This morning was my first San Francisco tornado warning. I was once again thankful to Patrick Blindauer, Rebecca Young, and St Louis puzzlehunt team WisBros for teaching me what to do in a tornado: get under something solid. (This morning's warning was before dawn; so I couldn't use the sky color to figure out if it was time to skeedaddle to cover. So I skeedaddled on general principles.)
2024-12-14T14:28:41.562785
I've been buying groceries more frequently, expanding my emergency food-and-water supply from ~three days' worth to ~seven.
My country re-elected a corrupt dimwit. I couldn't control that. He will appoint corrupt incompetents to positions of power. I can't control that. I don't know who he'll appoint to head FEMA (USA's disaster response agency). Probably either a Fox News commentator who doesn't believe volcanoes exist; perhaps an ex-senator under investigation for arson. I can't control that.
Anyhow, I think it's a good idea to expand my emergency food-and-water supply. I'm giving more attention to this simple task than it needs. But it's something I can control, so I'm focusing on it.
2024-12-12T20:17:13.953612
Today was my second consecutive Golden Gate Park walk seeing zero (0) coyotes. This was weird; I've become accustomed to seeing them. I'm not sure what changed. Mmmaybe: I see coyotes when I walk in the park around sunrise; nights are longer now, approaching the winter solstice. Sunrise is late these days; maybe coyotes don't want to stay up this late?
I guess I'll blog about it now; if I have the same question next year at around this time, I'll know that I guessed right.
I'm still learning the system. Ten years ago, I would have surprised-ly blogged about seeing a coyote; nowadays I'm more surprised by their absence.
Anyhow, here's a photo of some park bison. Because I don't know how to take a photo that illustrates a lack of coyotes.
2024-12-07T17:31:38.682500
2024-12-05T15:47:46.559662
It's all about USA policy for electricity generation, transmission, distribution, and retail. It gets complicated; there's some nation-level control, but mostly everything is different from region to region. From place to place, some organizations are publicly traded companies, some are government departments… just because you're an expert in your home town's system, don't assume you know how things work in the next town over, the next state over, in Texas (Texas is weird), and so on.
The book gets into the details. In my imagination, this book got started when the author, an electricity policy wonk, was explaining a spreadsheet of [ generation | transmission | distribution | retail ] organizations with a client who asked "You ever think of writing all this down?" I didn't retain those details; but some vibes remain.
E.g., public vs private power. Should your state/county/city/whatever have a power department or a monopoly power company? In theory, a power company could innovate more, motivated by profit, benefiting consumers. In practice, every time a power company makes a mistake, the state/county/city/whatever regulates the heck out of it. Take out some big loans to build a new thingy, fall behind on the loans, and go bankrupt? The state/county/city/whatever will make good, because they need to keep the electricity flowing. But they'll also apply a ton of new regulations so that it doesn't happen again. Sign up to build a few extra nuclear reactors because you believe that they'll produce power "too cheap to meter" and demand will skyrocket? When the state/county/city/whatever figures out that the power will, in fact, be more expensive than non-nuclear choices, they'll cancel the extra reactors and regulate big new projects. Oh, and the neighboring states/counties/cities/whatevers also impose regulations on their local power companies, because they saw the political disaster unfold and they don't want to get voted out of office.
So… In the long run, you probably won't get more innovation out of a monopoly company than out of a government department; voters will (reasonably) freak out when things go wrong and demand regulation.
Back when nuclear power was new, a lot of places really did sign up to build a lot of reactors, anticipating a lot of new demand that didn't materialize.
Back when nuclear power was new, the feds decided an interesting way to encourage it: if a company built a reactor and the reactor had a big accident, the feds put a cap on the liability. If I'm a citizen, this particular incentive makes me very NIMBY about nuclear reactors. What was the reasoning behind this incentive? If big, horrible accidents are rare, this guarantee doesn't affect anything, so why offer it? If big, horrible accidents are common, then I do not want a reactor near me, near my water supply, near anyone I care about…
Remember Enron? Remember how much you hated Enron? That hate was justified.
2024-12-02T23:07:25.752289
Today, I reached level 100 in Pikmin Bloom, a walk-around-with-your-phone game. That's the highest level achievable.
And Alexander humblebragged: "Oh woe is me, I just don't know what to do with myself now that there are no more worlds to conquer."*
Anyhow, I will soon un-install the app; I'm 🏁 done-zo 🏁.
*This is not accurate: There are more things to accomplish in the game, achievement badges to collect, etc; but I don't feel motivated to go after those. Also, when I looked up that Alexander the Great "quote" about no-more-worlds-to-conquer, I found out he didn't actually say that but rather kinda the opposite.
2024-11-30T16:22:32.521890
I continue to check my little dashboard of San Francisco COVID numbers each morning to figure out whether going into a cafe for professionally-made coffee is worth the risk or will be an embarrassing thing to explain to the medical professionals treating my long-COVID-induced heart condition[s].
This morning, my dashboard is b0rked. Good news: The California Department of Public Health has a new dashboard for COVID-in-wastewater data. Bad news: the old dashboard, whence my dashboard fetched its data, doesn't work anymore. I can't figure out how my automated dashboard is supposed to fetch data now. Ha ha, the timing: I get to find out its broken on USA Thanksgiving, and the relevant smart folks probably won't be checking their email for a few days.
Yay, computers.
This year I'm thankful that San Francisco's other COVID numbers are low; so I don't need to worry too much despite the fact that some data is missing at the moment.
2024-11-28T17:48:27.710715
A few months back, I played in SFPursuit, a San Francisco challenge hunt. Uhm, where "challenge hunt" is a phrase I made up just now for "Something like a puzzlehunt, but most of the activities aren't puzzle-y, but they are still challenging." One of the activities involved looking over a physical book that had been created for the hunt. A few bookstores across San Francisco had one copy each of the special book. Each player's phone directed them to the nearest such bookstore. There, they could find the special book, gather data from it, solve the book's riddle, and enter the answer on their phone to complete the challenge.
Thus I found myself in Fabulosa Books, negotiating with fellow hunt enthusiasts to snap pictures of the pages from a false birdwatching guide. Unsurprisingly to anyone who's participated in such events, everyone was very focused on solving the riddle; nobody bought any books.
I remembered back to when I helped run 'Terngame 2012 for Twitter interns. I'd set up one puzzle at Isotope Comics and then watched over the puzzle, in case any interns needed help solving it. All of the interns were hyperfocused on the puzzle; none of them dawdled in the store after solving to browse or buy. I felt pretty sheepish for having asked the store's propietor to let me set up the puzzle there. I might as well have used a conference room back at company HQ.
That's what was on my mind when I bought a book at Fabulosa. Oh? What's that you say? What did I think of the book?
In the novel The Fellowship of Puzzlemakers, the protagonist goes on a quest of self-discovery. He's been given a crossword puzzle; he's been instructed to seek out the answers to the puzzle along his journey. So one of the answers might be BEAUTY and the protagonist might encounter a really-nice sunset and contemplate its BEAUTY. And everyone reading the book is nodding "Yep, I saw that BEAUTY when I first solved the crossword, and now here we encounter it in the story; this whole story is satisfyingly interwoven."
The puzzle's kinda weird, tho. The book takes place in England, most of the characters are English. But the central-theme crossword puzzle uses American-style definition clues, not British-style cryptic clues. The book explains the reason thusly: The protagonist doesn't think of himself as smart enough to solve cryptic clues; the quest-giver knew this, and wanted the puzzle to be solve-able. But but cryptic clues are just a small part of why British-style puzzles can be difficult. A much bigger problem is that their grids are "sparse"; not every square is part of of an across-word and a down-word. In an American-style puzzle, if you don't know the across answer, the first name of whazzerface Świątek you might still be OK: Maybe you'll know the down-answers that intersect. But in a British-style grid, even if you have all the intersecting answers, you don't have all the letters.
Maybe you're looking at S□A□E and you're not sure whether it's SCALE, SHADE, SHAKE, SHAME, SHAPE, SHARE, SLATE, SNAKE, SPACE, SPARE, STAGE, STATE, or what-have you. If the clue is British-cryptic-style, you have some hope: the clue has two halves: one definition, one wordplay. If the definition is so vague that you're not sure whether the answer is SHAPE or SPACE, probably the wordplay will clear things up. (oh, a "change" of phase, ok)
So… our not-amazing-puzzler protagonist has to solve a crossword in a British-style grid and doesn't have an answer key, so he has to just hope that he picked the right words, that □E□T□E might be CENTRE or KETTLE…
Also, one of the crossword clues is wrong: it points you at a noun, but the answer is the verb form of that noun. It's weird that one clue was wrong. In a novel structured around a crossword, I either expect no mistakes (a puzzle nerd trying their hand at novel-writing) or many mistakes (a novelist who doesn't know the rules of crosswords but nevertheless wants to incorporate one).
Here's my best hypothesis: The author's first draft of this novel used a British-style crossword with cryptic clues. The author was delighted when an American publisher picked up the novel: His work would reach a big audience! But the publisher imposed a condition: Those cryptic clues had to go; you couldn't expect Americans to understand them. The author, gnashing his teeth at this slight to his culture, swapped out his cryptic clues for grudgingly-written defintion-clue replacements, not noticing that he'd messed one of them up.
Oh? What's that you say? What did I think of the book?
It was OK. It did a great job of structuring the story around the crossword. If the crossword had been a mistake-less cryptic, maybe I'd be calling this book masterful.
2024-11-20T21:59:34.077030
Much of the internet runs on ads; but maybe it's a house of cards. Apps, web pages, other content-thingies show ads to defray their costs. Advertisers want to show their messages to potential customers, and pay to do so. Everybody wins.
Except it doesn't always work so smoothly. A small business might want to show their ads on news sites and phone games and Google searches and other places; but maybe doesn't want to keep track of how their ads are doing at all those different places. So maybe the small business signs up with a couple of intermediary services, each service promising to show the ad to some desired demographic. Maybe everything's working fine? Maybe it's not? It's hard to tell.
The small business starts getting billed for ads. Thanks to creepy surveillance, that phone game thinks it knows its players' demographics, and thus will show the ad to the right people. But what if it has the demographics wrong? Or what if the intermediary service messed up and asked for the wrong demographic? Or what if the intermediary service is crooked, never actually places any ads, just sends a plausible-looking bill? Or what if…
It turns out, a lot of this stuff is pretty opaque. Publishers and advertisers have to take a lot on faith. We know ads aren't always shown to the right people, but we don't know how often. We know there's fraud out there, but we don't know how much.
The book's author compares the situation to the subprime mortgage meltdown: We know there are some bad assets out there. We know the rating system has incentive to be lazy about seeking out that badness. Is it all about to fall apart? Maybe. That could be darned bad news for news sites; even if the advertising markets only crash for a short while, it might be long enough for shoestring news organizations to go under.
2024-11-15T23:47:32.464406
November 5, 2024, I once again volunteered as a poll clerk for a San Francisco election. I clerked at the same garage as 2022, and it was still swanky: building owner Jay served up espresso drinks, Italian soda, donuts, snacks.
The crew was:
I mention that Victoria returned. Another returner: Kayden, our 2022 high school student volunteer. But he wasn't clerking this year; he came to vote.
Something new happened this year: We had a paper jam.
This ballot, physically, was four big sheets of thick paper. (thin cardstock?) Voters marked their sheets with a pen, tore receipt strips off the tops, then fed the big parts of the sheets into a scanner to be counted. But but one voter botched tearing the little receipt-strips off. They had ragged edges and tore into the ballot sheets. When they fed a sheet into the scanner, things didn't go great. The scanner pulled in the sheet, made some unhappy whining noises, stopped, and displayed an error message on screen.
I was standing next to the machine when this happened. I knew what to do next: We have a little booklet that explains the machine's mysterious error codes. Error 1 means such-and-such went wrong, the fix is this; Error 2 means this-other-thingy went wrong, the fix is this; etc. So I flipped through the little booklet; it didn't have an explanation for this error message. Fortunately, Poll Inspector and fearless leader Christina knew that our San Francisco Poll Worker manuals contained a second list of error messages and remedies. This list knew about the error message we were looking at: Yep, it was a paper jam. There was a remedy: Press a button to exhort the scanner to expel the jammed sheet. Alas, pressing this button caused the machine to give forth no sheet, but instead more whining noises.
We called up the Department of Elections phone hotline to find out what to do next. A nice Dept-of-E person asked for the error message, asked us to press the expel-jammed-paper-button again, and when that didn't work told us to wait for a technician to come visit and clear the paper jam. Until the technician showed up, we were to tell voters that they could either wait for the technician, or could hand over their ballots to be stored in The Auxiliary Bin: a compartment on the side of the ballot scanner. After the technician unjammed things, one of us could feed those stored ballots into the scanner.
So I, the clerk standing next to the scanner, had the not-so-fun conversation with voters: Unfortunately, we have a paper jam. You can't feed your ballot into the scanner right now. You can wait for the technician, or I can put your ballot in The Auxiliary Bin for now and scan it when the technician has fixed things.
To me, this felt sketchy. I knew that historically, sometimes poll workers are corrupt. Maybe that clerk by the scanner is lying about the "paper jam". Maybe if you, the voter, hand over your ballot to be scanned "later," that clerk will look over your choices, frown, and "lose" your ballot. (I asked many people to hand over their ballots. I wonder how many of them grew up in places with "elections" like this?) San Francisco has safeguards against this corruption: poll workers watch each other; an audit will catch sites that hand out more ballots than get scanned; probably other safeguards, too. Nevertheless, asking voters to just hand over their ballots felt bad. And this was during the morning rush—many people wanted to scan their ballots and then head to work. Now they had to pause and make a decision.
Most voters handed over their ballots without thinking about it. Some voters peered at me, thought, and handed over their ballots. A few hung on to their ballots. I found these few ballot-hanger-on voters comforting. They were keeping things honest.
Eventually, the technician showed up. You, experienced at clearing paper jams from printers might wonder: Do they really need a specialized technician to clear a simple paper jam? But this was an election, and this technician was, basically, tampering with election equipment. Most of what he did was keeping records. He stayed on the phone with someone back at Dept of Elections office, telling them "OK, I'm going to break the security seal. I'm opening up the panel." He spent a lot of time making sure nobody (including him) was committing election crimes, very little time clearing out the paper jam itself.
Then I took on my second sketchy task: Feeding those Auxiliary Bin-stored ballots into the scanner. Anyone walking into the polling place would have seen me hunched over a stack of ballot sheets, picking them up and feeding them into the scanner. It must have looked pretty suspicious. I kind of hope that some voter called up the Dept of Elections to report my actions.
A few days later, I ambled along a local retail street, seeking a sandwich. I spotted the voter who'd torn the ballot that started the whole mess. They spotted me at the same time; looked away; seemed embarrassed. I hope they weren't too embarrassed. I hope they keep voting, maybe hope they ask for help tearing off the strips next time. Nobody was born knowing how to do it, and I don't think standard high school Civics classes cover papercraft.
2024-11-11T16:00:09.569051
Archives:
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024