spondyl

I don't really know if anyone reads my blog because I don't use any analytics or care to know.

There are a handful of times that I've written a post about something and then when someone offline expresses interest in X thing by coincidence, I can say "Ah, I actually know a little bit and you're welcome to read what I know in your spare time!"

I also have a thing of writing about any thoughts and anxities I might have at various milestones (starting new jobs + 5 year increments).

That recently came in handy with a mentee who was a bit like "I don't know what I'm doing, is this normal?". Saying it's normal is one thing but even better is being able to point to a post where I blurted out my mental state 10 years ago :)

I also don't share posts anywhere but I have a hint that at least one person is subscribed to my RSS feed (I think they told me something broke once) but part of the fun is not really knowing who's out there I think

The other value too is that just like trying to explain something to another person, the act of putting together a coherent piece of writing forces you to sort through your own thoughts (even if you think you understand the topic) which is valuable even if no one elses reads your own writing.

I re-read one of my very first posts not too long ago that I had completely forgotten about and thought "This is good stuff! I wish I knew about it earlier" haha

To clarify for anyone wondering, 404 Media and WIRED have a partnership so WIRED will occasionally publish their content and they hope to collaborate together on articles (although I'm not sure if this was one of them)

Just an FYI that Pinboard pages aren't public anymore, only being visible to those who are logged in to Pinboard.

Marciej mentioned the following when I noticed myself and asked what the cause was:

> Unfortunately you've run into a necessary anti-crawling measure that keeps the site from getting overwhelmed. About four months ago I started getting hammered by distributed crawlers from Chinese IPs, with no real way to block them short of blocking the entire IP range. Putting things behind a login is a stopgap while I try to figure out a more durable and friendly solution that makes it possible to have public-facing pages again.

Another macOS native pie menu is https://www.pie-menu.com

You might be thinking of https://raindrop.io which is developed by a Kazakh developer?

> doesn’t work offline on mobile devices

To some degree, that isn't true anymore as of last week but of course, you meant when you used it, not generally: https://www.fastmail.com/blog/offline-in-beta/

I was looking for something else on this page and did a double take. Is that... /the/ Terence Tao? He's a Ranma 1/2 fan? Nice!

I'm only 30 and honestly, sometimes I sit down at a side project and think "How the heck did I just sit here and grind out this entire desktop application" when now I'll struggle to work on it for an hour at a time.

Part of it is motivation in that it went from being a personal project I use to something others primarily use but I also constantly ask myself "Is this really what I want to spend my time on" which was never really a thought back then when it was just fun (and something I needed)

Not exactly the same but echoes of the raw firepower thing where it's easy to fully commit if you either have nothing else or you're fully sure of your dedication otherwise you're sort of one toe in the pool and aren't sure whether to save your energy

> I mean, I am happy for them but this concept of growing a business to an exit is not going well for society as a whole

Oh yeah, I fully agree with the enshittification sentiment.

Admittedly, a DNS registrar for me personally is something I'd just swap without much thought but I can think of a few services I use where I wouldn't be so loosely coupled from the product if those founders were to exit.

It's a bit paradoxical on my part I think and I do wish we had more lifestyle businesses that don't have to become massive.

Mind you, I would have put iwantmyname in that basket now that I think about it.

I can't speak to any of those videos in a technical sense but personally, I don't feel like any of them are good?

The vibe they give me is similar to the iPhone photography commercials where yes, in theory, a picnic in the park could look exactly like this except for all the parts that seem movie perfect.

I guess it's really more of a colour grading question where most of the Sora colour grading triggers that part of my brain that says "I'm watching a movie and this isn't real" without quite realising why.

A few of the Hunyuan videos in contrast seem a bit more believable even though they have some obvious glitches at times.

The other thing I think Sora has is that thing in commercials where no one else except the protagonist exists and nothing is ever inconvenient. The video of the teacher in a classroom with no students reminds me of that as well as the picnic in the park where there's wide open space with no one around.

I suppose it depends if the goal is to generate believable video and how you define believable.

Oh damn, I didn't know this!

I've used their services for ages and even got to briefly meet the founders once in Wellington who gave a talk on Erlang.

Ah well, while it sucks that the good times may be over, I'm glad the founders got their exit :)

Discussed here previously: https://news.ycombinator.com/item?id=42284202

This is apparently a misattribution: https://www.macrumors.com/2024/11/17/tim-cook-didnt-say-that...

> The Wall Street Journal last month published an interview with Cook, in which he said he uses every Apple product every day. Soon after, The Verge's Wes Davis attempted to replicate using every Apple product in a single day. During that day, Davis said he mostly used the MX Master 3, but sometimes switched to a Magic Mouse or Magic Trackpad.

> In other words, it was Davis who said he himself used a Logitech mouse, not Cook.

According to the MacRumors article, the misattribution came from The Mac Observer misreading an article by The Verge: https://www.theverge.com/2024/10/22/24276142/tim-cook-wsj-in...

The author was attempting to interpret what a day with only Apple products would be like, written in a first person style.

They mentioned using a Logitech mouse but for someone skimming the article, it can be easy to misinterpret that as a pull quote from the mentioned WSJ interview

Shout outs to Henry Waxman who originally introduced H.R.3562 aka the Nutrition Labeling and Education Act of 1990.

It's wild to think that the mandating of nutrition labels is only 34 years old.

I can recommend his biography The Waxman Report which talks about this and other types of legislation that are taken for granted in daily life.

If you meant me, I live in New Zealand funnily enough.

I don't use it myself but I came across this open stats page which seems to suggest Chrome Extension Support is coupled to hitting $10K MRR.

The graph legend says

> To comfortably maintain our own build pipeline with Chrome Extension Support.

Anecdotally speaking, Spotify has become a sort of cultural default for a bunch of people.

If you're in a DM or a group chat, you may not hear "You should listen to X by Y" but rather "Check this out: https://open.spotify.com/blah"

At my current workplace and a previous one, both had Slack channels for sharing music and they were 99% Spotify links.

Partly because of genuine Spotify usage but even not using Spotify at the time, I would find the Spotify equivalent for my own recommendations (to reduce friction for the majority so they'd be more likely to actually listen to the recommendation)

Presumably for those few users not using Spotify, rather than having to find the equivalent song via text search, which may or may not contain a result for Provider Z, this service straight up just converts the Spotify link you've been given into all of the other provider equivalents.

Just to clarify a bit more for those who are new to strings and because the audience for the post may learn towards people fresher to reverse engineering:

While most of the time, you're dealing with variables and such in programs, at some point you have to hardcode some information such as URLs to query so something like

BASE_URL = "https://example.com" result = requests.get(BASE_URL + "/api/blah"

If we pretend this is in an Android app which is stored as an apk file (a zip file basically), running strings would spit out "https://example.com" and "/api/blah"

It'll also spit out anything that appears to be an ASCII character so plenty of junk but it's often quite handy as a starting point.

There are, of course, much more precise tools such as man in the middle proxying but that you'll only capture traffic for endpoints actually used by said app. The app may contain other endpoints let unused, rarely triggered and so on.

As someone who has been on and off the Degoogle train (I ran full LineageOS without Google Play at one point) and is now pretty deep in iOS territory, I'd say the main thing for me has been email.

I've used https://www.fastmail.com for a great deal of years now, which is also home to my calendar as well so there's nothing much of value tied to my Google account.

YouTube subscriptions would be annoying to lose but I subscribe to channels via RSS. At the moment, I'm using http://newsblur.com for that but sometimes I jump around RSS clients. Newsblur recently added backfilling of channel content from YouTube's API beyond what RSS feeds normally surface though which is great.

For plain text notes, everything is in https://obsidian.md compared to say; Apple Notes

One thing I would miss is photos which I currently have in Apple Photos stored in the cloud so I recently set up https://github.com/icloud-photos-downloader/icloud_photos_do... which runs nightly on a mac mini at my house and archives everything in a Backblaze B2 bucket. https://ente.io/ seems like a promising alternative in this space.

I don't actually have a good alternative to Apple Health now that I think of it. You can export backups but not automatically. I'm not entirely sure what format a lot of that data is even in.

Realistically, I don't look at it a lot, I just use some of the correlations when they rarely pop up. Something like https://exist.io/ is probably the closest that I actually used to (aspire to) get value out of but don't currently use.

The reality is that convenience is king and it's hard work staying independent from any centralised service, not to mention social pressures.

The middle ground I've accepted is still using it all but making sure you have a good escape hatch you periodically test + backups where you're solely reliant on large businesses who heavily lean on automated account banning.

When it comes to China, it's not entirely fair to single out Apple here given that non-Chinese companies are not allowed to run their own compute in China directly.

It always has to be operated by a sponsor in the state who hold encryption keys and do actual deployments etc etc.

The same applies to Azure/AWS/Google Cloud's China regions and any other compute services you might think of.

I'm not entirely sure why this is on the front page of HN but I can +1 Paranormasight after having played it shortly after it released. I thought the game was a treat :)

For anyone interested in a counter-point (ie; "Transparency is not inherently good and here's why"), the paper "Transparency is Surveillance" is a nice read: https://philpapers.org/archive/NGUTIS.pdf

It basically talks about how people will change their reasoning to suit public opinion when they work publically. You might imagine a scientist who gets a public grant and is required to share their findings publically.

In this hypothetical scenario, let's say their findings are required to be explainable to the public at large. The reality is that when you're building anything (whether it's a theory or a product), the reality is that causes and effects are not always clear nor will they ever be as we see with the human body for example.

As such, while we might think imposing transparency would increase trust, the reality is you'll often find ad-hoc justifications for why things are the way they are rather than just saying "We don't know why X or Y".

The author also presented the same ideas as a talk for anyone who prefers video: https://www.youtube.com/watch?v=ZcEJY61FKIc

Shout outs to Shiva who has appeared in some videos but is often behind the camera so doesn't get enough recognition as being the other half of their duo :)

If you'd like a counterpoint that complete transparency is not inherently good, you might enjoy this: https://philpapers.org/archive/NGUTIS.pdf

The Auckland Museum has an interactive exhibit where you can go into a little house and watch a (fake) news broadcast. It shows a suburb with a view of the ocean and then suddenly one of the underwater volcanos erupts which causes the house exhibit to shake and the lights flicker.

It's pretty neat and worth a try for anyone visiting Auckland. Having lived here my whole life, I can say that the news broadcast is refreshed every few years and is portrayed by our actual prime time news anchors (from our state broadcaster)

Yep, they're all just one div. Well, technically a div plus ::before and ::after pseudo-elements but still, it's all CSS

Yeah, it sounds pretty identical to a caching issue that happened to Steam way back on Christmas Day of 2015: https://store.steampowered.com/oldnews/19852

Despite the common confusion, it's not e-ink but rather a reflective LCD (rLCD) screen. It's the same principle as the original Game Boy where instead of using a backlight, the LCD reflects light instead.

There's an early review available here for anyone interested: https://arun.is/blog/daylight-tablet/

Oh neat, as a Transit user, I always wondered if they were using data from those train/bus stop questions

From memory, you definitely were able to install root certs pre-7.0ish but that functionality was removed, presumably to stop users from running man in the middle proxies and sniffing app traffic.

A huge irony in that alongside all the talk of conservatism, iOS provides much more user flexibility in this area of all things, and something that keeps me using the ecosystem.

You can still load root certs but as mentioned, it requires rooting your Android device. Some apps do have certificate pinning but the majority I would bet don't