Recent submissions

Comment · 20 days ago
Danfoss are quoted as naming two systems, AK-SM800 and AK-SC355.

AK-SM800 looks like a current product. Danfoss have a PDF dated 2026.06 about it: https://assets.danfoss.com/documents/latest/590332/BC3376298.... This mentions "All new Electronics platform supporting new Operating System (based on Linux)", and specifically advises against connecting it to the internet.. Reading between the lines, especially section 10.5, it sounds like there was a major security revamp in 'release 4.0'. The article mentions "outdated firmware", so perhaps that refers to things before 4.0.

AK-SC355 looks like a product from 2010 (that's the date in the PDF metadata, plus the examples in the manual mention 2009). https://3.imimg.com/data3/ID/LC/MY-3170595/danfoss-system-ma...

I haven't worked with supermarket control systems, so I only know what I've read in the PDFs, but it looks like this is not about throwing out entire fridges to update the control system. If you have the newer AK-SM800, it looks like the suggested action is to update firmware and follow instructions about keeping it well away from internet access. If you have the older AK-SC355, then keeping it away from the internet is even more important. Or, update to AK-SC355.

Comment · 7 months ago
Comment · 8 months ago
It took me a while to realise that you were responding to the article, not a comment here.

You're right in correcting the article, but I'd like to add that for probably around a decade, Erlang had 'sender punishment', which is what 'IsTom' who replied to you is probably talking about.

Ulf Wiger referred to sender_punishment as "a form of backpressure" (Erlang-questions mailing list, January 2011). 'sender punishment' was removed around 2018, in ad72a944c/OTP14667. I haven't read the whole discussion carefully, but it seems to be roughly "it wasn't clear that sender punishment solved more problems than it caused, and now that most machines are multi-core, that balance is tipped even more in favour of not having 'sender punishment'".

Comment · 1 year ago
Right, it's TX-only. The article says so explicitly halfway through:

"As before, this is a transmit-only proof of concept"

I didn't notice that on my first reading.

Comment · 1 year ago
Comment · 1 year ago
Assuming you're refering to Hövding, the recall was overturned on appeal, back in 2023, but the company went bankrupt anyway:

https://discerningcyclist.com/hovding-bankruptcy/

In some types of crashes, the Hövding had better crash performance than ordinary helmets. In others it performed worse.

More generally, it may be that someone comes up with a design which incorporates inflatable sections which is more successful.

Comment · 1 year ago
I looked at a London telephone book from 1979, when I think the record was released.

Most of the phone numbers were something like 01 361 1234, i.e. seven digits after the 01 area code. The _361_ part was bold, so I think that was an exchange number.

A few numbers were something like "Placename 12345". Wikipedia explains why that was.

Comment · 1 year ago
(Author here) No contradiction.

I think what happened is

1. The recording engineer dialled the operator. Could have been pulse dialling, could have been DTMF, doesn't matter.

2. Operator answered and the engineer said "I'd like to call London, collect, number 01xxx831".

3. Operator entered 044 1 xxx 831, and this was transmitted to another exchange in SS5 tones.

I didn't grow up in the USA, but a couple of people who did have said that, yes, they think that at least some of the time, you could hear the SS5 tones and also the initial conversation between the operator and whoever answered the phone. It may be that it depended on the operator, since they probably had a mute button, and maybe on the particular exchange the operator was in.

In the film, we hear from 3 onwards.

Comment · 1 year ago
659 8890

That one's just ordinary DTMF. I recorded the audio, trimmed it manually and then made a spectrogram like this:

    sox gun1.wav -n rate 4k spectrogram -m -y 500
The 'rate' switch is to cut down on how much of the frequency space we can see. I left the audio as stereo because there's less music power on one channel, making it easier to see the tones.

(And google finds quite a few pages confirming those digits)

Comment · 1 year ago
You're right; I think CCITT5 is just another name for SS5, because different groups were writing standards. Bell called it one thing, CCITT (an international standards group) called it another thing. And then in the 1990s, the CCITT renamed itself to ITU.
Comment · 1 year ago
The discussion is new to me.

I started analysing the audio because someone sent me a link to the film (The Wall) on youtube and asked me about the signalling. Once I'd decoded the telephone number, I tried googling it, to see if someone else had already figured out what it was (a US local number? the number to a US operator? the number a US operator called to talk to a UK operator? the number a UK operator dialled to get a London number?), but nothing came up. There's quite a bit of good discussion about that in the comments here.

A week or two later, I tried googling 'Pink Floyd Telephone Call', and found that the audio actually comes from the album, i.e. it's not just in the film, and a bit more information about how it was made, and put that in the addendum.

Comment · 1 year ago
Interesting.

Normally, films use deliberately fictious numbers, e.g. in the US it's always xxx555xxxxxx. Wikipedia says the UK uses various area codes for the same thing, including 011x and 01x1. The Pink Floyd number is a bit unusual---it's not made up.

According to a previous analysis, the call was the album's "Chief Engineer James Guthrie who called his own London apartment", with a neighbour answering the phone. Someone probably knows roughly where James Guthrie lived in 1979/1980 and what the area code there was. But I don't.

Comment · 1 year ago
Author here.

Yep, 44 is the UK country code. The problem I got stuck on is that the rest of the number, 1831, didn't make sense. I assumed the number was complete, since it had the right start and stop signalling (KP1/KF).

It's not long enough to be a London telephone number, and, today, I think London numbers start with 020. The UK numbering plan has changed several times since 1980, but I couldn't find a time between 1980 and now where part of 1831 was a London number.

Later on (in the addendum), it turns out that others took a look at the signal in the time domain and spotted a splice, i.e. digits are chopped out of the middle of the number, so the area code probably isn't there at all. It could be that the area code starts with 1, and then the phone number ends with 831.

Comment · 3 years ago
There's no smoke coming out.

Best guess: you're confused by the word "smokestack". Smokestack is a synonym for chimney. The power station is closed. The chimneys are still there, but there's no smoke coming out of them.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton