I listen to a fair number of podcasts. The Accidental Tech Podcast is my favorite by a mile.

It's the one that I look forward to every week. It comes out on Thursdays, often right when I'm leaving work. It's a sign that the weekend is right around the corner.

I'll get some of the reasons why out of the way quickly, the reasons particular to me and my tastes.

• The topic, the world of technology and all things around it, is right in my wheelhouse. It's my profession, but also my hobby. That's what got me in the door. If this stuff isn't your thing, then this podcast won't be either.
• They talk a lot about Apple products and that ecosystem, which where I spend my personal time.
• But they also cover a lot of issues that touch on the world of technology: business, law & politics, companies, social media. I like their descriptions, their take, their rants.
• They end up talking a lot about personal tech too. Like how to manage family photos and backups. The hard, fussy stuff that ends up taking up so much of our lives. For example, what's the best way to help parents deal with passwords. Hard stuff!
• I especially like tech-adjacent topics like home audio/video, home automation, and gaming.

But what I appreciate more than the content of ATP are the hosts and the care they take to produce a good show.

• They are friends. It's nice hanging out with people who like each other. It's kind of like we have permission to eavesdrop and be a part of that.
• They're not afraid to discuss their lives. Stuff like families and work/life stress. This is good stuff for me because I'm in the same life situation as them: middle-aged tech dads. But bringing their whole selves to the show is a bit of vulnerability that I appreciate.
• ATP is exceptionally well produced. I didn't appreciate this until I'd heard so many other podcasts that are produced terribly, with bad recordings or poorly mixed. Usually you have to turn to the corporate, fancy podcasts, but they also have lots of ads and are usually different kinds of shows.
• And the show is so well edited. They edit it without losing any of the content or the pacing, and sparing us all the awkward "um where were we" and technical futzing.
• They don't talk over each other. Part of this is discipline, part of this is just politeness. But also now that I've listened to the pre-edit "bootleg" a couple of times, I've come to appreciate how much of this is also fixed in the edit. Nice job Marco.
• They are respectful. They don't put people down, they aren't mean. When they kid each other it's in good fun.
• When (rarely) they wade into social justice or world events, they do so respectfully and thoughtfully. They understand that as three three cisgender white guys, it's good to have views to share, but also right to listen and help others.
• Their show is reliable. It's great that they keep to a regular format and schedule. It fits into the rest of your life and becomes something you count on.
• Finally, some small things: Yay for proper use of chapter markers, hardly anybody else uses them properly. And I like their their occasional forays into car talk, just because I like car stuff too.

Although they joke about how much of the show is devoted to feedback, it's one of my favorite parts. It shows that they listen and are learning. And they share that learning with all of us. A recent example was six minutes or so into Episode 570 when I learned how home power battery systems, when full, signal this by changing the frequency of AC power as a signal to solar panels to back off. Fascinating!

I like the members-only specials. It's OK with me that these are only available to members. They've been candid lately about ad revenue drying up and they could use the extra revenue channel. These episodes are a nice way to reward members. They've managed to do these without compromising the core show.

The best thing they ever did on ATP was getting a sponsor to send John toasters to review. And review he did! I didn't appreciate how many bad toasters there are, and how they can be bad for so many reasons. My favorite reason was poor knob feel.

Nice work John, Marco, and Casey. Please keep it up for a long, long time.

If you've traveled internationally, likely you've used a SIM card for local data and calls. There is a nice ecosystem around SIMs with a wealth of easy and affordable Pay As You Go (PAYG) options.

But the newer iPhones did away with the SIM slot if favor of some new eSIM hotness. Apple has all kinds of info claiming they have good international support, but I found reality falls short.

• Only a few carriers support eSIM's,
• The few that do require a contract. A tourist or student studying abroad is better served by a PAYG plan, and
• Even if you can stomach a contract, that would require a UK bank account; no way to easily pull that off.

We ended up falling back to international roaming. It works but is expensive.

I think Apple made the wrong call removing the trusty old SIM tray. Clearly the new models can be made to work well with it, since how they are sold in the UK. If you're unlucky enough to have bought your recent iPhone in the US, you're out of luck.

I this is an example of Apple bad tendency sometimes choose form over function, "courage" over usability.

I just finished the Big Dig podcast and it's worth a listen. It's about Boston's Central Artery/Tunnel Project, the most expensive highway project in US history. They cover the whole story, from conception, to getting it approved, to years of execution, and then the fallout from cost overruns and mismanagement. The podcast is well produced and has a lot of primary-source interviews.

I have a little bit of a personal connection, since I lived in Boston during the project's later years when it was on the news all the time. I even toured a part under construction.

But what I find most compelling is trying to connect to the question, can America build big infrastructure anymore? It's something I think lot about. We benefit from by the giant projects from the past (dams, bridges, interstates) but can't maintain them properly; we struggle to take on new things like high-speed rail.

They place most of the blame on headwinds that didn't exist in the public-works heyday of the twentieth century. Leaders are under more scrutiny; projects fall under a bunch of regulations intended to protect the environment and workers. Public commentary slows things down.

The problem is, while the old way was easier, it also caused a lot of harm. I learned the story of the Cambridge and East Boston families that stopped interstate projects that would have leveled their homes and neighborhoods. I know those places well. I used to own a home right where one of those roads was supposed to have been. I sure am grateful to those protesters who won!

It ends on a hopeful note. Not that it's easy, nor are we necessarily that much better managing big projects now. But there are success stories.

One part I especially liked was 20 minutes into the final episode, in the final interview with Fred Salvucci, where he told the story Saint Francis. God tells Francis to build a cathedral, but then tears it down. Francis builds another, God destroys it again. Why, Francis asks. Because it's not enough to build a physical cathedral, but you also have to build the support for it in the hearts and minds of the people. Maybe that's the part we're not doing well enough now.

Some bosses want you to bring them problems. They like to unscramble Rubik's cubes and are happy to work through it with you. Sure it's great if you have a proposal or recommendation, but be prepared to show your homework.

Other bosses want you to bring them solutions. If you bring them a problem, you'll get annoyance and "what do you want me to do about it?"

Most bosses have exceptions by domain. A common pattern there is the bring-me-problems boss who is also a techie -- they'll want to dig in planning, say, but prefer you to solve messy people conflicts yourself.

I've worked with both and each has their virtues. When you have a new boss, figure out their style and adjust.

The other day my daughter asked me about Software Engineering -- what do we actually do? She's eighteen and probably won't follow in my footsteps, which is fine, but I still want her to see my field.

I've always found this question hard to answer. I've been an engineering manager for a long time, and I'm happy to describe that job (emails, 1:1's, PRD reviews), but I don't think that's the heart of it. Plus these days I work for Google where things there are proprietary and deeply layered, not much help for answering questions like these.

Recently I've found my way to Mastodon for obvious reasons. I chose the Hachyderm instance because it was well run by people who shared my values. It turns out that the values of the owners and operators of the things we use matter, huh.

I then learned a cool thing, that the Hachyderm admins do much of their work publicly. They livestream debugging sessions on Twitch, they write postmortems, they share live graphs. @nova happened to be live-streaming at that moment, not surprising since the team's been busy absorbing thousands of new users and fending off attacks. My daughter and I watched a bit together.

Team Hachyderm (@nova @dma @quintessence @Taniwha @hazelweakly @malte): thank you for running this service well. But also thank you for giving me something I'm proud to use and proud to show my kid.

How many different New York Times Spelling Bee puzzles are there? Or more precisely, how many combinations of seven letters can be used to build Bee-type puzzles?

It turns out 7,742 different seven letter combinations can be used to generate Bee-style puzzles. There are more puzzles themseves based on what letter is chosen for the middle spot.

The majority of letter choices, about 62%, have just one pangram. That's lower than I expected, actually. It's not that uncommon to have two or three pangrams, which happens about about 25% of the time, and nearly four out of ten puzzles will have more than one pangram. The full output is here, the output of this program.

Watch out for the combination einprst. If this one ever comes up, good luck finding all twenty-seven of its pangrams.

The Bee Puzzle

A Bee puzzle has seven letters with one "special" letter in the middle. Make as many words as You can find with at least four letters, using only the letters given, but it has to use the center letter. Proper nouns aren't allowed. Every puzzle has at least one pangram, a word that uses all letters — this example's is amphibian. Wikipedia cites Frank Longo as the Bee's creator.

nytbee.com has more today's puzzle and some interesting stats about these puzzles in general. They don't seem to be affiliated with the NY Times but that seem to be OK and is's a nice site.