DailyWireless - Wireless Cable Infrastructure :

Cable operators are big fat pigs that threaten our future.
Let's take them out. [from: JB Wifi]

Two articles about Broadband and Wifi around the world.
Questions and Answers for Korea Telecom's Chief
IHT: Paris, the wireless wonder?

Two-thirds of all homes in Korea now have broadband. Typical speeds now are around 10Mbps. The major supplier has their sights set on 100Mbps. They typically have fibre to within 300m of the home and and anticipate fully fibre end to end connections to the home in the near future. They aim to convert their entire telephone system to VoIP by 2005.

Meanwhile Paris is intending to WiFi enable every metro station using existing fibre in the metro system. They're then reselling the bandwidth via people like T-Mobile. The aim is to turn the city into one big WiFi hotspot. They expect the whole project to cost E3m to E10m. Comparing prices with the typical US $10 per hour, a spokesman said "It must be much cheaper... The business model of trying to sell WiFi to people at these huge costs will fail".

Reading about these sort of things we really need to raise our sights in the UK. It's not about BT enabling exchanges in remote areas so we can have 512Kb or WiFi in half a dozen Starbucks any more. It's about getting real bandwidth (>10Mbps) available no matter where you are or what you're doing. We do actually know how to do all this. With 90% of all the fibre laid unlit or under-utilized, the backhaul bandwith actually already exists. It just takes the government and private enterprise to look beyond their current debt-ridden short term-ism and take the plunge.

Something happens when you switch from dial up to always on. Something happens again when you go from 50Kbps to 500 kbps. Does anyone want to speculate about what happens when you go to 5Mbps and 50Mbps? Or when this level of bandwidth is ubiquitous instead of being tied to a few sockets in the wall? Does anyone want to bet that these changes won't result in whole new market areas and some fairly radical social effects? [from: JB Ecademy]

Excellent FAQ on Wardriving. Well worth a read whther you want to play or not.

Has anyone got a good source for the Holux USB GPS in the UK? [from: JB Wifi]

DailyWireless - Long Haul 54G has a really comprehensive look at the US regs for power output and antenna gain. The contrast with the UK and EU is stark. [from: JB Wifi]

Buffalo Tech adds 3 for 11g

A cardbus card, a PCI card and a Bridge-Access point all running 802.11g. The cardbus card and PCI card are notable because they include MMX external antenna sockets. I think this is a first for G cards. [from: JB Wifi]

BBC NEWS | England | Devon | Man dies after drilling head

Let that be a warning to you all. Self-administered trepanning with an electric drill is dangerous! Can we now expect the obligatory safety warning on all Black and Decker products? "Do not attempt trepanning with this device except by qualified personnel" [from: JB Ecademy]

Wi-Fi Networking News: Blanketing Soho with Wi-Fi This is in the main Ecademy site rather than the WiFi SIG because we've already covered it there. One very interesting quote from a Westminster Council member as justification for the move. "For example, we could expand our CCTV coverage at a fraction of the current cost without the need for traditional connections. Not only that but officers would be able have access to CCTV and other data observation and collection points."

I've long thought that the answer to CCTV is to turn every camera into a webcam and give us all access. If Westminster start deploying WiFi enabled CCTV, this comes a step closer to reality. No matter how much encryption or security they put in place there's the potential for an enterprising hacker to turn them into webcams. Given this, they should just bite the bullet and make them easily accessible. [from: JB Ecademy]

DailyWireless - ORiNOCO Tri-Mode Card Supporting a/b/g, This should have Linux drivers available. However, it doesn't have an external antenna jack. So as far as I can tell there still isn't a single G card with an antenna jack. [from: JB Wifi]

One of the best sound bite analyses of the recent O'Reilly Emergent Tech conference. WERBLOG

1. Contrary to popular belief, innovation hasn't stopped. There are exceptionally exciting technologies and companies out there. Some are solving new problems, while others have new approaches for big old problems that haven't gone away. Why did we ever think the NASDAQ index was a proxy for the health of the technology industry?

2. The exciting innovations are inter-related, in ways we don't yet have words for. Social software, Weblogs, rich Internet applications, Web services, unlicensed wireless, grid computing, digital identity, broadband media. We keep seeing more connections pop up everywhere. I believe decentralization is the most useful prism with which to understand these developments, but it's not the only one.

3. We're experiencing a generational shift. "Yesterday" now means the emergence of the Web, not the PC industry. It's time for a new crop of innovators, leaders, and conferences. Of course, some of those who grew up in the prior era will make the transition and offer their valuable experience. But the reference points have changed.

My highlight. [from: JB Ecademy]

Quoted in it's entirety. Here's some antidote to the Register and Guy Kewney's downbeat comments about Westminster Council's plans to roll out WiFi across Soho, London.

Techdirt Wireless : In the US, it's always been good news when a local municipality tries to set up a "WiFi Zone" in order to attract more people or businesses. However, in the UK, it appears that people look at it differently. They complain that the Westminster council is being too aggressive in their rollout schedule and (even worse) are going to cause overlap problems between their own WiFi access points and those of other hotspot providers, like Starbucks. Of course, some of the complaining sounds like what you might expect from someone who feels that the government backed WiFi is going to compete with their own plans.

That last comment may be uncalled for, but in principle I think we should be celebrating this, not poking holes in it. [from: JB Wifi]

Revealed: How the road to war was paved with lies. Intelligence agencies accuse Bush and Blair of distorting and fabricating evidence in rush to war

The Independent is one of the few quality papers asking the hard questions about the Iraq War. Like where exactly are the "Weapons of Mass Destruction" (tm) that were the main justification we were given for the war? [from: JB Ecademy]

A quick update on my experiments with WiFi antennas. I've been working my way through the designs here.

The first attempt was the Cantenna I've already talked about here. This worked well and was dead easy to make. Recommended.

The Double Quad was amazingly successful. It's a little fiddly bending the wire into a figure of 8 of exactly the right dimensions but it appears to have almost the same gain as the cantenna in a much more compact form. I used some sheet tin from an old oil can for the backplane and I've ended up putting the whole in a tupperware box. Recommended.

The next attempt was the Patch antenna. Again using the steel-tin sheet from an oil can. Very compact, very easy to build and with a 70 to 120 degree beamwidth. Recommended. It's so small that it can go in a laptop bag and be used for getting hotspot access when you're a little out of range.

The toothpick and 5db Omni were less successful. The omni was fiddly to get the coils right and doesn't appear to give enough gain over the built in card antenna to be really worth it. It does mean that you can have the laptop on the passenger seat of a car and still get coverage by putting the omni on the roof or dash. But the actual improvement in gain doesn't really help in getting extra distance.

The final attempt was putting a cone on the front of the cantenna. I did this with aluminum foil covered cardboard held together with duct tape. It's all huge and unwieldy but gave the most gain of all of them.

I've been doing my testing against a home AP run by a neighbour 3 or 4 streets away. I've no idea what he's running or even exactly where he is. I can't get a signal from the built in card, however using any of the antennas from our bedroom window on the 2nd floor I seem to have enough line of site to get a signal. The patch, quad, cantenna and coned cantenna can all get a useable signal. The coned cantenna is rock solid with a best of 25db SNR. While I've been doing this I've had DHCP turned off and tried to avoid associating fully. I tried it once and discovered that they're running on NTL and successfully sent and received email. There's a lesson here that I've stumbled across someone who has their AP set with all the defaults and no obvious security in place anywhere. If I could find exactly where they are I might say hello. But without walking the streets and pointing cans at houses, I can't work out exactly which house it's coming from. The NTL T&C specifically say you shouldn't share your connection outside your home and with PCs that you don't own. Purely from a wireless connection point of view this is clearly difficult as the signal leaks out to a few streets away. [from: JB Wifi]

Some more on Linux APs

I guess what I was hoping for was something with the power of Locustworld, but with the price of current APs and that was no harder to setup than current APs. The current combo boxes are so nearly there. But the manufacturers have created closed systems that cannot easily be extended or fixed by 3rd parties. Or even examined. They're also buggy and the firmware upgrade process is awkward.

I'm concerned that one of these days a major exploit is going to be found in say the dominant Linksys AP and an exploit does actually get written. If there are very large numbers of these being used as home firewalls this could be as nasty as Slammer. I've got considerably less faith in Linksys (or Dlink or) being able to cope with this compared with say, Microsoft or the Sendmail community.

That all leads me to wonder why these companies want to be in the software business selling critical systems that sit in that crucial position on the boundary between Internet and internal LANs. And to do it with closed, proprietary software where they can't get any help from the community at large.

But mostly, I want to see some stupid cheap, stupid easy APs that make it trivial to put up a safe WLAN for one's own use, while also giving guests some access that is safe for you and your ISP. I don't believe we have this yet from anyone (except of course the infamous "Default Linksys Community Network!).
[from: JB Wifi]

After the recent discussions here about legal use of antennas in the UK and EU it seems clear to me that the current controls are broken.

- The restrictions take no account of the higher directivity of high gain antennas. Using a narrower beamwidth should reduce the likelihood of interference not increase it.
- The restrictions are limiting the use of the 2.4Ghz band for last mile broadband distribution since it makes it almost impossible to get enough distance to be useful.
- The restrictions are being widely flouted by both individuals and organizations using both commercial and homebrew setups.
- They are limiting the commercial exploitation of the band both in terms of setting up deliberately limited hotspots and exploiting distance connections. It may be ok to allow a situation where a blind eye is turned to non-commercial setups, but any organization that answers to shareholders cannot take this risk.

So the big question for me is how to fix this broken situation. Does anyone have any ideas or knowledge about how the restrictions can be made more realistic and opened up? Who should we lobby and who would have the lobbying weight to have some effect? [from: JB Wifi]

Lazywebbing a challenge to port Linux to a commercial WiFi AP.

The current crop of mainstream APs aimed at the home market typically consist of WiFi (b, g or a+g), 1* ethernet in, 4*ethernet out, Router-NAT-DHCP-Simple firewall. Sometimes they include an ADSL/Cable modem. These are available from Linksys, Dlink, Buffalo, Belkin et al for ~$100

As well as the WiFi and Ethernet hardware, they've typically got quite extensive software which implies non-trivial processing power and memory. The question is how much and is it feasible to port Linux to
them? The big thing missing is something equivalent to iptables/ipchains. I don't think the built in firewall capabilities on any of these are powerful enough to properly share bandwidth with guests, let alone run a community or mesh node. But it's so close it must be worth hacking them. So the rationale is to create a programming environment for their firmware that allows us to start porting in things like Nocat, hostap and meshap.

This has been discussed a few times. But I'm not aware of anyone world wide actually doing anything apart from Sputnik.

So I think this needs a challenge like the Xbox-Linux challenge to drag in the clever hackers and make something happen.

Getting an article published on Slashdot might do the trick. So that's the first task to document the target in a form that could get on Slashdot. Then to build a wiki/weblog to track progress.

Am I blowing smoke? If not, is anyone up for this? [from: JB Wifi]

For the People, By the People... is a website devoted to mapping WiFi in Manhatten. It does go further than a simple map because the mission statement is "To PROMOTE open access to the Internet for the benefit of all people."

We really ought to do something like this for London. Anyone up for it? [from: JB Wifi]

There's some new club functions. These mostly affect Club owners rather than members.

1. Clubs can be defined as "Private" This means that the forums and membership can only be viewed by club members. The club is still listed in the All clubs list. And it's web pages are still visible to everyone.

2. Clubs can be defined as Approved Membership. This means that after you join you still can't do anything until the club owner approves you. Club owners will get an email when people attempt to join. They will also see a "Pending" menu option on the Club and "My Clubs" pages. From there they can approve or deny the applications.

3. Clubs can be defined as "Business". All this means is that the Terms and Conditions entry that says "Do not sell or advertise goods or services" is waived JUST FOR THAT CLUB and it's forum. It's a decision by the club owner that this sort of behaviour is ok in their club. It DOES NOT MEAN that this is OK anywhere else on the site and doesn't give you carte blanche to do this anywhere else or to spam the members.

These definitions can be set when creating a new club or by choosing Admin from the club home page. [from: JB Ecademy]

And this is as good a place as any to start. WEBLOGSKY - industrial-strength weblog See also Joho, O'Reilly, Kottke.

You can tell this conference is bottom up rather than top down when digerati Esther Dyson and Howard Rheingold are sitting on the floor in the packed O'Reilly presentation. Could you imagine Alan Greenspan sitting on the floor of some financial conference? [from: JB Ecademy]

The "white stripes" twins of the weblog world,Six Apart: Six Apart Ltd. Announces New Weblogging Service, Investment, Executives and Board. They've taken the best weblog software and with some added VC created a hosted weblog service to rival Pyra and blogger.com. Good luck to them. [from: JB Ecademy]

Always On interviews Google's Eric Schmidt in three parts. The first touches on Google's acquisition of Blogger.com.

I believe that this notion of self-publishing, which is what Blogger and blogging are really about, is the next big wave of human communication. The last big wave was Web activity. Before that one it was e-mail. [from: JB Ecademy]