The customer service rep isn’t in control of the change. They’re in control of what they tell me, and he was bullshitting. Not really sure what else to tell you.
Have you never interacted in a non-business sense with someone who worked customer support? They DO NOT have control of what they tell you. They will get fired real quick if they don't follow the script. Especially someone working the first line of support in a chat box.
So the support person is a glorified bot, but we have to be careful not to hurt it's feelings? Honestly I feel for the support guy, but he represents the company. And companies purposely push their support onto social media platforms like twitter - where exactly are unhappy customers supposed to vent their anger?
It? This is an actual person we are talking about. And most people are perfectly capable of handling a mildly stressful situation without acting like a complete jerk to someone.
How is it misleading? Can you explain to me what the difference between "promoting an update to their own product" and an ad?
This is an ad, pure and simple. They're pushing a product they know I don't own with a clear call to action, and without a way for me to remove it or disable it.
Ad is a broad term, this confuses people. This shit is definitely worse than ads on the box, but it also definitely isn't the entire monstrosity we have come to assume when hearing about ads on web pages.
Somewhere an engineer was tasked to write software that does an inventory of my network to see what new shiny hardware of theirs I don’t yet own.
Then, the software that was written to manage a product that I paid for shoves a call-to-action that takes a considerable amount of the page to suggest I buy it. And I can’t dismiss or close it.
The difference with ads on the web is that I paid for the privilege of being served this garbage. In my mind, this is much worse.
I don't own the device but according to comments in this thread it has an ad you can't remove/dismiss and it uses Google Analytics. That is the same monstrosity as web ads.
I'm willing to pay for major updates that include new features, even on a per-device basis. I'm not willing to pay for things I already bought that don't work correctly (bug fixes).
Agreed. It's not like their hardware is cheap or subsidized either, like the Cloud Key, which is a tiny device that acts as a proxy and has some cache storage for firmware updates that costs upwards of $100 here.
One of the reasons I have used Ubiquiti over the last few years is their product quality over other networking hardware. The inclusion of ads leads me to revisit those assumptions before my next purchase.
Ads, even if self-promotion, are a bellwether for a company being desperate or greedy, neither of which bodes well for their long-term product quality, usually because it indicates sales people are making the decisions. They also often lead to a slippery slope of increasing user data collection
The funny thing is that I have a UDMP on the way, because one of my APs died and I figured it was time to upgrade the network a bit, but now I'm refusing delivery and sending it back for a full refund.
I was considering replacing my AP AC Lite with whatever 6E equivalent comes out when 6E is approved. I considered picking up a NanoHD for the extra MIMO speeds, but in the meantime I keep seeing problems with reliability, the 6.x Unifi version, UDM needing to be rebooted, etc.
Even if Ubiquiti was as good as it was pre-2019, if they pulled this advertising stunt then I'd not consider upgrading. There's now no chance I'll be sticking with Ubiquiti for my 6E upgrade.
My nanoHD reboots frequently with current firmware. Sometimes it needs readoption, sometimes a manual reset to be readopted. The AP AC Pro seems less affected.
Thanks for that input. My AP AC Lite has been absolutely rock solid so I wouldn't want reduced reliability; it's the reason I went with Unifi in the first place. Not that it matters; no chance I'll be going with Unifi in the future.
Yeah, and thanks to your comment it actually got me thinking that I can (at least temporarily) mitigate my current situation by taking the nanoHD out of the loop entirely, as I do have another spare AP AC Pro in my closet. I also got the nanoHD because of extra MIMO speed (among other things, it's feeding my home office setup wirelessly), but the egregious loss in stability is not worth it.
Downgrade your nanoHD back to 4.x firmware. That way you get to keep your MIMO speed with stable device, you will only lose WPA3 support (if you use it, it is a new thing in the latest controller).
We upgraded to Ubiquiti a few months ago. I was so impressed by the hardware for the price. The dream machine firewall had such a good interface compared to more expensive firewalls I’ve used in the past. I’ve been very pleased so far, but this and another article have me worried they’re going downhill.
Sorry! It was this one. I guess it wasn’t the article so much as the comments on it. Former employees were lamenting the management at Ubiquiti, and users in the thread mentioned they noticed some rumblings of a decline in quality. Like I said, I’ve been happy so far, and I’m hoping that lasts.
As a flip side experience, I have a dream machine too; it ignored my auto update settings and updated itself anyway, there's no way to roll back, and it now regularly stops responding, has high packet loss over wifi and supports response was effectively "go jump". Very unhappy.
I have all aps and controller blocked from the internet for that very reason. Don't tell me when to update heh. Really sucks about not being able to roll things back.
Yea, I've had a few problems with UDM as well. The management UI/controller container has died numerous times so to get visibility into whats happening I need to ssh in and restart that container or restart the router.
I've also had numerous cases where routing has stopped for no reason. This I find really bad as, "it has one job ...".
I recently purchased switches, wireless access points and cameras for a new home-office setup. I was looking at going full Ubiquiti because the dream machine (UDM) comes with a network video recorder (NVR) that looked quite good in comparison to Zoneminder, Shinobi and even Synology's paid product.
While researching I found-out that you must use a ui.com account to setup the UDM (the apologists call it "mandatory online registration"). While you can setup a local account and disconnect from ui.com afterwards, I was perplexed that they would choose to do this.
Their user forums had at least two flamewar threads with people pointing-out the security implications of linking your local network to a remote, internet-accessible service but zero communication from Ubiquity themselves. Fast-forward to January and we saw the disclosure of a massive breach on ui.com.
So I ditched the Ubiquiti cameras for Geovision (better bang for $), kept the switches, APs and have been running the controller in containers (https://github.com/jacobalberty/unifi-docker). Needless to say, I'm not going to upgrade from version 6.0.43, which does not include advertisement.
I hope it doesn't get any worse because I do love how easy it is to maintain VLANs, firewall rules and IPS services.
Technically it's an advertisement for their own product. I have around $1,500 worth of various routers, switches, and cameras from Ubiquiti in my house and don't really mind. Agreed, the "ad" is very large and there should be a way to close it and never show again (cookie or local storage). However I think this HN title is misleading (clickbait and outrage driven) as they are not making ad revenue as one would infer.
How is pegging and selling the successor to what I have not making ad revenue? The difference is that they’re not selling pixels for money, they’re trading pixels for sales.
The main difference in my opinion is that typical advertising is a lot more privacy invasive and malicious. Tons of cross-site tracking, ads of dubious quality, getting 3 ads per 5 minutes of YouTube, etc.
I’m not frustrated when I get an email or dismissible banner ad from a company regarding a new product. It’s possibly interesting to me, and easily dismissible if not.
Sure, technically, this is advertising. But the biggest problems people have with advertising (privacy invasion, seeing too many ads, or seeing low-quality ads) aren’t present in this example.
It’s advertising in the exact same way that GitHub might show a banner informing me of a new plan they have. While it can be annoying, yes, I would disagree that it’s “as bad” as typical advertising practices.
Even if they only advertise their own products, they will probably become interested in conversion rates and customer segmentation at some point. Or they want to advertise on other sites as well. One way to do this, would be to include trackers from ad-networks, such as Merkle M1, in the admin dashboard.
Reading the Glassdoor reviews of Engenius makes it sound like a clone of Ubiquiti minus Ubiquiti's weird dislike of automated testing. Still, I hope that there is a decent alternative to UI
It's been a decade or so since I paid attention to Engenius, but I don't think it's right to describe them as a Ubiquiti clone. Engenius for a little while was the vendor of choice for people trying to build out consumer mesh networks. Ubiquiti may have existed then too, but didn't become the more popular choice until a little later.
Back then, Engenius had a lot of nice features and solid products, but their radios had some issues. I think we had to replace about 60% of the Engenius devices we deployed over a five year period. We never had to replace any of the Ubiquitis.
I dunno. I'd be starting all over again and looking for different options depending on the application. One of the nice things about Ubiquiti was that they did several things well.
For plug-and-play home wifi, Eero might be a good choice, with some reservations.
For slightly larger scale prosumer stuff, I might look for a way to use Ubiquiti's hardware but with something like https://hostifi.com/ , maybe with older-generation hardware that I still trusted.
For the DIY hobbyist, pfsense, openwrt, and kin are still popular.
Honestly, mesh networks have been a tarpit forever. It's been exceedingly difficult to make them reliable, with all the features people want (like access point roaming), without spending a lot of time or money. I got out of that niche a few years back and have no desire to get back in to it.
> For slightly larger scale prosumer stuff, I might look for a way to use Ubiquiti's hardware but with something like https://hostifi.com/ , maybe with older-generation hardware that I still trusted.
For prosumer stuff -- why would you not just self-host?!? The average prosumer is never going to need more than the cheapest DO droplet to host. Or else just run it on a computer that isn't on all the time -- prosumers either already have a server on all the time or don't care if they have logs going back 5 years. This seems to cost $60/mo at the least.
Also, I do not have much faith in this company with their website being so slow (maybe too many HN people are hitting idk, never heard of them).
I tried EnGenius a few years ago and had no end of problems with them. Switched to Ubiquiti and they've been rock solid ever since. I really, really hope Ubiquiti isn't going down hill.
Run a dockerized or VM controller (so you don't need CloudKey or Dream Machine) and pin the version; I'm running 6.0.43 and don't see any ads. Also make sure you find the option to install using a local account (not available on UDM).
An advertisement for a company’s products on the login page is one thing - but a full on banner in the interface itself (with apparently no way to disable) is another.
Mikrotik user here, I find their software really clunky and buggy.
My dream networking gear for my house is whitebox switches that I can run Debian on, but I can't find any at an economical price. Anyone know of 5-12 port ARM switches?
I have an RB4011iGS+5HacQ2HnD, oddly enough made by the networking company Amazon acquired (Annapurna). In theory, the CPU is supported in the mainline kernel, but I can't find an OpenWRT release or figure out how to package a custom Linux payload for flashing.
Mikrotik support refuses to provide me the uboot config they use or guide me through this process :(
No the CLI too unfortunately. While it is consistent, the editing of row numbers and the display of settings is really cumbersome. Setting up VLANs in particular is really tedious. Compared to real enterprise grade networking (Juniper/Cisco/Arista) it's dangerously unintuitive.
I've also tried automating it via Ansible with no success, has anyone had any luck with that?
I wouldn’t go that far. There’s often leaky abstractions, sure. But at 30% (or less) of the price of the equivalent Juniper hardware, I can usually live with it
But if you're doing this for a home network you don't really need that much bandwidth out of the home and can use the 2-port one with whatever switch you want.
I ran openwrt on a laptop in a one-armed router network configuration and it worked quite well.
I have a pcengines box as my router/gateway (with OpenBSD, works absolutely wonderfully), but I need access points as well. Especially because I cannot lay cables in this house, so my home office is relying on a wireless uplink.
Got any recommendations for access points that support wireless uplinks? My Unifi ones unfortunately tend to reboot frequently, sometimes needing manual readoption/full reset. It's bad if it happens during a meeting.
I don't think I need many bells and whistles, at this point I'm just content if it's stable.
I didn't even think of it when I wrote the original reply, but if the locations share power grid wires internally you can always g.hn over that copper to get the connection through the wall.
Ah, I thought about that as well, but I hear conflicting things about what that means for ham radio. I don't have a license yet, but certainly plan to have one.
Someone else suggested Ruckus in another subthread here. If they support wireless downlinks, that may be a good option.
EDIT: If it's known that (quality) power line stuff does not cause interference, it does become a good option.
haven't noticed any interference from powerline stuff, but the link-quality is definitely hampered by interference from other stuff (ie. temporary packet loss when some electric motor starts).
i consider it somewhere in between wifi and an ethernet cable in terms of quality.
Interesting, Winbox or WebFig/CLI? I find this clunkyness to be a good thing, compared to shiny and inflexible products. I have some experience of quite a few brands and so far Mikrotik is one of the most coherent and unified experiences. WebFig and CLI are the same across devices running RouterOS. Winbox is not my bag though.
RouterOS is not perfect but it does most things without being overly complicated. It's DHCPv6 relay agent is pretty much useless and I really look forward to WireGuard in RouterOS v7, if ever released as stable :D
I just noticed PrivacyBadger is blocking Google Analytics on the management interface now, has it always been like that? I just did the update and I'm not seeing that ad.
Holy shit! I thought I was the only one. The updates terrify me. Watching devices at 150 sites all provision simultaneously makes my asshole pucker so tight it creates a vacuum in the office. Why Ubiquiti?
Same here. I switched to Ubiquiti gear a year ago, and by and large it's been very solid, running 2 AC-Pros, 2 AC-HDs and a bunch of switches and a gateway. My controller is a docker image running on a NAS, and having read about the problems with the v6 versions I have stayed on v5, and likely will for a long time. It's a relatively simple home setup so I'm not using any of the fancier configuration settings. For my purposes, it's infinitely better than the hodgepodge of equipment I had before.
I still like the UniFi experience and have not had any problems. I am increasingly concerned about the direction the company is going. I figure I get can a quite a few years of service from my current setup. I don't need to update or change anything for the time being, so there is no urgency. I might still recommend a Ubiquiti setup to others but it would be heavily caveated.
Ubiquiti has really fallen far. The 5.x series of controller with the AP AC (S)HD was great, but everything since has been hot garbage. The 6.x series of controllers is still basically a public beta and riddled with bugs.
The newer MediaTek based APs have constant problems with client device compatibility, and can’t even reliably support DHCP on the newest official firmware and controller versions (or any of the last 10 releases).
I strongly advise you to look anywhere else for your WiFi needs unless they really turn the ship around somehow.
It's really sad. I used to recommend UniFi and their other hardware without a second's hesitation to all of my friends and the software and hardware quality control has just gone so far down hill. Now they're putting ads and privacy violating code in the admin user interface? What a complete disregard for user security and privacy in one of the most important places for it. Ugh. Add that to their GPL violations ( https://web.archive.org/web/20170317174847/http://libertybsd... ) and it's just too much for me going forward. I won't be installing or recommending Ubiquiti gear any more.
I'd like to receive security updates for at least the warranty life of the device, but I wouldn't mind paying for new features/major versions of the firmware. That said, owning what I've bought is non-negotiable. I paid for what I bought and I'm not going to allow myself, or my friends, to be turned into a product.
I’m in the same boat. I think it’s pretty weird that these days I’m more comfortable setting up a new router on the OpenBSD command line than I am with a Ubiquiti GUI.
TP-Link has actually upped their game with routers and APs lately. The next time I’m doing an installation that doesn’t need big-boy stuff from Cisco I’m going with them.
Give Mikrotik a look, too. Their HW is solid and while the UIs are decidedly old-school, they also expose just about every configuration option you could want, and everything can be configured from the command line.
It's also incredibly userunfriendly and lacking features for home users. I have both UniFi APs and a Mikrotik router and configuring RouterOS reminds me of trying to setup PPPoE internet on a Linux machine in 1990s.
A lot of manual settings with obvious features (NAT acceleration, NAT loopback, good VPN client, easy QoS, reasonable firewall) being extremely hard to configure well in comparison to competing SOHO equipment.
Yes, Mikrotik allows tweaking everything, but if home user is not interested in that, there's Quick Set for configuring the common scenarios.
Opposite way is much harder, as we can see with Ubiquiti: they hide and remove options, that are needed. With USG devices, I could have config.gateway.json configuration for some of those hidden options (like having site-to-site VPN using hostnames instead of IP addresses... well, Unifi UI cannot do that, but strongswan doing the IPSec on Unifi has no problem doint that). With UnifiOS devices, they are not addressing the issues, they are taking away the workarounds instead.
So they can advertise as much as they want, until they make devices fit for the purpose, there's no point in purchasing them. I don't see the situation improving anytime soon, they don't listen on their own community forums.
Mikrotik definitely doesn't have a flashy UI if that's what you care about, but the configuration for those things you list didn't seem all that different to ubiquiti's edge line from my research. Maybe I'm wrong, but I'm also happy with my choice and had no problems setting it up
Adding my thoughts from being a Turris Omnia user since the release (2016).
I think it's one of the better standalone WiFi router/AP's out there. If you want a long term supported open source/openwrt one it might be the best choice out there.
The original 3.11 release track has been super stable for many years, and is still being updated with security fixes. The 4.0 branch was a bit bumpy ride but since the 5.0 release early last year I've had zero issues.
Note that the default WiFi chipset is not the latest generation, and although the dev team is actively looking at the possibility to get some wifi6 working [1], I think it might be a while before that is out and stable.
As a bonus it has some cool features like built in support for CZ.nic's honeypot as a service [2] and lxc.
I second that. Despite investing in their infrastructure and recommending to friends (so basically being biased to like it) they just don't seem to care anymore.
I'm waiting for some company to take their place. And I don't see how making router software open source for a company that is selling hardware is not a win.
> I don't see how making router software open source for a company that is selling hardware is not a win.
There are several business tactics that are foreclosed by making the router software open source, mostly around price discrimination.
For example, you can't sell the same hardware for different prices with some hardware features turned off in the higher-volume cheaper SKU (which lets you apply greater efficiencies of scale to producing the hardware).
Of course, most companies that use price discrimination are only trying to maximize their profits, but it is worth noting that since hardware vendors that keep their software closed are able to pursue this tactic, it can leave hardware vendors with open software at somewhat of a disadvantage.
Ive been using them since gen1 hardware on 2.6 controllers. And their advertised features like zero handoff were hot garbage. Their channel selection to this day still sucks and isn’t even self aware. I often have multiple aps on the same channel close enough they have their own co-channel interference. This even happens on 5Ghz where theres plenty of space, not even considering DFS and I live in the country.
Standards like fast roaming are fraught with issues.
And at one point they added call homes/telemetry. I think the 5.x code.
That is to say they have always been buggy in one form or another. More so than other vendors like Aruba or Cisco, I’ve managed both at campus scale in a previous life. I wouldn’t put them in an enterprise but a smb or soho deployment sure.
When they did the call homes/telemetry I created some firewall rules for the controller. The controller can only talk to the Ubuntu repos. Aps are on a private vlan with no internet. I only open the controller up for Ubnt upgrades and then shut down the rules and disable the ubnt repos in sources.list.
If I were someone new in the market I would look at the ruckus unleashed platform for wifi. You can get aps with extensive features cheaper, especially grey market.
I would and have avoided UBNT for all routing and switching.
Mikrotik has a great feature set and is mostly stable for the mainstream features. Where they really fall apart is wifi throughput. Most devices won't top 400mb/s.
There is some hope on the horizon with v7 and the newest devices like Audience. Have seen reports of 1.3gb/s on these. But it's barely even beta right now.
Ruckus is pretty solid and you can usually find them used for great prices. Many units have a stand alone firmware available that negates the need for a controller.
We used to operate a satellite office whose primary internet service was Mikrotik directional AP’s bridged at both ends and this was maybe a half mile line of sight uplink.
My testing even on gigabit chipsets mirrors your 400 mbit per AP comment, at least in my situation.
If you are talking long distances you need to start looking at directional antennas...But to say..span an A-Frame house...they would do fine as others as well.
You will need to have one AP wired (thats the root AP) and the next ones can be wireless connected/meshed.
Note: regardless of the vendor Meshing will basically 1/2 your available wireless bandwidth. Its handy to extend coverage, and you will still be able to stream netflix and probably even things like facetime etc, but you are going to be essentially inducing a half-duplex connection into your "backbone".
for wireless AP's? It really depends. I actually just went with Wave2 stuff. In the end i bought some UAP-AC-HD's because
1. I am familiar/setup to handle their quirks and had 99% of the setup already done.
2. I really wanted an AP with Dual 5Ghz AND dual 2.4 Ghz dedicated antennas for some very specific applications (multiple rtsp streams over wifi+client plex wifi streams+client traffic. With the first two largely being on 2.4Ghz.
3. I am lazy (lots of this, my wife is less tolerant of me re-designing the wheel and being locked into an upgrade for days)
4. The security stuff i had mostly addressed anyway.
If this were a greenfield purchase or I didn't really need/want dedicated radios.
Id go with a ruckus. the r700 unleashed and/or r500 unleashed can be found cheap and are chock full of features. and 2x2:2 is normally going to be perfectly fine in a home setting.
Going with ones that are "unleashed" mean that the controller runs on the AP's. not extra hardware needed.
For most home uses. wave1 is still more than enough and their Wave1 AP's can be had cheap. Hell you can probably even get one of the R710/R510's in a very good price range and be at wave2.
For switching...mikrotik is solid for cheap and works and stable. Even modems etc. For firewalls my preference is BSD, so something like opnsense...But mikrotik will probably do okay as well.
> Id go with a ruckus. the r700 unleashed and/or r500 unleashed
This is the problem with Ubiquiti, and maybe why they feel they can advertise in their UI and let their quality slip: a new r710 lists for $1295.
That's almost four times as much as a UAP-AC-HD. On sale for $810 it's still over twice as much. Some of the Ubiquiti WiFi 6 stuff can be had for $100-150.
Like, yeah, I'm sure Ruckus is way better, but I'd hope so for the price difference.
You can get the r710 on Amazon for $430. R510 are even cheaper around 230 or so. These aren’t the latest iterations of their hardware. Ruckus doesn’t require a contract to get updates firmware/software. Theses are around the same prices as a uap-ac-hd.
Regardless of list price. They can be had as cheap as ubiquiti and their software/firmware is not only better but doesn’t have all the telemetry crap built in.
No it is still new. Just not the latest line of production gear (Which would be the R750). Im not going to post direct links because i assume that's a no-no.
Ruckus refreshes their gear fairly frequently and the overstock can be had at a steep discount. Those not needing to do huge, high density deployments in things like auditoriums, stadiums, classrooms and the like but are just looking for something for their home can get a really solid piece of gear (imho better than ubnt) at a steep discount.
Its not really the same as even grey market Cisco stuff.
Like i said, i stuck with ubnt because of some specific things, including the fact that I could re-use mounting holes/measurements and could upgrade in an hour vs rebuilding entirely.
Thanks for the detailed information. Unfortunately, the UK prices don't reflect the US prices, both on Amazon UK and other retailers here. The only way to get Ruckus gear that cheap (or cheaper) is to get them used on eBay. I think I'll probably go down the TP-Link Omada route as someone suggested elsewhere in the comments here. The feature set and prices look to be about right for me.
The craziest part is they were poised to dominated the SOHO / MSP markets. There’s NO competition or viable products in those markets. I’ve searched high and low.
Instead of focusing on their strengths like great value with self hosted management infrastructure, they look like they’re setting up to force subscriptions on their users. All of the subscription BS for centralized management is so expensive that it’s literally cheaper to hire someone to manage everything by hand (the old fashioned way).
This.If I’m gonna have to waste time and effort to block my network gear sending its spy data back home, and blocking ads it tries to show, why wouldn’t I just do that on much cheaper throwaway Chinese hardware? I’m much more likely to buy stuff like TPLink gear that can be flashed with OpenWRT, or jump feet first into Mikrotik gear - than ever spend more money on Ubiquity’s modern shitware...
Was that running the original Firmware, or OpenWRT? In my experience, all vendor firmwares are to be regarded just as "proof-of-electrical-functioning", to be replaced with a OpenWRT running a recent kernel. Similar to how to you reinstall/upgrade a new laptop on the second boot fresh out the box.
tried openwrt, tried stock, and always non-dfs. what i do know is that after i plugged in my NAS using RJ465 the whole router collapsed and had to be force rebooted. every single time.
fast forward, and my UDM Pro handles everything, including ax.
What’s stupid is that ubiquiti could have easily advertised its hardware less intrusively in so many different ways.
For example, they could show “devices that integrate with <AP or Switch or Router>” or integrate the changelog of all their releases into the management UI for all products.
The list goes on for the more thoughtful ways that could have done it, but instead they did they took the absurdly lame way of advertising via banners. Jeeze.
This is disappointing. I've been waffling on buying a full Unifi setup for my home, but all the terrible reviews for the UDMP, and now this, have given me pause.
Are there any decent non-Chinese pro-sumer networking suites like Unifi that folks would recommend?
I was in the same boat. I've been running the older equipment that's currently doing just fine.
We're looking at the prospect of getting fiber this summer. Figured I'd use the opportunity to get all in on Ubiquiti. This is the nail in the coffin for me.
After some unexplainable sustained 100% packet loss (for minutes at a time) I switched from a set of eero routers to ubiquiti gear a few years ago and it's been a great experience so far. A single AP-AC-lite covered my entire 2600sqft house from the basement.
Somebody I know has stayed on the eero train (on my recommendation before I gave up on them) and he says the company has prioritized features he doesn't care about and deprioritized features he wished it had.
I think it's a common sentiment here,
but I'm turning into a more extreme anti-advertising zealot every day.
Even the large ad networks show paid content that violates what I would consider widespread ethical norms - presenting everything from hyperbolic claims to straight fraud to people with little repercussion. I'm reminded of a poignant example in television, where episodes of Jeopardy! are intertwined with their shilling quackery like Prevagen to an audience using the fear of aging. Perhaps I'm growing increasingly blind to any positive impacts of the discipline, but ethically I find it hard to separate the fraudsters from the engineers that enable them.
I guess advertising is saving me a lot of money in the end by cultivating this hatred, and though I can't change the world this screed was therapeutic.
Having less exposure to advertising, given that I watch less advertising-interrupted television than ever and have had a pi-hole running for a good couple of years, when I am exposed to advertising I'm reminded of how incredibly bottom-feeding it is.
I've said it before: Advertising is a psychological attack on human weakness.
> I find it hard to separate the fraudsters from the engineers that enable them
Paraphrasing because I can't find the original:
"What does it say of humanity to have so many of the smartest minds in the world working on making advertising more effective"
At the same time, we have created super computers that run on expansion cards that are being used for that same black-hat activity. Currently measured in the 10s of teraflops of processing power.
More and more I'm seeing a general public backlash against advertising. It's been so large and so prevalent that people find it irritating enough to avoid products that are getting heavy pushes. I've found myself having a visceral "stay the hell away from this product" reaction on occasion as well.
My stance on web advertising has always been "go back to static, not in-my-face, non-animated, no audio, no javascript ads and I'll consider turning off my ad blocker."
I'm absolutely not letting my computer get infected through the ad networks, which is a separate issue really.
I'm even more concerned about the second order effects of so much being driven by driving clicks to ad dollars. This is obviously what causes the attention economy and maybe many of our negative trends in society. How much less attention would BS articles on anti-vax and QAnon nonsense get without all of this? How many fewer content farms that are just trying to waste my time would we have?
I would love to see how an alternative reality without advertisements would look. I'm not sure if we could ever fully accomplish this given that there always will be people willing to pay for eyeballs and people happy to take money or goods to change their messaging.
Yeah, you're probably preaching to the choir here. The tracking and advertising industry has an unprecedented amount of influence and control over our lives, it's getting harder to pry your life away from their claws.
Sadly, I think things need to get much worse before they get better. People still don't understand the dangers of it, though I believe this has already become a sociopolitical problem instead of just an annoyance.
> a poignant example in television, where episodes of Jeopardy! are intertwined with their shilling quackery like Prevagen to an audience using the fear of aging.
Episodes of Jeopardy, now temporarily hosted by Dr. Oz.
This is not that simple, unfortunately. Advertising in itself is not bad. Why informing about a good product is something wrong? How we would have Tesla or iPhones, etc. without advertising. Nobody would know about all of those.
But something went wrong at some point.
Similarly with options - options were a great way farmers could offload risk to the products buyers, who, in exchange could get better prices.
But when options started to be used for something which is not much different from lottery bets, again, something went wrong.
This is important lesson. Capitalism is a great system, but it cannot be narrowed to greed. Capitalism requires working ethics since we are not able to regulate every aspect of life - it is practically impossible.
But ethics no longer works. Big IT companies owners were fixing employees wages, who cared. Amazon owner does not care about warehouses workers miserable work conditions. Facebook owner is ok to steer people into fights, as this increases "engagement", ads clicks and revenue.
Has it happened that Jeff Bezos or Mark Zuckerberg was asked to leave his country club or whatever place they are hanging out because of the unethical behavior? Obviously not.
> Capitalism is a great system, but it cannot be narrowed to greed. Capitalism requires working ethics since we are not able to regulate every aspect of life - it is practically impossible.
Whether it's still Capitalism at that point is an interesting question. Some people seem to call any market-based system capitalism. For others the lack of ethics is definitional.
My own view is that we probably ought to stop using the term "capitalism" entirely for the most part, and be more specific about what we mean.
> This is important lesson. Capitalism is a great system, but it cannot be narrowed to greed. Capitalism requires working ethics since we are not able to regulate every aspect of life - it is practically impossible.
The market is literally a big, greedy - in algorithmic sense - optimization engine. It optimizes away everything that stands in the way of making more money. It optimizes away quality, and it absolutely optimizes away any and all ethics. Marketing and advertising are a clear example of this in two separate ways: on the object level, they're already doing everything they can that isn't strictly illegal[0]; on the meta level, you see that competing on marketing and advertising has a better ROI than actually trying to make better products.
You cannot expect a sense of ethics from a system which requires ethics as a containment vessel, and continuously tries to erode it. Much like you can't expect water to just keep flowing in a straight line, in absence of a bank or pipe that would constrain its expansion.
--
[0] - And enforced. GDPR demonstrates that with insufficient enforcement, otherwise reputable companies will absolutely break the law.
You picked two terrible examples. I first heard about the iphone via word of mouth as a friend had one and was showing it off at the bar.
I hear about new ones by word of mouth from collegues (from people who go out of their way to watch product releases on the apple away day or whatever it is). I hear about new gadgets by paying for someone to go to CES and look at what's new.
Tesla, and starlink, again is word of mouth, or deliberatly seeking out information on it (by visiting tesla.com for example).
Advertising is the most toxic virus ruining everything it touches. TV, then Cable TV, radio, usenet, on and on. Can anything survive the destructive ravage of ads?
In my dream world, all forms of push advertising would be illegal, full stop. Meanwhile, I'll put whatever effort it takes to block everything.
Advertising is only able to exist when there's centralized entity controlling the presentation of the content. TV and radio are good examples — your receiver simply plays whatever is on air, and has no idea what's in there. Social media services like Facebook strive to control the experience end-to-end precisely because of that too, and that's the reason why public APIs for Facebook and Instagram got reduced from "enough to build a third-party client" to "you can post, getting any info for the current user requires a manual review, and you can't do anything else".
Contrast this with email and phone, the only successful federated systems. Spam is illegal in some countries, and there's a concerted effort by everyone to fight it.
When no one controls the whole experience end-to-end, there is simply no technical possibility to expose you to something against your will. This is why decentralized social media is the future. I just can't see how Facebook could survive much longer unless there's a drastic change of some sort.
I am very much anti-advertising, at least the way it currently operates on the web and the inherent invasion of privacy but I think from a technical perspective what's happening here is very different compared with how a traditional banner advert is being displayed and it's not constructive to directly compare the two.
The sentiment definitely resonates. The way I see it, advertising is a cancer on modern society[0]. The two big tech-related social problems so frequently discussed these days - surveillance capitalism and outrage culture - are directly caused by it.
> Perhaps I'm growing increasingly blind to any positive impacts of the discipline, but ethically I find it hard to separate the fraudsters from the engineers that enable them.
There are definitely useful spinoffs - like they always are, when you throw a lot of money and a bright mind into solving tough problems. But it's hard to argue the positives here where the primary driver is a huge net social negative, and spinoffs could've been done pursuing better goals.
My mother got a new laptop, and at some point I got remote access to help her get a few things set up (password manager/ad-blocker/backups/etc). I was really amazed at what her Facebook feed looks like. Actual Facebook advertisements for literal scams. 'Michelle Obama's investment tips' for some bitcoin platform. 'Patagonia Outlet' with items at 25% of the cost, at domains like patagonia-outlet.shjss72828.sub.domain.pl, etc. I scrolled for a while and more than half of the advertisements were like this. This is what senior citizens are dealing with.
This is just not ok. I understand that screening everyone in the world's personal posts/shares for 'fake news' is a daunting task, but KYC-ing advertisers and screening/auditing ads for blatant and obvious scams or false marketing is many orders of magnitude less resource intensive... and should be required by laws with teeth.
If the costs of doing so mean that it's impossible to make a profit as an ad platform, then that business model should die.
> but KYC-ing advertisers and screening/auditing ads for blatant and obvious scams or false marketing is many orders of magnitude less resource intensive
And very easy to automate. We know they can do it. But until there is sufficient threat of action if they don't then they are happy to let their users get fleeced in exchange for the advertising $.
There is an argument that "if they crack down on the easy to detect stuff, the scams will just get more clever" but that is not at all why they aren't making more effort.
One funny thing I've noticed with facebook, is that if you start responding to such adverts with comments along the lines of "this is obviously a scam because..." responses, they start showing you more and more of that sort of scam presumably because you have shown engagement with the previous ones.
For an extra dose of cynicism keep in mind Youtube is demonetizing countless videos supposedly because advertisers don’t want to be associated with such content. Garbage ads certainly aren’t exempt from that logic, yet there doesn’t seem to be as much of a crackdown on them.
"Not OK" is forgetting to clean up your table after the meal. What these scammers are doing is simply fraud and Facebook should be liable as they enable it.
I remember when I set up my first FB ad (maybe 10 years ago?) it took a few days because some of them were manually reviewed. Fast forward 10 years and they accept any fraudster who pays them a few bucks. Just low scum, that's all.
You're upset about regulators, not about advertising.
It seems like you're mixing advertising = fraud, when in reality it's just people who commit fraud that use advertising because regulators can't seem to act.
I got my first Internet connection when adblock was already a thing (though I was definitely an early adopter). I don't know how ad-filled Internet looks like and I will fight tooth and nail with every tool at my disposal to keep it that way: adblocks, Pi-hole, VPNs on phones, paying for ad-free versions, it doesn't matter. I do not watch TV and I will do my best to keep ads out of all my screens.
Good for you mate, just keep in mind that a large part of your purchase decisions are still influenced by paid media (not only advertising) - as long as you're aware of that, you're safer, that's the true adblock.
Because that is small potatoes compared to the tracking, matching and profiling of individuals. I believe that is the big threat to society (and possibly democracy)
One follows from the other. Tracking, matching and profiling of individuals isn't done for kicks, or because evil governments are asking - it's done for two reasons:
- Tune the advertising to best manipulate a particular person (think of it as a PID controller tuning inputs to a system);
- Ad attribution - i.e. determining which companies in the advertising chain should get paid when a customer buys something from someone.
Restricting the shape and form advertising is allowed to take addresses the driving factors behind privacy violations.
I recall an interesting interview I had with eyeo, which bought the original Adblock Plus extension and have since been pushing for their "Acceptable Ads".
- Do the ads still track users?
- Yes, but they are not unobtrusive
- Yeah, that is not acceptable.
At this point I'm just generally not okay with being marketed to against my will. Especially in places that have absolutely literally nothing to do with spending money — like, you know, every single news website and every single social media service.
If I want to buy something, and if I don't know exactly what, I'd like an unbiased list of things in that category to choose from. And when I don't want to buy anything — which is, again, when I'm reading an article for example — I don't want to be reminded that economy exists AT ALL. There is no place or circumstance in my life where advertising, as it is known today, would be anything but annoyance to me. Which is why all my devices have ad blockers installed and third-party cookies disabled.
That's not my concern, neither is it my responsibility to make sure whoever the website belongs to gets paid. Paywall it for all I care. Implicit contracts aren't a thing.
I have been watching Ubiquiti slowly go downhill over the past five years. I finally jumped ship to Mikrotik for a recent addition to my network and they seem like a good replacement for power users. Nothing they sell is quite as user friendly as the Ubiquiti gear, but I would rather have to spend a few minutes at a console then loose control over my network because a SaaS went offline
> loose control over my network because a SaaS went offline
That doesn't sound correct. I've always been able to access my controller without internet access as long as I was on the same LAN and was hitting the controller IP directly.
One auto update they broke my laptop's ability to connect. Turns out they'd auto enabled 5 GHz band steering, but my laptop's chipset only did 2.4 GHz. Even when it was enabled, the Android management app said the setting was disabled.
Customer support directed me to the forums. Ubiquiti didn't respond to my post. I posted on a thread where someone else was seeing the same problem, where a number of people gaslit me and blamed me for the problem.
I still have no idea if they've fixed their band steering. Or how they decided to force 5 GHz on a 2.4 GHz chipset. Or how to get support. Or how to report a problem.
Definitely won't be buying their products in future.
I have an old laptop that I was trying to fire up to run some lightweight games on. I could not get it to connect to wi-fi for the life of me. Must be the band steering.
We have 5 ubiquiti switches on our office network that are running 4 year old firmware. The next version after started dropping DHCP packets! I followed the thread for months and months and they never fixed the issue. I gave up caring, stopped watching the thread and swore never to buy any of their products again.
I do wonder if they ever fixed the issue. The switches lock up every couple months and need a hard reboot.
We also had all sorts of problems with their APs and switched over to Cisco Meraki. As much as I hate Cisco and their ridiculous cost at least the Meraki line works.
I'd been wondering about this.. Ubiquiti's wan and firwalling is super weak, especially for multi-site, when compared to some of the sd-wan products out there.
What's wrong with the ap's? Have you used the switches? How much of their kit can managed via a single console?
I had a Dream Machine (the all-in-one, not the pro) and it fried the very next day. Even after that I was willing to give Ubiquiti another try, but now I don’t know what to do.
I have two of them at two different sites. They’ve been completely fine for me, apart from the problems from the 1.8.x firmwares which probably should never have reached the ‘Release’ channel. (The recently-released v1.9.2 appears to have fixed all my issues from 1.8.x)
I’m definitely not pushing either of them very hard or doing anything particularly fancy, but they’ve certainly been a big step up from the aging Apple routers I’d been using before, at least for me.
Hmm, I'm still interested in a Dream Machine but I've been holding off for the so called SE[1], with the display on the front, similarly to the Amplifi Alien.
I see a lot of complaints here, but nobody offering any substantive alternatives. Mikrotik is mentioned several times, but always with the caveats that it has glitchy UIs or poorly implemented features. EnGenius... you've got to be kidding.
I've been a UI fan for many years and was also dismayed by the 6.x releases and the switch to Mediatek, but every time I went looking around, I saw the same tire fires at other vendors. At least with UBNT it's "the devil you know".
Cambium is interesting, I've been playing around with their cnPilot stuff and it's been pretty solid.
Same. There are enough annoyances with UI gear to keep us all here griping for days. But when you try to find an alternative company for your Wireless needs you realize quick that you're perfectly willing to put up with UI's stupid annoyances. Granted a few glitches have wasted days of my life trying to correct. But I'll be sticking with them for the foreseeable future. Over the years I've had satisfactory uptime and performance.
I usually prefer the regular line over the UniFi series of UI and they tend to have less UI polish but that‘s what usually tells that it‘s a product for technical people.
I won‘t switch to the UniFi series for now even though they would provide easier management. But I‘m a bit worried that it‘s going to be an Apple-style product line where the pros are somehow not respected as much as before.
Building your own router with pfSense is one prosumer solution I've considered, by shied away from due to lack of time[1]. Not sure how a pfSense setup would work with wireless APs though.
pfSense might have a lot of features, but the UI is terrible in all possible ways. They should stop letting their programmers "design" the UI; dumping the raw database into an HTML table is not a good user interface.
If you want a really nice UI then you need to stop looking at consumer gear and just buy enterprise cisco kit. All of these budget brands discussed here are consumer.
I’m curious, have you used EnGenuis before or just based on people’s responses? I was thinking of using them for a project but wouldn’t mind hearing an opinion from someone familiar with them.
three days ago smallnetbuilder reviewed a bunch of routers. I noted that the most recent router on the list that could run openwrt was the three year old r7800. someone mentioned building their own box for an ap and I discussed how hard it is to get wifi add-on cards[1], and oh, by the way, good luck running anything modern yourself.
the state of our wifis, right now, is extremely poor. upmarket expensive gear becoming a hostile rent-seeking shark is a sign of how late-stage the market is right now.
i will allow: there is a lag somewhat to be expected. wave2 and wifi6 and wifi6e have been all kind of sounding this new stuff stuff out. wifi6e feels a bit like a natural culmination, for a while, but we're only just making it across a long series of changes just now. the standards have been in flux, enhancing.
still, the driver situation feels terrible. hardware is hard to get except as consumer electronics running provided shrinkware. even in terms of stations not ap's, companies like intel can both do a good job, insure reasonable support, decent enough hardware these days, but also neglect basic love & care things like support for 802.11s mesh & other sincere basics.
anyhow. wifi6e looks like the culmination of a lot. the situation might stabilize some. the future is finally here & only now do we commence more evenly distributing use/usability of it. maybe. maybe not. maybe vendorization floggings will continue. i'm hoping somehow more ap class add-on systems start becoming in any way available again. hoping systems like wifi-p2p and wifi-aware and wifi mesh (802.11s) see general support. I'd love to see some pro-active movements in wifi world, some chipmakers really doing right, but wifi feels forever like the most cagey & slowest developing linux driver area. communication, not well supported.
324 comments
[ 3.0 ms ] story [ 87.7 ms ] thread* telling the customer they’re wrong...
* incorrectly...
* and with multiple typos.
Yes, very cool.
"This is unacceptable and this change makes me very angry. I will not be considering future products from Ubiquiti"
Have you never interacted in a non-business sense with someone who worked customer support? They DO NOT have control of what they tell you. They will get fired real quick if they don't follow the script. Especially someone working the first line of support in a chat box.
C'mon man.
This is an ad, pure and simple. They're pushing a product they know I don't own with a clear call to action, and without a way for me to remove it or disable it.
So far.
Yes.
> this confuses people
No.
> but it also definitely isn't the entire monstrosity we have come to assume when hearing about ads on web pages.
Slippery slope. I used to think of Ubiquiti as prosumer. Now it's definitely not.
Then, the software that was written to manage a product that I paid for shoves a call-to-action that takes a considerable amount of the page to suggest I buy it. And I can’t dismiss or close it.
The difference with ads on the web is that I paid for the privilege of being served this garbage. In my mind, this is much worse.
No. Somewhere a marketer queried a database using a simple marketing automation tool to decide which of a variety of banners to show you.
Once they have the ad channel setup, it would be dumb not to try to take advantage of it.
This ad-supported model isn't needed or wanted.
Ads, even if self-promotion, are a bellwether for a company being desperate or greedy, neither of which bodes well for their long-term product quality, usually because it indicates sales people are making the decisions. They also often lead to a slippery slope of increasing user data collection
Even if Ubiquiti was as good as it was pre-2019, if they pulled this advertising stunt then I'd not consider upgrading. There's now no chance I'll be sticking with Ubiquiti for my 6E upgrade.
https://news.ycombinator.com/item?id=25735032
I've also had numerous cases where routing has stopped for no reason. This I find really bad as, "it has one job ...".
That was a fun thing to explain. My respect for this company was stellar, but is now around rock bottom.
For me that was the "one job" problem.
While researching I found-out that you must use a ui.com account to setup the UDM (the apologists call it "mandatory online registration"). While you can setup a local account and disconnect from ui.com afterwards, I was perplexed that they would choose to do this.
Their user forums had at least two flamewar threads with people pointing-out the security implications of linking your local network to a remote, internet-accessible service but zero communication from Ubiquity themselves. Fast-forward to January and we saw the disclosure of a massive breach on ui.com.
So I ditched the Ubiquiti cameras for Geovision (better bang for $), kept the switches, APs and have been running the controller in containers (https://github.com/jacobalberty/unifi-docker). Needless to say, I'm not going to upgrade from version 6.0.43, which does not include advertisement.
I hope it doesn't get any worse because I do love how easy it is to maintain VLANs, firewall rules and IPS services.
I’m not frustrated when I get an email or dismissible banner ad from a company regarding a new product. It’s possibly interesting to me, and easily dismissible if not.
Sure, technically, this is advertising. But the biggest problems people have with advertising (privacy invasion, seeing too many ads, or seeing low-quality ads) aren’t present in this example.
It’s advertising in the exact same way that GitHub might show a banner informing me of a new plan they have. While it can be annoying, yes, I would disagree that it’s “as bad” as typical advertising practices.
[1] https://news.ycombinator.com/item?id=26630127
from 2003: https://www.theregister.com/2003/11/07/help_my_belkin_router...
I have never bought a belkin product since.
Good info here https://www.youtube.com/watch?v=4G2g7Txgzgw
Back then, Engenius had a lot of nice features and solid products, but their radios had some issues. I think we had to replace about 60% of the Engenius devices we deployed over a five year period. We never had to replace any of the Ubiquitis.
I wouldn't buy Ubiquiti now though.
For plug-and-play home wifi, Eero might be a good choice, with some reservations.
For slightly larger scale prosumer stuff, I might look for a way to use Ubiquiti's hardware but with something like https://hostifi.com/ , maybe with older-generation hardware that I still trusted.
For the DIY hobbyist, pfsense, openwrt, and kin are still popular.
Honestly, mesh networks have been a tarpit forever. It's been exceedingly difficult to make them reliable, with all the features people want (like access point roaming), without spending a lot of time or money. I got out of that niche a few years back and have no desire to get back in to it.
For prosumer stuff -- why would you not just self-host?!? The average prosumer is never going to need more than the cheapest DO droplet to host. Or else just run it on a computer that isn't on all the time -- prosumers either already have a server on all the time or don't care if they have logs going back 5 years. This seems to cost $60/mo at the least.
Also, I do not have much faith in this company with their website being so slow (maybe too many HN people are hitting idk, never heard of them).
The rest of the gear is working tiptop for me.
Guess we just got to step 2.
Sticking with Mikrotik for now.
My dream networking gear for my house is whitebox switches that I can run Debian on, but I can't find any at an economical price. Anyone know of 5-12 port ARM switches?
Mikrotik support refuses to provide me the uboot config they use or guide me through this process :(
The CLI, on the other hand, is pretty great. Everything is consistent.
I've also tried automating it via Ansible with no success, has anyone had any luck with that?
But if you're doing this for a home network you don't really need that much bandwidth out of the home and can use the 2-port one with whatever switch you want.
I ran openwrt on a laptop in a one-armed router network configuration and it worked quite well.
Got any recommendations for access points that support wireless uplinks? My Unifi ones unfortunately tend to reboot frequently, sometimes needing manual readoption/full reset. It's bad if it happens during a meeting.
I don't think I need many bells and whistles, at this point I'm just content if it's stable.
Someone else suggested Ruckus in another subthread here. If they support wireless downlinks, that may be a good option.
EDIT: If it's known that (quality) power line stuff does not cause interference, it does become a good option.
i consider it somewhere in between wifi and an ethernet cable in terms of quality.
https://www.crowdsupply.com/traverse-technologies/ten64/ https://www.turris.com/en/
RouterOS is not perfect but it does most things without being overly complicated. It's DHCPv6 relay agent is pretty much useless and I really look forward to WireGuard in RouterOS v7, if ever released as stable :D
Made a decision long time ago to go all-Mikrotik instead of ubnt, definitely glad I made that decision several years ago.
"Shit show" doesn't even begin to describe it.
Makes it almost sound unintentional.
I still like the UniFi experience and have not had any problems. I am increasingly concerned about the direction the company is going. I figure I get can a quite a few years of service from my current setup. I don't need to update or change anything for the time being, so there is no urgency. I might still recommend a Ubiquiti setup to others but it would be heavily caveated.
The newer MediaTek based APs have constant problems with client device compatibility, and can’t even reliably support DHCP on the newest official firmware and controller versions (or any of the last 10 releases).
I strongly advise you to look anywhere else for your WiFi needs unless they really turn the ship around somehow.
It's not like their hardware is cheap, I'd much rather pay a monthly subscription than be served ads.
Why not both?!
/s but probably for real eventually.
TP-Link has actually upped their game with routers and APs lately. The next time I’m doing an installation that doesn’t need big-boy stuff from Cisco I’m going with them.
[1] https://news.ycombinator.com/item?id=26629518
And that's it's biggest problem.
If only few sites used it, it wouldn't be that big of a deal
A lot of manual settings with obvious features (NAT acceleration, NAT loopback, good VPN client, easy QoS, reasonable firewall) being extremely hard to configure well in comparison to competing SOHO equipment.
Opposite way is much harder, as we can see with Ubiquiti: they hide and remove options, that are needed. With USG devices, I could have config.gateway.json configuration for some of those hidden options (like having site-to-site VPN using hostnames instead of IP addresses... well, Unifi UI cannot do that, but strongswan doing the IPSec on Unifi has no problem doint that). With UnifiOS devices, they are not addressing the issues, they are taking away the workarounds instead.
So they can advertise as much as they want, until they make devices fit for the purpose, there's no point in purchasing them. I don't see the situation improving anytime soon, they don't listen on their own community forums.
I think it's one of the better standalone WiFi router/AP's out there. If you want a long term supported open source/openwrt one it might be the best choice out there.
The original 3.11 release track has been super stable for many years, and is still being updated with security fixes. The 4.0 branch was a bit bumpy ride but since the 5.0 release early last year I've had zero issues.
Note that the default WiFi chipset is not the latest generation, and although the dev team is actively looking at the possibility to get some wifi6 working [1], I think it might be a while before that is out and stable.
As a bonus it has some cool features like built in support for CZ.nic's honeypot as a service [2] and lxc.
[1] https://forum.turris.cz/t/wifi-6-ax-adapter/10390/63 [2] https://haas.nic.cz/
I'm waiting for some company to take their place. And I don't see how making router software open source for a company that is selling hardware is not a win.
There are several business tactics that are foreclosed by making the router software open source, mostly around price discrimination.
For example, you can't sell the same hardware for different prices with some hardware features turned off in the higher-volume cheaper SKU (which lets you apply greater efficiencies of scale to producing the hardware).
Of course, most companies that use price discrimination are only trying to maximize their profits, but it is worth noting that since hardware vendors that keep their software closed are able to pursue this tactic, it can leave hardware vendors with open software at somewhat of a disadvantage.
Tesla doesn't release the software to their cars.
Standards like fast roaming are fraught with issues.
And at one point they added call homes/telemetry. I think the 5.x code.
That is to say they have always been buggy in one form or another. More so than other vendors like Aruba or Cisco, I’ve managed both at campus scale in a previous life. I wouldn’t put them in an enterprise but a smb or soho deployment sure.
When they did the call homes/telemetry I created some firewall rules for the controller. The controller can only talk to the Ubuntu repos. Aps are on a private vlan with no internet. I only open the controller up for Ubnt upgrades and then shut down the rules and disable the ubnt repos in sources.list.
If I were someone new in the market I would look at the ruckus unleashed platform for wifi. You can get aps with extensive features cheaper, especially grey market.
I would and have avoided UBNT for all routing and switching.
There is some hope on the horizon with v7 and the newest devices like Audience. Have seen reports of 1.3gb/s on these. But it's barely even beta right now.
Ruckus is pretty solid and you can usually find them used for great prices. Many units have a stand alone firmware available that negates the need for a controller.
My testing even on gigabit chipsets mirrors your 400 mbit per AP comment, at least in my situation.
https://docs.ruckuswireless.com/unleashed/200.1.9.12/c-MeshO...
If you are talking long distances you need to start looking at directional antennas...But to say..span an A-Frame house...they would do fine as others as well.
EDIT: The "Wireless Bridge Topology" is exactly what I need, here, but it's not entirely clear to me yet if I'd really need two APs on the gateway side as the picture suggests: https://docs.ruckuswireless.com/unleashed/200.1.9.12/c-Suppo...
Note: regardless of the vendor Meshing will basically 1/2 your available wireless bandwidth. Its handy to extend coverage, and you will still be able to stream netflix and probably even things like facetime etc, but you are going to be essentially inducing a half-duplex connection into your "backbone".
1. I am familiar/setup to handle their quirks and had 99% of the setup already done.
2. I really wanted an AP with Dual 5Ghz AND dual 2.4 Ghz dedicated antennas for some very specific applications (multiple rtsp streams over wifi+client plex wifi streams+client traffic. With the first two largely being on 2.4Ghz.
3. I am lazy (lots of this, my wife is less tolerant of me re-designing the wheel and being locked into an upgrade for days)
4. The security stuff i had mostly addressed anyway.
If this were a greenfield purchase or I didn't really need/want dedicated radios.
Id go with a ruckus. the r700 unleashed and/or r500 unleashed can be found cheap and are chock full of features. and 2x2:2 is normally going to be perfectly fine in a home setting.
Going with ones that are "unleashed" mean that the controller runs on the AP's. not extra hardware needed.
For most home uses. wave1 is still more than enough and their Wave1 AP's can be had cheap. Hell you can probably even get one of the R710/R510's in a very good price range and be at wave2.
For switching...mikrotik is solid for cheap and works and stable. Even modems etc. For firewalls my preference is BSD, so something like opnsense...But mikrotik will probably do okay as well.
This is the problem with Ubiquiti, and maybe why they feel they can advertise in their UI and let their quality slip: a new r710 lists for $1295.
That's almost four times as much as a UAP-AC-HD. On sale for $810 it's still over twice as much. Some of the Ubiquiti WiFi 6 stuff can be had for $100-150.
Like, yeah, I'm sure Ruckus is way better, but I'd hope so for the price difference.
Regardless of list price. They can be had as cheap as ubiquiti and their software/firmware is not only better but doesn’t have all the telemetry crap built in.
I'm assuming you're talking of used hardware as this isn't the case, at least in the UK.
Ruckus refreshes their gear fairly frequently and the overstock can be had at a steep discount. Those not needing to do huge, high density deployments in things like auditoriums, stadiums, classrooms and the like but are just looking for something for their home can get a really solid piece of gear (imho better than ubnt) at a steep discount.
Its not really the same as even grey market Cisco stuff.
Like i said, i stuck with ubnt because of some specific things, including the fact that I could re-use mounting holes/measurements and could upgrade in an hour vs rebuilding entirely.
https://imgur.com/a/a4LLnqH
Instead of focusing on their strengths like great value with self hosted management infrastructure, they look like they’re setting up to force subscriptions on their users. All of the subscription BS for centralized management is so expensive that it’s literally cheaper to hire someone to manage everything by hand (the old fashioned way).
every 3-6 hours the wifi just died and had to be again force rebooted.
will not touch linksys again. they're a consumer brand, and my requirements are clearly SMB.
imo their best product was the WRT54GL.
fast forward, and my UDM Pro handles everything, including ax.
Has the makings of an internal power struggle where the people who actually hold things together we’re not on board and bailed.
For example, they could show “devices that integrate with <AP or Switch or Router>” or integrate the changelog of all their releases into the management UI for all products.
The list goes on for the more thoughtful ways that could have done it, but instead they did they took the absurdly lame way of advertising via banners. Jeeze.
Are there any decent non-Chinese pro-sumer networking suites like Unifi that folks would recommend?
We're looking at the prospect of getting fiber this summer. Figured I'd use the opportunity to get all in on Ubiquiti. This is the nail in the coffin for me.
Somebody I know has stayed on the eero train (on my recommendation before I gave up on them) and he says the company has prioritized features he doesn't care about and deprioritized features he wished it had.
I hope Ubiquiti doesn't go down this road.
Even the large ad networks show paid content that violates what I would consider widespread ethical norms - presenting everything from hyperbolic claims to straight fraud to people with little repercussion. I'm reminded of a poignant example in television, where episodes of Jeopardy! are intertwined with their shilling quackery like Prevagen to an audience using the fear of aging. Perhaps I'm growing increasingly blind to any positive impacts of the discipline, but ethically I find it hard to separate the fraudsters from the engineers that enable them.
I guess advertising is saving me a lot of money in the end by cultivating this hatred, and though I can't change the world this screed was therapeutic.
I've said it before: Advertising is a psychological attack on human weakness.
> I find it hard to separate the fraudsters from the engineers that enable them
Paraphrasing because I can't find the original:
"What does it say of humanity to have so many of the smartest minds in the world working on making advertising more effective"
> this screed was therapeutic
+1
My stance on web advertising has always been "go back to static, not in-my-face, non-animated, no audio, no javascript ads and I'll consider turning off my ad blocker."
I'm absolutely not letting my computer get infected through the ad networks, which is a separate issue really.
I would love to see how an alternative reality without advertisements would look. I'm not sure if we could ever fully accomplish this given that there always will be people willing to pay for eyeballs and people happy to take money or goods to change their messaging.
Sadly, I think things need to get much worse before they get better. People still don't understand the dangers of it, though I believe this has already become a sociopolitical problem instead of just an annoyance.
Episodes of Jeopardy, now temporarily hosted by Dr. Oz.
We should be charging vendors for traffic that is only designed to forward their business, when it happens on our own local loop.
But something went wrong at some point.
Similarly with options - options were a great way farmers could offload risk to the products buyers, who, in exchange could get better prices.
But when options started to be used for something which is not much different from lottery bets, again, something went wrong.
This is important lesson. Capitalism is a great system, but it cannot be narrowed to greed. Capitalism requires working ethics since we are not able to regulate every aspect of life - it is practically impossible.
But ethics no longer works. Big IT companies owners were fixing employees wages, who cared. Amazon owner does not care about warehouses workers miserable work conditions. Facebook owner is ok to steer people into fights, as this increases "engagement", ads clicks and revenue.
Has it happened that Jeff Bezos or Mark Zuckerberg was asked to leave his country club or whatever place they are hanging out because of the unethical behavior? Obviously not.
Whether it's still Capitalism at that point is an interesting question. Some people seem to call any market-based system capitalism. For others the lack of ethics is definitional.
My own view is that we probably ought to stop using the term "capitalism" entirely for the most part, and be more specific about what we mean.
The market is literally a big, greedy - in algorithmic sense - optimization engine. It optimizes away everything that stands in the way of making more money. It optimizes away quality, and it absolutely optimizes away any and all ethics. Marketing and advertising are a clear example of this in two separate ways: on the object level, they're already doing everything they can that isn't strictly illegal[0]; on the meta level, you see that competing on marketing and advertising has a better ROI than actually trying to make better products.
You cannot expect a sense of ethics from a system which requires ethics as a containment vessel, and continuously tries to erode it. Much like you can't expect water to just keep flowing in a straight line, in absence of a bank or pipe that would constrain its expansion.
--
[0] - And enforced. GDPR demonstrates that with insufficient enforcement, otherwise reputable companies will absolutely break the law.
I hear about new ones by word of mouth from collegues (from people who go out of their way to watch product releases on the apple away day or whatever it is). I hear about new gadgets by paying for someone to go to CES and look at what's new.
Tesla, and starlink, again is word of mouth, or deliberatly seeking out information on it (by visiting tesla.com for example).
In my dream world, all forms of push advertising would be illegal, full stop. Meanwhile, I'll put whatever effort it takes to block everything.
Contrast this with email and phone, the only successful federated systems. Spam is illegal in some countries, and there's a concerted effort by everyone to fight it.
When no one controls the whole experience end-to-end, there is simply no technical possibility to expose you to something against your will. This is why decentralized social media is the future. I just can't see how Facebook could survive much longer unless there's a drastic change of some sort.
> Perhaps I'm growing increasingly blind to any positive impacts of the discipline, but ethically I find it hard to separate the fraudsters from the engineers that enable them.
There are definitely useful spinoffs - like they always are, when you throw a lot of money and a bright mind into solving tough problems. But it's hard to argue the positives here where the primary driver is a huge net social negative, and spinoffs could've been done pursuing better goals.
--
[0] - And I mean that - http://jacek.zlydach.pl/blog/2019-07-31-ads-as-cancer.html.
This is just not ok. I understand that screening everyone in the world's personal posts/shares for 'fake news' is a daunting task, but KYC-ing advertisers and screening/auditing ads for blatant and obvious scams or false marketing is many orders of magnitude less resource intensive... and should be required by laws with teeth.
If the costs of doing so mean that it's impossible to make a profit as an ad platform, then that business model should die.
And very easy to automate. We know they can do it. But until there is sufficient threat of action if they don't then they are happy to let their users get fleeced in exchange for the advertising $.
There is an argument that "if they crack down on the easy to detect stuff, the scams will just get more clever" but that is not at all why they aren't making more effort.
One funny thing I've noticed with facebook, is that if you start responding to such adverts with comments along the lines of "this is obviously a scam because..." responses, they start showing you more and more of that sort of scam presumably because you have shown engagement with the previous ones.
Glad I deleted mine in 2010.
"Not OK" is forgetting to clean up your table after the meal. What these scammers are doing is simply fraud and Facebook should be liable as they enable it.
I remember when I set up my first FB ad (maybe 10 years ago?) it took a few days because some of them were manually reviewed. Fast forward 10 years and they accept any fraudster who pays them a few bucks. Just low scum, that's all.
It seems like you're mixing advertising = fraud, when in reality it's just people who commit fraud that use advertising because regulators can't seem to act.
Because that is small potatoes compared to the tracking, matching and profiling of individuals. I believe that is the big threat to society (and possibly democracy)
- Tune the advertising to best manipulate a particular person (think of it as a PID controller tuning inputs to a system);
- Ad attribution - i.e. determining which companies in the advertising chain should get paid when a customer buys something from someone.
Restricting the shape and form advertising is allowed to take addresses the driving factors behind privacy violations.
If I want to buy something, and if I don't know exactly what, I'd like an unbiased list of things in that category to choose from. And when I don't want to buy anything — which is, again, when I'm reading an article for example — I don't want to be reminded that economy exists AT ALL. There is no place or circumstance in my life where advertising, as it is known today, would be anything but annoyance to me. Which is why all my devices have ad blockers installed and third-party cookies disabled.
I use ad blockers.
Pay for YouTube premium. Use a sponsorship skip extension.
Only watch ad-free TV / streaming.
Pay for Spotify.
Use tweetdeck for Twitter.
Don't use Instagram or Facebook.
I certainly don't listen to the radio.
Where else can I possibly see an ad? A billboard? That's easy to ignore.
Imo it's all worth it.
That doesn't sound correct. I've always been able to access my controller without internet access as long as I was on the same LAN and was hitting the controller IP directly.
but i have to agree that it is no far fetched for ubiquiti to fuck this up.
Customer support directed me to the forums. Ubiquiti didn't respond to my post. I posted on a thread where someone else was seeing the same problem, where a number of people gaslit me and blamed me for the problem.
I still have no idea if they've fixed their band steering. Or how they decided to force 5 GHz on a 2.4 GHz chipset. Or how to get support. Or how to report a problem.
Definitely won't be buying their products in future.
https://community.ui.com/questions/Intel-Centrino-2200-N-cli...
https://community.ui.com/releases/UniFi-Network-Controller-5...
https://community.ui.com/questions/Default-AUTO-OPTIMIZE-NET...
I have an old laptop that I was trying to fire up to run some lightweight games on. I could not get it to connect to wi-fi for the life of me. Must be the band steering.
I do wonder if they ever fixed the issue. The switches lock up every couple months and need a hard reboot.
We also had all sorts of problems with their APs and switched over to Cisco Meraki. As much as I hate Cisco and their ridiculous cost at least the Meraki line works.
Obnoxious.
What's wrong with the ap's? Have you used the switches? How much of their kit can managed via a single console?
The switches have been pretty solid. Everything can be managed off the fortigate, I think they also have a cloud management platform.
I’m definitely not pushing either of them very hard or doing anything particularly fancy, but they’ve certainly been a big step up from the aging Apple routers I’d been using before, at least for me.
[1] https://external-content.duckduckgo.com/iu/?u=https%3A%2F%2F...
I've been a UI fan for many years and was also dismayed by the 6.x releases and the switch to Mediatek, but every time I went looking around, I saw the same tire fires at other vendors. At least with UBNT it's "the devil you know".
Cambium is interesting, I've been playing around with their cnPilot stuff and it's been pretty solid.
I won‘t switch to the UniFi series for now even though they would provide easier management. But I‘m a bit worried that it‘s going to be an Apple-style product line where the pros are somehow not respected as much as before.
[1] https://www.pcgamer.com/i-built-my-own-super-router-out-of-o...
OTOH not sure what i'd recommend to others - there must be 20+ unifi networks out that via recommendations i've made to friends + network
the state of our wifis, right now, is extremely poor. upmarket expensive gear becoming a hostile rent-seeking shark is a sign of how late-stage the market is right now.
i will allow: there is a lag somewhat to be expected. wave2 and wifi6 and wifi6e have been all kind of sounding this new stuff stuff out. wifi6e feels a bit like a natural culmination, for a while, but we're only just making it across a long series of changes just now. the standards have been in flux, enhancing.
still, the driver situation feels terrible. hardware is hard to get except as consumer electronics running provided shrinkware. even in terms of stations not ap's, companies like intel can both do a good job, insure reasonable support, decent enough hardware these days, but also neglect basic love & care things like support for 802.11s mesh & other sincere basics.
anyhow. wifi6e looks like the culmination of a lot. the situation might stabilize some. the future is finally here & only now do we commence more evenly distributing use/usability of it. maybe. maybe not. maybe vendorization floggings will continue. i'm hoping somehow more ap class add-on systems start becoming in any way available again. hoping systems like wifi-p2p and wifi-aware and wifi mesh (802.11s) see general support. I'd love to see some pro-active movements in wifi world, some chipmakers really doing right, but wifi feels forever like the most cagey & slowest developing linux driver area. communication, not well supported.
[1] https://news.ycombinator.com/item?id=26597238