I don’t know about you but I’m still waiting for fiber optic connectivity to my house. If you’re one to dwell in the past you could check out “The $200 Billion Broadband Scandal” aparently written by someone with a strong dislike for the Regional Bell Operating Companies (RBOCs) but that’s unlikely to change where we are today.

Perhaps it’s time for a paradigm shift in communications facilities. Telephone companies are whining that it’s simply too expensive to install fiber connectivity to the premises and the payback is too slow. Of course these are the same companies trying to extort money from web service providers such as Google and Yahoo claiming they aren’t paying their fair share and the same companies which have scammed us in the past (see above). But what if the telephone companies didn’t own the fiber network? What if your local municipality installed the fiber from your house back to a central point, leased it to you and allowed you to make a service agreement with any ISP who would be responsible for “lighting up your fiber”, maintaining the circuit and providing access?

This is just the proposal made by Bob Cringely in a recent column. It appears that someone may have been listening too as the Register (and later Cringely) pointed out the Berkley, California city council tabled a proposal for citywide WiFi to expore a fiber optic proposal instead. If the council moves ahead with the fiber plan and it is sucessful this could result in a string of such projects nationwide, speeding fiber deployment from coast to coast.

It’s worthwhile to look at where this idea originated and some of the details. It seems Bob got the idea (via Bob Frankston one of the VisiCalc inventors) from Bill St. Arnaud, a Canadian researcher. In this presentation St. Arnaud outlines a proposal for muncipal fiber leased to customers. SUch proposals have existed in the past but are often fiercly fought by incumbant carriers worried about loosing marketshare. What makes this proposal different is that the city only installs the fiber loops from a common access point to individual subscibers, they do not provide any services. Customers are responsible for contracting with a seperate service provider to receive voice, data or TV service over the fiber. The idea is that incumbant companies, along with startups, would provide these services and have little to loose as they have not made the large invenstment in running cable to individual subscribers. Another benefit is that the local loop would not be owned by a company removing the need for an incumbant to lease local lines to a competitor as is now common practice in the telephony industry and increasing competitve choice by reducing the barrier to entry.

St. Arnaud further suggests that the customer’s contracted service provider would be responsible for maintaining the fiber (either itself or, more likely, subcontracted out) removing ongoing maintenance expense from the municipality. In addition, the municipality would be able to use fiber connectivity itself to connect various city, county and school buildings to a high speed data network. The presentation suggests that a fiber system such as this using either optical ribbon fiber or micro-conduits would cost only about $1000-1500 per subscriber at 25-40% take up. St. Arnaud also points to a similar installation in Sweeden which he claims has been extremely successful.

I would love to see how this idea would play out in the real world, and would love for my community to be the one to try it yet that seems unlikely. In the meantime I’ll monitor resources such as the Fiber to the Home Council and FiberFirst Minnesota for details about fiber connectivity proposals in my neck of the woods.

One of the best ways to distribute phone lines in residential and small business applications is to use some kind of punchdown block in combination with a star (also called home run) topology. While the site is not completely devoid of commercial aims (they do sell equipment) the Phone Man site has great explantions of how to wire 66 and 110 blocks for phone distribution (when you don’t have a PBX type system and simply wish for all phone lines to go to all jacks).

For what it’s worth I recently bought a DSL splitter from the site for a recent project and highly recommend installing a centralized DSL splitter such as the Wilcom PS15-I3S and doing away with all those individual DSL filters around your house.

Last year I made a post about some of the options for inexpensive DMX control. For those not familiar, DMX is a serial control protocol used in most modern professional lighting systems for theaters, concert halls, rock tours, etc. I continue to receive feedback on my original posting, some of which I still plan to respond to. In the meantime I would like to pose a few more ideas for you to think about and resources to investigate.

One of the interesting aspects of professional sound and lighting work is that, over time, it has become more and more integrated with computers and more recently, computer networks. Modern sound consoles can use “digital snake” technology and digital control technology which allows for much more flexible installations. In the lighting arena it is incresingly popular to run DMX over (relatively inexpensive) ethernet networks instead of dedicated serial cabling.

As with much of the technology found in production environments DMX over ethernet has been way overpriced. Be it the small demand for the technology compared with “consumer” technology or just the long track record of paying a premium for production technology DMX over ethernet has been out of a reasonable price range for most schools and non-profit theaters.

While poking around on the internet one day updating myself on current production technology I stumbled across the work of Simon Newton. Simon has created an open source driver for ArtNet (one of the popular and standardized protocols for DMX over ethernet). His project, called libartnet is specific to POSIX based systems, including both Linux and Mac OS X.

But what can you do with it, you might ask. Right now if you wanted to control some lights with a PC (using Martin’s LightJockey for instance) you would run a program which would output DMX on some kind of hardware interface (either proprietary or open source) which you would run into an ArtNet converter box and convert to ethernet. That seems kind of silly to me. After all, most PCs already have an ethernet port so why buy two more pieces of hardware? (Remember what I said about production technology being too expensive…) Anyway, with libartnet (assuming your control software recognized it) you could output DMX straight from your ethernet interface. But what software has libartnet support? Well, DMX4Linux does for sure and that means almost any lighting control software that runs on Linux will. Specifically, the GUI based Q Light Controller will. See how useful having a driver abstraction layer such as DMX4Linux is, but I digress. Another benefit of coming directly out of an ethernet port is the support for wireless.

Wireless ethernet is increasingly prevalent and compared with exisiting wireless DMX products it is downright cheap. Now we have solved the problem of wanting wireless control, after all with technology like this you could focus lights using your PDA with it’s built in 802.11x capability. But what about the other applications for wireless DMX? Simply routing the DMX signal between trusses or pipes in a large theater can present a problem. As I mentioned traditional wireless DMX is incredibly expensive and not all that great a technology. It turns out I have a solution for this as well. A few weeks ago I suggested to a hardware engineer I work with that he look into the possibility of using a pre-certified 802.11x transmitter (such as those from Digi International) to create an inexpensive wireless DMX product. After looking around on the internet a couple of weeks later I discovered that wireless ArtNet tranceivers already exist, but as can be expected with production technology, are wildly overpriced.

One of the most interesting solutions I found came from the same Simon Newton who developed libartnet. Simon was able to replace the OS on a Netgear router with a version of Linux and libartnet. Then, using the USB port on the router he was able to connect up some USB-DMX adapters and create an extremely inexpensive wireless ArtNet node. See this page for more information and specifics. The solution isn’t esxpecially elegant but it does show how inexpensively this could be done and gives some hope for future inexpensive ethernet to DMX adapters based on embedded system technology.