bxparks

I read this all the time from other people, but for me, Selection sort is the easiest to remember and implement. My next easiest would be Insertion sort.

Bubble sort doesn't click for me easily. I think it's because the terminating condition seems uglier than Selection sort or Insertion sort. I always have a little voice in the back of my mind, "Is this outer loop guaranteed to terminate?"

On 8-bit and 32-bit microcontrollers (e.g. 8-bit AVR, 32-bit ESP8266/ESP32), Insertion sort is 6X faster than Bubble Sort on random data. I have tested this up to about N=1000.

Both Insertion sort and Bubble sort are O(N^2). Both are stable sorts. Insertion sort consumes only about 10-20 bytes more flash memory than Bubble sort. It's hard to think of situations where Bubble sort would be preferred over Insertion sort.

Shell sort is vastly superior if you can afford an extra 40-100 bytes of flash memory. (It's not too much more complicated than Insertion sort, but sometimes, we don't have 100 extra bytes.) It is O(N^k), where k ≈ 1.3 to 1.5. As soon as N ⪆ 30, Shell sort will start clobbering Insertion sort. For N ≈ 1000, Shell sort is 10X faster than Insertion sort, which in turn is 6X faster than Bubble sort. Unfortunately Shell sort is not stable.

Comb sort has a similar O(N^k) runtime complexity as Shell sort. But it seems slower than Shell sort by a constant factor. Comb sort is also not stable. I cannot think of any reason to use Comb sort over Shell sort.

Quick sort is not much faster than Shell until about N ≈ 300. Above that, the O(N*log(N) of Quick sort wins over the O(N^k) of Shell sort. But Quick sort is not stable.

Merge sort is stable and runs in O(N*log(N)), but it consumes an extra O(N) of RAM, which may be impossible on a microcontroller. You may be forced back to Insertion sort for a stable sort.

[Passed my edit window, so replying to self]

Looks like the original repo of syncthing-android, (https://github.com/Catfriend1/syncthing-android), which was maintained by @Catfriend1, now redirects to the one maintained by @researchxxl (https://github.com/researchxxl/syncthing-android).

The problem seems to be that no one in the syncthing community knows who @researchxxl is. The account was created only 3 weeks ago. There was no communication about how the transfer took place. Was the transfer actually authorized? Did @Catfriend1 get hacked? People are worried about a backdoor hijack attack similar to the XZ libary.

There's a long discussion at https://forum.syncthing.net/t/does-anyone-know-why-syncthing.... It is too long and complex for me to follow. (I tried to get an LLM to summarize that thread for me, but the output was not helpful.)

It is unclear what sycnthing-android users are supposed to do right now. I am staying at version 1.30.0.4 until things become more clear.

I don't follow the Syncthing ecosystem, so it's difficult to understand what is happening.

[Edit: The GitHub repos are called "syncthing-android". The Android apps are called "Syncthing-Fork" or "Syncthing-Fork Wrapper", which adds to the confusion.]

If I recall, there used to be a syncthing-android app on Google PlayStore. That was discontinued by @imsodin in Oct 2024 (https://forum.syncthing.net/t/discontinuing-syncthing-androi...).

There was a version of sycnthing-android on F-Droid. I don't remember who maintained that. I have version 1.30.0.4 installed. But I cannot find any information about that version anymore.

The current version on F-Droid is 2.0.12.1. That seems to be maintained by a fellow named @researchxxl. Apparently @researchxxl claims to have inherited the source code and signing keys from a person named @Catfriend1 (Not sure who that is, the maintainer of version 1.30.0.4?)

There is another fellow named @nel0x who seems to be maintaining a different version of synchthing-android? (Edit: Here it is, https://github.com/nel0x/syncthing-android, which says that it is a fork of the one maintained by @Catfriend1).

Could also be French speakers. They would say "J'utilise le format .avif depuis quelques années." I think the "depuis" throws off the French speakers when they translate that literally as "since some years" instead of "for some years".

Another common tell: I wake up in the morning in the US/Pacific time zone, and see the European writers on HN using "I have ran" instead of "I have run".

I tried eBay like 4 times. But they were all fake or defective. At least one of them was dead on arrival.

I finally bought one from iFixit.com. Far more expensive than eBay, but the battery actually worked great for about a year. Then about a month after the 1-year warranty expired, it degraded noticeably (maybe to 80%) with only about 100-200 charge cycles. Even iFixIt cannot source a battery as good as the original Apple.

Right now, it's at 65% capacity after 584 charge cycles, after 4.5 years of service. I will probably go back to iFixIt. At least I'll get one year of full capacity from them, instead of the fake or DOA ones from eBay.

The last time I tried this in 2022, I encountered a bunch of rough edges. I tried Ubuntu or Mint on a MBA 11 (2013 or 2015 models). Maybe Linux has gotten better since then?

For example, the fan did not work out of the box. I had to install something called 'macfanctld'.

Sleep did not work, it would wake up as soon as it went to sleep. I had to install something called a 'suspend-fix.service' hack.

Brightness control was flaky. It would always wake up at the lowest brightness setting. Disabling auto-dimming caused it to wake up at the highest brightness.

Yes, it was infuriating.

In addition to the Classic option that is shown only after hitting "Cancel", they also had a secret "Microsoft 365 Basic" option for $20/year. It includes no Office products, but provides 100 GB of OneDrive. Which is all I needed. So Microsoft is getting $20/year from me that they don't deserve.

Why do I pay them even $20/year? It's insurance against the same kind of BS from Google. I back up my Google Drive to OneDrive.

I'm not going to write things like Django, flask, requests, urllib3, numpy, scipy, mathplotlib, dateutils, and psycopg2 on my own. That would take multiple lifetimes.

Sure, you can choose to write everything yourself, but you are limiting the size, scope, and complexity of the problems that you can solve. We have to stand upon the shoulders of others to get stuff done.

I use SyncThing for that purpose. It syncs across my phone, my laptops, and my Synologies. But I don't sync all my photos.

I don't like the thought of providing Google thousands of personal photos for their AI training. Which will eventually leak to gov't agencies, fraudsters, and criminals.

Do you happen to know why? A package manager system seems essential for sharing and reusing code with other people.

For example, I don't particularly like Python as a language, but I will use it because it has 677,633 packages in PyPI.

From what I can tell (since I am just finding out about this today), they stopped manufacturing the old Chromecast hardware, and at some point, will stop supporting the old devices. The old devices may stop working in the future, for example, because they sunset the servers. Like their thermostats. Who knows?

That was probably me, when I stopped using Google Search some years ago. :-) Got tired of the ads, the blog spam, and AI-generated content crap floating to the top of their results page.

Strava? :-) Half-joking, half-serious, I haven't used Strava in years, I don't remember all its capabilities.

Edit: Missed the "locally" part. Sorry no suggestions. Maybe Garmin has something?

Hmm, good to know. But given Google's history, I assumed that it would stop working.

I also need to sell my Google Chromecast with Google TV 4K. Brand new, still in its shrink wrap. Bought it last year, to replace a flaky Roku. It was a flaky HDMI cable instead. I trust Roku more than Google for hardware support.

I also have low-E windows all around, and I still have good cell phone reception. I think it's your aluminum siding.

As I understand, low-E reflects solar thermal infrared radiation (3-8 microns, 37-100 THz), while letting through visible light. I don't think it affects 5 GHz radio waves very much.

But yeah, it would be very satisfying to finally wire the house with ethernet.

A lot of things on https://killedbygoogle.com/ . I used to use 30-40 Google products and services. I'm down to 3-4.

Google Picasa: Everything local, so fast, so good. I'm never going to give my photos to G Photos.

Google Hangouts: Can't keep track of all the Google chat apps. I use Signal now.

Google G Suite Legacy: It was supposed to be free forever. They killed it, tried to make me pay. I migrated out of Google.

Google Play Music: I had uploaded thousands of MP3 files there. They killed it. I won't waste my time uploading again.

Google Finance: Tracked my stocks and funds there. Then they killed it. Won't trust them with my data again.

Google NFC Wallet: They killed it. Then Apple launched the same thing, and took over.

Google Chromecast Audio: It did one thing, which is all I needed. Sold mine as soon as they announced they were killing it.

Google Chromecast: Wait, they killed Chromecast? I did not know that until I started writing this..

You are lucky. In 2025 I have to run most of my 20-30 wifi devices on 2.4 GHz because 5 GHz won't penetrate the walls in my house, especially diagonally.

My dev laptop is about 10 m (30 ft) away from the wifi access point, but goes through about 6 walls diagonally, due to some weird layout, and 2.4 GHz is way faster.

The house has some thick walls.

Same with phones. As soon as I'm in a different room, 2.4 GHz is faster. So I just keep things on 2.4.

Yeah, I've been planning to wire the house with Cat-6 into every room and add some access points. It's been on the backlog for 6 years..

No, unless you have a laptop used by their developers. Every 2 years, I try to install FreeBSD on some of my Dell laptops, find that the wifi doesn't work, then give up. Been doing that for about 8 years..

In my opinion, a language that requires a programmer to have a "revelation" to understand basic control flow is not a language that is useful or practical for solving real world problems.

I would prefer to write in assembly language than write in Forth. Which is what I have done with one of my current projects.

With assembly language, there is a good chance that a random person with some minimal programming skills would understand my program if I were hit by a bus. With Forth, I think the chances of that are close to zero.

> I've actually never worked with a "pure" interpreter in Forth, only compilers of various levels of complexity. Threaded code compilers are (in my experience) by far the most common way to deal with forth -- and they are very much 2-pass. Even when used as an "interpreter," they generate (trivial, usually) machine code, then jump to it.

Lots of good info, thank you. I don't think I will fully understand what you wrote until I implement a Forth interpreter myself.

So a side question: If most Forth "interpreters" are compilers, how does a Forth interpreter work in a Harvard architecture microprocessor (with separate memory space for data and instructions) instead of a Von Neumann architecture with a unified memory layout? In other words, in a Harvard architecture (e.g. AVR microcontrollers), the Forth compiler will live in read-only flash ROM, and it cannot generate machine code into RAM and execute it, because the data memory is not executable.

From the comments in this thread, it seems that to understand how Forth implements a simple IF-THEN-ELSE control-flow, I have to understand the difference between non-immediate and immediate words. I also have to understand the difference between outer and inner interpreter. And I have to understand how Forth generates snippets of machine code (where does that get stored? I thought Forth only has 2 stacks, does it also have a general heap?). Then understand how the THEN token goes back and patches the placeholder address generated by the IF token. And understand the difference between the parsing phase and the interpreted phase of the Forth interpreter/compiler.

But you are saying that the Forth version is simpler than C version which will kinda look like this after it's compiled (Z80 assembly code, it's in my head right now):

    ld a, (variableN)
    cp 42
    ld hl, StringHey
    call z, Printf
    ...
 StringHey:
    .db "hey!", 0

I find that hard to believe, but I accept that you believe that.

> Nowadays Forth is probably most useful on embedded systems.

Nowadays almost nothing is written in Forth. That's the problem with Forth. Even on 8-bit Arduino microcontrollers with 2kiB of RAM, we write programs in C++ (with a little bit of C, on top of hand-coded AVR assembly).

Wow, that's going to take some time and effort to digest, but thank you.

Yes, I think control-flow is easier to understand in assembly language than the implementation you showed in Forth. :-)

I will definitely look into that.

If understanding this special IMMEDIATE mode is required to understand the Forth interpreter for something as fundamental as control-flow, it seems fair to say that Forth is not a simple language. It's not just an advanced programmable RPN calculator An RPN calculator has a program counter, which makes control-flow easy to understand.

In comparison, C is a high level language, but the mapping from C code to assembly language is relatively simple. (Yes, compiler optimizations against the C "abstract machine" can make the resulting code completely obscure. But if we turn off optimization, the resulting assembly code matches the C code fairly directly.)

Ah I see, this is peeling back a few layers of obscurity about the Forth interpreter for me. Let's stick with a Forth interpreter because that seems easier to think about for me.

Are you saying that the Forth interpreter is a 2-pass interpreter? Or does the interpreter go into a special IMMEDIATE mode upon hitting the IF keyword, then it just consumes subsequent tokens without doing any dispatching, until it hits the THEN token? It sounds like nested IF-THEN-ELSE becomes tricky to handle.

How does the FORTH interpreter handle loops? Does the interpreter hit the WHILE token, goes into IMMEDIATE mode, remembers the location of the WHILE, then dispatches all the subsequent code, until it hits the REPEAT token, then branches back to the WHILE?

I kinda know how it works at the user-level, I meant to write that I don't understand how control-flow is implemented in Forth. I say more about that in my reply to a sibling comment: https://news.ycombinator.com/item?id=45334556

Thanks for that. I kinda know how it "works" at the user-level. I meant to say, I don't know how it is implemented.

My mental model of Forth is that there is a simple parser that consumes space-delimited keywords. The interpreter looks up that keyword in a dictionary, which gives the address of the machine code that handles that word. The interpreter either makes a subroutine call to that address (subroutine threaded), or jumps to that address (called "direct threaded" if I recall, where the handler jumps back to the interpreter instead of executing a RET).

But that's where my mental model of Forth breaks down, because IF-THEN-ELSE cannot be implemented in that model. So there must be something else fundamental in the Forth interpreter that I don't understand.

As I like to say: "C is a language that solves a million problems. Forth is a million languages that solve almost nothing." :-P

I've been reading about Forth for 30-40 years. The dual stack is easy to understand. My problem is that I cannot see how control flow works in Forth, e.g. a simple if-then-else.

I think that something as fundamental as an if-then-else should be obvious in a useful language. Heck, it's obvious in assembly language. But not in Forth.

Yup, this is the way.

Regarding Shorts, I have no YouTube history, but I still get Shorts. I hate them. You must be doing something else to disable them.