This seems really simple to store as a graph database. If for some reason it does matter the model can just study the graph to contextualize its response? I don't get why Monica would struggle with this...
On VOR: It's also a PITA to use, and for inexperienced people, or someone who rarely uses it, they can confuse flying towards and away from the station.
If you are a general aviation goober like me, flying a 1960's propeller airplane, sure, that is true.
However, in a modern commercial aircraft, all position sources including VOR-DMEs, GPS, and IRS are cross-compared by the flight management system and turned into a latitude, longitude, and circular error of probability. There is no TO/FROM flag to worry about in a 777, and no reverse sensing on the back side of a localizer approach.
Yeah that makes sense. My experience is mostly in 70s to early 90's prop planes, I have had the opportunity to use a glass cockpit a few times and it was much easier. I guess i did the typical thing and contextualized this on my experiences and ignored that 99.9% of flights are not GA haha.
Time was, if you didn't understand how to use VORs, you couldn't really fly. We used them as primary en route nav, as well as for approaches, for many decades. As a CFI I think most of the confusion around VORs these days comes down to poor or inadequate training.
That's fair. I replied to someone else with a similar remark, but essentially doing most of my flying in 70s to 90s prop planes, I imagine the tools I'm most used to are just dated. Your point around training is likely very accurate, I had some great teachers but I also had a few instances of being given inaccurate information and having to teach the teacher so to speak.
A lot of those planes still exist and have not gotten panel upgrades, so IMO you still should learn how to navigate purely via only visual references and radio navigation. I believe you are still required to know how to use them to pass your private pilot check ride, too, but I may be wrong.
I don't fly and I wouldn't call my self someone that plays a lot of flight simulator. But I do enjoy navigation methods and I did learn the basics and practices of VOR navigation. I got an Android app that simulates flying courses, holding patterns etc, but only the navigation part. I would say I learned some intuition to it.
I found that surprising too! This probably isn't a completely accurate number, as the sensor is black and likely experienced some heating from the sun. Something also worth noting is that different layers of the atmosphere experience different temperatures www.noaa.gov/jetstream/atmosphere/layers-of-atmosphere
Amazing. I love that i didn't know this. When i was studying weather as part of my pilots license i noted the basic mental math that i'd lose ~2 degrees celsius per 1000ft, but never did i take a second to ask, what happens when i go up 28,000 km.
They have 10% our population and had twice the cases last year (admittedly, we're starting to catch up). But the truth is that there are huge lag times when it comes to measles / antivax. This is a symptom of a problem that started 10-20 years ago.
When it is cheaper, and the "lower quality" model is adequate for the task at hand.
Plenty of problems have a low(er) skill/intelligence floor, anyone who uses the dual-mode agent paradigm (plan, then act) figures out the second phase can be completed by a less capable model. Even when disregarding costs - speed is important here because the agent can rapidly iterate without human supervision, based on compiler errors, lint and test failures
A coding agent driven by a large LLM can delegate smaller tasks to a faster model. For example searching through the codebase for references, examples, or established patterns. They are treated as tools and don't pollute the main agent's context.
So i've done similar things in the past - and my justification for 'app' over website has been offline.
Does the PWA state of things resolve that in the modern days? If it did, yeah I'd agree, no need for an app at all. In my case the app was being used in rural Ontario. I cant even make a phone call here without wifi.
This actually handles that, at least for my use case. Specifically, on first load of the page (in a browser, or the stripped-down browser that runs the app-like PWA install) a service worker gets installed, that caches the entire site/app/map locally, and runs it from there. Then every 24 hours, or on new page load, it checks the live site for content changes and pulls down the changes if they exist. If there's no network connectivity it just silently runs the map.
Each map is 16MB - 20MB in total, so this is all nice and simple to do. Even on a slow 3G connection it's only a minute or so for a full map update to stream in.
The whole point of this system was to take a snapshot of data (mostly OSM), add on some local things that can't really be represented in OSM (like WHICH parking lots are most appropriate, stylistic overrides, system descriptions, etc) and display them. Because of issues I've had in the past with well-meaning-but-misguided OSM mappers wrongly editing trail systems I did not want anything that pulls live.
And then by having purely static content the hosting is very cheap and easy, there's no security concerns around... well... anything dynamic on the site. And each map is portable were I to want someone else to host them. And literally in a couple of years if I haven't updated the map it won't change yet still will work, and that's fine and accepted for this use. Sort-of like a mobile version of a traditional print map. Kinda like the print workflow of editing/design/etc and then rendering the PDF, but web.
This all aligned nicely for me to have a tool that works this way, with each map generated by a tool.
(Sort-of disclaimer: It was also a big personal project in learning to work with AI stuff for development. I knew and understood the inputs and outputs, was able to design the UI, handled/managed all the testing... But I didn't have to worry about the actual-code part. I was able to make pretty quick progress and iterate nicely on my ideas.)
Happy to talk, etc, more about it too. Either here, or contact info is on the site.
That was an amazing reading. I just implemented something similar a couple days ago. And now I'm building an android app based on this PWA (no tool for that, claude fable can code a good custom webview wrapper).
> my justification for 'app' over website has been offline.
Gosh, I wish more apps did some kind of progressive "enhancement" and let people read already cached messages and do deferred sends like the old days, instead of being completely useless without a data connection.
I used to run a company (2010-2018) where our customers were almost always using our product inside of grocery stores, which are almost always metal sheds with some finishings on the walls. Not so great for cell service, it became essential for us to support offline mode if we wanted to keep those customers happy and engaged.
I spent many late nights trying to debug reachability bugs. It's frankly a nightmare trying to build a reliable app when the user has /some/ cell service, but not enough to operate the app reliably.
Gosh, having tried to use the Home Depot app (to find products in store) while IN their stores... It's so bad. I can't even imagine what you dealt with in trying to handle those things.
(And so, so many little bugs in my map thing from upthread were these odd timing quirks when a user didn't have good service and one check would run and leave something else hanging resulting in a blank map. <sigh>)
Yeah, home depot is a tricky case. They have a huge array of products, and in my experience when you cant find something you need to search beyond the store you're in....so the database you need access to isn't exactly small. Makes a practical offline experience tricky.
Home depots website sucks anyway, slow, clunky, terrible touch space, and the search is awful.
Aside, they should ad cell repeaters inside to fix all this.
Maybe in the US, in Canada i find their online services are pretty average. A lot of the time i search for something and i get a US link, much to my disapointment.
As an Australian, Apples voice models have always sucked. I've tried using stt (again) more recently and its improved, but i'm so tired of having to Americanize my voice to get it to figure out what the hell i'm saying.
As a Texan first, American second, I sympathize with this statement. Siri can't understand me probably 25% of the time. I use STT for iMessage while in the car, and half the time it will take 3+ times to either get it right or me give up, and hope to remember to text them by hand when I next stop.
What does this mean? When I pledged my oath to become a citizen, I had to promise to put America above all other allegiances. Is it in relation to allegiance in the general sense or some weird statement that only relates to STT?
Texans (generally) consider themselves Texans firsts. Similar to born and raised New Yorkers (NYC). That said, I really meant it tongue-in-cheek that my accent was too thick, and doesn't work well for Siri's STT model. ChatGPT's does way better, but I really don't like talking to an LLM.
The Two Yoots problem. Do you use d's in place of t's such as dees/dems/dose/dere? I have a heavy queens accent so you'll hear me say things like "deres tree uh dem ova dere."
So for about 10 years my main interview question was:
"Write me a function in any language of your choosing, that takes an array of integers and returns the sum."
I loved it. Here is why:
1. I'd get to see them write code, in a low pressure way, but they'd have to write something
2. A shocking number of people would struggle to write the code. That was my signal to end the interview early.
3. I'd get to ask "So tell me how it works" and they'd sometimes look at me like I'm a moron, but others would be respectful and kind, and that would tell me how they'd answer other people who ask questions they felt had obvious answers.
4. I'd ask "what could go wrong at runtime?" - this would be where most people got surprised by their own responses, but it was a fun conversation to have about a seemingly simple function.
5. I'd ask how they would fix any potential runtime exceptions or potential undesirable behaviours
6. I'd try break it, and ask how they would handle that (if i could, i often could)
7. If we got this far, then we could move onto other questions and they're warmed up and generally feeling safe about how the conversation would go. I'd like to switch from coding into data structure related questions normally.
I hate high pressure coding interviews, also, who the hell doesn't just sit there and Google / LLM the answer anyway. The real question I want to know is "How curious are you? Do you want to learn? What kind of person are you? Will I enjoy working with you when things get hard". That's hard to figure out, but you're not going to do that if you just try stump someone in an interview. I think it's on the interviewer to find a way to ask questions that are revealing and accessible in an interview environment...and frankly, I think you get more out of it if you make the effort to keep it simple.
I used to do phone screens with a similarly simple question. I am still blown away that something like 60% would fail to write working code. I had more parts for the 40% that got it the first one, but it was crazy how many people couldn't do the most trivial task.
Interesting approach, thanks. Yeah, I don't mind a question and conversation like that to start things off. But I do think getting into something a little deeper (which you also mentioned) later in the interview is important, too.
For sure. I mean, i think naturally you'll end up there if all goes well. Have some tricky questions prepared of course, but starting simple gives you so much so quickly and often tells you where their strengths and weaknesses will end up being.
I'm not sure if its best to focus on strengths of weaknesses, but i did prefer to focus on strengths. I found convincing myself why I do want to work with someone was a better experience than trying to find reasons not to. Also it just tended to get better buy in from the other team member that way, and i'd know how to assign work once they joined.
ha, i'd have loved that. I'm out of the world where I hire people these days, solo founder life. Maybe i'll go back there, but it's a real mission objective to just be a one man show right now.
reply