My car came with a cheap low-quality aftermarket combination car stereo and GPS. It would only play music from discs (no USB support), didn’t even support variable bitrate MP3s, and the GPS didn’t know the car’s location half the time.
I finally replaced the head unit with a plastic bracket and a standard old school in-dash stereo system. Much nicer setup.
It’s the cars which can’t have their head unit replaced with a simple stereo which worry me.
> It’s the cars which can’t have their head unit replaced with a simple stereo which worry me.
I agree. There's a lot of room for innovation, consumer choice, and competition that was previously guaranteed to be available in those one or two DIN units of space.
I haven't seen them present in any new car I've looked at in nearly ten years, but they were in every car I saw from the 50 years prior.
Denso head units for Toyotas can play MP3s, but the order of the files is when when they were copied to the SD drive. And you can't sort them any other way.
Yes, I can move the files, but it's also ridiculous in the 2020s that I have to resort to doing this workaround instead of the software reading ID3 metadata and sorting by one of the well-known fields.
Thanks for the tip on fatsort, it does save me a bit of trouble.
Embedded controllers of that class are usually not able to do such things. And if they are, it becomes a bottleneck when you plug in say a 64GB device. Five minute drive? No music because the stereo takes six minutes to sort the files on startup.
Another drawback of the death of DIN aftermarket stereos is that they stopped innovating there, I last updated my stereo in ~2015 and not found a more advanced replacement since. So never got one to accept a >256GB drive, though trivial elsewhere today. Raspberry Pi level tech.
My car stereo, in theory, cannot accept a drive over 32gb in size. In practice, I have a 128gb drive in there. I used a third party utility to format it as a FAT32 filesystem to be 128gb in size. The car stereo only claimed a 32gb limit only because that’s the largest fat32 filesystem Windows will make from the GUI.
Yeah, that's often the case. It's not really the size of the filesystem I think, my description was simplified. Believe when I got over N thousand songs it stopped working, which defeated the whole purpose of having my entire music collection on a drive. Still works on the home stereo though.
Fun fact: the FAT32 limit - which, by the way, was literally just added by Dave Plummer because he thought anything larger would have too much wasted slack space - made its way into the SD card spec. It is perfectly valid behavior for a device to refuse to read a >32GB card with a FAT32 file system, or a <32GB card with exFAT, or a card with multiple partitions, etc. None of those are actually SD cards as they are formatted.
In practice, nobody in their right mind would design a device's firmware that poorly, unless they were deliberately trying to force users to stay within the bounds of the device's QA test procedures. I don't have any examples of this. There are devices that will misbehave with such a card, however: for example, the Nintendo 3DS will happily use an oversize FAT32 card, but the Data Management screen will take way longer to load for some reason.
Also fun fact: the first time you plug a correctly formatted SDXC into a Nintendo Switch, it "downloads an update" and restarts. Presumably Nintendo wanted to save money on exFAT licensing fees so Nintendo is only paying for each unit that actually uses the exFAT library?
3- Yup, nailed it, same reason WMA and Flash require "activation" on the PSP and why newer MPEG codecs like H.264 and AAC require an "optional media update" on the Xbox 360
2- The slowdown usually comes from scanning the FAT to count the free space, with the PSP being notorious for this
Another related point is the varying support for the "semi-official" 64K cluster size, which gives a curious effect on the 3DS: in GBA mode (officially only used for 10 games only distributed as compensation to those who bought the original model at full launch price), a strip of "junk" (coming from the virtualized GBA's VRAM, IIRC it's just wraparound) will be shown at the bottom of the screen...
I understand — I am only counter-pointing the following statement:
> Five minute drive? No music because the stereo takes six minutes to sort the files on startup.
With FLACs, you'd have even fewer files in 64GiB, so sorting a full list should be even quicker — Raspberry Pi HW (any generation) would do that instantly for all intents and purposes.
I don't have a raspberry pi in the car, I have an older DIN car stereo. RPI was used as example that this isn't hard technically. Also left 64GB behind many years ago for non-car applications.
The counter argument to that is well designed modern cars have great info-tainment systems which get regular updates. My Tesla gets regular software updates, has upgradable hardware, and won't be subject to these kinds of issues for decades. Old cars encouraged exchanging and needed aftermarket parts because they didn't care or try, so the products were low-quality and often failed.
> My Tesla gets regular software updates, has upgradable hardware, and won't be subject to these kinds of issues for decades.
Teslas from four years ago with Intel Atom processors and HW3 are already getting different and lesser updates than new Teslas, without any hardware upgrade plan in place.
A Tesla you buy right now is running HW4 which is inferior to current Cybercabs and HW5 has already been announced.
Legacy automakers are still using the "ship it and forget it" approach. If I want my 2023 Hyundai to have the same features as a 2025 (manual battery preconditioning, Google Maps directions on HUD, wireless Android Auto, etc), I need to root the head unit and install a bunch of vibe coded packages from randos on GitHub.
Although to be fair the vibecoded packages are probably higher quality than what Hyundai would've shipped.
Correct. Without root, the only way to heat the battery was if you use the cars built in nav to get directions to a fast charger
The problems being:
1. It didn't always recognize a destination as a fast charger
2. The built in nav is missing recently opened fast chargers
3. It only does this if you have the "preconditioning" setting enabled, but the car has a bug where if you use the (maybe only US version) Hyundai app to change the AC or DC charge limits it unchecks that setting.
Ah, I have an Equinox EV and it will automatically precondition based on the on-board maps, which, in my case, is Google Maps. I don't have the luxury of Android Auto or Carplay unless I jump through some hoops and self-publish an app in the Play Store. And even then it is janky.
Even so, I have a button in the battery settings which can start the preconditioning whenever I want to. It's been a fantastic car, but I am envious of your DCFC capabilities. But not of your ICCU. :-)
> well designed modern cars have great info-tainment systems which get regular updates. My Tesla gets regular software updates
And tracks you everwhere you go, resulting in that data being available to hackers, marketers, and any sleazeball that wants to extort you by suing you.
Not unique to tesla, of course, but a property of all those 'well designed modern cars'.
One legal way to share news stories which generate controversy is to have an AI make a summary of the article, rewording it to remove any copyright from the article (this legal trick only works when the news site in question is reporting on factual events, since facts can’t be copied as per Feist vs. Rural 1991, and I make sure that the summary has “a different sequence, pace, and setting than the original article” as per legal judgments on whether an AI summary of factual events as reported by a news source is subject to copyright).
What about just a copy and paste into an HN post comment? What is the legality of that? What if someone who reads a share, and then pastes it as a comment?
While Righthaven LLC v. Hoehn did rule that copying and pasting an entire newspaper article is fair use, this ruling was vacated on appeal and instead the lawsuit was nullified because they ruled Righthaven never had the right to bring the lawsuit in the first place. So, yes, one can be sued for copying and pasting a newspaper article verbatim.
It might be useful for your own memory since you can decide that you trust what was written. IE you trust that when it was written, you agreed at the time that it was an accurate summary, and later you only have to trust your earlier self.
But "My ai said that the site said ~essentially~ this on this date." is useless for anything else.
But it suggests an interesting idea. Maybe if we had some ai equivalent of trusted vetted accountable court reporters who write down notes and draw sketches in rooms where cameras aren't allowed.
That is purely second-hand reporting that we more or less trust, partly because there is usually more than one of them from competing venues reporting on the same event, partly because they are people with something to lose if they are unfaithful or even merely inept, partly because the facts are ultimately vettable with some effort for those who really need to.
There's no such ai yet. Even if one exists that could do the job, there is no consensus or basis for it, for everyone to recognize it as a trustworthy source of second hand reporting. But maybe if there were more than one and operated by different camps, they could grow to become recognized as a valid source of info.
Personally, as scummy as archive.today is at times, I can generally trust them to have an honest archive about, say, someone’s drunk post before said drunk deletes the post.
Old still exists and gets heavy use a decade or so later exactly because regular reddit just doesn't work at all, even when you're logged in.
But even aside from opinions on their UX, they're aggressively experimenting with the login more... Even on the new UI I now semi-regularly see that I can read the initial post and one or two comments but as soon as I scroll past or try to expand a comment thread get a login prompt.
I have moved my more complex scripting from Perl/PHP/Python to Lua for a lot of reasons. One being that Lua (even PUC Lua) is a good deal faster than Python (or /bin/sh); another is that, if I want even more speed, I can use LuaJIT (which keeps me stuck at Lua 5.1, but Lua 5.1 is perfect for my needs).
I wish Lua 5.1 could become part of POSIX; there are a lot of implementations of it out there.
Yeah, this is sort of the disconnect. Lua is a fast dynamic interpreter, sitting well above Python but rather below e.g. V8 in its ability to run scalar code over "interpreter types" like doubles and strings and vectors and hash tables.
The upthread comment was about an idea for "inline Cython", which was then hijacked because that's largely what numba is. These are technologies aimed at optimized native code running over native types to do stuff like numerics and ML and other HPC tasks. Lua isn't any more responsive to this domain than Javascript has been, it's not really the subject under discussion.
LuaJIT handles double, strings, and dictionaries at near C level performance. Also, arrays, but Lua’s data model has dictionaries and arrays be the same data structure. Lua 5.1 (i.e. LuaJIT) only supports double numeric types,[1] and its strings are immutable.
Lua 5.1’s edge over Javascript and Python is that it’s a lot smaller and simpler.
[1] It’s possible to recompile Lua 5.1 to have the numeric data type be an int, but the standard compiles of Lua5.1 only have double precision floats for numbers. Bitwise operators are technically not part of Lua 5.1, but a library to add support is easy and is usually done: Most Lua5.1 forks including LuaJIT and MediaWiki’s version of Lua5.1 have such a library available. Later versions of Lua have both int and float support.
> LuaJIT handles double, strings, and dictionaries at near C level performance.
Yeah, but again you don't reach for C when you have a string or hash-based logic problem, which is the reason that other platforms are favored. You reach for Python[1] to solve those problems, and it works fine.
You reach for C when you care about cache footprint or memory layout. You reach for C when you need SIMD parallelism. You reach for C when you care about doing the math in single precision or 32 bit integers and will handle the precision worries yourself.
You emphatically don't reach for python (or Lua) in those situations. But notably you can reach for numba, and implement your small core of performance code in python next to the other logic you already have. And there's value there.
Screaming about using Lua instead is just nonresponsive. Lua is arguably better at performance code than Python. Lua is absolutely not faster than numba, nor will it ever be (absent a numba-like translation layer) and repeatedly trying to argue so is very much missing the point.
[1] Or fine, Lua or JS or whatever. But I'm trying deeply not to let this turn into a frothing language flame, and this is after all a python discussion.
> Screaming about using Lua instead is just nonresponsive.
The point of those comments is not to hype Lua, it's to show what a JIT can do.
The idea is that a python JIT could be just as good as Lua for the parts that aren't suited to numba.
> Lua is absolutely not faster than numba, nor will it ever be (absent a numba-like translation layer) and repeatedly trying to argue so is very much missing the point.
Nobody is arguing that. People are agreeing that numba is fast, but they also want other non-number-crunching things to be fast. It's "yes, and". They're not missing the numba point, they're agreeing and you're missing the additional point they're responding with.
If you're saying V8 is quite fast why did you say the world is "waiting for a counterexample" to JITs on very dynamic objects being slow?
When OP wrote "inline Cython" they didn't seem focused on numerics, they talked about "proper C speed", and they mentioned modifying the language. I don't think it's getting off-topic to suggest a deeper integration that lets the JIT be very intimate with the interpreter and truly inline things in a way that bypasses language-crossing and type-dispatching overhead. As long as you're in the hot path, let it look up object fields in a couple instructions.
> When OP wrote "inline Cython" they didn't seem focused on numerics, they talked about "proper C speed"
Isn't that basically what I said? Lua isn't "proper C speed", nor is PyPy, the cancelled Python JIT in question, nor even V8. That's a different problem area with different technologies. The fact that you like coding backend stuff in Lua or whatever isn't really saying much about whether numba is great.
I agree Numba is great. The reason I stay clear of Python is because the Python2 → Python3 transition burned me; the bureaucrats at Debian decided that, since my open source project had one Python2 script in it I wrote many years ago, I had to update my code to make that script a Python3 script. A very toxic and arrogant attitude; old versions of languages should not be killed like that when there is so much code written in the language, and when it can be very difficult [1] to convert a script from Python2 to Python3.
Lua5.1 is much more stable by the virtue that the only updates it code gets are very minor changes to fix any lingering security issues it has (the last Lua5.1 CVE is from 2014).
[1] I guess these days a LLM can do the grunt work of converting a Python2 script to Python3.
V8 gets a hell of a lot closer than this experimental JIT. Numba speed for number crunching and V8 speed for non-number-crunching would satisfy OP, I think.
Numba is great at its niche. But OP directly said Numba is "not really" what they were looking for because it's so narrow.
Well done. It’s important to know the difference between a real world exploitable bug and hot air; as an open source maintainer, over half of the bug reports I get are “hot air” bugs.
I’m surprised this level of surveillance from a smart TV (if this is real; right now the only sources are social media) doesn’t open up Samsung to a lawsuit.
To wit: There was a lawsuit where Susan Clements-Jeffrey sued Absolute Software (Lojack) for a stolen laptop she acquired; Lojack used surveillance to track down the stolen laptop and Susan successfully sued for invasion of privacy (Lojack settled instead of trying to fight it).
A couple of days ago I set up a samsung smart, ai, yadda yadda tv for a family member whose technical expertise boils down to "I unplug the computer to turn it off".
It refused to let me set it up for cable without granting it internet access (as in the bundled remote control didn't work):
First I ended up on a bunch of websites that hallucinated non-existent physical buttons on the tv
I then tried to scan scan the modal qr code which didn't work because the phone I used didn't have google play services.
I had to get another samsung phone (which immediately picked the tv up and asked me if I want to set it up without scanning any code) and after I agreed to share the data with some 450 partners, I got an error saying "setup did not complete, please try again"
Though at that point the builtin remote started working beyond on/off and mute and I was able to finish setup that way.
In the end, I can't even change the wifi password because the phone will share it over bluetooth with the fucking spybox.
Fuck samsung. You were shit in the galaxy S1 days, you're even worse now.
I've never once signed a contract for a TV. They're routinely set up a single member of the household. They're the single most plausible place for "maybe my kid / guest / cat hit OK on the remote".
Very different from a single-player video game, or a website with a login form.
It has hit the point where I’ve told people reporting bugs with my software that I won’t acknowledge the bug unless they provide a patch to mitigate the issue.
A lot of these bug reports are incorrectly filed. Let’s look at one issue, which I am sure was found with AI assistance:
CVE-2026-40719
Here’s the issue: An IP authorized to perform recursive queries can cause the DNS server to be non-responsive if it sends hundreds of specially crafted packets a minute (this can be changed to millions of packets a minute if one sets a parameter in the configuration file). Normal service is restored within a minute of the attack being stopped. The workaround would require me to essentially rewrite the recursive part of the DNS server.
Since the DNS server, like any server in a LAN, is already vulnerable to a flooding attack, this is just a flooding attack with extra steps.
I told the guy that was interesting, I’m not being paid enough to rewrite the DNS server to be immune to the attack, and people shouldn’t be using its recursion (it has very limited IPv6 support, for starters), and if they do use its recursion, only authorized IPs (e.g. 127.0.0.1 or 192.168.1.0/24) should be allowed to perform recursion.
It wasn’t a serious enough attack to merit a patch, and anyway the guy reporting the bug didn’t even have the decent courtesy to give me a patch.
Point being, a lot of these CVE reports out there which supposedly are high impact bugs are trivial things like “someone can overload a server for about one minute if they send hundreds to thousands of specially crafted DOS packets”.
(Today, I got an issue which, as far as I can tell, looks to be a Linux kernel issue and not an issue with my code. I told him to show me how he would patch my code to “fix” the issue, but so far have gotten nothing but crickets)
“Forced” is a little strong. RedHat/Fedora filed a bug report, and I felt it was prudent to have a patch for them to use, even though I explained why they didn’t need to use the patch.
I’ll bite. I have written a recursive DNS resolver. It’s buggy, mind you, it doesn’t handle improperly configured domains bug-for-bug the way BIND9 does, it doesn’t support IPv6 for resolving records (it can still resolve AAAA records, mind you), but it’s about 80 kilobytes in size, runs nicely on 32-bit systems, and does full DNS resolution starting from the root servers.
Does the fact I have written this resolver make me liable should it not block the domains they listed which may contain copyright infringing content? Am I now required to add these domains, and make them part of a mandatory block list? Will I still be held liable if people can still access these allegedly copyright infringing domains by modifying the source to my program?
Keep in mind that when Quad9 had a similar lawsuit in Germany, they blocked queries from German IPs, but the idiots said Quad9 was still liable because people could use VPNs in Germany to bypass their geoIP blocking.
Quad9 being forced to block these domains will achieve nothing. Will these copyright enforcers now go after Google for 8.8.8.8? Cloudflare for 1.1.1.1? Level3 for 4.2.2.1? Will they go after ISC for BIND? NLnet Labs for Unbound? CZ.NIC for KnotDNS? Me for MaraDNS? Even if they enjoin me from making a recursive resolver which can resolve these domains, that won’t stop anyone and everyone from forking my code (see also: Twitter trying to take down Nitter).
> Does the fact I have written this resolver make me liable should it not block the domains they listed which may contain copyright infringing content?
I use unbound and have a similar question. Shouldn't we just encourage everyone to run their own resolvers (it's turnkey for the most part!)? Will Sony/beIn start going after the root servers and enforce a global block?
I have used Fontforge to remove all ligatures from a programming font; e.g. my monospace font of choice is a aggressively subsetted variant of Fira Code with no ligatures whatsoever.
For example, ligatures in Fira Code cannot be disabled when using the Cygwin terminal in Windows (mintty). In that application, for example, “===” looks like three continuous long lines atop each other in Fira Code, with no way to make it look like three equal signs again.
“Fira Code is not only about ligatures. Some fine-tuning is done for punctuation and frequent letter pairs.” Note that without ligature support, a lot of that fine tuning is lost.
I finally replaced the head unit with a plastic bracket and a standard old school in-dash stereo system. Much nicer setup.
It’s the cars which can’t have their head unit replaced with a simple stereo which worry me.
reply