Graham Sutherland / PolynomialI am fully here for the move to RGB LEDs on network switch status ports. it's one of those features that sounds like gamer gimmick bullshit but is actually super useful.
Alesandro Ortiz 🇵🇷🏳️🌈@gsuberland Also, disco racks.
@AlesandroOrtiz waiting for someone to do Bad Apple on a 42U
EveyPN@gsuberland Yee some DC switches had it too and like makes life so much better with remote hands etc
@evey even stuff like marking tagged ports by colour is super useful. makes it easier to avoid plugging stuff into the wrong place.
23n27@gsuberland @evey What I've always wanted is an eInk display next to the ports for labeling them automatically
R@gsuberland yup, it took us ages to realize that putting one (1) 2020-size Gamer™️ LED on a PCB actually makes for a really good status indicator
@r the one thing that usually stops me from doing that is they're almost all MSL3 and above so it usually bumps the design into the higher assembly cost bracket.
@gsuberland oh is _that_ the reason? TIL
a bunch of them are tagged MSL5, probably even worse
@r yeah usually MSL4/5 is guaranteed to get you bumped up, and MSL3 is dependent on the part size.
AMS@r @gsuberland These days you can get them in 0404 (1mm square) and even with integrated controllers.
🗦new🗧 FireFly@r @gsuberland it's silly to me that smartphones stopped(?) doing them (idk, mine still has one, but I understand they're not in vogue anymore)
Zimmie@gsuberland It’s about time, yes. Would be nice if they also allowed the LEDs to be turned off, either entirely (maybe with a button to turn them back on) or when everything is going fine (no detected faults, link is duplex, 1g or faster, etc.)
Darrell Harmon@gsuberland I'd really like to know what VLAN a port is tagged with before I plug something in but mine doesn't light up until there's a link so I have it set to indicate speed. Sticky notes to indicate VLAN tags. Would be nice if it was more programmable.
@dlharmon yep, VLAN tags were exactly the use-case I was thinking of
@gsuberland If I had my way, link down would be the VLAN color at minimum brightness, link up would be medium brightness.
Gary 
@gsuberland hard nope on that: utter shit UX for people who are #colourblind.
There are three guys with CVD in my network team, a statistical anomaly for sure, but 1 in 12 men are colourblind, and 1 in 200 women.
0x5c@WiteWulf @gsuberland even with very mild red-green colourblindness, bi-colour link speed LEDs are useless to me
@0x5c @gsuberland best hardware interface design I ever saw in this respect was a Cisco ISR voice gateway chassis that had a single LED for each Ethernet port. Off meant no link, 1 blink was 10mb, 2 was for 100mb and 3 was gig.
Very simple, very accesssible.
@WiteWulf @gsuberland yeah that's nice, though depending on the switch and what it (might) need to also indicate I'd want a secondary led
But even that is very doable on standard two-led ethernet ports
But even that is very doable on standard two-led ethernet ports
@0x5c @gsuberland recent Cisco gear (that's the majority of my experience, we don't use any other manufacturer at work) has a single LED above each port, and a Mode button at the end that allows you to cycle through a number of functions such as activity, link speed, PoE status, stack status, etc.
That works perfectly for me.
@WiteWulf I'm aware of the stats having rather heavily studied human visual perception. achromatopsia is extraordinarily rare, so CVD-safe colour sets are entirely doable here, especially given that the subset of CVD that makes up 70% of CVD prevalence is served with a single common distinguishable colour chart. the feature itself is fine, it just needs to be used in a way that's inclusive.
@gsuberland fair point well made, but I've yet to see it done well/inclusively on any equipment that uses, specifically, LEDs to indicate status.
Manufacturers typically just don't care enough, and accessibility gets value-engineered out of the final product.
There's something weird (that you likely know far more about than me) about LEDs that make colours far less distinguishable than white light through a filter. To me there are two colours of LED: [red|green|amber|yellow] and blue.
@WiteWulf true, but the point of RGB here is that it's adjustable to any colour by whoever configures the switch.
RGB LEDs are narrowband (in the case of direct emitters approaching three Gaussian monochromatic sources). you get less radiant flux between the peaks so when you have anomalous trichromacy you aren't getting as much information as you would with a broad spectrum light source of the same colour. the effect will still be apparent for dichromats but to a lesser extent.
@gsuberland ah, this is along the same lines as why purple (which I can't distinguish from blue) isn't a real colour, right?
https://www.sunglassscience.com/post/purple-it-doesn-t-exist-how-do-we-see-it
Purple: It doesn't exist. How do we see it?
I first came across this topic when I was looking into how rose lenses work and why they are so contrast-enhancing in some environments, but downright useless in others.What I discovered is that magenta, rose, purple, etc. do not exist at all.Magenta is an ‘extra-spectral color’, meaning that it is not found in the visible spectrum of light, which is why it is not in a rainbow.You can “see” an object when light reflected from an object enters your eyes and strikes the photoreceptors inside them.
@WiteWulf that's actually a super misleading popsci nonsense thing. purple is a colour, just the same as brown, black, white, or grey. it's just not a colour you can express as a single wavelength (monochromatic source). and even then that's doing a disservice to the extreme complexity of our visual systems and the qualities of human perception.
@gsuberland whatever, purple literally doesn't exist to me, it's purported being is a lie 😂
@WiteWulf yeah fair :P
@WiteWulf for example most people with normative vision can perceive that a deep violet colour is due to UV light reflected off stuff instead of just a violet light, despite us not being able to perceive the UV itself, and many people can even distinguish narrowband far-violet visible light from a UV source. the reason is that our perception includes all sorts of subtle details about the material surface, light scattering, changes with movement/angle, and prior perceptual hypotheses.
@WiteWulf essentially your brain figures out colours based on the relative bleaching of the opsins in the L, M, and S receptor cells, which are sensitive to long, medium, and short wavelengths respectively. so if you have a deficient opsin expression (anomalous trichromacy) or one missing opsin protein (dichromacy) your brain has a narrower gamut of signals to work with, and distinguishing colour gets harder when the light source is narrowband and isn't bleaching multiple opsin types.
@WiteWulf there is, of course, a slightly smaller subset of colours that are distinguishable by everyone but the achromats, which works for literally 99.9999% of humans, which is pretty cool.
also, random fun fact, of the 0.0001% of people with achromatopsia, the majority only have it in one eye because it tends to only affect one hemisphere (bilateral achromatopsia / hemiachromatopsia)