But it may well be that the shambles the slate makes of the Hugo Awards will do all the convincing necessary.

  12. So as I understand it 2 Year Eligibility has been postponed indefinitely, but Best Series and 4/6 go for a full debate?

    Hmm. I’m glad that 2YE is out, I don’t think it made much sense to drastically increase the nomination pool. I still have no idea about the other 2 though, though with 4/6 I do have concerns that people might view it as an acceptable compromise between EPH and doing nothing, without quite realising that it’s value as a mitigant is severely limited and specific.

  13. @CPaca

    Assuming, of course, that the executable run is the same as that derived from the source code, an assumption too many people are willing to make without due consideration in my experience.

    Not a problem if you compile the executable yourself from the source code. Or if it’s implemented in an interpreted language like Java or Python, such that there isn’t an executable per se.

    Or what Keith or Mark said. Really need to learn to read the whole thread before commenting . . .

  14. Is it possible to include a motion to release nomination ballot data in the near future from as many years as possible to study EPH’s effects over the next year?

    Being a programmer, I know despite all the explanations of how it works under the hood and running hypothetical sample data through it, some people need to see it side by side with stuff they are familiar with (beyond just one year unfortunately a while back). Much more grounded for them that way.

  15. @Mike Glyer: Morris: Iā€™m sitting here taking notes for a post while you are liveblogging in the comments. Should I shoot you or shake your hand? The latter I have decidedā€¦.

    @rrede: You canā€™t do either unless thereā€™s a second.

    I move to Postpone Indefinitely.

  16. Snowcrash, second the motion to postpone indefinitely the motion to shoot Morris and/or shake his hand….

  17. P. J. Evans: On EPH, I did read all that material, some of the material at Making Light, all the comments on it here, and you shared the 1984 data with me. What’s more, I am inclined to vote for EPH because it appears well worked out, capable of diluting the power of slate voting, while keeping those voters on an equal footing with everyone else.

    My desire not to disclose the 2015 Hugo data set (or that of any other year) is not, then, because I am dissatisfied with EPH. I do not want to set a precedent of sharing the transcribed ballots outside the committee, or a precedent of a concom doing such a thing to facilitate one party’s rules change proposal, and I am also concerned that the data release will inform future hacks on the Hugos, not just efforts to protect them.

  18. “1141: Member tries to raise amendment to motion thatā€™s been sent to committee. His punishment for this breach of conduct is being put on the committee.”

    obSF – Reminds me of Parliament in The Moon is a Harsh Mistress. šŸ™‚

    Who is Rachel Acks? Loved ‘TEA FOR THE TEA GOD!!!’

  19. @CPaca : Assuming, of course, that the executable run is the same as that derived from the source code, an assumption too many people are willing to make without due consideration in my experience.

    Jon F. Zeigler : Not a problem if you compile the executable yourself from the source code..

    Still a problem if the target doesn’t (i) check the source code and (ii) recompile it each and every time they run the program. And most users can’t. It’s a REALLY good idea to consider the security in who has access and can overwrite files on a computer, especially in the small user environment. Speaking completely hypothetically, of course.

    An interpreted language would be my preference too.

  20. CPaca–

    Not that Iā€™m ever going to admit to having exploited this flaw in human nature for fun or profit, because that would be illegal.

    Heh.

    @Kilo —

    The easy answer to that is have the admins compile their own executable from the source codeā€¦

    But how do you guarantee that the compiler wasn’t hacked to recognize the EPH algorithm and insert bogus code? We must go deeper…

    cf The Ken Thompson Hack. I remember stories from the 90’s about compiler vendors adding code to their compilers to recognize the Dhrystone benchmark and inserting hand-optimized code. (Didn’t find a good link in the amount of time I was willing to spend).

    Disclaimer: I’m just being perverse! I strongly doubt that the Hugos are worth this degree of effort when there are so many higher-value targets available. Besides, this applies to all vote-counting software, not just EPH.

    @John F. Ziegler–

    Or if itā€™s implemented in an interpreted language like Java or Python, such that there isnā€™t an executable per se.

    Java source code isn’t directly interpreted. It is compiled to byte code which is then interpreted by the Java VM (Virtual Machine). Byte code can be thought of as the instruction set of the JVM. Some JVMs (JITs) further compile the byte code into machine code, i.e. the instruction set of the real hardware (note to other geeks: please don’t bring up microcode unless you want peoples’ heads to explode).

    The Python systems I’ve used do something similar. The .py file is compiled to a .pyc file (compiled python) which is then interpreted by a VM. The .pyc is executed unless the .py source file is newer.

    And a general comment–

    Don’t underestimate the amount of work EPH will involve.

    The time it takes to code a prototype is the tip of the iceberg. Even if the core algorithm is pretty simple, a tiered web app with a DB isn’t. Production quality code with error handling and logging is often several times bigger than prototypes, and test code can be even bigger than the application code. Good testing is a lot more involved than normal people may think.

    With all the contention, this is going to need good testing with data sets designed to hit corner cases, not to mention audits.

  21. Ken Marable: As I understand it, the Business Meeting has no authority over the Hugo Administrator as to whether they release voting data or not, short of putting it into the WSFS Constitution (which would, of course, take at least two years). The most they can do is respectfully ask them to release data, but it’s entirely up to the Admin if they actually do.

  22. @P J Evans–

    Rick K, why would it need to be a web app?

    I’m just repeating Kilo’s comments at Making Light:

    If EPH passes, we’ll need a web database app for the real system. I’m not qualified to write that type of code, but for demonstration proposes (which I think may be what you’re looking for), what I have works fine.

    Original comment.

  23. Can’t see any reason for it to be a webapp. Unless you want to integrate it with the system for entering votes.

  24. @SJW75261: Kent Bloom was among many other things chair of Denvention 3, 2008’s Worldcon. Suffice to say, he has deep reserves of cred on WSFS matters; anyone with common sense will listen carefully. (In this case, I did and then respectfully disagreed.)

    @Ken Marable: Please be aware that the Business Meeting has always (in my experience) voted to destroy site-selection ballots upon results getting certified and announced. Of course, the Business Meeting could decide to make anonymised full 2015 data available before destroying ballots, but that would need to be discussed and voted. Anyone wanting that to happen should be working right now on detailed wording for the required motion and how to convince attendees.

    @Wanderfound: Ben Yalow would make a singularly unlikely member of Faction Canidae.

  25. @Wanderfound: 100% herding cats. There was one guy sitting next to me muttering about how EPH was “a fake!”; he was the only plausible Puppy there, and even he seemed more crank than Puppy.

    Ben Yalen is a cool guy–45 years of Worldcons, former Parlimaentarian (and very helpful to young folk like Kate Secor who want to propose new things), and very aware of, and unhappy w/, Vox Day. But he takes a very long view (one additional year to study won’t make a difference in the long run, in his mind); he also believes that several of the Sad Puppy voters were operating in good faith, and that’s important that WSFS feel open to all of fandom. I disagree with his positions, but he’s coming from a very good-faith place.

  26. I watched several sections of the business meeting. I would guess that for the old hand “parliamentarians” making those maneuvers at the preliminary stage is a way to set out one’s case clearly and be on the record with one’s objections. I mostly agree with the things Ben Yalow and Kent Bloom said, of course.

    Moving the debate on 4/6 and EPH to Sunday was a terrible idea. Folks won’t have had time to think objectively about any new information, and some might come to the meeting angry about whatever the outcome was. It is harder to think clearly when you are angry.

  27. And holding the vote early allows people to think that there isn’t a real problem. Heads Vox wins, tails, we lose.

  28. @NickPheas
    Both the mess and EPH have been highly visible for a while now. So I would expect most people willing to educate themselves and make a rational decision (as opposed to the equally valid gut decision) to have done so already. Anyone else has enough time to gather information over the weekend.
    The way I see it, the only people for whom this years results matter crucially to their vote on EPH will the the one’s who would have gone with their gut feeling anyway. And for them, it’s very useful to have the information that is most relevant to their vote before they vote.

  29. Nice little blog text about the puppies latest war front:

    “Looking at it this way, an associated group of writers has made an effort to get their stuff on the ballot in order to increase both their prestige and their sales. This may be correct, but if so, the main effect has been to expose the writing and the writers to criticism, some of it literary and some of it otherwise, leading to episodes of trolling and generally bad behavior that reduces their standing in the community. Likely some of the Puppies entered into the deal with this idea.

    On the other hand, there have been a number of complaints about how this battle has ā€œbrokenā€ the Hugo Award system. We can figure this is from people who have a lot invested in the awards, for example, past and aspiring winners, their editors, publishers, agents, etc. This suggests that some of the Puppies actually have the aim of destroying the Hugo Award system in order to break the hold of the group maintaining it as a literary award. This would reduce its prestige value and level the playing field a bit. Nominating lower quality and offensive works would be a part of this strategy. As I understand it, this is the split between the Sad and the Rabid Puppies led by Vox Day.

    Looking at Hoytā€™s statement that this is about getting some ā€œfunā€ stuff on the ballot, I didnā€™t see anything especially fun. Leckie, Heuvelt and Monette came closest, but thatā€™s the literary stuff. Looking around, I canā€™t find any reviews that say of how fun these nominees are. Reviews for a couple of the pieces actually say ā€œavoid.ā€ Applying the scientific method, this means I have to reject this claim. So, I conclude there may be an actual attack on the Hugo Awards.”

  30. I disagree with his positions, but heā€™s coming from a very good-faith place.

    Yalow may be acting in good faith, but his attempt to prevent an EPH vote this year prior to debate by pushing it to a committee would have been a black eye for the Hugos. Thousands of people are familiar with the slate controversy. The EPH proposal has been publicly drafted for months and received many comments in feedback.

    Even if Yalow does not support it, he should have let it get a fair hearing this year when the integrity of the nominating process has been attacked. As a listed proponent of EPH and a Hugo voter for six years, I want to see that WSFS is taking the idea of protecting our vote from slate manipulation seriously.

    Win or lose, I’d like to see the business meeting show that it is evaluating EPH on the merits and not buying into FUD like what Brian Z was spreading here for weeks.

  31. Proposers Kate Secor and Ben Wolfe might have had an easier time with B.2.4 if theyā€™d written it to recommend viewable/auditable source code rather than open source, as the latterā€™s definitional requirement to permit any party to fork covered code doesnā€™t sit well with many programmers for a variety of reasons.

    Since it’s just a recommendation, I’d prefer it to favor open source. Open source is a familiar concept to many people today and something that builds confidence in the integrity and reliability of software. The concept of “viewable/auditable” source is more vague.

    Would a programmer who doesn’t like her code to be open like it much better if people interested in the Hugos could ask to view it?

  32. Would a programmer who doesnā€™t like her code to be open like it much better if people interested in the Hugos could ask to view it?

    Quite possibly, yes. “Open source” is generally understood to mean that the software is licensed with one of a specific set of licenses, all of which have the effective result that other organizations would also be free to use the software. I could very well imagine a programmer who’d be very happy for WSFS to use their software for free, and to inspect and audit and even edit it any way they’d like – but who wouldn’t want other organizations to be able to do the same (for free).

    I’ve watched all of the business meeting videos, btw, and am amazed (and a little bit scared) by how much I enjoyed the whole procedure…

  33. @LunarG

    Nothing dramatic! Just some boring health/government bureaucracy stuff on top of the usual stacked up just enough that the spoons I typically assign to having a social life were sacrificed to the cause of not living in squalor. I’m trying to claw them back now – and I couldn’t miss this weekend. The house will cope somehow I’m sure. šŸ™‚

  34. Greg,

    Ben Yalen is a cool guy . . . he also believes that several of the Sad Puppy voters were operating in good faith, and thatā€™s important that WSFS feel open to all of fandom.

    I know it’s an unpopular view here, but I agree with him on those two points.

    One of the Sad Puppies major complaints is that a small group of insiders have been imposing their tastes on the Hugos over the past N years. One of the major complaints against the Sad Puppies is that slate voting allows a small group of voters to disenfranchise the majority. EPH addresses both of these concerns (which, of course, represent the same underlying issue). Everyone involved in this kerfuffle should support EPH.

  35. Keith “Kilo” Watt,

    And speaking of data, unless they release ballots (and I donā€™t know that they will), I donā€™t see any chance of generating a statistically meaningful ballot set based on this yearā€™s nominations in just a few hours. Jameson might be able to do that, but I just donā€™t think there will be time. I havenā€™t talked to the rest of the group about it yet. But I think the results from the Hugos on Saturday will highlight the problem, so I think thatā€™s what some people are referring to. If they release the ballots, I can definitely run then under EPH, though. Iā€™ve got the code with me, so itā€™s trivial to do. Weā€™ll see, though.

    If you need more technical assistance over the next few days, don’t hesitate to ask. Since I’m not schmoozing at Sasquan, I probably have more free time available this weekend than you do.

  36. I don’t even know if it’s that unpopular a view. I think he’s right. Some of the sads were operating in good faith, some of them are the classic ‘Useful Idiots’.

    Was Seanan McGuire’s five nominations a sign of a perfectly functioning system? Probably not. Is trying to exclude anyone who likes Seanan McGuire’s works from the shortlist an appropriate response? Not really. Is inventing a stream of insults for anyone who likes Seanan McGuire a sign of good faith? No freaking way. Do the people who didn’t want Seanan McGuire to win five Hugos deserve to be squeezed out entirely? No, and I seem to remember that she only got the one, and that was part of a podcast collective.

  37. @rcade:

    Since itā€™s just a recommendation, Iā€™d prefer it to favor open source.

    Well, if you were in sole charge of the Business Meeting, that would settle it, then. ;->

    The concept of ā€œviewable/auditableā€ source is more vague.

    Au contraire, I meant something highly specific, and greatly easier for Business Meeting regulars to understand: program code being open to inspection. (And, if you think open source is a familiar concept to Business Meeting regulars, I beg to differ on the basis of long acquaintance.)

    Would a programmer who doesnā€™t like her code to be open like it much better if people interested in the Hugos could ask to view it?

    Wrong question. Would a programmer who doesn’t want his code to be able to be freely adapted to any purpose by any party without his permission and without fee — which is what the Open Source Definition denotes — be nonetheless OK with Hugo voters having the ability to view its source code? I have not asked Ron Oakes about the nature of his concern, for example (though i saw him a few times yesterday, and didn’t want to bother him), but many coders distrust the lack of personal control inherent in OSD-compliant licensing.

    I fear that the proposal as written will be voted down quickly as we work through this year’s crowded agenda, which is a shame, doubly so because its core aim could be achieved with a small edit.

  38. @Patrick May

    I know itā€™s an unpopular view here, but I agree with him on those two points.

    I don’t think it’s unpopular at all. By and large, as I’ve experienced, the consensus has been that if the Puppies of any stripe have work that enough people are willing to vote for, there’s no reasons that they shouldn’t get nominated and even win. The issue has been, from the start, the fact that they voted on a slate and intentionally exploited a system that they knew ran on goodfaith tradition of voting in order to defranchise the rest of the people involved.

    As for good faith of Puppies voters, I’m sure many did see it as works that they felt were worthy and that didn’t get the attention or respect that they should. The caveat is that I don’t for a moment believe that the Puppies Leaders had any other agenda other than what happened.

Comments are closed.