Guilt if not denied, not guilt on accusation

Once again, many New Zealanders are saying incredibly incorrect things about copyright law – most of them because they are parroting other people in a twisted Chinese whispers style of journalism. People are calling the latest changes “guilt on accusation”. All you have to do is actually read the amendment (which is not something that many do) and you can tell that is not the case.

The relevant part of the amendment is this:

122MA Infringement notice as evidence of copyright infringement

(1) In proceedings before the Tribunal, an infringement notice is conclusive evidence of the following:

(a) that each incidence of file sharing identified in the notice constituted an infringement of the right owner’s copyright in the work identified:

(b) that the information recorded in the infringement notice is correct:

(c) that the infringement notice was issued in accordance with this Act.

(2) An account holder may submit evidence, or give reasons, that show that any 1 or more of the presumptions in subsection (1) do not apply with respect to any particular infringement identified in an infringement notice.

(3) If an account holder submits evidence or gives reasons as referred to in subsection (2), the rights owner must satisfy the Tribunal that the particular presumption or presumptions are correct.

What this says is that if you are accused and you do not deny the accusation then the tribunal does not need to consider additional evidence – they can get on with the rest of their work. As soon as you say “no, this wasn’t a rights violation”, or “there’s a mistake here”, or “this notice wasn’t in the right form”, the burden is explicitly on the rights holder to prove guilt. In case this is “TL;DR”, it’s the final words: “the rights owner must satisfy the Tribunal that the particular presumption or presumptions are correct.

This is doesn’t mean you are guilty until proven innocent – this means that if you can’t be bothered denying the accusation, then the Tribunal doesn’t have to waste a lot of time looking at evidence. If you’re innocent, all you need to do is point that out, and you are presumed to be correct until proven otherwise.

Possible gmail break-in

When I logged into gmail this morning, I saw the message that I dread the most – detection of an unusual access. There was a connection (two days ago) from a server in Malaysia, although it’s actually an AWS server (Amazon web services).

There are two possibilities: the good one is that this is something that I’ve previously given access to my account, accessing it via an alternate method (e.g. Backupify can access my mail to back it up, and they use AWS) so that it showed up an unusual. The bad one is that someone was using AWS to bulk-attack accounts and got in.

In favour of the good one, as far as I can tell, no email was sent – I can’t see anything amiss at all. The email account is the central lockbox for everything, of course, so it’s possible that it was just used to break into other things, or the email content was retrieved. My password (changed now, of course) was a random 8-character string of lower-case alphanumeric characters, so not particularly simple to break (although not difficult either, given sufficient resources). I never give out the password to anything that I do not completely trust, and nor do I give out access via other methods (e.g. oauth, openid) unless I trust those services too.

I had intended to turn on two-factor identification, but hadn’t got to it yet. I’ve done that now, for the main account at least. My password is now over 30 characters long, including upper and lower case and punctuation – I probably should have changed this a while ago too.

For now, I’m leaning towards the good possibility, so I won’t be completely resetting everything that can send a password reminder to my gmail account. I’ll be keeping an eye on things as closely as I can in the next week or so, though. If you see anything suspicious come from me, please let me know.

More pitching (scifi.stackexchange.com)

The elevator pitch discussion died down, but I’ve been thinking about this over the last few weeks. As I noted earlier, I’m not a great pitcher, but perhaps I can get something good enough together that others can then work on.

DVD extras meets user generated content.

I like the idea of starting the pitch with a mashup of something that the pitchee is likely to be familar with (harking back to Donaldson’s thought that all good ideas come from the collision of two separate ideas). In addition, “user generated content” is fairly buzzword-y, which I generally dislike, but is probably appropriate for a pitch. I’m referring only to the best DVD extras, of course, but I think that is implied (also the best of user generated content), and “Book extras” if there was such a thing, as well.

Imagine if you could set the questions on the DVD extras for your favourite TV show or movie – or if you had extras for your favourite book. You get answers about what, how and why things happened both in-universe, and in reality, from people who are intimately familiar with not just that one work, but the entire science fiction and fantasy genres – people that can pull together expert and interesting answers about how this work relates to other genre fiction and to the world. They’ll even explain what stories you should move to next if you loved particular aspects of this one.

This doesn’t include story identification – but the pitch doesn’t need to include every topic, and it doesn’t fit with the “DVD extras” analogy. It hits some marks that I think are particularly important:

It’s not just about focusing in on one universe – it’s about having knowledge across the entire genre/genres.
It’s about the in-universe world, but also about how the fiction impacts reality.
It punches the question words “what”, “how”, and “why” (“who” is probably better answered by IMDB, “where” probably by Wikipedia, and “which” could cover too many things). This emphases (subtlety) that this is a Q&A site (as does “answers” later on), but also what sort of questions are most appropriate: especially “why”.
DVD extras are generally narrated by experts (cast, crew, authors). The site isn’t necessarily going to have the foremost expert (e.g. the author) for every question, but it is about getting expert opinion.

It does include recommendations (although I try to make it clear that they need to be very specific). My opinion follows the original meta discussion: as long as they are specific enough to invite good (subjective) answers, then they’re ok. Actually, they’re not just ok, but the type of question that users will really love the site for. I’m sure many people will add books to their reading lists by reading interesting, detailed, answers on the site – not just these ones, but certainly including them.

Figuring FAQ (scifi.stackexchange.com)

The scifi.stackexchange.com community (or more accurately, the meta.scifi.stackexchange.com community) is still trying to figure out what’s on-topic, even though it doesn’t appear that the meta consensus directly influences the actual reality of the site.

When I last left the search for an elevator pitch, I wondered whether the FAQs of the other (launched) sites would be a fertile ground for inspiration. So, here goes – this is the same list of sites as last time:

Web apps is fairly straightforward (although they interestingly single out “adult content” sites as off-topic). Gaming is nice and short, with an all-inclusive policy (with two exceptions: recommendations and shopping). Ubuntu’s FAQ barely says anything about what’s ok – I guess the implication is that anything related to Ubuntu is on-topic. Webmasters is similarly short, with no exclusions. Game development has a brief list of sub-topics that are considered acceptable, and an explanation of how to choose between StackOverflow and that site.

The three that I think do the best job (in terms of something that scifi.stackexchange.com can emulate) are photography, cooking, and mathematics.

Photography is an interesting case – they link to a few meta discussions, and they have some off-topic examples that seem obvious (programming, website development, graphic design) but must have caused problems at some point.

Cooking reads very nicely – there are clear examples of what’s on-topic, and some examples of what’s not on topic. There’s a link to questions tagged “faq” on meta (scifi.stackexchange.com has used “on-topic-discussion” for the same purposes I think).

Mathematics doesn’t just have on-topic and off-topic suggestions, but also suggestions for topics that are on-topic but might get better answers elsewhere. I think this is a great addition. The off-topic examples are quite limited, but it’s probably quite obvious what’s ok on the site.

Does this help with figuring things out for scifi.stackexchange.com? Not as much as I hoped. The results of the ‘on topic – off-topic’ meta-meta discussion can probably be turned into the on/off topic bullet points that are common; we should try and include a link to an appropriate meta tag as well, and links to other sites (like writers.stackexchange.com) for some examples would be great too.

Pitching (scifi.stackexchange.com)

There are currently 15 StackExchange sites that have launched (i.e. made it past beta), excluding the original trilogy. The “elevator pitch” questions on their meta sites are:

I hoped to find some inspiration for the scifi.stackexchange.com elevator pitch by examining these successful sites. Unfortunately (or perhaps not?) the majority of these questions focus on “taglines” rather than pitches – i.e. single sentences or sentence fragments that would take under a second to “pitch” (perhaps elevators are much faster in the US! – the blog post does say single-sentence).

It seems like this is mostly because of the history of StackExchange around this time – a decision was made to change from unique domains for each site (e.g. seasonedadvice.com) to generic subdomains of stackexchange.com (e.g. cooking.stackexchange.com) with unique branding. As a result, the focus is typically on the branding – and in some cases is filled with complaints about that decision (maths, for some reason, has a bunch of answers concerned about commercialisation of the site in their elevator pitch question).

So Apple, Unix/Linux, CS-theory, English, TeX/LaTeX, photography, game development, webmasters, Ubuntu, and web applications aren’t really of any inspirational use. There are some short pitches at gaming.stackexchange.com, but not very many.

I’m guessing that another reason that the pitches are so short is that it’s more clear what the site is. If you’re familiar with StackOverflow, then it seems reasonably obvious what apple.stackexchange.com is. This isn’t always the case – for example, is cooking.stackexchange.com only for professionals? (No).

So my hunt must continue. In the meantime, interest in this question on scifi.stackexchange.com has died down a little – which is probably bad, not good. I’m thinking now that the best way to get this resolved is to indeed answer it myself (but community wiki, so that others can improve and especially shorten!), with perhaps one sentence and one paragraph versions.

Perhaps the FAQ’s of the launched sites will be more inspiring – that’s where I’m heading next (I’m also reading through several of the other meta sites, so that might also offer some illumination).

Dealing with recommendations (scifi.stackexchange.com)

One of the problems that scifi.stackexchange.com faces that isn’t unique to that site is “recommendation” questions. These are a specific type of list question (which are generally ill-suited to the sites), where each answer offers one possible recommendation, and the votes cast are not just “that’s a good answer” or “that’s a bad answer”, but votes for (and less often, against) a suggestion. StackOverflow has a long history of problems with these sorts of questions (e.g. “favourite programming cartoon” and “great programming quotes“). With few exceptions, these questions are closed, often not before they gather huge numbers of votes, answers, and answer votes.

I don’t really see a huge problem with these questions myself – if they are “community wiki” then they aren’t just a way to gather huge amounts of reputation, the voting has merit (if something is popular, it is likely for a reason), and the format does fit (unlike discussions, for example). I can understand that a specific site might choose to disallow them (e.g. StackOverflow), but others (e.g. programmers.stackexchange.com) could allow them. I don’t think that having the majority (or even just a non-small percentage) of a site’s questions be ones of this type would do much for the quality, but in moderation they seem ok.

I doubt this would ever be allowed by the powers that be at StackExchange, but I think the following would be a great system for scifi.stackexchange.com:

Once a month a new meta.scifi.stackexchange.com question is opened that asks what this month’s recommendation (or perhaps more generally, poll) question should be. People make suggestions and up/down vote the ideas that they like (yes, a poll about a poll – however, poll questions seem to be acceptable on meta sites).
One a month, the most popular answer of the previous month’s meta question is created (e.g. by a moderator, ensuring that the question is community wiki).

I think this strikes a great middle-ground. These types of questions are generally incredibly popular, and that would help get people coming to the site (something that scifi.stackexchange.com needs, but others, e.g. StackOverflow, do not, since it’s already at near saturation). Since there would only be one of these each month (others would be close-voted, with comments pointing to the relevant meta discussion) they wouldn’t dominate the site (even one a week would presumably be only noise compared to the number of other questions). Many more users would find themselves interacting with the meta site – hopefully some of these would explore more than just the one question they came for, and end up participating in community building, support, moderation, and so forth – perhaps these users would otherwise not have ever visited the meta site.

Other arguments against these types of questions tend to be: they age badly (this could still be true, but hopefully the pre-asking meta stage would assist with that, and there’s always down-votes), they provoke discussion (discussion answers/comments can be flagged, discussion in chat is a good thing), and they don’t provide interesting answers (I totally disagree – especially in the context of a Q&A site dedicated to fiction).

The Internet was built by sci-fi fans (scifi.stackexchange.com progress report)

scifi.stackexchange.com has been in beta for a couple of weeks now, and has gone through quite a few struggles. Although I’ve been a StackOverflow user since the very early days (2 years, 4 months now), this is my first experience with a StackExchange site in the beta stage – it’s possible that all of the sites have these troubles, although it does seem like they are particular to a topic like “science fiction” more than more concrete topics like mathematics, cooking, or theoretical computer science.

Most of the issues boil down to two: how should the moderators interact with the site, and exactly what is on-topic.

In an effort to save the site from its community, the (SE employee) moderators have chosen to take a very hard line. Rather than let the community police itself, they’ve decided that it’s better to act as quickly as possible to remove any questions that aren’t suitable (in their opinion) and hope that if there’s any disagreement that it will be take up on meta. This policy does appear to be working, so perhaps that’s enough validation. In my opinion, letting the community close the bad questions would be better in the long term (and only slightly slower in the short), and the moderators’ time would be better spent educating the community leaders in the meta site. Hopefully, this will become less of an issue as the more major question of what’s on topic is answered, and pro tem moderators are elected from the community.

I’m not sure exactly how much of the graduation of a site from definition to commitment to beta in the Area51 system is automated. It’s possible that no human was involved in the progress. I think if there was, that would help – the majority of the example “on topic” questions for the site are, in fact, deemed off topic, and it seems likely (given their comments) that the StackExchange staff would have been able to point that out before the site opened. Perhaps the 5 “on topic” examples for each site could be examined when “commitment” is reached, and if the StackExchange staff can see questions that just won’t be acceptable in the StackExchange model, the proposal can be bounced back into definition (with an explanation).

The much more pressing issue is to decide what’s on topic for the site. There seem to be some clear-cut decisions: questions about writing are out, questions that could be answered by a basic check of IMDB are out. There are many more that are unclear. There’s a meta question asking to define the “elevator pitch” for the site, which will help a lot. Unfortunately, the answer that’s voted highest at present isn’t one that I think would make an interesting site (and it’s certainly not one that I’d be interested in spending much time on). There’s another answer that’s much closed to a good site – I think one of the main weaknesses of that answer is that it’s not “pitchy” enough. I’ve considered having a go at a pitch myself, but the pitch has to be concise (by definition), and conciseness isn’t a strength of mine, which means that even if I got the topics right, it wouldn’t be a good pitch.

What I’m hoping to see scifi.stackexchange.com become involves these sorts of questions:

Story identification. These have gone through a rough ride: one of the (SE employee) moderators mistakenly thought that these were outright-banned, when in fact they are merely held to a higher standard than average questions. This lead to a lot of confusion that was only resolved in the last couple of days. It’s clear now that the community can choose to allow them (again, assuming they are well asked), although it’s not clear whether everyone agrees that they should be. I think they will form an invaluable and essential part of the site. Reading the questions and the answers is an excellent way to find new material (without opening up poll/recommendation questions), and is the way to entice new users into the site/community. Once these questions are indexed, identifying the material will be simple via a search engine, as well.

“Trivia”. The problem is that it’s hard to define “trivia”; literally it’s something of small importance – this is of course, extremely subjective (the majority of StackExchange questions are probably of small importance to most people – that’s what addressing the long tail is all about). Questions that only allow for uninteresting answers clearly aren’t of value (although uninteresting is again subjective). The voting system provides a good mechanism of determining this – especially combined with reputation and tags. A good metric is that good answers are generally long answers. I do think it’s possible to have something that some people would call “trivia”, but be long and interesting.

Answers found elsewhere: one of the reasons that StackOverflow was such a success was that there was a huge vacuum that it filled. This isn’t so much the case with a science fiction Q&A site (although hopefully there is some vacuum). Practically every Science Fiction TV show, movie, or popular book series has some sort of wiki site dedicated to it, usually filled with encyclopedic quantities of information about everything to do with that series. For those few that don’t have their own site – and those that do – there’s Wikipedia entries on everything as well. The Internet was built by scifi fans – this shouldn’t be a surprise. I hope that the eventual consensus is that there’s no way to draw a line here, and questions have to be judged on their own merit. If you can figure out the answer to your question with a single Google search and skimming a wiki page, then there’s no real value duplicating that content. If you have to read through a lot of wiki pages, or they’re really hard to find, or if you need more conjecture or opinion than a wiki page allows, then that should be ok. I would like to see these questions left alone, and (e.g.) if they get go below -1 votes, then vote to close.

Something in the source material: related to the previous, obviously. Here you can’t find the answer online (or at least it’s very difficult to), but if you had the source material (the book, the TV episode, the movie), then you’d be able to figure it out. I think these are borderline, too. What if the book is out of print (or otherwise rare), or the TV episode isn’t available on DVD (or purchasable online)? Not everyone keeps everything that they watch/read. I think that makes it too hard to create a blanket rule about this sort of question. If the question’s not good, then it’ll get voted down (e.g. if anyone that has every read the book would easily remember and could explain in a sentence). If it is good, then it’ll get voted up.

Real world: questions that relate sci-fi to the real world are interesting, but to be useful I think they need to be very specific, otherwise it’ll just degenerate into discussion/argument.

Sci-fi for dummies: not what I’d call it on the site, but face it – sometimes the science in the science fiction is complex (especially in ‘hard’ sci-fi). Getting a bit of help figuring a plot point or allusion out seems like a great use of the site.

Industry information: questions about writing are apparently out (go ask on writers.stackexchange.com). However, there’s still industry specific questions that could be useful – although it’ll be hard to ask a truly interesting question (e.g. “how do I contact author X” is answerable by a single link to a website generally, and “what convention is good” is considered bad form on StackExchange sites). I haven’t seen any questions like this yet, but I suspect they will turn up eventually, and perhaps be rare gems.

I think a good elevator pitch would start with: scifi.stackexchange.com: a place for fans of all science fiction, not just one series, to help each other out, with … Perhaps I’ll manage to figure out the rest of that and submit it as an answer in the next few days.

scifi.stackexchange.com

The Science Fiction StackExchange site came out of private beta today. It’s not the first proposal that I committed to that has made it to beta (that was Card/Board games), I’ve found it more interesting (so far) – Card/Board games has so far focused on a lot of games that I have no interest in (and I’m not so interested or have enough time to ask a lot of questions myself).

The current state of scifi.stackexchange is a little worrying – as might be expected, there are a lot of list/opinion/subjective/discussion questions, which aren’t really a good fit for a SE site. A lot of questions are “community wiki”, which reflects this, and that means that reputation is hard to come by for many users. It seems like there might already be a lot of ‘definitive’ factbook-type sites for many of the major scifi stories (e.g. wookiepedia for Star Wars), and there’s little point just duplicating that information (even Wikipedia has a lot of data, presumably because of a sci-fi bias among many of the editors).

However, there does seem to be a lot of potential for the “long tail” type of questions that SE is designed to address. There’s certainly a lot of lesser-known scifi novels/TV shows that don’t have a lot of information about them online. It looks like “identify this book” type questions will be acceptable as well, which definitely seems like it would be valuable (even when coming from a Google search). Overall, I’m hopeful – so go check it out!

I’m still waiting for the Parenting StackExchange site to reach the beta phase. It might end up being a huge mess of subjective opinion, but it might also end up being a truly valuable resource. If you’re interested, commit!

Trac to iCal

One project I work on uses Trac and has a custom “due date” field (it doesn’t really have milestones – updates are more granular). While this is useful, one problem is that I don’t check this Trac instance every day, and so sometimes I’ve missed deadlines because I haven’t noticed that they are due. However, I do check my calendar every day (multiple times a day). It seems like exporting this “due date” value into my calendar will help with this.

This little script creates a calendar that can be subscribed to. I run it once a day (due dates don’t change very often), and have iCal set to update once a day, so it should work fine. If it’s of use to you, use it (no restrictions). Let me know if there are things that can be improved! It requires the Python iCalendar module, which I already use for parsing public holiday data.

#! /usr/bin/env python

import sqlite3
import datetime

import icalendar

cal = icalendar.Calendar()
cal.add("prodid", "-//PROJECT NAME Trac Deadlines//trac.tonyandrewmeyer.com//")
cal.add("version", "2.0")
cal.add("method", "publish")
cal.add("x-wr-calname", "PROJECT NAME Ticket Due Dates")
cal.add("x-wr-caldesc", "Due dates for PROJECT NAME tickets")
db = sqlite3.connect("/trac_location/db/trac.db")
c = db.cursor()
c.execute("select t.id, t.owner, t.summary, c.value from ticket t, "
           "ticket_custom c where t.id=c.ticket and t.status!='closed'")
for ticket_id, owner, summary, due_date in c.fetchall():
    if not due_date:
        continue
    due_date = datetime.datetime.strptime(due_date, '%d/%m/%Y')
    due_date = datetime.date(due_date.year, due_date.month, due_date.day)
    event = icalendar.Event()
    event.add("summary", "PROJECT NAME #%s (%s): %s" % (ticket_id, owner, summary))
    event.add("dtstart", due_date)
    event.add("dtend", due_date)
    event.add("dtstamp", datetime.datetime.now())
    event.add("url", "https://trac.tonyandrewmeyer.com/ticket/%s" % ticket_id)
    event["sequence"] = datetime.datetime.now().strftime("%Y%m%d%H%M")
    event["uid"] = "ticket_due_%s@trac.tonyandrewmeyer.com" % ticket_id
    cal.add_component(event)
c.close()
db.close()
open("/var/www/subfolder/project-name-trac-due.ics", "wb").write(cal.as_string())

Sleep Cycle Alarms for Two

My body clock works well – I almost always wake up each morning on time, feeling refreshed and well rested. I only use an alarm when there’s something I absolutely can’t be late for, and even then I set the time to be the last possible moment because I’ll always wake up before it goes off [1].

If this was also true for Olyvia, there would be no problem. If I wake up before her, she doesn’t wake up – I quietly spend 30 minutes or so going over email, news, FaceBook, Twitter, work, etc in bed, and even if I get up before her it’s unlikely to wake her up. Unfortunately, Olyvia does not wake up well naturally, and so she does use an alarm whenever she needs to get up at a particular time (which is often at the moment).

Recently, she has started using an iOS app to wake her (with an alarm) at the ideal point in her sleep cycle (i.e. the app is helping her do what I do naturally). She says that this works extremely well for her. Unfortunately, I believe it does work, because it’s waking me up too and I feel terrible (I’m clearly not in the right state to be waking up).

It seems a significant flaw in these sorts of applications – whenever you wake up the user, it’s quite likely to be a poor time to wake up whoever they are sleeping with (unless there’s some sort of odd synchronising of sleep cycles that I am unaware of).

However, this seems easy to solve (and if anyone can take this idea and implement it, please do so – free! – so that I can sleep well again). iOS provides simple-to-use APIs for communicating with other local iOS devices. I, like many people, have many iOS devices lying around. If you can solve the “when to wake” problem given a fixed point and a sleep state, it seems likely that solving the problem given two sleep states (presumably overlapping somewhat at times) and a fixed point cannot be too difficult (probably less optimal, but better two near-winners than a winner and a loser).

This would be simple to use – you set the alarm to coordinate with another local device, and when it is ready to wake you up, it confirms with that device first. They negotiate the ideal time given the two sleepers, and the alarms go off simultaneously (I’d set my alarm to silent since I don’t need it). They can communicate with Bluetooth or WiFi, whichever is less likely to fry my brain by having it under the pillow.