Weeknote and links: 2 October 2023
Links at the bottom.
Work #
On Friday I announced that you can pre-order the print edition of Out of the Software Crisis, which should be released 7 November 2023 (or thereabouts).
Pre-orders get a $10 USD discount and:
… if you pre-order the hard-cover edition before October 15, 2023, your name will be added to a list of subscribers and supporters that will be printed in the book. Your support will be recorded there for posterity.
I’ve had a lot of fun working on this. Hopefully that’ll come through in the final product.
Reading #
Complete “meh” week when it came to reading.
Movies #
Focused on remakes this week.
That is, I watched the original movie, followed by its remakes, to get a sense of studios have handled remakes over the years.
Fright Night #
The first was Fright Night (1985) and Fright Night (2011).
This was a good start because the 80s original is a genuinely fun bad movie while the remake is just fun. The original has awful acting, relies a little bit too much on the novelty of its self-awareness—something that has no novelty value today—and only really works if you’re a fan of the Hammer Horror series of vampire movies.
The 2011 version with Colin Farrell isn’t great, but it’s good enough. It follows the original plot pretty faithfully but with better acting and modern aesthetics. David Tennant does a good David Tennant, Colin Farrell feels made for the role, and seeing Anton Yelchin hammers home just how much of a tragedy it was to lose him so young.
Skip the original, but watch the remake if you like a competent vampire movie.
Cat People #
The 1942 original Cat People is a true horror classic. It’s spooky in the right places, uses the black and white lighting to maximum effect, and has a surprisingly nuanced take on the relationship between monstrosity and sexuality—something that most horror movies lack.
Because it adds a second love interest—and one that’s implied to be more sexually active than the first—it makes it much harder for the viewer to interpret Irena’s transformations as a representation of the dangers of female sexuality. Instead it lends itself more to an interpretation that it’s repressed sexuality or non-consensual sexual advances that are harmful. It’s nuanced enough to lend itself to a broader range of interpretations than is common for horror movies.
Thoughtful, well-written, and with genuinely scary moments and a bittersweet ending.
The 1982 remake tries so hard but ends up being a complete mess because the filmmakers almost certainly let their boners do their thinking. Its take on sexuality is a much more simplistic “all sexuality is dangerous” and the screenwriter and director seem to have thought that this take was an improvement on the usual “female sexuality is dangerous”. They seem to have been too excited about the opportunity to film Nastassja Kinski naked to think the plot through. It’s like they forgot they were making a horror movie at times.
And their “bittersweet” ending just misses the mark for me.
Great soundtrack by Giorgio Moroder, though, and a title track by David Bowie is always a good thing. Doesn’t save the movie though.
Invasion of the Body Snatchers #
This is right on the edge of qualifying as remakes but the novel The Body Snatchers by Jack Finney has been adapted four times, spread out over sixty years. They make for an interesting study in how Hollywood has evolved.
The first, 1956 version, is my personal favourite by far, although I can see why some might disagree.
The movie itself is genuinely great. It’s theme or premise is incredibly well drawn: what would it feel like if everybody you knew, everybody in the world, suddenly turned against you. It is a compelling idea and the alien invasion is just a mechanism used to illustrate it. It’s portrayed from the perspective of a small town doctor who sees the community around him change. Friends become strangers. Patients turn to him frightned before themselves being transformed. The pods are only introduced after the theme has been established and serve mostly to hammer home the inevitability of the change. The climactic scene where the main character is shouting at passing cars—original supposed to be the ending—is powerful enough to be called back to by most of the other adaptations.
And once you realise that it was made during the McCarthy witch hunts—one of the USA’s more notable swerves into authoritarianism—it becomes even more relevant to the modern age.
So, off to a good start.
The 1978 movie takes a different tack. It frames the story squarely as a creepy alien invasion and follows the conventions of what is by then an established formula. We see the arrival and first growth of the aliens. None of the characters that get taken over early are well-drawn prior to their “snatching”, which serves to emphasise their role as adversaries. It is well-made and genuinely great. I can see why some would prefer it over the 1956 take as it’s more faithful to the genre. It also, as an indicator of the different attitudes of 1970s Hollywood, both quite unsubtle about making the lead couple an adulterous one (she is married to someone else), and has the unambiguously bad ending that the 1956 director wanted to have but was forced to undercut by the studios.
The 1993 version, directed by Abel Ferrara, is less effective as it’s a by-the-numbers horror movie—complete with a ‘mad person makes incoherent warning at a gas station early in the movie’ scene. This makes sense once you realise that the story and screenplay is by horror-legends Stuart Gordon and Larry Cohen. Decent executions of the basic horror formula is what they do. It’s not a bad horror movie and it even has a few really effective moments, but it’s not in the same tier as the first two.
What’s interesting about the first three movies is that they could, thematically at least, exist in the same universe. The 1978 movie has an overt callback to the 1956 version, which could be taken to imply that some version of those events took place in that world as well. The 1993 version uses the same design and mechanism for the aliens and takes place in an isolated military base, so it could easily be thought of as happening in a variation of the world of the 1978 version.
This makes the first three adaptations complementary in many ways—each with their own take but in a way that builds on what the previous movies attempted.
The fourth, however, is just bad.
Not badly made. It’s slick and has some of the best actors available in 2007. The effects look good, and the cinematography is engaging. Its problem is that it’s constantly sabotaging itself. It begins as an invasion movie, much like the 1978 version. It does this well until it loses interest. Then it becomes a parable about having to conform for survival, leaning much harder into the “behave like them to survive” message than any of the prior versions, before it loses interest in that too. Then it becomes an action-filled near-post-apocalyptic commentary on the violence we do to survive under extreme circumstances until it loses interest again and summarily solves everything by waving a magic wand before ineptly trying to pretend that “maybe humans are worse than the alien infection” was its theme all along.
It’s a muddle. The first three adaptations each have their own reason for being. The first is a great piece of psychological horror. The second is a classic alien invasion move. The third is a competent horror movie.
But the fourth? Skip it. Pick whichever one of the first three fits your interests the best.
Links #
Software dev #
- "The invisible problem – Scott Jenson". “Text editing on mobile isn’t ok. It’s actually much worse than you think, an invisible problem no one appreciates.” Are there actually people out there who think that mobile text editing is ok? If so, then that explains a lot.
- "React is slow at its very core | Go Make Things". “The golden rule of JavaScript: more abstractions means worse performance.” More abstractions also make the code harder to reason about and harder to fix.
- "How to escape CSS selectors in JavaScript | Stefan Judis Web Development"
- "Speeding up the JavaScript ecosystem - Polyfills gone rogue"
- "Patterns for Reactivity with Modern Vanilla JavaScript | Frontend Masters"
- "Building like it’s 1984: A comprehensive guide to creating intuitive context menus - Height". Interesting overview of what context menus do. They’re one of those things, tho, that nobody gets right if implementing from scratch. I’ve never seen a web app pull it off.
- "Don’t prematurely optimize for performance | Go Make Things"
- "Salesforce CEO Marc Benioff hiring boomerang employees". I’ve been saying over the past year that the layoffs were irrefutable evidence tech CEOs are incompetent. Well, after hiring people and laying them off, these clowns are rehiring the people they laid off after hiring.
- "Why not React? - DEV Community".
- "CSS Shorthand Syntax Considered an Anti-Pattern – CSS Wizardry – Web Performance Optimisation". This is good advice.
- "Accessibility is not binary | Erik Kroes - Accessibility Consultant"
- "‘I follow you for tech, not politics’ | Go Make Things". “My friend, all tech is political.”
- "table-saw Web Component—zachleat.com"
- "Compromised Microsoft Key: More Impactful Than We Thought". “This includes managed Microsoft applications, such as Outlook, SharePoint, OneDrive, and Teams, as well as customers’ applications that support Microsoft Account authentication.” A reminder that MS is bad at security.
- "The Gravitational Pull Of Business As Usual – Paul Taylor"
- "Meta in Myanmar, Part I: The Setup - Erin Kissane’s small internet website". Fucking hell.
- "Open Source does not win by being cheaper"
- "NODE_ENV Considered Harmful". Not a huge fan of “considered harmful” these days but, in addition to the points in the post, using env variables is also a great way of making otherwise platform-agnostic JS completely dependent on node.
- "Why multi-page apps? | Go Make Things"
- "Robots.txt - Jim Nielsen’s Blog". What he said. Robots.txt is a convention based on a social contract and the industry has demonstrated that it doesn’t believe in social contracts.
- "Failing Without Knowing Why: The Tragedy Of Performative Content - DaedTech"
- "Optimizing for Taste".
- "Classic rock, Mario Kart, and why we can’t agree on Tailwind"
- "Introducing Railway V2". “They say the day you delete all the Kubernetes code from your codebase is the day you truly start living.” The few experiences I’ve had with Kubernetes have all been nightmares.
- "Software is What We Learned Along the Way - Jim Nielsen’s Blog"
- "13 super rad web dev tricks you don’t know about (but should) | Go Make Things"
Interesting #
- "Writing Is Thinking - by John Warner"
- "Writers and talkers and leaders, oh my! | everything changes"
- "Google Books – search results confuse reviewers with authors – Terence Eden’s Blog"
- "Writers: use the Spoken Content accessibility feature to have your Mac and iPad read words and help you proof text". This is a useful tactic for those who can’t just read the text aloud themselves because it would disturb the coworkers/family/cat.
- "They Studied Dishonesty. Was Their Work a Lie? | The New Yorker". Certainly starting to look like behaviour economics is mostly bullshit.
- "Why We’ll Never Live in Space - Scientific American". I’ve always been a big fan of space research, but I think we’ve reached a point where some of the money should be funnelled into climate work instead. At least for as long as the crisis is playing out.
- "Summary of the 2023 WGA MBA". So, most of the details here are outside of my wheelhouse, but the “AI” clauses seem like a sensible compromise, for the most part.
“AI” #
- "Google’s New Gmail Tool Is Hallucinating Emails That Don’t Exist"
- "Elegant and powerful new result that seriously undermines large language models". Like I’ve been saying for a while now: LLMs do not think or reason. They are not on the path to AGI. They are extremely limited correlation and text synthesis machines.
- ’“AI”, and the trouble with inaccessible SaaS - localghost’
- "Statement on Generative AI | Ben Myers"
- "ChatGPT Caught Giving Horrible Advice to Cancer Patients". Like I’ve said in the past: this AI bubble is likely to kill people and everybody hyping it is contributing to that risk.
- "Why I chose Codepen Pro over Copilot" So, the standard advice in any creative industry on improving your work is that step 1 is to read/watch/listen to a lot of other people’s work. I am really strongly of the opinion that the same applies to software dev. Read code. Lots of it.
- "AI data training companies like Scale AI are hiring poets - Rest of World".
- "Adactio: Links—Just Another Music Monday - by Rusty Foster - Today in Tabs"
- "Google accidentally leaking Bard AI chats into public search results". Mass layoffs followed by deathmarch development of trendy new features leads to mistakes. Odds are that we’re going to see more like this.
- "We Don’t Need CEOs Anymore and AI Could Easily Replace Them". Can’t be any more dysfunctional than the executives we have.