- It was nice seeing my 2025 reading list represented.
I started the year reading the first five books of the Foundation Series (book #1 on the list). A must read for anyone who hasn’t read it. I couldn’t believe how well it held up 70+ years later(!!)
I just finished the 3 Body Problem trilogy, and think it’s appropriate book #2 (The Dark Forest) is on the list as it’s probably the best — but all three are great.
I’m now ready Project Hail Mary. It’s been a long time since I read the Martian,but Andy Weir’s writing style is fast paced and practically a screenplay already. It’s obvious from the first chapter why it was picked up for a movie.
- I agree. I started using Foot based on a recommendation from the notcurses library author, who has deep expertise in terminal emulation and collaborates with maintainers.
A tip for new users: The default theme is a bit harsh. I was able to port my Alacritty theme and other config by feeding the config file to an LLM (along with the Foot documentation). It generated a configuration that was 80-90% correct and only required about five minutes of manual fixes.
The result is now visually identical to my Alacritty setup, but Foot feels faster.
- “A language is a dialect with an army and navy”
https://en.m.wikipedia.org/wiki/A_language_is_a_dialect_with...
- Tell that to Tex Johnson:
- I read Isaac Asimov’s foundation series a few years ago (side note for anyone who hasn’t read it: it still holds up incredibly well with a small suspension of disbelief and some grace for when it was written).
In the preface to the 4th or 5th book (which were written 30+ years after the “original” trilogy) he discussed how the originals parts of the trilogy were published as a set of short stories in a SciFi publication over 8 years, and later compiled into the books.
I was astonished.
Perhaps everyone else already knew this. But such a clear narrative through line to be written in discrete short stories. Very impressive.
It sounds like this may have been common prior to this era as well.
- I've just read through the entire documentation end-to-end. It's very clear and well written, thank you for the time and love.
One thing that's unclear to me in how I might manage this "side-by-side coexistence": how you manage potential route conflicts, e.g., You have a Hologram "page" that defines `route "/hello-world"`, and the Phoenix application router.ex also defines a route, `live "/hello-world', MyApp.HellowWorld, :index`.
I could imagine the lack of a central router may be offset through conventions of how you organize pages (perhaps a directory-based-routing convention). I saw in the Installation section that it's common to organize under `app/`, and `app/pages`. Maybe an expansion on best practices for larger project organization, and how you keep Hologram routes straight, would be interesting to other people as well.
Regardless, I've just started dipping my toes in the elixir and phoenix ecosystem about 3 months ago, and it's been a real joy. I'm throwing every hobby project I can think of at it to learn more. I'll give hologram a spin -- thanks again for such clear documentation.
- It’s available as a VSCode extension and documented here:
- Thank you for sharing! After reading the article, I played your game and really enjoyed it. My first text based game in ~30 years.
I was inspired and went on to play one of the games listed on IFDB tagged “short” and “recommended for beginners”. (Suveh Nux).
Also super fun and satisfying to beat!
- Absolutely. And I’m finding the same with “agent” coding tools. With the ever increasing hype around Cursor I tried to give it a go this week. The first 5 minutes were impressive, when I sent a small trial ballon for a simple change.
But when asking for a full feature, I lost a full day trying to get it to stop chasing its tail. I’m still in the “pro” free trial period so it was using a frontier model.
This was for a Phoenix / Elixir project; which I realize is not as robustly in the training data as other languages and frameworks, but it was supposedly consuming the documentation, other reference code I’d linked in, and I’d connected the Tidewave MCP.
Regardless, in the morning with fresh eyes and a fresh cup of coffee, I reverted all the cursor changes and implemented the code myself in a couple hours.
- > I can’t put my finger on why.
To me it feels inauthentic.
Based on early comments, which say it makes sense, I’m an outlier.
However, this shifts from something that had semantic value (not saying Apple used Semver, but there were linear major and minor versions) to marketing driven version numbers.
- I was in Mexico City two weeks ago and a few uber drivers were driving BYD Dolphins. I was impressed. At a $14k price point, I can’t imagine better value for money. I got sucked into research and there are no near term plans for expanding to the US market. It’s too bad. They’re expecting to sell 500,000 units in Mexico this year. With the current (lower) emissions standards, I’m optimistic such affordable EVs will make a big difference to air quality in the coming years.
- According to the FAA’s website, it is not in use today.
https://www.faa.gov/faq/faa-getting-rid-air-traffic-skills-b...
- I love this. I agree with the "about" that it's visually compelling, and I'm mesmerized.
This doesn't detract from my enjoyment the site, but for trip planning I'm a little skeptical of the results around the edges, especially when I'm assuming multiple transfers would be required (e.g., Local -> Express -> Express -> Local).
With a caveat this was over 10 years ago (~2012-2013), and train frequencies may have changed:
I used to live pretty far up on the upper west side, and took the 1 train from the 103rd street station daily. My weekday route was 1 -> 2/3 -> 7 into midtown. The 20 minute radius is accurate at peak times, when it only takes 2-3 minutes to catch a transfer. However, the website "about" makes an assumption this is for noon on a weekday. I don't think I ever made it to Brooklyn in under 40 minutes.
- Does anyone have experience if these provide tactile feedback?
There’s lots of discussion about the actuation point, but is there a “click” feel?
I’ve had a lot of wrist strain issues over the last 20 years. I tried many “ergonomic” keyboard layouts, but ultimately switching to a standard ikcb CD108 with Cherry MX brown fixed my issue. The click stops me from mashing the keys too hard, which seems to be the primary root cause (for me).
- Absolutely. I used the number of administrative personal * administrative salaries because it's a quick mental visualization. I just reviewed United Healthcare's 2022 10k. In 2022, they had $324B in revenue, and paid $210B in medical expenses. Of the $114B remaining, $85.5B was various operating costs, $28.5B (25%) was corporate profit / operating income (before interest & taxes). Even if all 40,000 employees of UnitedHealth Group were making 200k (they're not), that'd only account for $8B of their $85.5B in non-medical costs. https://www.unitedhealthgroup.com/content/dam/UHG/PDF/invest...
- Only 10% of the increase in medical cost in the US is attributable to doctor salaries. The remainder is administrative overhead, Rx costs, insurance administration, etc.
There are now 10 administrators for every 1 doctor.
Since 1970, the number of physicians has increased by 2x. The number of administrative staff has increased by 38x.
If each administrator is making 100k+ - 150k+, that's significant cost.
Some of these crazy billing practices are due to the private equity buy-ups and consolidation, creating workflows and fee structures that are hostile to patients and doctors, but benefit the PE firms.
https://www.commonwealthfund.org/publications/issue-briefs/2...
Lots more data here: https://www.healthsystemtracker.org/
- Banking and financial services average 8.1% of revenue on IT Spending. Industries like industrial manufacturing, food and beverage, retail, and energy average 1.2% - 1.9%.
Many security leaders (CISOs or otherwise) do not have the budget or authority to meet their board's or CEO's expectations, but it may be outside your sample of big banks.
I'm otherwise aligned with your comment. The successful CISOs I've interacted with, regardless of industry, educate their leadership team on the trade-offs and risks of investment levels, set realistic response and recovery times expectations based on those investments, and turn it into a business decision, rather than promising the impossible.
- Four requests per day seems a little low to verify the quality of the responses before shelling out even a few bucks.
That said, I've already gotten great quality formula and VBA generation directly in ChatGPT (GPT-4.0), with only a sentence of prompting, so I'm not the target customer.
- Will you control access to qualified professionals?
I’ve observed a large part of specialization and sub-specialization in medicine is understanding what to ask and how to ask it in order to uncover the fine differences between similar symptoms.
I am not a doctor, but have many friends in medicine. I’ve found they aren’t very good at understanding what to ask or look for across the boundaries of their specialties, and if they try to search medical journals online based on symptoms they come up with the wrong results.
The people I know are in Family Medicine, OB, plastics, Anesthesia, derm, and optho.
- The points are not "free" to airlines, though. (Without looking at every airlines 10k, at least one mechanism is to): Account for them as a liability on the balance sheet as "deferred revenue." They then recognize the revenue when the points are redeemed, meaning they incur the same blended Cost per Available Seat Mile (CASM) as a purchased ticket. That's in addition to the significant costs of managing a loyalty program (IT, Partnerships, Legal, etc.)
There was a cool fact in there that he sees every price tag as the opportunity cost with compounding interest.