MetroFi installed a great municipal wireless network. Seriously.

Or at least that’s how it looks to me — from my experience in the last few days.

The background: MetroFi built an extensive wireless network for the City of Portland. It didn’t work often, and when it worked, it didn’t work well. So the City opted out of the contract paying for bandwidth for core services. There’s lots of chatter in the blogs about the bad decisions made along the way, the cost to taxpayers, etc. Lots of gazing at trees.

But in the last few days, I’ve had what may be a remarkable glimpse of the forest.

I’ve connected to MetroFi nodes in numerous parts of town, and noticed two new things each time:

  • There were no advertisements, and
  • The connection WORKED.

Has anyone else had this experience? If you’re gonna check, do it quick, because the scuttlebutt says the whole thing’s getting shut down Friday. After that, who knows — sold for scrap?

But as we Portlanders contemplate what to do next, what about this possibility:

Was MetroFi’s network killed by a flawed technical approach to advertising? Did they shut it off after the contract was voided, only to reveal a perfectly functional municipal wireless network under all the junk?

And as we think that over, keep in mind that municipally-operated Internet access has been proven to work, with enormous benefits to citizens as ‘net consumers and citizens as taxpayers, alike. Since the late 1990s. In Kentucky.

20 Responses to “MetroFi installed a great municipal wireless network. Seriously.”

  1. The network was technically sound. It provided ubuauitous outdoor coverage. However, indoor coverage of 2.4GHz will always be spotty unless you have AP’s every 500′. That piece of information was never conveyed to the public and most municipal WiFi companies never wanted to really admit it to themselves. If they had used this 500′ rule in their numbers then they wouldn’t have been able to make a financial case for it with AP’s costing $3000 or more each installed.

    There were 2 ways to adjust this business model from a technical standpoint and several ways from a financial/marketing viewpoint. However, the free advertising idea was blatantly stupid from the beginning. When MetroFi set that standard, it was game over for everybody else until they inevitably failed. Now we can get back to business. That means realistic statements of work, performance expectations, and business plans that aren’t like the old concept of “we lose $1 on every book we sell but we make it up in volume”.

  2. The network did *not* provide ubiquitous outdoor coverage. We estimated, based on random sampling, that in order to have a 90% chance of a connection outdoors you needed to be within about 300 feet of a SkyPilot. Most of the time outdoors, you aren’t within 300 feet.

    The solution to free wifi is for people to give free wifi. In most of Portland, there are about 2500 *existing* wifi networks per square mile, 100 times the density MetroFi was ever going to deploy. If people decided to share even some of those, we’d have ubiquitous free wifi *today*. So, why not do *that*? It has been what Personal Telco has advocated all along. It’s not hard. Personal Telco can help. No business model needed.

  3. snap cat Says:

    This blog states:
    “it didn’t work well. So the City opted out of the contract”

    That is not true. Yes, the city did not ever become an “anchor tenant”, consuming bandwidth and paying for it.

    Maybe it does not matter in the end, but I dislike reading false items as they seem to always get repeated.

  4. Snap cat, thanks for the correction. Fixed now, I believe..let me know if the current phrasing is still inaccurate in some way?

    Rory, during the last year I frequently tried to connect, while outdoors, within sight of a connection point. The connection usually failed, or connected briefly and then dropped. Even if I was displaying a full-strength signal.

    My observation in the last week has been that connections have been solid, without advertisements. Can you explain that? Is it an anomaly?

  5. Russell, Personal Telco does great stuff, but any advice to simply bypass NAT protection is not something I can buy into. I hope Personal Telco steps up to fill this need, and I have a lot of faith in the people behind it to do so. But advising people to opt out of an important part of network security, without fully informing them of the consequences of that decision, is irresponsible, and not something I can get behind as a core strategy.

  6. threedegrees Says:

    I used Metro’s WiFi for a bout a month in the inner east side in April/May and thought it worked brilliantly. I feel it’s a huge asset for the city in the long run and would hate to see them pull the plug (or router…)

    Think of enlisting homeless people camped out in front of City Hall to come together for impromptu meetings to work on wikis/blogs to map out solutions (the laptops will be provided ;) Now if that’s too much to fathom, seriously think of what free (and fast) internet could do for equity in the age of the interactive Internet.

    Keep Portland wireless!

  7. Huh? Who said anything about NAT?

    It isn’t Personal Telco that needs to step up, they’ve been around for years, standing ready to assist people setting up community wireless networks. The people who need to step up are the people who have capacity that they aren’t sharing. Personal Telco can do *nothing* without the hospitality and support of node hosts. Personal Telco knows how to set up networks to operate safely.

  8. Just another correction, Portland did not void the contract. MetroFi has been in default of various terms for at least six months. MetroFi decided to pull the plug. The entire responsibility for this failure rests on MetroFi. The City wanted MetroFi to continue building out the network and to do what MetroFi said they would do. For various reasons, MetroFi decided it didn’t want to. At that point, there wasn’t a lot that the City could do about it.

  9. Russell, your point about the contract was pointed out above and corrected a couple days ago — is there still a problem with the way it’s phrased?

    Apart from that, I’d submit that you’re the one looking at all this through the lens of “who should we blame.” Not me. I’m interested in ensuring that Portland has ubiquitous wireless coverage going forward, not in pointing fingers. (Though I do think there are worthwhile questions about how the City allocated tax dollars, I’m not sure I’m informed enough to comment on that myself.)

    Sorry about the NAT comment though, I got our discussion crossed up with one that wasn’t on this blog. I’ll try to summarize it, because it was an interesting one: basically, the premise was that this is a very “teachable moment” for Portland, and if Personal Telco wants to raise its profile and become a highly effective advocate for the sort of personal sharing you mentioned, now is probably the best opportunity to do so.

    But there was some disagreement about how to best approach it. My position is this: PersonalTelco should come up with a set of instructions that is (1) easy for non-technical users, with a variety of equipment, to follow; and (2) documents an approach that is responsible, security-wise, and makes it clear to the non-technical end user what the security consequences are. It was point (2) where NAT came into play.

  10. Search for the phrase: “Did they shut it off after the contract was voided” in your post as of the last time I refreshed.

    As far as “teachable moment” and this being a tremendous opportunity for that, I agree.

    The thing that is standing in the way though, I don’t think is the lack of clear instructions. The primary impediment has been the meme that someone else was going to do this for them and they could relax and wait for it to happen. We are trying to replace that meme with one that says “it’s going to take *you*”, it isn’t hard, it can be done safely, and that Personal Telco is here to help.

    The easiest way to safely share is with two routers. Plug your broadband into router #1’s WAN port, set router #1 up as the public wifi, plug router #2’s WAN port into the LAN port of router #1, if desired set up an encrypted wifi on router #2. Plug in any local wired machines into the LAN ports of router #2.

    There are _better_ ways, that take less hardware, but they aren’t the _easiest_ way. Usually, they involve a particular router that can be reflashed with custom firmware, like OpenWrt. Again, Personal Telco Project is knowledgable about these things, thinks that sharing is good, and is standing by ready to help.

    Can you point me at this other discussion? It sounds like it involves some misinformation that ought to be cleared up.

  11. By the way, if anyone tries my “easiest” way, they might run into trouble with conflicting IP networks. Router #1 and router #2 need to have different LAN networks. If they are both 192.168.1.x (netmask, then you’ll find that the router #2 network won’t get to the internet. If the two networks conflict, you can change one to something that doesn’t conflict, say 192.168.0.x. Again, people at the Personal Telco Project know this stuff, are happy to teach you about it, and are willing to set it up for you as long as you are sharing.

  12. Russell, the discussion was a mix of in-person conversations and email followup, mostly with Don Park, who I think was one of the founders of PersonalTelco (though I only met him last week). I don’t know what mistakes you think were made, though.

    The technical outline you present is precisely what I think would be the best approach, I think we’re in 100% agreement on that. I think we’re in 95% agreement on how best to accomplish it — I would simply propose making a printed document, an emailable document, and a press release based on that, and “branding” it as the “Personal Telco-endorsed Wifi hot spot,” or something like that. I wouldn’t say that Personal Telco has a responsibility to do anything — it’s just a buncha volunteers — but I think an extremely simple, attractive, and useful message would be the most effective way to get widespread buy-in.

    Anyway- I’m planning to attend the next PT meeting, if you’ll be there maybe we can hash this out a bit more there?

  13. For what it’s worth, we’ve tried the printed brochure route. As far as we could tell, the people we pointed at the printed brochure never became node hosts. The people we engaged with personally were more likely to.

    There are reasons to prefer a “better” approach than the simple one I outlined above, for example to use less hardware, or to provide for a splash page that can give users the terms of service and a way to know who is sponsoring the node. That implies more complexity, some individual consultation becomes appropriate, and the simple brochure solution tends to fail.

    For reasons of simplicity, the stock firmware of every off-the-shelf wireless router I have ever seen bridges the wireless and the LAN ports. From a security point of view, that’s the wrong choice in a sharing context. You can compensate by using two routers, as I outlined in the “simple” solution. If you are using OpenWrt or some other alternative firmware, you can use a single router and split the networks so that the wireless and LAN ports are on separate networks and the public wireless cannot access anything on the LAN. We do precisely that at some of our nodes. However, installing OpenWrt might not be possible on arbitrary off-the-shelf routers. Even for those routers that can run OpenWrt, different instructions might apply for different devices.

    We don’t want to dwell on the complexity, because to us it’s not *that* complicated. We don’t think it should be an obstruction to implementation. We are here to help so that for the node host, the complexity becomes transparent.

  14. Russell, can you clarify — in your first comment on Personal Telco, you refer to “they” — but now you say “we” tried something.

  15. I volunteer some of my time with the Personal Telco Project.

    Also, for what it’s worth, Personal Telco is having it’s monthly meeting at Jax, 826 SW 2nd Ave, at 6:30pm on Wednesday June 25th.

  16. Thanks Russell — that’s pretty much my story too, though I haven’t done anything in several years. See you Wednesday night!

  17. Ah yes. You (or someone borrowing your name) are credited for helping make this happen:

    Several years indeed! ;-)

  18. Wow. The Internet really never does forget, does it?? :)

  19. I sort of fell of this but I’ll finish what I started.

    The reason you may not have been able to connect to the network even though you could see the access point is one of 2 reasons:

    1) Either your signal was too low or it wasn’t at least 10dB above the area noise level at the AP

    2) The AP may have been off line. Older firmware may have been the cause of that.

    As for Portland ever getting another shot at a truly ubiquitous system, I don’t think it’s going to happen. The revenue is too low and the vertical asset cost way to high for that level of revenue. The system we deployed in Suprise only works because we don’t have to pay for vertical assets. However, by leveraging a PTMP system off the wifi system, it should be profitable. In some cities, we may be paying $150 per year per AP which is reasonable. Portland was as high as $1200 per month for a single rooftop which was ridiculous.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: