What I'm most concerned about, as somebody switching back to a Windows Phone soon, is the apparent lack of a web interface for the OnHub. According to this page [1] the OnHub requires an Android or iOS device, with the app installed, to setup and configure the router. I don't blame Google for not supporting Windows Phone, the market share is still very low, but I'm not convinced that requiring an app instead of using a web interface like every other router is what I want from this.
Apple Airports all require apps. There's no web interface. They have Windows, OSX, and Linux apps.
I agree with you I prefer a web interface. Just pointing out that there is a pair of very popular routers that don't have one.
Fortunately they also don't need internet which it sounds like the OnHub currently does? In other words if I'm doing some kind of installation that users a router without internet I can still configure an Airport but if understand correctly I could not configure an OnHub ... yet?
That's sort of a crucial distinction. You have to run the application, but pretty much everybody has something that can run it. That's very different than requiring an application that only runs on mobile devices, and even then only some mobile devices.
I'm actually not seeing a Linux version of the Airport utility anywhere, although apparently the Windows version pretty well with WINE.
I spent an hour fighting with the Airport utility in WINE a couple months ago. It launched, but I had issues discoving the Airport. Eventually I caved and had a non-technical friend with a Macbook configure my router for me. </shame>
But even fans of them have to admit that they do attract a certain demographic. Namely people who want Apple-only products throughout their home, Macs, iPads, iPhones, Airports, Time Capsule, Apple TV, etc. I can count on my hands how many people I know who have an Airport, and they're all Apple fanatics.
Most people either seem to utilise whatever their ISP gives them (the vast majority), or buy whatever is popular on Amazon and has the features they want. Even then power users are a small subset of consumers.
Airports requiring an app isn't a good justification for why Google can do the same. I'd argue that Airports are in the wrong here, and that the market has shown that routers which provide a web interface do better.
I don't think people don't buy Airport because it doesn't have web interface. I guess the real reason is the price. If you looks at all routers at $200 price range, Airport probably has the worst looking spec.
I do agree that the specs do not look enticing on paper, compared to say the Asus ones, but in my experience they last longer.
Whenever I visit my parents I usually bring replacement hardware for whatever is broken. When I first got them their Airport Extreme, I replaced it after 5 years. I got them an Asus router with a really impressive wifi range, but it broke after just 1 year. Now they're back with the latest Airport Extreme.
It's a bit pricey for the specs but it's very reliable. I use 2 at home, that are bridged via gigabit ethernet, and they're pretty good. My only complain is that they don't have VLAN support (IPTV) and their wifi range is really bad.
But have you used one? Spec and usability are not the same thing. I have one of the high end Asus routers and several Airport Extremes. I think you know where this is going.. I'm using the AirPort Extreme right now. Dead simple to set up bridges, everything works, no need to do any overclocking to get enough power to penetrate my 17th century stone walls. If the iPhone taught us anything it's that specs matter far less than user experience. Of course there are probably still guys hacking away on Windows XP that would disagree of course.
I don't have one. But I never said Airport is bad. The point I'm trying to make here is that the "profile" of Airport is not very attractive to non apple fans and experienced user. iphone can draw attention by user experience rather than cpu numbers because people care about their phone. But the importance of router is at a position that usually ignored by average user. Take a car as example, cellphone and laptop are like engines and suspensions which draw every car buyers attention. While router is like tire, and non car guys pursuit 4WD instead of good tire when they need traction.
I've been having a similar experience. I wrote some software (http://docs.happyfuntimes.com) that connects lots of users to a local web server. I've personally used 5 different routers with it and others have used more. The AirPort Extreme is the only one that hasn't had any issues so far.
I'm hoping the OnHub will be good but I need a way to configure it without Internet
My opinion probably won’t be popular, but for me the Airport configuration app is a selling point. It’s a simple, sane, nicely designed native app that works wonderfully for configuring either a single router or whole network of routers from a central point, as opposed to having to hop around between each router’s config page. I find it much nicer than most router web interfaces, which are often badly designed and look like they’re stuck in 2005.
What’s more is that the same app works for both my Airport Extreme from 2004 as well as the one I bought two months ago. The consistency is awesome.
I agree. I have two Airports - one in my house, and one in my detached garage/office. Being able to bridge the two together into a single seamless network was easy and reliable. Doing it with my previous wireless routers was a total pain.
I don't think the lack of a web interface has anything to do with Apple's option selling less, most people buying these simply will not know the difference.
Also Apple's Market is not really anyone that would need a web interface as it would generally be marketed to existing Apple customers.
I am surprised Google don't have a web interface, If they are able to make all those other fancy features but no web interface, what other corners did they cut? To say that it is coming soon just seems to me they released this product before it was quite ready, like why is the usb port not active yet?
That's sort of a crucial distinction. You have to run the application, but pretty much everybody has something that can run it. That's very different than requiring an application that only runs on mobile devices, and even then only some mobile devices.
I'm actually not seeing a Linux version of the Airport utility anywhere, although apparently it works pretty well with WINE.
Off topic but I am quite tired of "apps" replacing websites. I have 2 friends that used wedding apps as opposed to websites and it really made it difficult for many to know what was going on.
This would also be an easy win for a company such as Google that is finally 'getting' design since every router web interface I have ever seen has been beyond appalling. It wouldn't take much to be the industry leader in this regard by a country mile.
They're functional and usually better looking than the default interface, but that's a pretty low bar. I still wouldn't classify them as looking "good".
Ugly. Not intuitive unless you are an expert. I wouldn't expect someone's grandmother to be success using it; constraining that with the airport setup that is super simple.
I forget which one, but a review I've read in the past 24 hours said that they plan to eventually have a web equivalent to the phone app for basic setup.
They also emit a constant, loud fan whine because they're meant to live in a rack. They also take multiple minutes to reboot. Automatic updates are delivered by trained humans via archaic incantations of "tftp get".
Worth noting that Ars also reviewed the OnHub [0] and weren't all that complimentary about it.
Particular negatives were the price tag, single LAN port, poor wifi performance (particularly important if you're not supposed to use wired Ethernet...) and no web interface.
Not sure I would put much weight in an article with a click-bait label like "Google’s smart home Trojan horse". I know there are many that are suspicious of this, but a site like arstechnica putting this in the title when it's, at best, speculation seems a bit sensationalist.
In context the Trojan part of the title is: Google’s smart home Trojan horse.
It's a Trojan horse because it is their way of rolling out "connected-home" devices without people having to decide to "move to a connected home". That's what all the non-enabled sensors and hardware are for.
It's a smart play on Google's part, and Ars is correct to note it.
Smart or connected home stuff has never been about moving to such a home, its always been ways to enhance an existing home, so its bizarre to describe what you put out as unusual in anyway.
And, "Trojan horse" always implies malicious deception; its an inappropriate metaphor for any other intended message.
>Smart or connected home stuff has never been about moving to such a home, its always been ways to enhance an existing home, so its bizarre to describe what you put out as unusual in anyway.
[0] would disagree:
"more elaborate systems could easily reach up to $1000 if you have a lot of hardware to install and don't shoot for the cheapest units you can get. Putting in a smart switch in three bedrooms, a living room and a kitchen can be $200-250 by itself, and that assumes a fairly spartan set up and excludes any power outlet installations. Be sure to tally up all of the parts you'll need before you start buying anything."
>"Trojan horse" always implies malicious deception; its an inappropriate metaphor for any other intended message.
It's commonly used in this way. [1][2][3] are good examples.
It also doesn't mean it's not malicious. Google doesn't have the best track record on privacy. Spend some time on their advertising platforms and you can see just how deep they go. Full text email scanning with the ability to advertise to people not just in the content of their emails, but based on the content of their recipient's emails. They've cooperated on frequent occasions with warrant-less surveillance requests. Google makes some nice products but they have long since jumped the Don't be Evil shark. They don't make all of these "free" web products for nothing. It's all around one singular purpose: to gather information that they can sell about you. No matter how people want to frame it, Google is an advertising company. Buying a router from an advertising company.. Have fun with that. What is the motivation for Google to suddenly care about routers? Is that really a pain point that's profitable to solve? Not really, unless you have a long term plan to leverage that into something that serves the core business which is advertising.
> They've cooperated on frequent occasions with warrant-less surveillance requests.
If you're not talking about NSLs, please provide a reliable citation for this claim. :)
Every company that receives an NSL that passes the laugh test is more-or-less obligated to comply with it. It's my understanding that Google fights as hard as they legally can to protect the private information of their users.
Fun fact: Did you know that from 1986 until very, very recently the law of the land in the US was that emails stored on third-party servers for longer than 180 days were considered abandoned and could be retrieved -in their entirety- by law enforcement with nothing more than what is effectively a note that declared that the information was relevant to a law enforcement investigation? It's true! [0]
"United States v. Warshak is a criminal case decided by the United States Court of Appeals for the Sixth Circuit holding that government agents violated the defendant's Fourth Amendment rights by compelling his Internet Service Provider (ISP) to turn over his emails without first obtaining a search warrant based on probable cause."
So, yeah, I fucked up. :( It's a circuit court decision, so it's not law of the land, but it is promising.
Maybe I'm being cynical, but it seems like the motive behind that wording was more likely to attract the privacy-conscious users -- as this phrase often implies -- rather than it being a hook for their other products. That's why I feel it's click-baity.
Especially considering the talk around the device collecting data when it was announced.
I think you're being overly cynical. After getting part-way through the article I understood that the "trojan horse" was inbuilt hooks for Google's (presumably) impending home automation hardware.
If you're partway through the article, you've already taken the clickbait. The point is to confuse you with possibly misleading terms, so you click to read the article and see what they really meant.
Competing routers at similar price points have more LAN ports, are faster, often include a cable modem, have a web interface and app, and usually can also be used as VoIP server & answering machine (same usability as Google Voice, but self-hosted), and often even can be used as NAS.
For this price point, the OnHub has to do something special, or it will be a loss, like the Nexus Q.
Can you point me to these competing routers you're talking about? The one Ars used in their review, the Asus RT-68U is priced[0] the same as OnHub. The main pieces I see that the Asus router has is a web interface and more LAN ports, but nothing else you list.
If you look at OnHub, it has a lot of little hardware components that aren't active yet. The light sensor and a 802.15.4 sensor.
I don't see OnHub being priced that far away from competing wireless routers (for 3x3 antenna with 802.11 a/b/g/n/ac support). My guess is the routers you're talking about don't have AC or 5GHz support, both of which seems to add a lot of cost to wireless routers.
The one I am talking about has both of that, costs 219$, has integration with a whole ecosystem of smart home devices, has -AC (1.3GHz), 5GHz support, DSL vectoring, can be used as proxy endpoint, all that nice stuff.
The Fritzbox is €219, not $219, and doesn't seem to be available in the US, so the comparison isn't really fair. At $200, there still aren't many obviously-better options, although there are plenty of good ones
Interesting, thanks for the link. I do wonder why they cost the same as other popular router companies for a lot more built in functionality. Either they are able to do a lot in software, they are using cheaper components, or they are pricing so their profits aren't as high so they can get some market traction.
The TP-LINK Archer C7 is probably the best deal on AC routers right now. It's half the price of the OnHub and includes more LAN ports, is faster (if the Ars review is to be believed), has a web interface, has NAS features, and has good custom firmware support.
It doesn't have all the fancy extras that the OnHub has, but if OnHub's wireless performance is in fact as poor as the Ars review suggests, you're probably better off with a traditional router for now and waiting a few generations for Google to work out the issues. I mean, most wireless N routers I've owned could do better than 32 Mbps...
Regarding other routers "are faster", none of the reviews I've seen have tested LAN-WAN throughput (i.e. how fast can your internet connection be before the router starts locking up mid-download because there isn't enough memory for the NAT table/stateful firewall/packet inspection).
Given Google want to quietly add features down the line (which are bound to eat up CPU time and memory), that's probably relevant.
Sadly, I agree. Over the last year or so, the quality of content produced by The Verge has noticeably decreased. I find it difficult to take their opinions of tech seriously.
I agree. They're quality has gone downhill considerably. Which site(s) do you recommend? I like sites with strong community/lots of comments. Ars is pretty good. But I'm looking for a site that's a little more pop tech than sci/tech.
While a bit of a sidetrack, that review is horrible. Aside from the unnecessary narrative, the complete hackery of their tests leaves me fondly remembering the old days when tech sites actually did regimented, repeatable, bonafide tests among multiple competitors (the Anandtech gold standard, once even shown by sites like Toms Hardware and even rag PC magazines). Now someone hangs a couple of routers off each other, does some non-descript random test (from the web?) at some cherry picked position, and calls it a day.
Maybe this is the result of the ad-supported and then ad-blocking world, but the raw, profound laziness is just too much.
You seriously think that's going to help? The entire industry is hemorrhaging like crazy already due to low add revenue. All it means is that they can't afford real reporting and the experienced reporters and editors (if they even have editors) get replaced with recent grads who are wooed in by the old glamour. Buzzfeed might survive adblocks but good luck to anyone that needs real content.
No. I don't. I wish it would, but the theory that a huge collapse in ad revenue that leads to widespread extinction among the bottom-feeders who rush to publish crappy junk -- or simply push someone else's crappy junk out wrapped in a few lines of useless "editorial" -- the theory that this will somehow magically be a return to a golden age of journalism; this theory is almost laughably naïve.
~3 weeks ago I upgraded to an industrial-quality Cisco Aironet 1252AG access point (no -ac but it does -n 300Mbps just fine), that I got off eBay with power supply for a grand total of $60.
I'm a big fan of separating the NAT/router functionality from access-point functionality, and my core "router" is a Ubiquiti EdgeRouter (X right now, used to be the Lite).
I'm also a big fan of tweakable routers/APs (I formerly wouldn't buy something unless it could run DDWrt or OpenWRT) and have a couple storage tubs full of various wireless gear, including an Asus AC68U and the Ubiquiti Unifi-LR (running OpenWRT) that was formerly my main AP.
Despite all of this, I have an OnHub on preorder that should get here tomorrow, and I'm excited. There's tons of promise for this hardware, including home automation features that haven't been implemented yet - and despite what one review said, you can use it just as an AP - that's what "bridge mode" is.
However, I'm going to run it in parallel with the Cisco AP for a while before moving all my stuff over. Always have a fallback plan...
Sorry this is a bit off topic, but - What made you move away from the Unifi LR?
I'm actually going to be upgrading my networking equipment soon and was considering doing an edge router x and a unifi LR - I've deployed a few LR's professionally and found them very stable with far reaching quality signal.
Was there an issue you ran into with the AP, or did you just want something more that it couldnt do?
I've heard that UBNT is going to release some new UniFi APs soon-ish-ly. You might want to consider waiting to see their specs and price point before performing your upgrade.
Thanks for posting this, the Aironet looks really nice! I was considering the Unifi-LR but the Aironet looks like it accomplishes the same, with the added benefit of supporting 802.11n.
Would you please post which Aironet power supply did you got? It's not clear from the manual or eBay listings which one I would need.
It is worth saying, that with those Aironet APs, if it doesn't expressly say it comes with the antennas and power supply, assume that it doesn't. Cisco actually sells them that way, and you're meant to mix and match which antenna configuration you need (but only certain ones are supported).
They're a good buy, but aren't designed for consumers at all (or to be deployed as small business/home APs).
You can get IOS 12 (autonomous) for it with only a Cisco support login (no contract), but I had to ask a friend for the IOS 15 image. Got huge performance improvements with the newer version.
You'll want the c1250-k9w7-tar.152-2.JA1.tar IOS load to run it on "autonomous" mode (not needing a central controller) and have Chrome/Firefox compatibility for the Web UI if you use it.
Vaguely off-topic: Why'd you move from the ERL to the ERX? If my ERL burst into flames today, would you suggest that I replace it with an X? Assume that I have $120 dollars to spend and I don't need (but wouldn't particularly mind) extra switch ports. Is there anything that the ERL does better than the X?
What kind of uplink/downlink bandwidth do you have/need?
As the other poster said, the ERX doesn't have hardware offloading (UBNT beta forums says they're working on it), but for my use case, there's no difference in performance between the two so the cheaper ($50 vs $99) ERX is fine.
ERL has the advantage of the onboard flash being a USB flash drive plugged internally and a dedicated serial console port; ERX is permanent onboard flash and no serial console unless you open it up and find the points on the board.
I'm going to dodge your question [0] and ask a different one that cuts to the heart of the matter [1]: "Unless I'm mis-reading the datasheet, it seems like the device has a worst-case switching speed of ~8.3 mbit/second. Have you played around with it to try to determine its performance when filtering or routing packets? (As opposed to just switching packets.)."
[0] Thanks much for the other info, though! That's useful stuff.
[1] Sorry for not asking this question in my original post. :(
It looks like X doesn't have any hardware acceleration for forwarding. With 64b packets, X can only push 130k PPS while ERL should do at least 1M PPS. This doesn't matter in a home network though.
kornholi, do you have an ER X? If you do, I appreciate your first-hand information. If you don't, I asked mrbill the question that I asked because I was looking for his first-hand experience.
For my needs, they're performance-identical (I only have 50/10 Business internet from Comcast), and well, I wanted to try out the ERX just to see how it compared. So much feature! So tiny box!
So far so good, and I have three OpenVPN tunnels nailed up 24/7 to various company locations.
This review seems based purely on one anecdote. I'm curious whether the antenna positions on his Asus were partly to blame.
The Wirecutter review [1] was a bit more negative on performance.
But, even with Wirecutter there's not much consistency in review quality with APs. I usually trust the methodology of Wirecutter but Smallnetbuilder has the most comprehensive coverage I've seen and they rate the Wirecutter top pick (the Archer C7/C8) quite middle-of-the-road for its class. I'm interested in what SNB's test results say once he tests the OnHub.
Seems pretty lacking. The best dual band routers have MU-MIMO, the ability to talk to multiple clients at once without making them each take a turn, or at least send an extra spatial channel to help reception with more limited SU-MIMO clients. This which would go nicely with the support in the Nexus 6, it's a nice advantage iPhone 6 doesn't have, but I don't see it mentioned or listed for the OnHub.
The reviewer points out the disabled USB, so no storage options, and lack of ethernet ports. For routers that don't have MU-MIMO, the best ones right now are tri-band (one 2.4GHz, one low 5GHz, and one high 5GHz channel) and the OnHub is only dual band.
There's also no mention of open source firmwares and distros like some routers make sure to mention they support ("open source ready" WRT routers by Linksys). So this product seems like a failure in the $200 router segment according to features power users care about.
I guess maybe a rich consumer willing to pay the price tag despite the lack of features might go for it, though. The claim is that the design is unique, but D-Link has had similar design for a long time. Here's a round router of theirs from 2 years ago:
http://www.pcmag.com/article2/0,2817,2422521,00.asp
Disclaimer: I work for Google, but not for OnHub and know nothing about it that hasn't been made public.
TBH, this comment reads a lot like "No wireless. Less space than a nomad. Lame." [1] I'm not sure why you would buy an OnHub if you wanted to put your own distro on it (that's clearly not the target audience), USB is only disabled for now (why have it there if there were no plans to use it?) and The Verge notes that the connectivity is very good, which is what I think the target audience very much does care about.
I think the big difference is a the the iPod offered users a superior UI, a simple syncing experience, was vastly smaller, and weighed a ton less. What does the OnHub offer?
It offers a simplified setup - something you do extremely rarely. That's a lot different from a UI that you use every single day, a size and weight that change it from something non-pocketable to something easy to pocket, or a syncing interface that you might use weekly. Google is solving something that is annoying, but done so infrequently that it might not matter.
The iPod made music easy and portable which the nomad didn't. The OnHub has a cool setup process that I'll run once every 3 years?
It does have great specs, but the question is whether they help me. Seems like they don't help speed. AC1750 routers are more than enough for my 110mbps internet and can be had for less than half the price. Why not put a microphone and chromecast in it and make it a voice capable casting device? It has a dual core 1.4GHz arm processor with 4G of storage and a gig of ram.
The thing is that it solves something that is crappy, but that you almost never have to deal with and doesn't seem to offer a lot on other fronts. As someone doing the Boston apartment shuffle today, that's a crappy experience and I've done that more often than I've set up a router over the same period.
My hope is that it might push users away from the $40 routers to $100 ones. Google could price anchor people toward more expensive routers and that's good for Google with more reliable, fast connections. Google might also add more OnHub features in the future that might make it a lot more attractive, but I just don't see it solving ongoing pain points in the way that the iPod did. It solves something as rare as moving and just like I hired people to schlep my stuff, people can buy a router and give the neighborhood techie kid $20 to get it going.
This router [1] can do everything yours can, costs the same, integrates with an existing ecosystem of smart home devices, can be used as NAT, VoIP server (with their own app), is moddable, etc.
Your connectivity is worse than competitors (especially the LAN-port is an issue), and it has less features.
If you don’t change something in either functionality or price point, the OnHub will end up like the Nexus Q – beautiful device, practically useless, economic failure.
You've written the same comment several times now, so I'd like to point out a few things:
1) The Fritz!Box seems to be much more focused on telephony, rather than home automation.
2) Many of us do not like modem functionality integrated with our router (I have a Docsis 3 cable modem).
3) Fritz!Box doesn't have a Zigbee transceiver, which seems to be the technology ideally suited to home automation(replacing de facto standards that have been around for years, i.e. zwave).
Just so we're clear, I have no interest in Onhub, and would never consider letting Google have access to my router firewall.
Yes, but now you have to see that in some countries, like Germany, the standard router you get from your ISP is the Fritz!Box 7490 – if you already have that, the OnHub is literally a waste of money, because it does everything slightly worse.
One thing the Fritz! System does to retrofit existing devices with "smart" capabilities is that it also allows you to use socket adapters [2] to old stuff, so you can automatically turn on / off devices like your sprinkler, etc.
In general, it uses DECT (or WiFi) for home automation, which also allows you to connect normal phones, babyphones, wireless cameras, etc to it.
> [I]f you already have [the Fritz!Boz 7490], the OnHub is literally a waste of money, because it does everything slightly worse.
Don't take this the wrong way, but this is pretty much equivalent to saying "If you already have a good router that you're happy with, a similarly priced router with similar features and performance will not interest you.", which is a statement that should surprise no-one. :)
Well, it wasn’t a "this device is shit", more a "well, the OnHub isn’t really filling any niche, nor disrupting the market – it’s a mediocre product in an oversaturated market".
The situation in Germany appears to be very different than the situation in the US. If folks get a router and/or AP from their ISP it's almost always a flaky pile of shit.
If this doesn't become yet another abandoned project with in a year (or vaguely-rubber-stamp compliance effort inside of two), the Google folks will likely put a sizeable chunk of R&D effort into ensuring that they push the state of residential wireless APs forward.
As it stands now, there are ALLSORTS of really cool things that WiFi APs should do, but just fucking don't for reasons that are beyond my comprehension.
Well, except for the Fritz!Box, all other routers we get here are also shit. But the Fritz!Box is something special xD
AVM, originally a huge advertising company on German BTX (a kinda internet before the internet), became large with modems and home phones, so their routers are an extension of that.
And, well, as all devices are assembled in Germany, the price tag is usually accepted, as you get the best of the best out of it – the routers are prolly gonna last a lifetime.
But yeah, Google OnHub might have some potential, but if the ISPs in the US didn’t give their users a Fritz!Box in the past decade, they won’t give them an OnHub either.
And only for Google Fiber or self-bought seems pretty... wasteful.
> And only for Google Fiber or self-bought seems pretty... wasteful.
Eh? If your current wireless router either sucks or dies, then you're in the market for a new one. When it comes to networking gear, it pays to pay for quality. Not everyone knows that, but many people do.
> ...if the ISPs in the US didn’t give their users a Fritz!Box...
It seems that the only people I hear talking about Fritz!Boxes are Germans. I've known many techies with many strange devices, but I've never seen a Fritz!Box in person. Their fame might be restricted to Germany and maybe neighboring countries. :)
There are many routers out there that fit the bill of what you're looking for. I think Google's aim here is for the generic consumer that has wifi signal problems but doesn't want to fiddle with it. This seems more a competitor to Apple products than tri-band, WRT-ready routers. You're right though, it doesn't (can't?) compete in that space.
> But I’ve had zero issues with the OnHub: coverage has been shockingly good, I haven’t needed to use the wireless repeater at all, and the interference I saw with the baby monitor has disappeared. In areas that were formerly dead zones, the OnHub has given me enough bandwidth and throughput to stream 4K video, even in my basement, a full two stories below my office. This is despite putting the router in a less-than-ideal location: upstairs, in my office, far from our general living areas.
This is what I was hoping for. It seems every router I've tried -- and I even have the ASUS RT-N66U, which is supposed to have great coverage -- never fails to have multiple bad spots in our home (and it's not that large).
If OnHub has the advertised coverage, it may set a bar that hopefully other manufactures will be forced to follow.
Have you considered either doing actual ethernet or failing that powerline ethernet?
A lot of these dead spots are caused by the physical materials or distance of the property, which often cannot be overcome WELL unless you used directional antennas, or exceeded the legal power of your WiFi equipment (which can sometimes cause overheating or signal issues).
Instead of trying to break through the dead spots with better and better WiFi, maybe you should just cut your losses, get wired internet onto the other side, and then set up an inexpensive AP there to complete the coverage picture. If your budget is $200, then you could easily do powerline ethernet and a cheap AP.
I've tried multiple solutions, actually. Re: ethernet, everything that can be ethernet in our home, is. All of our consoles & desktops are on ethernet, as is our printer. Fortunately our home was built with ethernet installed.
The problem is my laptop and handheld devices (which don't support ethernet obviously). For these, we've tried various routers, multiple locations for the router (though this is somewhat limited), and even setting up multiple access points.
The problem with access points, is that WiFi devices are notorious for hanging on too aggressively to the wrong (weaker) access point. So we found we were frequently having to manually switch between the access points. So I gave up with that, and decided to go back to a single router.
Which is why, as you say, I keep shopping for better and better WiFi.
This feature may be present in other manufacturer's devices, but I know that Ubiquiti's UniFi APs let you set a minimum client signal strength [0], below which the AP will disassociate the client. [1] Folks in the support forums report that this works really well, but that you do have to play around to find the right setting.
The three downsides to UniFi APs are:
* There is controller software which is required to configure (but not run) the APs.
* The price of the UAP-AC v1 and v2 models.
* Every AP setting change requires that the controller reboot the affected AP. The APs take ~60 seconds to reboot.
The controller software is reasonably well done [2], easy enough to use, and makes managing a swarm of APs pretty easy. I fire it up when I need to change channel settings or add an SSID or whatever and shut it down afterwards.
I hear that Ubiquiti is going to be releasing some new, lower-priced 802.11ac units soon-ish-ly that aren't much slower than (and are much less power hungry than) their existing 802.11ac units. So, if you want to give their hardware a try, you might want to wait until the new units are released. (Edit: Unless -of course- you don't care about 802.11ac, in which case I hear good things about the UAP-Pro, which does 2.4 and 5Ghz 802.11n.)
If you have questions about the APs or controller software, feel free to ask.
[0] AKA RSSI.
[1] The idea is that the newly-disassociated client will search for a new AP to attach to, find that the one it was ignoring has better signal strength after all, and attach to it, rather than the previous one it was attached to.
[2] It's some Java software that presents a web page on localhost that you log in to to manage your existing APs and adopt new ones that you've attached to your network.
Since I bought a Ubiquiti Unifi AP, I have no bad spot at all.
I actually have my ISP wifi downstairs, I put the Unifi upstairs thinking I would switch from one to the other, but even closer to my ISP router I still get a better signal with my Unifi.
I've had a similar experience with my Unifi. I had planned to ceiling mount it upstairs in the middle of the house. But coverage is so good throughout the entire house with it sitting on a bookshelf in my office in the corner of the house, I haven't gotten around to moving it.
Several people are complaining about the lack of ethernet ports.
Back in 2012(?), Apple came out with the MacBook Air, which did not have an optical drive. Everyone was up in arms: NO OPTICAL DRIVE?!? Has Jobs lost his mind??!? I was the same: what the heck? How will I get along without an optical drive?!?
Then I thought about the last time I used the optical drive on my MBP. After a bit of thinking, I realized that I had basically never used it in the ~4 years I had that MBP! I had been lugging around an optical drive, which I had used zero times! Then it dawned on me: I've been getting along fine without an optical drive.
Something tells me that the story will be similar with ethernet ports. Most of our devices today (laptops, tablets, phones) don't use wired ethernet. Ye cranky old desktop does, but..... what if you could put a $50 ac card in it, and presto, no wires!
Yup. Another feature competitors have, that we won’t see here, is integration with VoIP and phone apps – if someone calls your home number, you can respond on your mobile. You have an app where you can see who called when – and listen to the recordings of the answering machine. Essentially competitors have self-hosted Google Voice, Google has... well, US-only Google Voice.
Since you're quite obviously an AVM fan (given that you promote their product all over the thread) and it seems you're referring to them here as well (the "phone app" integration is a historical emphasis/accident of that company, given that they got into the DSL-Modem business by expanding from their telephone systems and ISDN cards, and much more of an issue for them than for any other vendor of dsl/cable/wifi modem/router thingies):
Outside the German (speaking) market AVM doesn't matter much. And looking at the tiny selection on the German version of http://store.google.com, the German market doesn't matter much to Google.
That's market segmentation at work, and OnHub likely won't be a threat to your favorite DSL modem brand (and it isn't a DSL modem, btw)
Well, it’s not my favourite brand or anything, the issue I’m just seeing is that I’d like to see a good deal from Google.
And what we’re seeing here is that a lot of people say "this router is awesome", "it does so much and is so cheap", "there’s no other router with that performance", and I’m just here, facepalming.
It’s an upper-mid-range device in an oversaturated market, not something worth hundreds of reviews and articles.
Tbh, I don’t care about AVM – the Fritz!Box 7490 was just the first router I could find with a similar featureset and similar price.
Also, the Google Store is even more empty in many other countries – US, Canada, sometimes UK, NZ and AUS get the nice products. Some european countries still have no Google Books, Google Movies, Google Music, or even the ability to buy Nexus devices.
Google voice is amazing! I travel all over the world, in fact, haven't been to the US for two years, and I haven't found anything as good and cheap (free).
Well, it’s not available for people outside of the US, so it’s natural that competitors spring up. And, usually not having the same server infrastructure Google has, it usually ends up as a self-hosted solution.
Agreed. However, Apple has a nasty habit of disabling features (such as power settings) and making the control software less and less friendly. I love my Airport Extreme it's software doesn't do everything that the OnHub software does. Some things I would really like is the ability to quickly see who/what is connected to my network, some history of connections, and even a way to tune QOS for specific devices (i.e. my Roku boxes should have more bandwidth than my phone).
Sorry, but this device is idiotic. It gives Google the ability to entirely remotely control your network from outside, is entirely designed to facilitate their own services, and will become a privacy nightmare ... because if they can access it, someone else can, and law enforcement will be able to go to them and say "OK, we need access to that network, you have to give it to us".
This is the "bend over and take it" device which puts control of your home network in the hands of Google .. primarily to benefit Google.
This is a terrible idea, and it's not something I'd trust even a little. This is all about locking you into Google, and making it easy for them to manage your home remotely.
I would be reluctant to ever treat my home network as secure. I just don't have time or expertise to validate that. Instead I want my router to be maintained by someone responsible to try and mitigate threats. Google may not be perfect, but they are probably more competent than an ISP or companies like Linksys or Belkin. For Google to be able to update the firmware they need complete access to the system and they could certainly use that to invade my privacy. And the more you trust your home network the more that is a risk to you.
"Importantly, the Google On app and your OnHub do not track the websites you visit or collect the content of any traffic on your network. However, OnHub does collect data such as Wi-Fi channel, signal strength, and device types that are relevant to optimize your Wi-Fi performance."
DNS Settings
While OnHub doesn’t track the websites you visit, your DNS provider can associate your web traffic with your public IP address. OnHub sets your default DNS provider to Google Public DNS. (This can be changed in the Advanced Networking settings of the Google On app.) Google does not associate Google Public DNS information with your Google Account. Learn more about Google Public DNS.
We are constantly working on new features and controls that help you get more out of your OnHub. It’s possible that, in order to implement these new features, we may need to change the way it collects, stores, and uses data.
For what they keep permanently. (Network AS and requested DNS Site, with my home business account, that is enough to uniquely identify my houses traffic)
(Tedious disclaimer: my opinion, not my employer's. Not representing anybody else. I'm an SRE at Google, and my team is oncall for public DNS.)
I believe you have misunderstood what a network AS number is. This identifies your ISP, not your account. So the permanent logs would say something approximating "a level3 user queried news.ycombinator.com".
Unfortunately this was the first thought that came to me as well.
It looks like a beautifully designed and efficient product. But given Google's pedigree in privacy invasion, any product of theirs comes loaded with doubts.
Given the pricing, it feels like Google could generate
sufficient profits without having use personal data to augment bottom lines.
I wonder if they are bothered to address such fears of a few of their potential consumers or is there a sufficient majority who are willing to ignore the privacy invasion and enjoy the product for it is.
> ...given Google's pedigree in privacy invasion...
Do... you have a citation for this that isn't 75% conjecture? Before answering, remember that every web site you use has to collect (but not necessarily store) some of what is potentially PII.
I see. Makes sense to me. I'm glad there are people who love to tinker, test, and modify. FWIW, no steps were taken to prevent installation of DD-WRT or any flavor of Linux. There are no carriers to de-certify the device, just the FCC. There is a developer mode. More details to come. You can install whatever you want, just like with an Android or CrOS device.
I have an Apple Extreme which is good, but getting a bit up there in age. Might hand it off to my parents and upgrade to this. My experience has been that investing in good routing hardware makes my life much more pleasant. Internet when I need it, and none of the nonsense. When I've had cheap hardware in the past (growing up, in college), I feel like I was constantly battling it to get it working -- ddwrt or not. I actually inherited a nice Apple Extreme when my startup graduated to pro-grade networking hardware, and I've loved using it.
I am not worried about the lack of extra ethernet ports, or internal storage as most content is moving to the cloud and off-site backup and video security is a best practice.
What do I find to be missing is a backup WAN, ideally 3G/LTE cellular modem via USB. This is important if you are setting, say, Dropcams for security while away from home -- in addition of backup power.
That's one thing I do like about Peplink/Pepwave routers, such as the Balance One or Pepwave SoHo -- I am curious to hear anyone's opinion on these devices. Granted, Peplink goes a level further with extra WANs to increase bandwidth and allow business continuity.
I assume Google may actually add this functionality on the USB port, and I think this would be more important than storage.
My take has been that this is basically a combinatition of a smart hub (like wink and smartthings) and router/access point. The real question I have is whether radio level integrations for IoT really makes sense. The good news I suppose is that a lot of heavy hitters are following google. The bad news is that a lot aren't. As a consumer, there aren't that many use cases where I really need a radio level integration as opposed to something over the web (or even if it stays in my local network) so I would argue this is going to be a bit of a mess. With that said, if I were Wink I would be more afraid of this than if I were Linksys...
What's the difference between routing a coaxial cable to the middle of a room versus an ethernet cable? At least you can get long flat ethernet cables.
I think there's insufficient difference, and that's the original poster's point.
The fact that the device is aesthetically pleasing isn't a misfeature (other than perhaps bumping it in the cost bracket, if it even did so). But as long as it's connected to a bulky, obnoxious Surfboard by a physical cable anyway, it's not something you'll be using as a room centerpiece unless you intend to put an unusual amount of work into cable-routing.
Even if it wasn't connected to a surfboard (assuming the OnHub was also capable of being a modem), you would then need to route a coax cable to the middle of the room. There's 0 reason to keep the surfboard physically close to the AP, and IMO, it's preferable to not keep it close (because that way I can keep my network switch near the modem instead hidden behind all my electronics never to be seen). Either way, you will be routing a cable to the center of the room, you're just deciding whether you're routing a coax cable or an ethernet cable.
I think the main reason I probably won't be getting an OnHub (for myself at least -- it might be perfect for my parents) is my love of tinkering and learning networking. I like that firmware like DD-WRT lets me poke around at the internals, and actually calls features by their standardized names instead of some made-up copyrighted marketing name.
The interface is definitely not going to win any Awwwards. The feature set, though, is one you'd usually have to pay a considerable amount more to get directly from a router manufacturer.
I'm not talking about Google, just different manufacturers in general. I think the specific router I had helped configure before was a TP-LINK small business router, which may have contributed to the feature naming. I wanted to set up a server in a DMZ, but that feature was called something unexpected like "Virtual Server."
If it works as well as advertised, with as little fuss - this is absolutely the go-to Christmas gift for techies to get their parents/grandparents/uncles/etc. It's worth the price tag for me to get my family on a device that I don't have to fiddle with every time I visit.
hmmm, just bought the Nighthawk X6 AC3200 and it is still unopened. I think I will keep it and wait for another rev on this.
Anybody has any opinions on the X6 AC3200? Bought it for $399 based on "I don't want to have Wi-Fi problems at home" (tons of devices, works from home, etc. etc.)
There are 237 critical reviews of the Nighthawk on Amazon, with problems like needing daily reboots and the 2.4GHz network maxing out at under 15mbps repeated across reviews, where the cheaper router it's replacing had no such problems. All of the consumer AC3200 products have similar high percentages of negative reviews, probably because they largely share the same chipset. I wouldn't spend $300 on something that'll create more problems than my current $20 N router, and it appears there's a high probability of that being the case with that product.
No specific knowledge of the AC3200, but I bought an AirPort Extreme as a fairly pricey "I'm sick of having network problems" fix. Couldn't tell you when the last time I had to reboot it was, and I consider it money well spent.
I'm glad DD-WRT/Tomato existed for my string of Linksys/Asus/Belkin type consumer routers in college, and I'm even more glad to not be tinkering with it anymore.
You might read through the AC3200 reviews on Newegg so you at least know what you're getting in to. Reception is mixed (28% 1-star reviews), with complaints mainly about dropped connections.
It's about being realistic, you are the product for most of Google's services. This will no doubt be any different.
If you see Google as your "friend", then your minimal and non-contributing comment could be considered of value. I see that you just made this HN account, and appear to be astroturfing for this new device. Mind explaining to HN why that's the case?
1. In most cases, Google services are free to the user and leverage ads to cover the R&D and operational expenses of the service, and of course for profit. The difference between that model and a device you purchase for what certainly appears to be a non-subsidised price seems pretty stark to me. As a result, I balk at your blanket, unsupported statement such as "This will no doubt be any different". Are documents such as https://support.google.com/onhub/answer/6246642?hl=en outright lies? If so, please provide some citation or reason.
2. HN is not an entity to whom I need to explain. The accounts are throw-away by design. It exists because HN requires I have an identity in order to operate in its context. Requirement set. Requirement met. Here we are. Freedom.
3. I am sad to hear you found my comment "minimal and non-contributing". This concept has resurfaced in many conversations I've had in regard to this and similar topics. Allow me to expand and hopefully add value.
I think of opposition to Google (or other corporate entities) as having two distinct brands: people who consider them as adversaries and people who disagree with or dislike their business models. The difference in my mind is that an adversary is an entity that is expected to be malicious and should only be trusted if bound by strong technical controls. If you believe that Google is an adversary in this sense, then your position makes sense to me. Deliberately avoiding Google's products/services/devices is one of the very few technical controls available to you as a private citizen. I disagree with your assessment, but the course of action you're advocating is consistent with your world view. You can can hardly be blamed for discarding the policy I referred above as patently false. That's the long winded version of what I said. Fair?
There is an alternative: you disagree with Google's business model, and don't trust Google to act with your best intentions in mind. It is not evil, but its actions may have effects you do not condone or expect. Yet, from time to time it releases products/services/devices that solve some problem you have. In that case, I'd argue that you can afford to operate in "mixed mode", where you make decisions product by product based on some analysis. This analysis could be something like auditing Google's products where feasible (e.g. AOSP, CrOS, Chromium, etc.). Hurrah for open source. The demands this option places on you are quite high. The return is comparatively low. I claim the return is low because meaningful analysis requires a lot of time and less than common skills, and in return you may get a tool to solve some primitive problem (e.g. browse the web, make a call, send an email, etc). I expect that the the vast majority of people will not go down this path. Certainly nearly no one will review all post-facto software updates for adherence to initial audit criteria/standards. Another flavor of "analysis" is to delegate the decision to trusted parties. For example follow EFF's recommendations to minimize risk. (e.g. HTTPS everywhere, Privacy Badger, and their "Who Has Your Back" reports, etc). This approach scales quite well. It allows the user to do some basic inspection (e.g. reading reviews, articles, maybe skim the source, etc) and make up their mind. They feel informed, but are in fact removed from underlying details and reasoning.
The questions I have for you are as follows. Which camp do you belong to? If it's the first, I'd be curious to understand why. Is Google your sole adversary or is Cisco/Juniper, Asus/Netgear, etc in the same boat? They all play a role in mangling, mirroring, and forwarding your packets. The vast majority of network devices are upgraded, patched, reconfigured without your input or knowledge. Is that not a concern? If not, why not?
If you're a mixed mode sort, then why advocate a response tied to the more extreme position? Also, unless you've done first hand analysis or can cite some, I don't grasp how you could have arrived at your conclusions already. OnHub code will be open sourced at some stage. It can be audited. You can even wait for someone to audit it and defer to them. Why jump the gun and call to abandon ship?
What does it mean for your LAN to stay up if your devices connected wirelessly lose connectivity and the other devices on your LAN lose Internet access when your router goes down?
The majority of the devices on my LAN are wired and -even though this might not apply to OnHub users[0]-, my wireless APs are also attached to a switch. So, what it means for my LAN when my router reboots is that the LAN loses Internet access [1] for about a minute.
[0] I guess we don't know how sophisticated the link handling is in the OnHub. It might be able to keep the AP active and pass traffic through it (and its LAN port) while restarting the OS and do similar things while restarting the AP. The marketing copy does make it sound like you'll get near-constant connectivity, no matter what the software on the thing is doing, and I have seen (ages ago) NICs that would do basic packet handling even if the machine they were attached to was powered down.
[1] And -I think- all uPnP-initiated port forwards.
When I saw this, the first thing that came to my mind was Amazon Echo and whether Google had come up with a smart router that embedded Google Now features.
All these comments, while interesting, miss the point: how on earth does one trust Google with their router? Google makes some great products but their entire business model requires them to hoover up more & more data from customers.
That's not a company whose products I want to use anymore. Not for email. Not for search. Not for a thermostat. Not for a realtime JSON database. And not for listening in to all the data in my house.
> That's not a company whose products I want to use anymore. Not for email. Not for search. Not for a thermostat. Not for a realtime JSON database. And not for listening in to all the data in my house.
While I can't say I disagree with your point on Google's incentives being exactly the opposite of absolute privacy, you're clearly in a tiny, tiny minority. Even assuming that the majority of users simply aren't aware of this (which is a faulty assumption in my opinion), there are many (myself included) who have simply accepted this and go about our lives, preferring to have a simpler, easier to use life than obsess over every single privacy concern.
I'm guessing you were born after the Berlin Wall came down, and missed all of the palpable Western demonstrations of just how fragile freedom is, and just how dangerous it would be to build a turn-key surveillance state.
In my experience, the people who "don't care" (as opposed to simply not understanding the insanely complex topic) are the people who:
1) benefit from it, e.g., SaaS/startups heavy on analytics, and
2) lack the context to understand that the current state of Western civilization is not in any way immune to regression, and
3) don't understand how their own actions can in fact induce that regression by creating a high-value turn-key totalitarian system ready for the taking, and
4) don't understand that their actions are already serving this purpose in more totalitarian states outside of the USA.
That's even ignoring the implications of __massive__ information asymmetry when it comes to negotiation between corporations and individuals.
You think you care, but you have no idea who gives you privacy. Maybe you trust Apple, but have you seen their code? What if snarfing your data leads to better data for siri? Then let's not talk about the millions of small sites snarfing your data as you browse through the web, none of them beholden to sarbanes-oxley, your cell phone conversations which are snarfed by your carrier and aggregated with geo data and sold to the highest bidder, providing billboard locations that maximize demographics. You say you care about privacy, but you have no idea that you have no privacy. You may hate google / think you're maintaining your privacy by warning people about google, but it's just the beaten horse that you beat to make you feel better. It's the bright colorful logo in your face that you see, but that's the tip of the ice berg. In short you are ignorant and lashing out and making no difference whatsoever. Google's just easy to attack because it's more up front regarding how it uses your data, you can login to your dashboard and see everything. Your cell phone carrier and friends, Apple, not so much. Don't remember seeing their dashboards with the data they collect on you. So, continue on pretending like you are making a difference.
This is the best comment on this topic I've read to date.
I agree with everything you say and I still do business with Google. Let me go through your list.
1. No. Not me.
2. I hope I'm not that naive. I've made similar arguments to other people in related discussions.
3. Yep. I understand this.
4. Also true.
So - I understand those things and agree with you and I still do business with Google et al.
The thing is I also drink, fail to exercise, drive a petroleum-consuming vehicle and recycle poorly.
I do many other things that harm myself directly or perpetuate or accelerate trends in society I perceive as negative.
You might consider this apathy and that is probably true to some degree.
But I am also running a continual cost/benefit calculation on every action I perform.
Some of my conclusions are out of whack and some are frankly irrational but the list of things I consider every moment of every day is extremely long and I've decided that the cost vs benefit of doing business with Google is in my favour. I've evaluated the costs you set out above and decided they fall far enough down my list that I'll tackle them at some point after I've started jogging and voting, stopped drinking scotch and begun regularly volunteering at my local worthy cause.
soooo, no words spent on privacy? this is akin to paying someone to build a fence and gate for you, and then making sure they have a copy of the key at all times, and never asking them what they do with it... rhetorical question here, is it common practice prefer Internet servers which look cool and cost a ton and are managed by Google, and not discussing privacy issues? the consumer is being transformed into a marionette, and the issue isn't even mentioned in this article. beyond frustrating that it has come to this.
"You agree that Google may collect and use technical and related information, including but not limited to information about your computer and/or mobile device, operating system, peripherals, applications, connected devices, network traffic, and data use to facilitate the provision of the Software and Services, including support and other related services. The OnHub Privacy FAQ describes the categories of data collected and how you can use privacy settings to change what data is collected by the Services. "
I'm no lawyer, but from my semi educated reading, that tells me Google reserves the right (which to me means they absolutely will) to collect your network traffic and actual data, send it to the Google databases and use it for whatever purpose they so choose. Does anyone read it in any way more legally limiting than that?
I think it fair to hold "journalists" to a standard which does not count verbatim repetition of their subjects words without any inspection as "ink spent".
I've seen a lot of complaints about this router only having one ethernet LAN port. This really could just be an issue of different use cases, but are there a large amount of modems out there that don't have plenty of ethernet ports on them? I have never needed to plug in anything to my router that I can't just plug into my modem, but I'll be the first to admit that I don't connect too much to either to begin with.
If your modem (which isn't really a modem if it has >1 LAN port) is plugged into the WAN port on the OnHub (or any other router), anything plugged into the ports on the modem will have trouble seeing things on the LAN because of NAT and firewall rules.
Has nothing at all to do with either NAT or the Firewall rules; it has to do with routing. Assuming your modem/router has this capability, all you should need to do is set up a static route pointing to the network/subnet that the devices behind the other router. You should not need a corresponding route on the other router because since it's on both networks, it already has a route to the other network. Even if it didn't, the default gateway on the router is the modem which can take care of routing the packets correctly. This is all assuming that the modem/router combo is assigning an IP on a private network/subnet to the additional router.
> I don't connect too much to either to begin with.
Inclines me to believe that you haven't been burnt, yet. If your WiFi setup works stick with it until the day you die, possibly pass it on to your descendants - they'll appreciate it.
Exaggeration aside: WiFi has to be the most horrifically inconsistent technology that we have. A friend's setup will spontaneously hit 200ms latency, which can only be repaired with a restart of the device. Other friends have lesser horror stories to tell, but most people I know do have them.
I only use WiFi when I don't care about the quality of the internet connection. In truth: my setup is probably excellent but I simply no longer trust the technology at all and I suspect that many of the people complaining about the ethernet port are in the same boat.
It is a spy device to peek inside your home and your internet usage to 'improve' the quality of ads that you see.
The fact that it has sensors scares me.
Sure they can adjust the brightness of that ring with the ambient sensor.
What I would do if I ran an ad company is use that sensor to find out what time you turn off the lights and go to bed. And what time you wake up. And then send you more relevant ads.
Or use the bluetooth in it to find out what devices you have at home, or what car you drive. Or what IoT devices you are using.
And I would use the microphone to find out how many people are in your home. Or if you have kids, and how many.
If I understand correctly, aggregating and utilizing as much data as possible is the essence of building better products via a Big Data approach.
This isn't inherently a bad thing. It turns out a lot of assumptions are demonstrably false if you have enough data to start making statements about the population instead of non-representative statistical samples.
Someone should break this sucker open and find out what else is in there. Microphone was a guess actually. But I would place a bet on it that it is there.
[1]: https://on.google.com/hub/support/#get-started