Wireless is a trap (2020)
benkuhn.net
benkuhn.net
I tied a ping pong ball to a piece of string and used a vacuum cleaner to suck it through the pipes, then used the string to pull a stronger rope and then the Ethernet cable. So now my Ethernet goes from my office to the wireless router via the vacuum ducting.
It would be pretty expensive to turn every coax run in my house into a hardwired ethernet switch but well placed hardwired APs obviates the need to do that.
This is assuming you have coax rg6 or better already in the wall
In my case, I didn't build a MoCA network, just used these devices point-to-point. If you go full MoCA, think of it like a distributed hub - no switching, no routing, no security. In fact you should put filters to prevent neighbors from listening in using their own MoCA adapters. For me, point-to-point worked. YMMV but Ethernet over Coax is definitely worth looking into for upto 1gbps speed, which is plenty for wifi, netflix etc.
Wow. Who knew that inbound cable internet has outbound data capability? Thanks for posting.
Is this a suitable filter (-70db)? https://www.amazon.com/GLP-1G70CWWS-Filter-Eliminate-Multi-R...
The physical medium isn't directional, so there's no reason to assume it would not transmit and receive. A coax cable is equivalent to three wires with better isolation from the environment (and with one "wire" usually used for ground).
There's an asymmetry in commercial broadband over coax, but Im not sire why that is precisely. Perhaps someone familiar with DOCSIS etc. could say.
Ethernet runs are cheaper in parts, but MoCA adapters are crazy effective at avoiding large drywalling jobs.
G.hn is fast, but all of the G.hn devices I've found seem to use the same chipset (and presumably similar firmware, since the configuration pages have identical options, but different branding), and the chipset seems to buffer a massive amount, which means a single hiccup while using a high-bandwidth application makes the network unusable for a while.
Example: while streaming video over a G.hn adapter, I was running ping. There was a brief hiccup, and then ping times exceeded 20 seconds (!!), finally recovering after a few minutes. I eventually ran a program that ran on devices connected to each side and would reboot the adapter if ping times exceeded 500ms. This made the network far more reliable.
The problem with P/L is that it can work, until suddenly it doesn't.
I ran a set of TP-Link powerline at my last 2 places: in both, I rented a room in shared flats (apartments).
In the first I got a very poor intermittent wifi over 3 floors from my landlord's wifi: 2-5 Mb/s if that. With powerline I got a consistent 60-70 Mb/s.
Yes, they needed rebooting regularly, but it was totally worth it.
2nd place, here in Prague, I bought new units, because the tech had updated. My old landlord still uses my old setup as a wifi repeater, over 4Y later.
New place, my flatmates had put Ethernet across the floor, so it was in very poor shape: walked on, trapped in doors, insulation frayed, etc. Connection was very intermittent.
In that place, powerline allowed me to remove the messy dangerous ugly cable and gave me a steady reliable 200-250 Mb/s connection.
I think the complaints here may be from Americans who are famous for having mains power like wet string, running at half the voltage of the rest of the world. Across Europe it's all 220V, with mandatory earth (ground) pins, and every dwelling is on a separate ring main, so nothing your neighbours do affects your home and there is zero leakage.
Also, electric kettles work, and as I run on tea, that's important. :-D
I could have put Ethernet in, but adding a wire under the floor was vetoed in the interests of not making a mess.
Don't listen to them! Run that Ethernet!
Lucky you, you had a preinstalled system of cabling conduit :)
Before I had my basement finished, I installed a bunch of PVC conduit for the network wiring, so it would be possible to upgrade or add to it in the future.
It was more of a challenge than expected. The ducting had ramps, multiple corners, debris/rubbish, and a mouse skeleton to navigate around. All of this wasn't helped by the terrible latency and needing to move the phone to different places to reduce distance and attenuation through the floor. I got stuck quite a few times but could always use the string to pull back a bit when I got wedged on a rock. Sometimes I just needed to gun it to get over an obstacle.
In the end, the car and ethernet cable were cheaper than a good wireless card or POE kit, and I got to give away the car to a kid.
My uncle worked for a company that designed and installed commercial ductwork. The joke he liked to tell while they were going in: "Make sure Bruce Willis can get around in there."
This is what I have and it's great. I also have a toe kick adapter, so I can use a regular broom and sweep the dust to a little port under a kitchen cabinet and it automatically sucks it up.
I don't have allergies, I just never liked the way it smells so dusty after using a vacuum cleaner inside.
I grew up in a house with one. I haven't lived there since 1998, but, yes, we used it all the time. As my sibling commenter lreeves says (https://news.ycombinator.com/item?id=31361950), it's good if you have allergies.
Old vacuums were heavy.
Now, they're more useful to repurpose into shop vacs.
I have a Miele and it is much quieter than any other vacuum I have heard.
On the minus side the long-ass hose is a huge pain.
In the 2020's you can just get a robovac for daily cleaning and a smaller unit for manual stuff.
Of course these days you might just give up the land line entirely to get gigE over all 4 pair and just remove the old rj11 jacks.
10base-T and 100base-Tx both use two pair even for half duplex.
You need only one pair to run half duplex 10/100bt (contrary to what's stated above), and two pair to run full duplex 10/100bt. In FD, there's one pair each for tx/rx. 10/100bt are very similar in this regard.
The different cable categories mostly relate to the wire quality and the number of twists in each pair.
gigE is very different and uses all 4 pairs.
How do you wire your one pair for half duplex so it works with commercially available Ethernet equipment? I'll grant you that only one pair will be active at a time, and that 10base2 works with two wires, but you need two pair for 10baseT-HD. I'm very happy to be wrong, please point me in the right direction.
There's also 10Base-T1S, but it's a new specification (2020), and seems limited to automotive and industrial applications, not generally available.
With 10baseT, the T stands for "twisted pair," as opposed to coaxial cables. Everything we're talking about here is going to be baseT, for twisted pair.
In 10baseT full duplex there are only 4 of the 8 rj45 pins used. In 10baseT half duplex there are only 2 of the 8 rj45 pins used. I don't remember which pins off the top of my head, but you could easily experiment to figure it out. I'm sure it's using one of the two pair used in full duplex.
You don't need to wire anything differently. When you select half duplex mode in your driver it will start communicating over only 2 pins. This sometimes happens naturally with faulty or broken cables, when it looks like the other pins aren't connected properly.
100baseT is more or less the same as 10baseT in terms of pinouts, but at a higher frequency.
If you get traffic on the RX pins when you're sending on the TX pins, in full duplex, that's great. In half duplex, it's a collission and you stop sending your current packet and send a jam signal instead through the minimum packet time.
In 10base2, the RX check is more nuanced, because it's reading from the same bus that TX happens on, collision is only signalled if the RX differs from TX. In 10baseT, I think (but again, would love to be corrected), the expectation is that an ethernet hub will isolate a sending port from receiving its own transmission to the shared bus inside the hub, and of course switches are switches.
For the simple case of two Ethernet devices connected with a crossover cable, there is no possibility of actual collision, but if they run in half-duplex mode, recieving a packet while sending will be processed as a collision anyway.
If you connect only one pair, you can connect RX to RX and TX doesn't go anywhere, TX to TX and TX goes (subject to collissions), but nobody listens, or Peer A TX to Peer B RX and get a one way connection; peer B may detect link up because it gets link pulses, but peer A won't. I'm not sure what Auto MDI-X does in that situation, but it still won't work.
Autonegotiaton usually doesn't do anything smart about failing pins or poor connections either. If you've got two pair connected, you can autonegotiate (at 1Mbps, with the 10baseT link pulses) to the best connection offered by both ends, which will then fail to work if you negotiated to GigE and only have two pairs, or you negotiated to 100M and the cable is really terrible. It's getting somewhat common for OS drivers to detect this negotiation succeeds/connection doesn't behavior and restart the negotiation with fewer options though.
If I have some time today, I'll make a single pair cable and confirm, but if it were as easy as you said, there would probably be a web page telling people how to do it.
If 10bT really does always use 4 pins then I wonder what would happen if tx+/rx+ and tx-/rx- were shorted? I wouldn't be surprised if there's no page illustrating it because why would you ever want to?
One solution was to use the existing phone line in each room and put an ADSL modem on it. They got Internet (and kept analog phones) in every room by just upgrading the endpoints.
I use it to get my mm-wave 5G 2Gbps signal from my antenna on my garage over the same coax my cable TV runs on to my lab in my basement and have a wifi AP for the rest of the place.
Ya, sorry, I stream but I still like my cable, shrug :)
very famous diagram !
10Gbps !
:)
Also, for the cost of either powerline or MOCA adapters you can buy a big roll of CAT6 CMR and a fish bit and run ethernet through the walls (just do a good job on the patches if you're renting).
I'm also just laser focused on good consistent latency though after years of suffering awful cable internet and trying to minimize all latency that I could control (https://i.imgur.com/i84mIsD.png was what it was like during the day in 2020 in the first wfh period. Luckily I've only been getting consistent spikes up to ~200ms throughout 2021 /s).
edit: That ping was on hardwired ethernet with no load as well, and Rogers (realistically the only ISP where I live in Canada because my available Bell DSL connection is too slow) was throttling uploads to 4mbps.
It worked great. I got a consistent 80Mbps through them. I was happy and probably commented here or elsewhere about how they can work, or, "don't dismiss it, it might just be the trick."
Well, now they're pretty flaky. They still work just enough for me to keep using them, but other WiFi issues I'm having are convincing me that I gotta stop being lazy and run some cable to all my rooms. (I even have a crawlspace! It will be relatively easy if I can just get cracking on it.)
I don't know what changed between then and now. Phone chargers on the circuit? That random smart plug I bought last year? My A/C cycling in the summer? Doesn't matter. Powerlines are a hostile environment for high-speed data.
They basically turn all the electrical wires in your house into a giant antenna and radiate broadband noise.
I honestly don't know how the damn things are even legal by the FCC given how noisy they are.
And they do it very poorly, I've never been able to get powerline networking to work well. I had trouble reaching the Wifi node in my office from the bedroom on the far side of the house, I thought powerline networking was the answer, but the link was not that fast (around 10mbit), with very variable latency, anywhere from 5msec to 100msec, and around 2% packet drops. Both outlets were on the same leg of my home power panel, I verified it at the breaker box.
I finally ended up putting up a couple Unifi Nanostation M2's (the bigger ones with the 10dBi antennas) aimed at each other through the walls to act as a point to point network, and it was faster and more reliable. Still not super fast, I get like 30mbit, but latency is a nice constant 7msec, with little packet loss.
But in a previous house, the walls were real plaster with metal lath behind acting as a faraday cage, and I had no choice but to use powerline networking (it was a rental house, so I couldn't easily run ethernet)
They are also the other issues people reported.
IMO, they are fine in a rental where you don't have a lot of options and don't expect too much, but if you have a choice, use a better alternative.
The big problem is that it might work just fine one day, then fail catastrophically if anything in the environment changes.
Man that would be perfect, except the main coax is already being used for my internet as is. I wish HFC for the NBN would hurry up and die already.
While my highly-tuned Wifi 6 setup at home is honestly quite excellent (even for my competitive FPS gaming needs, and even wireless VR has been brilliant), it would be nice to ditch it for higher reliability.
Problem is, I'm pretty sure my rental has a single coax socket for cable TV (which is used for my NBN internet connection now). I'll have a look though!
https://www.techreviewer.com/tech-answers/moca-vs-ethernet
I think it is a pretty decent review of the pro's and con's of using Ethernet cable vs. Ethernet cable. Perhaps others not reading this thread could also benefit from the knowledge.
I was looking into MOCA, but eventually I decided against it due to pricing and latency concerns. In the end, I opened up the coax socket at my router, tied an ethernet cable to the coax cable, went to my office (room above the room where the router is), opened up the coax socket, pulled out the coax cable, and voila, I now had an ethernet cable in the duct.
So here I'm feeling adventurous for pushing 10 GbE over a run of Cat5E for the next decade or more.
yeah, sure, if i got a bunch of really fancy wireless equipment, that could probably swing it too, but would also require getting new WiFi cards to actually be able to take advantage.
Then Later I moved to this place and there is no FiOS but I have mm-wave 5G (2Gbps) but there was no Ethernet cable in the house so I thought, hey, I know, I will use that MoCa thing, and it worked like a charm, got my signal from the CPE antenna unit mounted on my Garage (best signal) down to my lab in my basement.
I have Fiber to the house and then an Eero first generation to provide Wifi. I also have an AmpliFi HD. And we still had regular (3x per week) problems with streaming video.
THANK YOU !!!!
This is more of an issue in a commercial building where you have lots of people who need to move through potentially narrow escape routes than it is in a home where every major room will have multiple means of egress.
I just ran mine up near where the wall meets the ceiling. When inside I look down a lot more than I look up so the ugliness of the cable doesn't bother me.
Here's part of my Ethernet cable run [1] through the living room. The other cable in that picture is a speaker cable for one of the home theater rear speakers.
I had a cable going diagonally across the wall for ages before the SO put their foot down and demanded something slightly more aesthetically pleasing. Now it follows the trim at least.
Held with screws of course, or otherwise removable and replaceable.
(Sometimes the clips can be a bit too cheap, in which case tearing off the original crappy adhesive and using some of that insanely-strong double-sided 3M tape helps.)
[1] https://www.youtube.com/channel/UC8uT9cgJorJPWu7ITLGo9Ww
There's also decorative ones that run along your floor, I forgot the English name, but they're a bit bigger than the usual ones and can hold a number of cables out of sight.
In theory it would be a matter of tying an ethernet cable to an existing coax cable and pull through, but in practice, eek. I don't believe those lines are continuous, or there may be debris or kinks in it somewhere in the wall. In the end it took an electrician half an hour and an improvised cable pump to get it through.
I have one of these, advertised as 2000Mb/s but actually doing about 80, and in a small apartment. Every 6 months or so it connects to my neighbour's system without me doing anything.
Overall I found the standards of electrical work in Europe is significantly higher than US counterparts. In modern buildings every outlet and line has mandatory grounding, and GFCI is built into the fuse box on every power line for any building that's less than 30 years old. That really came in handy once when I accidentally hit a power line when drilling through the ceiling. I can't imagine doing any kind of renovation work without GFCI on everything.
I've seen none of the problems all these (presumably N American) commenters are reporting, but our power systems are famously much better in Europe. I'm from England but live in Czechia and apart from different plugs everything works the same.
I bought a rice cooker from a Taiwanese woman a couple of years ago and was shocked (fortunately, not literally) to discover that apparently Taiwan runs on American sockets and American (crappy) voltages. Luckily I had an adaptor plug and it seems to cope with 220V mains without releasing the magic smoke.
Wireless is just temporary, if you have wireless headphones, you have to plug in a wire to charge them. If your computer has a battery built in, you have to plug it in every day to charge. So it's just an extra chore that makes your life worse. Not that exciting.
Maybe everyone fantasizes about being a digital nomad. I do not. I have a great desk, chair, keyboard, and 3 monitors. I couldn't work without them. If I want to see the world, I'll go do that without bringing my computer. A vacation with your computer is just work.
A question I'm asked is "what if there's a power outage" but we have had about 10 total minutes of power outage here in Brooklyn in the 10 years I've lived here (during Hurricane Sandy). My thought is I just won't do any work if there isn't any power. We waste an hour a day dealing with wireless not working, 10 minutes every 10 years is no big deal. ("But what if it's the big one!?" Then my ISP's backup batteries will also be drained and I still can't work.)
This can't be possible (at least practically) with a wired setup. I've almost never ran out of battery on my laptop or headphones and never have issues with Wi-Fi (worst case I have wireless hotspot on phone anyway).
I don't see any problem with a wireless setup other than convenience, where the extra capacity/reliability of wires isn't needed 99.9% of the time anyway.
> I don't see any problem with a wireless
One of the problem explained in the article is that you may not realize the problems.
---
So the idea would be to use wires by default (when sitting at your desk), and switch to wireless when you want mobility.
Literally never.
I'm actually quite surprised that people still have problems with wireless devices in this age.
Even if I had a desk, I'd probably only plug my laptop to prevent battery cycles (and to keep ready for when I wanted to move) but would never bother with Ethernet, or a wired mouse, or headphone jack for example. Any additional latency/performance problem is imperceivable for daily tasks (unless I am a professional esports gamer or trying to download a huge blob over a gigabit connection as fast as possible, which is probably not 99.9% of people do).
You basically say that you don't care about latency/performance loss, but are surprised that people still have problems. Maybe because these persons do care?
While on a Zoom meeting the latency introduced by Bluetooth, while slightly perceivable, isn't of an issue over the latency of being connected over the web itself.
Any wireless system would have more or less be less performant than a wired equivalent. What I'm saying is that for 99% of the people in 99% of daily use cases, that difference wouldn't matter.
On the second floor the difference in speed tests going through cable vs 10 ft. away through one thin wall on my gaming pc is 250 Mbps -> 24 Mbps. There can absolutely be a practical difference in daily use.
Most people do not have multiple APs, most people are not getting APs other than what is provided by the ISP.
While definitely not ideal, it's a solve-once-and-forget problem.
I think a good balance is a laptop with a single wire to USB-C/Thunderbolt dock, with external monitors and wired peripherals, ethernet.
If you move around, wired headphones are a PITA, Airpods have pretty much solved wireless audio for me (especially consistent device switching).
Bluetooth on the other hand is an absolute no-go. I have had uniformly bad experiences with it, and refuse to use any device that depends on a bluetooth connection.
You're missing the point. Wireless technology failing is a fantastic excuse to get out of Zoom meetings!
Landlords never notice or care. Just don't make it look shitty, they will just think it's a phone jack anyways.
I still find wired Ethernet to be a better experience all around though I do have a pretty decent wireless setup and I do roam quite a bit as well.
When you move, just remove them and install them in the next place - or have a chat with your landlord and leave them for the next tenant.
When I was poorer, even the "minor" cost of cables would have meant even more missed meals. (Fortunately that wasn't an issue because residential internet wasn't a thing back then, but you get the idea.) Losing a security deposit would have been an absolute disaster. I went to court twice to avoid exactly that, and could well have ended up homeless had I not prevailed. Not every landlord is forgiving, and not everyone has the option to prioritize freedom to make modifications over things like location, price, square footage, etc. Being able to shrug off costs and risks that loom larger for others is practically the definition of privilege.
Even among the more affluent, not everyone would put up with visible wires in every room and on every side of the exterior, or with the effort/expense of "doing it right" by fishing through walls with multiple obstructions. They might well prefer the tradeoffs represented by wireless, even if it's not perfect. Do you use landline phones because you never have to worry about being in a dead zone? Different people, different choices. That's all I'm saying, but somehow people who can't conceptualize that other people aren't exactly like them think that any choice other than their own deserves an inquisition.
For the good stuff! Half or less for the last gen 5e. This is Menard's retail pricing too.
This doesn’t seem to have anything to do with your “wires are more reliable” thesis, does it? A laptop with onboard devices drawing wall/battery power would be drawing more power than if those devices each have their own separate batteries, right?
Not necessarily, because the laptop has to do additional work transmitting Bluetooth/whatever for the wireless devices. In any case, his thesis appears to include preferring a permanently wired desktop computer to a laptop, as well as preferring wired peripherals to wireless.
Personally I use mostly wired peripherals, but I am a sucker for a wireless headset. Being able to stay on a call while I walk over to the other side of the room to grab something is a great freedom, and plugging the headset in at night is not a big deal.
Agreed. My last job had a bi monthly all hands meeting at 4pm on a Thursday, I used to just stick my headphones on and make dinner during it.
My current setup is two USB-C laptops, one for work, one for personal use. When I switch from work mode to My Time Mode, I just swap one USB-C cable from one to the other - all the peripherals follow. Trackpad is wired, keyboard is wired. Headphones wireless - I broke way too many wired headphones by snagging them on crap while moving around.
The relevant comparison would be the power required to transmit and receive Bluetooth vs the power required to power a wired peripheral (e.g. mouse, keyboard, headphones) where the wireless version would have its own batteries.
Wireless is nice and all. Powerline ethernet is great when it works out of the box, and a Kafkaesque nightmare when it does not.
I am on my second home, the first was a condo where I had no access to the "above" or the "below", only the "within".
Phone lines had been run, so I thought I could attach Cat5 to the phone lines and pull them through -- but the previous owner had decided to wall up the phone jacks. No idea where they were, and no access to the "above" part.
Let's just say - I tried a lot of things to figure out where the walled up phone lines were, but never could get to the level of confidence you would want before you start knocking random holes in the wall in a property you own.
Powerline Ethernet was a godsend - for a while. Then one day (7ish years) it started struggling, and eventually couldn't stay connected. To untangle that was going to be hundreds of dollars of hardware and research and/or contractors (side note: have you tried asking an electrician to assess Powerline Ethernet issues?) -- I went wireless at that point.
New house maybe a year back after getting married (yay!). Ran ethernet. Used these guys:
https://www.amazon.com/gp/product/B00404UATS
Yeah, I drilled some holes in the wood floor. I am not that rich, nor is my neighborhood, it's a non-issue. Game consoles, NAS devices and work computers have ethernet. Everything else less hungry (or important) like a tablet or mobile has wireless. Life is good.
The fact that they need to be charged doesn't matter. My current ones have a swappable battery with the battery that's not in use able to be charged by slotting it into the wireless reciever that sits on the desk. But there's also newer wireless headsets that have 30-40 hours of battery life. So you only need to plug them in overnight (or every few days depending on usage).
That's also true for me. Ultimately, I think it's a matter of using good judgment. All my major devices - even laptops - run on cables by default, tablets/phones wireless of course. For keyboards and mouses (OK, not WiFi in a strict sense) it depends.
seems unlikely
Not in my experience. Ethernet cables have bern reliable, but most have connector latches that self destruct.
HDMI cables are a lottery, and the prize is in no way related to the ticket price.
mini/micro USB connectors go out in about year, depending on use.
All 3.5 mm audio cables seem to have half-life of around 4 months, regardless of price. The only survivour cable came as an extension with old Sennheiser earbuds.
In comparison, I have yet to experience those bad bluetooth earbuds just stopping to work.
[edit] not to mention all the equipment with soldered-in cables that are either too long or too short and cannot be replaced. Hi, homepod mini!
I'm still using the USB-C cables from my Nexus 5X today, no problem with them.
With HDMI cables I just buy them, connect them to my GPU and wait for the AMD driver to tell me the maximum bandwidth they support. Super easy way to measure what they'll support and what not. So now I've got a dozen that work at 4K 60Hz HDR just fine.
On the other hand, I've never been able to get a single bluetooth device to sync reliably for more than 5 minutes.
That sounds pretty unbelievable to me - even the cheapest crappiest Bluetooth devices can pair with the cheapest crappiest USB Bluetooth adapter reliably for more than 5 minute's.
Headphones snag, the jacks get wobbly, theyre fixed lengths (standing desks mean they need to be long enough to move now so they're messy unless you do a bunch of extra work). Ive got problems with interference with my wired mouse that seems to cause stuttering if I use a specific usb port, but maybe once every 2-3 weeks or so the port I do use just doesn't recognize the device so I have to un plug it and plug it back in. Bluetooth may not be perfect but it's tidy, easy to use, well supported, reasonably reliable, and good enough quality for 98% of my use cases.
I’m intentionally already using https://ark.intel.com/content/www/de/de/ark/products/189347/... as WiFi/Bluetooth card on all my computers, but I still can’t get anything to sync reliably.
In the end I just gave up and now I use the dualshock 4 with cables. And for headphones, I’ll continue using my wired sennheisers from 2008 (I checked and updated the year in the previous post as well).
The Dualshock 4 gets about a minute at best before I get suddenly random input from it and it stops working. That’s with a genuine Sony DS4, second version.
I actually still use wireless mice, but only the ones with custom dongles.
Me too. My problem is going to store and finding such cables, that will last 10 years.
> My headphones are in fact from 2008
That reminds me. When I was complaining about 3.5 mm cables in one music store, how my only good cable is from earphones I bought ~20 years ago, the store person went, like, "of course it won't break if it was made back then! why do you expect to buy something like that now, they ain't now making any"
I think the mouse that feasted on cable insulation in my backpack may actually have made it to second place. That's an entirely different story...
"Get a dongle" -- maybe I will when I finally give up my original iPhone SE, but I'm not going to be cheerful about paying an extra $30 for the privilege along with the overhead of keeping track of an additional thing.
$9. Not $30. If you're going to grind a three-year-old axe, at least get your figures right.
https://www.apple.com/shop/product/MMX62AM/A/lightning-to-35...
Google clarifies: Lavalier(aka Lapel) Mics. I know my mental model should drop my first jumped to conclusion but I find it childishly amusing.
So now I assume you have a lapel microphone specifically set up in your toilet for zoom calls that happen to catch you out.
The axe grinding, however, was brought up by Mr Kuhn. This is simply one instance of it, and the economic fraction isn't even the biggest part.
That rating comes in handy for those who have butter fingers like me and drop their phones in water once in a while.
Couldn't the jack itself be made to be water resistant? Seal it up and use relatively inert metals for the contacts.
https://www.gsmarena.com/sony_xperia_1_iii-10712.php https://www.gsmarena.com/samsung_galaxy_s10-9536.php
If it was possible then it's possible now.
The comment I made doesn't require that I push, or be for against, some higher narrative that you seem to see me as having in my head. It's perfectly valid to criticize the technical merits of an individual comment. If you see that as weaselly, then tribalism has infected your mind.
Does it work? Well, one of the phones used to be my daughter's. She forgot to take it out of her pocket when she put her trousers in the washing machine so it ended up going through a full wash and spin cycle. It still worked, there was no water ingress in the phone itself. The earpiece did seem to have gotten damaged so I replaced it (5 new earpieces for $2.50 incl. shipping, I still have 4 left...). This phone is now used as a media player, running mpd which can be remotely controlled from other phones in the network. It connects to my jobsite radio (an oversized and overweight boombox I made around an old car MP3/CD/Radio with 2x40W speakers with bass reflex and a SLA starting helper/power pack) using that 3.5mm jack.
a lot of their phones still have headphone jacks as far as i know. seems like they are the only decent company left in that regard
I've stuck with the Pixel A series precisely because of the headphone jack. I'm sad to see that once again Google is trying to remove it: https://www.theverge.com/2022/5/11/23067702/google-pixel-6a-...
I bought a totally cheap no name one when I got a company phone without audio jack. Since then I have used it on many different machines as well as (Android) phones. The adapter stays with the headphones as I just would loose it otherwise.
Edit: Typo
I usually just go wireless, although it has its problems especially with Spotify.
But if I decided to go wired, I would have no qualms about going wired with the Apple USB-C audio jack dongle.
There's a quirk that requires a script to uncap the volume on Android but it's the best IEM DAC I've heard under $200. I don't think they subsidize it. It's just economies of scale at work. Here's a Head-Fi review and analysis of it: https://www.head-fi.org/showcase/apple-usb-c-to-3-5-mm-headp...
Edit: More context on 'fixing' the Android volume capping issue for it (quirks are common for devs to add at usb-audio driver level in kernel, but this is something different with Android's defaults): https://forum.xda-developers.com/t/guide-magisk-apple-usb-c-...
At my desk, i do have wired headphones, but they are plugged into my TB dock.
Depends on your usecase. Music and Zoom is fine, but for competitive gaming I'd always go wired.
So did that work prior to M1?
Though I do use a wireless mouse (G Pro X Superlight) and it's brilliant, though it's wireless adapter is attached to a cord that lets me plug it in to keep going if I forgot to charge it, so its the best of both worlds.
Really, its only my mouse that I move around enough for wireless to be worth it in competitive gaming.
Some advice for Bluetooth headset gamers - Bluetooth completely breaks the audio for some games like Battlefield 1. The only solution is to go into device manager and disabling Handsfree Telephony on your device.
Away from the desk though? Bluetooth all the way. Even if I'm just lounging on my bed bluetooth is better because there's no wires for the cat that will most assuredly be in my lap to chew.
I now use AirPods, and it's, mentally, very different, since my head is no longer part of my phone.
¯\_(ツ)_/¯
I also use air pods now and am fairly certain this water fountain i walk past daily has a short causing some sort of arc that interferes with bluetooth as they always cut out walking past it.
Same, an embarrassingly large number of times actually.
I got the dang dongle. Now, I want to listen to a podcast while doing to bed. Of course, because cellphone batteries only last like 24 hours, I have to charge overnight.
So now I've got some big fancy inductive charger, so I can leave my charging port open, to attach a dongle. But hey at least Apple saved $1 and got to remove the only non-proprietary port on the thing.
https://www.apple.com/shop/product/HLJV2ZM/B/belkin-35-mm-au...
I had some adapters with the plug built in like that, they were notably lower quality than the one without the plug, but it could be that the ones I got were just some no-name junk brand.
I'm willing to use the equivalent of a dongle with my PC (audio interface), but obviously thats a very different use case. I'm just glad laptops haven't decided to take the "brave" action yet.
I have some fancy Sony Bluetooth noise-cancelling phones I bought just before my kid was born. They are _amazing_ but they should be for over £300. The noise cancelling is superb... so I use them when flying. Otherwise, too much faff.
My £20 Sony wired bass-boost earbuds sound nearly as good and live in my pocket and I use them far more, because of the convenience factor. They never need charging, pair 100% instantly the second I plug them in, latency is a couple of nanoseconds (a metrr of copper at the speed of light?) and quality is superb.
They also are excellent for voice calls, while the fancy Bluetooth cans are rubbish at picking up my voice.
1) wireless degrades by degree and is hard to debug
2) it fails silently or just in non-obvious ways
3) people generally have no idea how good it is supposed to work, until they experience it working properly
I think there are many things, in life generally, but especially in tech, where these qualities conspire to give people a rotten time.
People will complain when things fail, or take action. "Graceful degradation" seems like a feature, but it can keep you on a slippery slope of increasing tolerance toward ever worse performance.
The complexity, and fuzzy "connection to everything" means that no one has to take responsibility. Maybe it's your neighbour? Or the trees? Or the alignment of Venus and swamp gas on a full moon?
Unless you have a full spectrum RF field meter and other very expensive test gear microwaves are "black magic" to even great engineers.
I'm sure when Apple met to discuss remotely degrading iPhone performance the words "They'll never notice" were spoken. Companies that sell complex services and oversell resources love this combo of fuzzy, hard to measure, diffuse, naturally erratic behaviour, broad tolerance, mixed with vague customer expectations.
They prevent you from knowing, and comparing, what you should be getting, and what you are getting.
Many recent development directions, 5G, secure enclaves, encrypted updates and exfiltration (telemetry), are set to increase the "random magic factor" and make gear even more precarious and mysterious to the end user.
I suspect Apple would prefer their products to be reliable rather than erratic.
This was likely a compromise from that design decision.
Since the slowdown was based on battery health and not absolute age of the device it was (and is) still possible to replace the battery.
But “the market” (ugh) seems to have decided that it wants phones with integrated batteries, and touch screens, and no physical keyboard, that are too big to use with one hand.
So that’s all we get.
Unless it’s extremely niche.
edit: disregard this, I don't see the other thread anymore. Maybe I was the one experiencing magic fuzz factors.
You actually have that, at least for the frequency range that matters: the wireless hardware in your computer.
Your wireless card absolutely knows how good/bad the RF environment is - it wouldn't work without it. The problem is that none of this information is exposed to the user.
The Wi-Fi "signal strength" meter is consistently useless on basically every device I've tried. In fact, I wonder what would it take to actually make it go down, since it's happy sitting at max level even if you can't get a single packet through. I guess it must be measuring RSSI, when the one that matters to the user is "how many packets am I actually missing" which can be obtained by keeping stats on TCP flows or pinging the default gateway.
I think this is usually recommended for safety reasons, to prevent mistakes from causing high voltage to get into your low voltage runs and become a shock hazard or potentially start a fire. But, I have noticed empirically that running a bit too close with the wrong cable does absolutely result in some serious interference issues. So… your wired connections are not free of signal issues either, especially if you want to neatly run them in-wall.
edit: also, worth noting that this is a highly U.S. centric piece of advice. I mean I’m sure it applies elsewhere, but both electrical systems and construction differences probably make some of the details very different, and I’m not cultured enough to know exactly which and how.
I mean, this might be different for fiber in some regards, but I doubt that is terribly economical either.
[1] Or more accurately, as laptops dropped builtin ethernet ports, the kind of ethernet you use got decoupled from the laptop model.
Then you need 10Gbit network equipment. A 10Gbit switch with more than 8 ports is going to be expensive and consume a pretty ridiculous amount of energy. It would be difficult to argue that there is any truly “consumer grade” 10GbE equipment; almost all of it comes in sizes designed to fit in racks… Wifi 6 APs exist, but my AP-HD isn’t, because it wasn’t purchased that long ago, still works, and I have had no good reason to replace it. Not to mention the actual speeds I get on WiFi are never even close to the theoretical speeds, making it seem unlikely it could even saturate my 1GbE setup in practice.
Although I do have some 10GbE equipment, even directly connecting devices together I have found that many of my devices, of which many are consumer-grade, already bottleneck elsewhere and can’t even come close to saturation. So as of today, 10GbE does not seem to be a huge win for most users, even if they do own some equipment that can do it.
- It is true that a modern HDD has over 120 MBps sequential read, but not by much: 4GiB WD Reds are probably somewhere around 150-160 MB/s.
- Sequential speeds aren’t really that practically useful for me. While I do have some large files, most workloads are going to involve smaller files and a lot more seeking.
- When using a filesystem with parity, like btrfs or zfs, you definitely have some overhead for writing and checking parity. I’m sure it’s still very fast, but it is overhead, and NASes don’t tend to be the most cutting edge compute boxes.
- Even if some RAID configurations might improve performance, the configuration I am using has a fair bit of overhead. I suspect you need something like RAID 0 or 10 for maximum performance, but those setups are inflexible compared to RAIDZ or Synology SHR.
- As it is now, my NAS is unable to saturate the network bandwidth using Samba alone. It just never reaches a gigabit. Part of this is probably related to Samba or the CIFS protocol, but honestly I don’t know if it has the raw CPU muscle.
All in all, I am not too concerned about it. It is possible that I could yield better performance with 10 GbE, but I don’t think it would matter for most of my use cases. Plus, disk resources are definitely being consumed by background tasks as it is.
The transfer rates go up roughly hand in hand with capacity.
Monoprice sells double-shielded (SFTP) enterprise-grade cables directly.
If you think about this, of course, wireless will be limited spectrum, there are only so many Mhz that is made available for whatever wireless standard you are thinking about, Wifi, 4G, 5G, Bluetooth, whatever.
And, obviously, wireless will struggle with Noise and Interference.
Wired solutions, whether shielded twisted pair (e.g. CAT6) or Coax, or Fiber or Waveguide, whatever, will do much better on SINR, you have thermal noise sure, but generally, it would be a much quieter medium.
Also, you would not have the FCC saying you only get X Mhz, whatever you can get across that medium you get to use.
Thinking about all of these factors then of course, Wired will generally win out, all things being equal.
The Trade off, is that wireless is more convenient.
On the long term, if you look for reliability and hassle-free setup, you lose the benefit of convenience.
6 months later, I got a notice that the cable I used was recalled for a bad UL rating. Monoprice had to pay a contractor had to come out and replace it. When coordinating the work, I warned the company that it was a very dirty job that would require two people. They sent an older guy out all alone, and he poked around for half an hour then left, saying he would come back with PPE and a helper. A couple weeks later the same guy came back alone and unprepared again. Finally the third time he came back with a much younger guy and bunny suits, and they redid the runs.
I had stapled the cables up every couple meters, and followed a Manhattan path based on where I could crawl to. they just fished it in a straight line, and left it resting in the dirt. They presumably billed Monoprice several thousand dollars...
Also, did you try pushing back on the replacement worker's work? It's a bitch to do, but I've gotten companies to either redo their handyman's bad work or just let me choose a vendor (which I would ask for after the 2nd time he showed up.) Stapled Manhattan distance wiring is not the same as diagonal through the dirt.
(For reference, I searched "bunny suit ppe" to find an example.)
The LATENCY is noticeably better. If you are on a zoom call with both parties on headsets and on wired connections (so no echo cancellation needed) or other audio connections I find it noticeably better (less stepping on each other etc).
I have AT&T's fiber product (2 Gbps) and the office has their biz product (300 Mbps) and latency between is good and to google et al is great (< 10 ms in almost all situations).
But, at work, there's hundreds of ms difference between ethernet and WiFi. I assume it because most offices use mesh networks now, and none here are WiFi 6.
So, network setup and router quality definitely play a part. My AP was ~$200, used for streaming VR.
So, yeah, wifi can require some management, but the thought of running all those laptops, phones (you couldn't rely on cell service - which of course is another form of wireless - inside most of the buildings - worked some places, not others), and portable medical devices only on wired ethernet is a non starter.
As for other forms of wireless - Bluetooth and BLE have their issues, but you get an awful lot in return.
These were the issues I had:
1. SO's cheap wireless mouse produced WiFi interference -> massive packet loss
2. Small 2.4Ghz USB dongles plugged into USB 3.0 port (e.g. Logitech's wireless dongle or USB WiFi for RPi) -> USB 3.0 interference render the dongle unusable
3. The HTC 10 has a metal casing -> incredibly poor BT RF performance, causing massive drop outs (especially bad with my FiiO Q5 DAC/headphone amp, which has a massive aluminum case)
4. Client stays on AP with very poor signal (barely in range) instead of jumping to different AP -> retransmissions cause massive degredation for all other clients
5. Lots of APs running at full power in an apartment complex, paired with a router/AP that had low WiFi performance anyway -> bad everything
All but the last can be solved relatively easy (okay, I didn't swap the HTC10 due to BT, though that really annoyed me; it's time eventually came when the battery lasted half a workday without using the phone for anything). Especially coordinated WiFi APs (like Unifi) are great. And even the last point was a non-issue both in our last apt (one well placed UAP for testing) and our new old house (four UAPs).
Except for the Wi-Fi widget I click on when I want to choose Wi-Fi network, what software needs this access? I can think of two use cases:
- A Wi-fi diagnostics application.
- Data collection to co-relate Wi-Fi networks and locations.
I don't see why the OS doesn't simply refuse going into to polling except for white listed applications, because out of the box there is only one application that needs to ask for polling (my wi-fi widget) and I don't care if Google Maps wants to geo-tag the wifi in my surroundings, so just give it the list from the last whitelisted poll. And if I need a Wi-Fi diagnostic tool, then I am an advanced enough user to figure out how to white list it.
Pretty valid use case IMO.
Can confirm. I work on videoconferencing infrastructure, and easily 90+% of the end-user quality problems we investigate are caused by the end-user's WiFi. It's very difficult from a customer support perspective too, because it's common for WiFi to have latency and loss problems that are severe for real-time video or audio but not so noticeable for bulk TCP transfers (like most speed tests) or web browsing. So communicating the issue to users is hard.
I'd also suspect people on HN might tend to have paid more attention to their WiFi setup than the average user.
A wireless mouse is still worth it though. Logitech G305 is essentially perfect IMO. Extremely reliable, extremely low latency, months of battery life, reasonable price.
I've never heard a single person complain about a wireless mouse. The author is just a crank.
I would advise people to not get a Logitech mouse, however. Their button switches are poorly made and fail much sooner than any other manufacturer.
Also from what I've read, usb-c has a lot of interoperability problems because the connector is a different standard than all the different communications standards that it can be used with.
It's a bit like the butterfly keyboard of speakers. A wonderful idea in theory. But it makes you wonder if anyone at Apple has ever tried using it an apartment building, a dusty environment, or some place that isn't a pristine lab or upper middle class suburbia
It still boggles the mind why there is no wired input though.
What I'd really love is if my HomePod Minis stopped assuming the primary HomeKit Hub role and instead let my Apple TV connected via Ethernet handle all of the requests. There's no way to select a device to be the primary hub as far as I can tell.
(I'm genuinely willing to believe it could, because I've seen stupider problems personally, but it does seem like a weird one.)
I've been meaning to take an old Netgear router, put dd-wrt on it, and then put that between my cable modem and the mesh router.
Also, being conscious of what's in the air space is as crucial as the quality of your wireless equipment. It's all for nought if your microwave, AV sender, bluetooth headset or crappy 2.4Ghz wireless mouse is shitting into the same airspace as your network connection.
I say all of this as someone that pushed HARD to switch our office from wired to wireless when we moved. Aside from the direct cost saving (~200k in wiring alone) the side benefits of things like not having to use a crappy 'presentation PC' in each meeting room was a complete revolution. You just bring your laptop, and it's not even going to shit itself roaming from wired to wireless because boom, you were on wireles the entire time anyway. Same principle about wiring in fixed items applied though, we put fancy ipad meeting room control stuff in, and dropped extra on ethernet adapters for them because they're fixed devices, wireless is not a good choice for that despite ipads being wireless devices primarily.
Lack of visibility is probably the biggest shortcoming for enthusiast-level home wireless. I've had good outcomes buying ex-business Aruba gear and repurposing it for home use, but it's power hungry and you need to know what you're doing to get it off the ground. But being able to actually see what's sitting in the spectrum and watch the impact of things like turning on the microwave for ten seconds is a game changer.
Now the problem is just how to manage all the cables that come from having a gaming PC and work laptop that use the same keyboard and monitor. I have a KVM switch but I can't hide it under the desk if I want to press it.
Why not? I bolted mine to the underside of the table roughly 20 cm (about 13/20 ft) back from the edge, it's both invisible and the button is trivial to find without fumbling.
OTOH he got the most fun using the power tools drilling the holes.
https://www.extremenetworks.com/extreme-networks-blog/how-do...
I replaced a high-performance n/ac router with years old leftover enterprise AP. Price-wise the enterprise AP would've cost as much 2nd hand.
It improved the situation in a very very crowded airspace in every measurable metric (~10-50%). Better latency, average, minimum and maximum speeds, connection stability, packet loss and even security (PMF support). All that on the same standards and other conditions.
My suspicion is that consumers are being sold bottom of the barrel hardware and new protocols just rise the bar on what s*t they can get working.
If the device is expected to be used in a fixed spacial or temporal manner, or you care about the data conveyed across the link, constrain its path as much as possible with a waveguide.
The radio is wireless, but that is public broadcast and is not a connection to a specific device. Sometimes it is unreliable, but that is not much of a problem in this case.
Anecdata. On my Mac the MX Master 3 works flawlessly via Bluetooth. However, it has janks when I use it on my recently purchased ThinkPad running Linux.
> Similarly, when connected to Mac bluetooth, my Jabra Evolve 75 headset would frequently have the mic or sound drop. It (mostly) worked fine on its own dongle.
Again, all my bluetooth headphones work fine on my Mac, but on some (not all) of my Linux computers it sometimes work and sometimes doesn't. I don't know if it's Linux or the hardware.
Also, on the aforementioned ThinkPad, when I connect a new device via bluetooth the wifi disconnects and immediately reconnects 40% of the time. Don't know what's up with that.
I've been using Linux long enough to take things like these in its stride :)
(Btw, it was after I was issued a Mac by my employer that I started using Bluetooth devices regularly; before that in Linux it was mouse via IR-usb-thing, wired headphones etc.)
I agree with the general sentiment that wired devices are inherently more reliable. But I don't know if we should collectively move towards wired, or improve the reliability of wireless.
The issue I'm currently dealing with is that my Bluetooth ZMK keyboard works flawlessly on Linux but won't pair with a Mac. I guess devices are nuanced.
Once the usual quick checklist items are exhausted,suggesting that maybe we ought to just run a cable before getting into troubleshooting buggy wireless drivers, old hardware that doesn't support modern 802.11 amendments, or conducting a full spectrum analysis are often dismissed, or even if they are open to it, no one seems to own an Ethernet cable these days or their gateway is in a closet, basement or some other inconvenient location.
Someday I hope that installing Ethernet jacks throughout the house becomes as commonplace as it once was to do the same for phone lines, but to many it is an afterthought as they are used to their crappy wireless.
Partly because of the physics involved. The AP can't magically sense the best channel for the clients, since it can't sense the spectrum at the client's.
- I can move houses and just put all my WiFi mesh nodes anywhere and I’m online again.
- I can play music from my phone while cooking, and use the phone besides my stove to view the recipe simultaneously
- I can browse photos and create a photo book online sitting next to my wife on the couch, instead of sitting at an uninspiring desk in the working from home room
But since then I've realized things have changed. It's incredible how much wireless and battery technology have improved in the last 10+ years.
If you want to perform at the top level in most games, you need a wireless mouse. It allows for significantly easier and freer movement than a wired mouse, even with a bungee. Charges can last for weeks, and due to faster charging can charge in an hour or two.
And as my list of equipment and peripherals grows, I find myself wishing more of them were wireless so I don't need to deal with them getting tangled, needing to re-route them, them being too short if I raise a desk up, etc.
However, there are some exceptions:
* Internet. The reliability problem has largely been solved if you have decent quality equipment. (I have a Ubiquity Wifi 6 lite, and it works great) However, the delays are still there, and some houses/apartment layouts and materials just aren't workable. I think people overall underestimate WiFi though, as most setups are very poorly done. The crappy WAP in your comcast router/modem/switch/wap combo box is terrible and will never perform well no matter what. But a good dedicated setup with WAPs spread and placed properly will do amazing.
* IEMs. There are no standards for transferring high-resolution uncompressed audio wirelessly right now (BlueTooth cannot do it), so the only way to get the best experience is a wired IEM. This isn't true for all digital audio transfer, like streaming. That can be done wirelessly just fine over wifi. (IE, wirelessly sending audio to a Raspberry Pi, when then sends it to a DAC, etc)
The moment such a standard exists for IEMs, I will gladly swap over. The wires constantly get tangled, are too short, get damaged, or rub against things while moving leading to undesirable sound coming into the ear.
At home switching to Wi-Fi 6 has solved some f dead zone problems. I notice a stark contrast between performance of Wi-Fi 5 and Wi-Fi 6 devices.
That said, I adore wireless headphones and would never go back. Whether it's my earbuds to listen to podcasts while doing the washing up, or my BT headset that I use during the day for work. My BT headset (a Bose noise cancelling one) is definitely the most frequent cause of call issues (now that I've fixed a motherboard problem causing wifi dropouts), but it's still worth it to be able to move around, still hear what's going on in the call while I get a drink etc.again).
I imagine concrete houses or congestion might be different but I can’t even find a reason to get a mesh system in my house.
Wireless mice and keyboards also work great. I use them even for gaming where many swear by wired.
You can’t hide a mouse cable thing like you can hide an internet cable. It’s literally on your desk no matter what you do, so the upside of wireless there beats the downside. Same with wired internet to a smartphone…
I still don’t trust BT audio though.
The 2.4ghz ISM bands are just noisy, sometimes all it takes is a poorly shielded microwave to take out everything in a 50 yard radius.
I don't care much about heat from electronics when I'm trying to warm the space anyway, but I'd rather minimize it in the summertime.
> It was the epithet of King Harald Bluetooth, who united the disparate Danish tribes into a single kingdom; Kardach chose the name to imply that Bluetooth similarly unites communication protocols. [https://en.m.wikipedia.org/wiki/Bluetooth#Name_and_logo]
Another thing, not mentioned in the article, is that microwave ovens tend to interfere with Bluetooth in my experience.
Seriously. There were many times that I was working on something that suddenly had a spike in the receive failure rate and I'd look at the time and realize that somebody was heating up their lunch in the microwave.
Wifi on the other hand... I have no idea if its interference or the modem. It worked fine for months, then all of sudden my connection went to 2Mbps. And my Chromecast can no longer connect to it, etc..
For whatever reason this isn't a problem for me. I work with a laptop connected to my WiFi, use bluetooth headphones, and have 2-3h of meetings per day. Not a problem whatsoever.
This intrigues me and makes me wonder:
1. Why should it need to do this?
2. Why should this degrade the wireless network performance? (I don't recall the details of wi-fi but shouldn't it be able to do this passively without disrupting anything?)
2. The radio in most (all?) WiFi radios can only tune to one channel at a time, scanning all channels requires retuning to each one, which takes a little time.
I vaguely remember from the time (~10 years ago) is that this was a requirement for phones (eg, when Qt is used on a Nokia phone) that the network request go to the interface preferred by the user (so not to use the cell network when wifi is available). Anyway, the whole thing was implemented in a "bearer manager" plugin, and that thing was quite buggy. https://doc.qt.io/qt-5/bearer-management.html
I think apart from people running Qt on phone, this had virtually no use. And yet this was enabled by default on all platform, and causing lots of troubles. What we ended up doing for the application we were shipping was to make sure that this plugin was not included in the final package.
I must have an incredible $10 BT speaker (used for quiet TV listening at night). I don't see any delay versus the regular audio (plan old 1/8" audio jack out of Android box, to stereo). You would notice 150-300 ms as serious lip synch issue.
For interactive audio like games or video conferencing, this doesn‘t work of course.
Working on Logic (just hobbyist here) and gaming are the only two activities that I connect my wireless headset via USB C. The latency (because of Bluetooth protocol itself) is too much.
For Zoom calls it doesn't affect me much.
I've noticed this with YouTube where sometimes the audio gets out of sync when listening on bluetooth headphones and I have to pause and re-play to force it to sync again.
Work at a desktop computer.
Listen to CD's, not broadcast radio.
Never use GPS.
Never use WiFi.
Never use RFID.
Never use Find my iPhone to recover devices.
Never use automotive remotes.
Never use TV remotes.
Never explore outer space.
Never explore underwater.
All must be wires.
Because wireless is a trap.
Also, this guy's "cable management", haha.
If only people listened earlier.
The phone antenna isn't a straight piece of wire entwined with your cable (probably not even parallel with it), and your cable isn't connected to the antenna port.
How would the cable block the signal?
Also, a lot of SoCs with FM antennas use the 3.5mm ground as the antenna, so they are wired into the SoC in such a way that they can receive undesired interference.
I got my paranoia from working with military SIGINT - the things we routinely harvested from radio and telephone of the USSR was a major wake-up call.
I don't need cabled connections of any kind 99.9% of the time, Wi-Fi, Bluetooth headset, and my laptop "just work". Any stability and performance gain from using a wired setup is negligible in daily life (apart from very specific tasks) and convenience of wireless is far outweighing any gains of wires.
Been this way for many years with zero issues.
So yeah, I like this "trap".
It boggles my mind that any company wouldn't do that. Wifi sucks so bad and offices are so dense with people. The cube at my office have several network drops each, across many different cubicle systems. I figured they were standard equipment.
although wifi works pretty well in professional installs these days and bluetooth earbuds were a godsend with masks.
bluetooth connections are still unacceptably flakey though... even with top shelf name brand gear.
Your Wi-Fi and Ethernet are both network interfaces with their own IP addresses - those addresses don't move when you switch between interfaces. If you established an SSH session while on Ethernet and then unplugged it, that session will not carry over to Wi-Fi unless you change your Wi-Fi interface's IP address to the wired one's and that both networks are on the same L2 segment (typical for home networks, but not always the case for enterprise).
A potential solution is either MPTCP (which will establish multiple connections over all interfaces and can tolerate the loss of all but one of them) or to VPN into your router and use the VPN link as your default route - this means that the VPN connection might drop and reconnect but your IP address to the other hosts (which is actually the one of your VPN gateway) will remain constant, so traffic flow will be restored as soon as your VPN reconnects even if your local IP changes.
Also IETF's envisioned post-TCP transport protocol (SCTP) supports multihoming natively but has failed to take off.
Sun/Solaris machines used to have a machine specific MAC address. Then DHCP, bootp etc would give it the same IP independent of interface. Wonder what would happen in a current *BSD/Linux system if you gave switched the MAC address over.