November 2005 Archives

A metered download policy is not one that the average American is going to easily accept, but an metered upload policy would probably work quite well. The telecoms face several problems, all of which are of their own making. They do not charge enough money to sustain proper broadband services, and they do not impose sensible policies on extreme bandwidth users who are typically using bandwidth for illegal purposes.

In order to make VoIP services, and possibly at some point, a video-enabled upgrade, possible, there needs to be a sufficient amount of upload bandwidth to allow for audio equivalent to 32kbps MP3 to stream reliably. Nothing less than 128kbps would be enough to provide that, and ideally they should be providing at least 256kbps. 512kbps of upload bandwidth is more than enough for the current generation of home network applications that need a lot of upload bandwidth.

Since most a significant amount of uploading is illegal, telecoms and cable companies could penalize their illegal users by making a lower amount of upload bandwidth the norm, and then charging metered uploads on bigger connections since those are what P2P users often go for. If you want 1-2mbps of bandwidth, fine, but that's going to start costing you an additional $2/GB of bandwidth after your first 1GB of bandwidth use. 1GB of upload bandwidth at this point should be more than enough of a cap at such speeds to allow a VoIP user to use their service unmolested without making them subsidize the cost of the P2P free riders.

The telecoms and cable companies are victims of their desire to dominate at all costs because they know that they cannot make proper infrastructure investments with $15/month for DSL service or $25/month for cable service. They are going to have to accept the fact that metered bandwidth is unacceptable to the average broadband user, regardless of whether they use P2P applications for copyright infringement or not. People like me, who use it to buy off of the iTunes Music Store, are especially not going to accept this, and no amount of academic philosophical tripe is going to convince me that all of the blame falls in the laps of the file sharers when it has been obvious for a while that they have been courted by the broadband companies as they tried to build their services.

Besides, once the current generation of file sharers really get out into the workforce, they're not going to have the time to do these things. When you have a steady job, it's much easier to pay $15 for a CD than it is to spend a hour or two hunting down all of the tracks online. Not only that, but there is a certain degree of danger in doing this as an adult. Juries are not going to be sympathetic, even if you have kids, and employers would immediately look askance of anyone who continued to spend so much time taking content when they clearly had the money to pay for it.

Just wait a few more years, and a lot of this will probably start to correct itself. In the mean time, telecoms and content producers alike need to get out of their idealistic, philosophical world where they can rant to Congresscritters about these problems and actually find cost-effective means to fix them. The content producers would, unfortunately, be the most to suffer from this suggestion, if for no other reason than it would force them to develop more cost-effective delivery mechanisms so they could reduce their production expenses while producing content that is compelling enough to buy.

Before you can use Ruby on Rails you have to learn Ruby so I've spent part of the night trying to learn Ruby. What a fascinating language, and it's amazing how cool it is from just a syntactical perspective alone. Here's a simple example with an object in Ruby and a comparison with the equivalent in Python:

class PointKeeper

def initialize(name, count)

@name = name

@points = count

end

def Name

@name

end

def Points

@points

end

def Points=(points)

@points = points

end

end

pk = PointKeeper.new("MikeT", 4000)

print pk.Name + "t" + pk.Points.to_s + "n"

pk.Points = 5000

print pk.Name + "t" + pk.Points.to_s + "n"

<strong>Python version:</strong>

class PointKeeper:

def __init__(self, name, points):

self.name = name

self.points = points

pk = PointKeeper("MikeT", 4000)

print pk.name + "t" + str(pk.points) + "n"

pk.points = 5000

print pk.name + "t" + str(pk.points) + "n"

Now what's the big deal about the difference? Seems like the Python version is simpler and thus more elegant, right? Wrong. I have a soft spot for Python given that it was one of the first languages I studied, and I have even used it at work to get stuff done, but it is a bit lacking in some respects compared to Ruby and this happens to be one of them. Python doesn't allow you to restrict object member variables and that is a problem because it makes it harder to achieve proper encapsulation. Who knows what variables should be accessed and which shouldn't be accessed? That of course requires more documentation, which requires more reading, which ultimately means less time spent working on the actual code.

For a while I was resistant to learning Ruby, but now that Rails has come along, I have to learn it in order to feel like a proper geek. It's amazing how cool this language really is. I really do wish that I had been forced to get familiar with it a while ago.

If the broadband providers want to actually assuage the fears of the supporters of mandated net neutrality as it applies to bandwidth charges, they need to begin tossing out the numbers that they realistically see themselves charging. It would be far less threatening to their customers to hear that they Verizon, for example, would be more than happy to drop their basic DSL service down to $10 for 5GB of download bandwidth and 512MB of upload bandwidth at 768kbsp. Then, for $25 a month, they give you 3mbps DSL with 8GB-10GB of download bandwidth and 1.5GB of upload bandwidth. Each additional 1GB of download bandwidth could then cost $0.25 and each additional upload would cost $1 per GB. That way, the people who are buying off of the iTunes Music Store end up not taxing the broadband provider's infrastructure since spending $0.99-$1.99 per product would control their bandwidth use, but they would end up charging those who use 600GB of download bandwidth for illegal purposes many times more than the average user to compensate for their use.

At this point, I am willing to consider a metered bandwidth policy since I do not download enough anymore to worry about $100 bandwidth bills. Even downloading several CD images of Linux in one month plus buying $50 worth of goods from the iTunes Music Store would not be a problem here. Vonage's service doesn't use enough upload bandwidth to really be a problem either, unless the telecoms want several dollars per GB, or even worse, for every few hundred MB of upload bandwidth.

As I have said before, the broadband providers are grossly underselling their services and that is why they are getting hit so hard on their profit margins. $15 for unlimited DSL is not sustainable, but $100 a month might be. Either way, as long as the rates are cheap, I'm open minded. The telecoms just need to shut up with their free rider arguments against companies like Google and Amazon, and start tossing out the rates that they would want for their consumer broadband services.

I am reluctant to say this, but I agree with Patrick Ross of the PFF, that what is in the best interests of many users is a metered system these days. The more that I think about it, the more that I cannot help but think that as long as the bandwidth rates are cheap, and the base service rates are cheap, that it might actually lower costs to go to a liberally metered system. The only thing that would have to be in place is a committment on the part of the providers that as they invest in their services, and begin to roll out more bandwidth-intensive features and competitors do the same, that they will increase that amount of bandwidth that every dollar buys. I do not want to be paying $1/GB to download when a TV-on-demand episode is 3GB-4GB of HD video data. The best thing that the providers could do is to remember that their service is enabling delivery of content, not actually providing the content that is being delivered.

Dozens of law-enforcement officials fanned out across Orange County on Thursday to arrest 23 suspected members of a white racist street gang that includes nonwhites and is allegedly involved in identity theft, drug sales and other crimes.

More than half of those arrested were women, including a Vietnamese woman and a Latina. Authorities did not list the charges, but Sheriff Michael S. Carona said most suspects were picked up for parole and probation violations.

They can't be too racist if they are accepting non-whites into their gang. Sounds more like an equal opportunity gang that just targetted certain races or ethnic groups for its crimes. So what does this make the Bloods and the Crips? Black racist paramilitaries by comparison? The actual crimes that they had listed seemed to be petty compared to the stuff that other gangs like the Latin Kings and MS-13 stand accused of. But, we can rest assured that they are as bad as the LA Times makes them out to be because the ADF says so. It's not like they've ever been prone to wild-eyed extremism on these types of things before.