Halogen to be banned

It's interesting to view a narrow spectrum light - like say yellow low pressure sodium - then match it by adjusting the RGB drives to a monitor. No two people will get the same result.

Is your experience is with CRT? These are associated with poor purity wrt to LCD

BICBW

You can use a light box if you prefer. Any method of generating a variable colour light.

it is to demonstrate that we see colours differently. Two people with nominally perfect colour vision will disagree about the best match. And it's not subtle, either.

Obviously, we agree on the name of a colour by experience. But if we could swap brains, would that still hold true?

Thanks.

My incorrect belief was that cone response would be largely non-overlapping and uniform from person to person. And that colour blindness would be primarily caused by missing or mixing of cones' responses.

This is an interesting article regarding tetrachromacy in humans.

I can only guess you don't have a clue what you're talking about.

Not sure about that. You seem to know a whole lot more than T i m. Thanks for the info.

But one he makes over and over about pretty well everything (especially if Google tells him so).

And much of my stuff is Zigbee (as is Philips Hue I believe).

But did you see the fuss the DickHeadxx made (and is still making) about my *choice* to move away from the RPi as a platform for Home Assistant because of his *insistence* that you could set one up to not write to the uSD card, even though logging is very much part of what I wanted my system for! Because *he* didn't see the point, I was wrong (apparently)? Of course, I take no notice of his narcissistic drool ...

It's now currently running very smoothly on an 4GN Odroid N2+ with a

Cheers, T i m

I very much saw the point. You could have saved the log file to any attached or network storage but didn't seem to know how.

Yes you were wrong to make the statement, "I think the RPi is not considered the best choice for a serious / long term system is that it runs on an SD and that's not good for long term writes".

You didn't think and anyone with knowledge about the Pi would have known this was a lie.

Not particularly cheap. You could have fitted a 500GB portable hard drive to your Pi for greater life-expectancy and reduced cost.

I'd always thought (and I could well be very wrong!) that Hue lights (and other remote-control bulbs) communicated by signals modulated onto the mains wiring, rather than by RF (wifi/bluetooth). I suppose the acid test is to connect one bulb by a very long mains cable so it is a long way from any other bulbs, and see if it can still be controlled.

I may be about to learn something... ;-)

I suppose within a house, there is the reasonable assumption that every bulb will be close enough to at least one other bulb that a daisy-chained wireless communication method is always viable, even when the end-to-end size of the area covered is large.

Does anyone know why Philips Hue still have a maximum of around 50-60 bulbs that can be controlled by one Hue logon (and maybe one Hue hub)? It's the one thing that customers are crying out for on the Philips forum, especially with the increased use of a larger number of ceiling-mounted GU10s to light a room rather than a much smaller number of brighter omnidirectional bulbs. It's a shame there isn't a technology which allows a set of bulbs (defined by their MAC address) to be controlled as a single unit (only consuming one "unit" out of the 50-60).

Agreed that you don't have to write a log file to a Pi's SD card. But if you write it to a networked drive, *that* drive needs also not to be SD, otherwise you are no better off. And writing to a physically-connected disk (whether SD card or USB-connected spinning or solid-state disc) is probably more reliable than writing to a network drive, where writes (and reads) will fail if the network server (probably using SMB) fails or the network infrastructure fails.

How does the Pi (or indeed any computer that has a solid-state disk) manage to read/write/overwrite a swap file, where the contents will be changing very frequently as different memory pages are swapped between RAM and disc, without rapidly degrading the disc? Am I correct that it is the number of writes (as opposed to reads) that causes the gradual damage to an SSD?

Actually people have measured the responses of the cones and have found that they are not RGB; one lot are blue sensitive and the other two are in the yellow; yellow and greeny yellow. Rather close, but the brain can make sense of them and detect subtle differences in colour. Non-primate mammals such as dogs and cows just have the blue and the yellow.

It appears that, in (most) primates, the gene for the yellow cones duplicated and one copy mutated to change its frequency sensitivity.

The RGB idea came from Maxwell who was a Famous Scientist who made the first colour photograph (of a tartan scarf) by taking three pictures through red, green and blue filters. But, in those days, photographic emulsions weren't sensitive to red at all, so the "red" filter must have been passing some other colour, perhaps UV.

But he was a Famous Scientist.

The Pi is able to use RAM and doesn't generally need a separate swap file. Although it is a facility you can enable but is not recommended as it severely shortens the life of the SD card.

Most OSes with sufficient RAM don't need a swap file, or should I say page file?

you can stop logging altogether or write it to ram disk

you can eliminate swap altogether, or write it to a ram disk. Pointless, but easy to configure if you cant get rid of swap altogether

I cant answer for the NVRAM chips used in pis, but on a 'real' SSD the lifetime seems to be *better* than a real disk

Erm, it wasn't the 'electronic starter' bit that was unusual (or unknown / function to me) but this particular 'built in' one, as seen top left of the picture:

I knew what it was and what it did, just that I'd not seen one like that before so wondered how common they were (in that format, and partly why at a quick glance I thought the fitting was 'different' to all the others I've dealt with over the years and that the difference might have meant it was HF (It wasn't obviously, once I'd got it down and could see inside it easily).

Cheers, T i m

Some 7 or 8 years ago when we had the downstairs shower room refurbished, I specified that the 4 halogen downlighters that SWIMBI had decided to go for, use 12v lamps fed from transformers discovering that the "transformers" I was expecting to see fitted were in fact 60W rated smpsus which go into CC mode once the current demand tries to exceed the

This made them even kinder to the 35W 12v halogen lamps that I'd been expecting even the older 12v transformers to be. The switch on surge in this case being limited to just a 67% overcurrent rather than the 100 to

Have you got a photo of it before it was disassembled?

That does look/sound like something I have never come across before.

Most unusual.

No, sorry, but it was just a single, steel, white, Fitzgerald, slimline 5', florry fitting with std keyhole cover retaining screws all the components you see in the picture inline inside and (short of the chock-block mains connection around the middle) nothing much more?

Strange, given how many you must have seen the insides of whilst fitting.

FWIW I bought it new from a local / reputable / trade-counter 'old skool' electrical parts supplier (now gone) and probably about 30 years ago?

I opened the starter up for you OOI:

Cheers, T i m

You really should sent that to Big Clive and ask him to do a tear down.

The must be a "tomorrows world" version he mentioned.