Filho is really good at setting a scene and immersing you in a time and place, and just as I was with the titular town in Bacurau, I was fascinated by The Secret Agent’s recreation of Recife during Brazil’s military dictatorship in the late 1970s. There are so many details and characters, and I didn’t mind that many of them are superfluous to the plot. I also didn’t mind that the plot really takes its time—80 minutes in, we still only have an inkling of what’s going on. That first half’s self-assured assembly put a spell on me, but it was broken in the back half, when a variety of frivolous tangents are given priority, the central conflict’s murderous animosity is only vaguely justified, and its most consequential moment is shunted off to an afterthought of an epilogue.

• Tagged: July 2026, film diary, film and television, review, Kleber Mendonça Filho
• Reply via email

what emotional malfunction makes me so relish being brutalized in this way

• Tagged: July 2026, film diary, film and television, review, Oliver Laxe
• Reply via email

• Tagged: July 2026, politics
• Reply via email

With the exception of Hollow Man, Paul Verhoeven’s studio sci-fi extravaganzas are all slathered in sociopolitical satire, and this one is arguably the most overt, with Verhoeven, recalling his childhood in the Nazi-occupied Netherlands, saying in an interview:

If I tell the world that a right-wing, fascist way of doing things doesn’t work, no one will listen to me. So I’m going to make a perfect fascist world: everyone is beautiful, everything is shiny, everything has big guns and fancy ships, but it’s only good for killing fucking bugs.

Starship Troopers’ fascists sure do kill a lot of huge alien bugs—and vice versa—and there are more than a few not-so-subtle hints that the militarized autocratic utopia they’re fighting for is not such a great place. But the film has always felt lopsided to me, its focus on action spectacle diluting its satire. If the runtime of maybe 20–25 percent of the frenetic, hyper-violent battle scenes had been instead devoted to additional ideological world-building, I think that might have been a better balance.

That aside, this is the first time I’ve seen this movie in a long time, and I was impressed by how well most of the digital effects hold up.

• Tagged: July 2026, film diary, film and television, review, Paul Verhoeven
• Reply via email

And now we’ve lost Jack too soon. Fifty-fucking-one years old. Glioblastoma.

Jack was loudly in attendance both times I competed at the US Air Guitar National Finals in Chicago, and we’d hang out most other times I was in town, too. Same when he was in NYC. Every time, we’d try to make arrangements to go to concerts together, usually without success. My attempts to visit Jack after he entered hospice in late May were likewise unsuccessful, and it’ll be a while before I can forgive myself for that. But we texted a bit, just like we sometimes texted each other from concerts.

I looked for him in the crowd while watching the livestream of Low’s set at Pitchfork in 2022. “My earplugs are getting a workout,” he said. That was Mimi’s second-last performance before she succumbed to her own cancer a few months later. Jack added the show to his concert history on Setlist.fm. Can I tell you what a treasure that is? He was the only other person I’ve known who shared my unreasonable degree of Low fanaticism, and being able to retrace his steps with the band and compare them to my own is so special to me. Cult fandoms, religious resentments, paternal regrets—the similar paths we walked were ones of great significance to both of us.

Geography never gave us the chance to be as close friends as we might have been, and there’s still so much about Jack I don’t know, but I think we nevertheless shared an uncommon understanding, and I’m grateful for that. I don’t subscribe to any of the usual conceptions of an afterlife, but if there is one, I hope Jack’s hanging with Mimi and his dad, and giving the Holy Trinity many, many notes.

Rest well, Jack.

• Tagged: July 2026, design, music, personal, Low
• Reply via email

After finally completing the gargantuan task of manually adding metadata to thousands of my old tweets so they could be more integrated with the rest of my site, I wanted to keep the momentum going. So I turned to the last large personal data source that was still missing from my site: Last.fm, a service for tracking and sharing the music you listen to. As of a few days ago, I’ve been a Last.fm user for 21 years, and it’s rare for me to go even a day or two without interfacing with it in some way, making it perhaps my most-used internet thing of all time, even if most of that usage happens in the background.

In late 2005, Last.fm merged with Audioscrobbler, which was the service I originally joined, and so sending listening data to Last.fm is called “scrobbling.” To date I’ve scrobbled over 150,000 tracks from over 8,000 artists. Given my propensity for using my site to document my music purchases and concert-going, I was naturally eager to give my listening data a home here as well. But the breadth and depth of that data were intimidating. It’s fun to explore my Last.fm profile as a sprawling database, but what shape would that information take on my blog, where all my various online activity is oriented within a timeline? Making each track an individual post would be ridiculous, and even posts collecting a day’s or week’s listens would make a lot of clutter. Ultimately I decided that breaking it all up into monthly posts was sufficiently granular: When looking at one of my blog’s month archives, knowing what I was listening to at the time might add some interesting context to whatever else I was up to then.

But the volume of data was still an issue. Did the hundreds of tracks I listened to each month all need to be accounted for? Did the three minutes I once spent with an Osmonds song really warrant an Osmonds creator tag? Nope! Collating each month’s top 10 artists and top 10 albums seemed like a reasonable compromise, and for added context, I’d list the month’s total number of artists, albums, tracks, and plays. When I got wistful about the ton of data I’d be leaving on the cutting room floor, I reminded myself that the data isn’t comprehensive anyway, as I’ve listened to plenty of music over the last 21 years that for one reason or another never got scrobbled. And anyway, I’d still have the data and I could always do something more with it down the road when Last.fm inevitably goes the way of the dodo.

With all that decided, I started building some new site infrastructure to accommodate the incoming data, including overdue improvements to my bar-chart styles. When I started, I still wasn’t sure how I would even get all the data, as Last.fm doesn’t let you just download it, and I still don’t really know how to query an API. But less than a day after mentioning the project to my buddy Jon, who previously helped me with my Letterboxd data, I had a huge JSON file with everything I needed, plus one more addition to the pile of reasons I like Jon very much. As with past data excursions, I spent some time cleaning it up and hacking with Node.js to get everything neatly formatted into Markdown files, and less than a week after I started the project, my listening diary was born.

I’m glad to have it. Some entries are documents of how I self-soothed during a challenging time. Many make space for artists I kept coming back to who didn’t previously have a presence on the site, often because I streamed their music without writing about them or buying any of their records or seeing them live. Hopefully going back through the listening diary will help me start to rectify that. The diary also gives further emphasis to some artists who were already established favorites, sometimes in sad moments when they passed away. And I’m sure over time it’ll prove itself valuable in other ways. If you click around and find anything interesting, please let me know!

• This post is part of a series
• Tagged: July 2026, design, music, projects
• Reply via email

• Tagged: June 2026, listening diary, music, Sweepers, The Magnetic Fields, YHWH Nailgun, Horse Lords, Mastodon, Upsammy, Valentina Magaletti, Widemouth, Hoavi, Microwaves, Intensified Chaos, Pest Control, Morphine
• Reply via email

My first Tortoise show since I very belatedly learned that Doug McCombs’s distinctive secret weapon is the unassumingly mighty Fender Bass VI.

• Tagged: June 2026, concert diary, event, music, Tortoise, Evan Chapman, Upper Merion Township Building Park, King of Prussia, PA
• Reply via email

One web browser default I’m seldom satisfied with is the appearance of list markers. Their style can be customized with CSS using li::marker, but the options there are limited. The best way I’ve found to fully customize them is to remove default list styles and use a li::before pseudo-element as a list marker:

ol,
ul {
  list-style: none;
}

li::before {
  content: "•";
}

I’ve used a plain old bullet in this example, but content’s value can be whatever you want, and you can also use position, background, and other properties to customize the list marker to your heart’s content.

This works great for unordered lists (ul), whose markers are all typically uniform. But for ordered lists (ol), whose alphanumeric markers are variable and context-specific, something more is needed. Enter CSS counters, which are essentially numeric variables CSS keeps track of and increments and decrements as needed. You can use CSS counters in lots of ways with pretty much any HTML element, but I’ve only ever used them with ordered lists, beginning by instantiating a counter using counter-reset and giving the counter a custom name, like list-position:

ol {
  counter-reset: list-position;
}

Next, I’ll tell list-position, whose default value is 0, to increment with each list item and make the list item’s ::before content whatever the current list-position value is:

ol li::before {
  counter-increment: list-position;
  content: counter(list-position);
}

For the first item in the list, list-position’s value will be 1; for the second item, the value will be 2; and so on.

This approach does the trick for most ordered lists, allowing you to add styles to li::before to customize the list markers however you want. Great! But note that most lists is not the same as all lists, and this particular implementation of CSS counters isn’t bulletproof. Say your ol has a reversed attribute and is meant to count down rather than up. Or it has a start="2" and is meant to count up from 2 rather than from 1. The above CSS doesn’t account for those cases and will still just count up from 1. CSS counters have additional capabilities that can address these issues, but it means writing even more CSS to make a hacky recreation of a thing HTML already does by default. There has to be a better way!

Well, it turns out there is! I just today learned I can forego counter-reset and counter-increment entirely, and in place of the content reference to the counter I had named list-position, I can just use an existing CSS keyword, list-item, which the counter will increment or decrement automatically according to the needs of the list. Boom.

ol li::before {
  content: counter(list-item);
}

I still wish ::marker could handle all my ordered list styling needs, but as compromises go, CSS counters aren’t bad, and simplifying them further with the discovery of list-item has brightened my day. Here’s my new list-styling boilerplate:

ol,
ul {
  list-style: none;
}

ul li::before {
  content: "•";
}

/* Top-level ordered lists get standard
decimal markers, starting with "1" */

ol li::before {
  content: counter(list-item, decimal);
}

/* Nested ordered lists get lowercase
alphabetical markers, starting with "a" */

ol ol,
ul ol {
  & li::before {
    content: counter(list-item, lower-alpha);
  }
}

/* Further nested ordered lists get lowercase
Roman numeral markers, starting with "i" */

ol ol ol,
ol ul ol,
ul ol ol,
ul ul ol  {
  & li::before {
    content: counter(list-item, lower-roman);
  }
}

Further reading

• Using CSS counters
• An in-depth guide to customizing lists with CSS

• Tagged: June 2026, design, code
• Reply via email

Last week? Not my favorite week. My dog injured her leg while playing with other dogs in the park. My basement needed some emergency masonry work. My acid reflux graduated from intermittent warning shots to full-on assault. But somehow, in the midst of all this, I managed to finish up a project I’ve been chipping away at for countless hours over the past 10 months: manually adding metadata to the nearly 6,000 old tweets I’ve republished on my site.

The tweets have been on the site since my redesign launched last September, but any metadata beyond what was provided by Twitter’s data export wasn’t originally part of the republishing plan, since adding it was obviously going to be an enormous hassle. After the launch, however, it really bugged me that my blog’s numbers didn’t tell the whole story. For each topic, creator, and location listed, there was an untold number of relevant posts left uncounted.

So I started counting them. One tweet at a time, looking for anything taggable within my existing taxonomy. Any of my limited set of topics relevant to the tweet, tagged. If I tweeted about a place, the location was tagged. If I quoted or otherwise mentioned a song or movie, the musician or filmmaker was tagged. If I linked to an article, the author was tagged. (It’s worth noting that of those links, the New York Times deserves credit for being the only mainstream site to almost completely avoid link rot, and for most of the rest of the author names, the Internet Archive was reliably miraculous.) I also took the opportunity to stitch together many series of tweets that would have already been threaded if they hadn’t predated Twitter’s threading capabilities.

For me, this much fuller accounting of the site’s contents validates this laborious cataloging endeavor, especially for the 1,100 creators who were previously untagged. Seeing the updated topics numbers all together provides a decent hierarchy of my public preoccupations, and apparently I spent enough time tweeting about my dreams to allow pretty much anyone to cobble together a workable psychological profile. Perhaps most importantly of all, Dick Clair and Jenna McMahon, the creators of the 1980s sitcom The Facts of Life, have finally received their flowers.

• This post is part of a series
• Tagged: June 2026, design, projects
• Reply via email