Michigan Telephone, VoIP and Broadband blog

An article entitled “Cisco weighs in on net neutrality” in Telephony Online begins as follows:

Cisco Systems weighed in on the issue of network neutrality in a letter to Congress last week, affirming the importance of neutrality but urging lawmakers to refrain, for the time being, from enacting legislation on the subject.

“We strongly support the principle of an open Internet,” Cisco CEO John Chambers wrote in a letter to Congressman Joe Barton, who chairs the House Energy and Commerce Committee. “We must, however, balance the fact that innovation inside the network is just as important as innovation in services and devices connected to the Internet. Broadband Internet access service providers should remain free to engage in pro-competitive network management techniques to alleviate congestion, ameliorate capacity constraints and enable new services.”

Huh? Let’s lay out what’s at issue here, since the executives at Cisco don’t seem to get it. Nobody is saying that an Internet provider can’t manage their network, as long as it’s done in a fair and equitable manner. What is NOT fair is to discriminate against some services, while giving preference to other, similar services.

Or to put it another way, lets say that there’s a lot of traffic on the network, and it appears to be divided between e-mail, web browsing, streaming audio, and VoIP (this is just a simplified example, so bear with me for a moment). Now, it doesn’t matter much if the e-mail is delayed a few seconds, especially since a good percentage of it is probably unwanted spam anyway, so it doesn’t hurt to give that a relatively low priority. With web browsing, you want snappy response, but customers are not really going to notice if packets are delayed maybe half a second, or even a second during high usage periods, so you can give that medium priority. With streaming audio, the main thing is that once the packets start flowing, they aren’t delayed by more than what the user’s audio buffer holds (typically 2-3 seconds of audio), so while speed of delivery isn’t that important, consistency is. And then we have interactive voice and video over the Internet, which needs both fast delivery and consistency.

Few would object to prioritizing traffic in this manner. Some things do benefit from faster delivery. But when a customer pays for broadband service, it could be argued that part of what they are paying for is for their provider to give them the best possible online experience. Customers don’t give their providers $40 a month or more just so the provider can buy more TV ads! Instead, they expect to receive something of value in return, and intelligent traffic management in a way that does not give the customer a bad online experience ought to be the desired goal. In fact, if that’s NOT the goal, why have traffic management at all? Just let the packets fall where they may. But of course, providers can’t do that because a fair percentage of customers might be unhappy enough to go back to dial-up, or seek some other broadband option.

What people object to is the favoring of one equivalent experience over another. Let me give you a hypothetical example to help you visualize what the the phone and cable companies are wanting to do: Suppose that UPS had a monopoly on parcel delivery service in your area, and supposed that they were owned by Staples. As long as you ordered your office supplies from Staples, you’d get them promptly, but if you dared to order them from OfficeMax or Office Depot, they might arrive a week or two later, or they might not arrive at all. Were something like that to actually happen, you’d expect the Federal Trade Commission or Congress to step in and do something to insure that the other office supply places could get their packages delivered.

Well, that’s what’s being asked for here. The phone companies and cable companies want to be not only the transporter (and the monopoly, or one-half of a duopoly transporter in many areas) but also one of the users of their transport, and they want to show their own services favoritism. What we are talking about is unequal treatment of similar packets, based on who is paying “protection money” and who is not (except that, of course, the phone and cable companies don’t have to pay themselves for the preferential treatment, so in effect their own services get a free ride on their networks!).

It’s all well and good for the Cisco CEO to say that he supports the principle of an open Internet, and I for one am glad he does support it. But just saying you support something, with the absence of some affirmative action to make it happen, often results in the exact opposite of what you support. It’s like saying you support open roadways while you let trees grow up in the middle of the roads – sooner or later, the inaction results in the same effect as if you didn’t support it at all. Support “in principle” is often synonymous with “we’ll say we support it, but we don’t really care.” I don’t know if that’s the case here, but I sure would have liked to have seen a stronger show of support out of Cisco, not the wishy-washy statement contained in their letter to Congress.

Unfortunately, Congress may be listening – News.com reported today that Senator Ted Stevens is now saying that Net neutrality may not happen. It would be a real shame if Cisco’s apparent lack of enthusiasm for legislation was what killed the momentum, and results in a very non-neutral Internet.

Technorati Tags: net neutrality, legislation, regulation, Cisco, MichiganTelephone

This is copied verbatim from David S. Isenberg’s blog:

An anonymous comment today points out that Vonage-Comcast complaints continue, dramatic data here. Paddy Link, the Comcast PR person who told me No, No, No, it is a Vonage Problem has not been back to me with any backup of that claim. She said that Vonage admitted it, but could point to NO VONAGE STATEMENTS to confirm her claim. She said she’d get specifics from Comcast, but nobody from Comcast has presented specifics to me to date.

In the absence of specific explanation to the contrary, it sure looks like Deliberate Discrimination to me.

David Weinberger points out that Shaw, a Canadian Cableco is purposely degrading Vonage packets too. Only Shaw has the stones to offer its readers un-degraded service for a $10/month de-degrading fee.

All I will say is, whether the problems are caused by deliberate action, or just plain neglect in not fixing the problems, Comcast will more than likely lose customers if they continue to play this game. I can envision some corporate executive rubbing his hands together and thinking, “Heh, heh, heh, we can force these suckers to buy cable TV, broadband service, AND telephone service from us.” But that is a dangerous assumption, because if customers get angry with Comcast, they may well not only take their broadband business elsewhere, but also their television business. And people who get satellite service rarely go back to cable, unless their “professional installer” did an exceptionally poor job (which, unfortunately, does happen).

Comcast has been running their “ditch the dish” ads, which seem to me to have become more and more shrill. So as a public service, let me explain some things about satellite viewing. The main reason people have trouble with their dishes is because they are improperly installed. A dish should NOT normally be mounted on a roof, because then all the things that are said about snow collecting in the dish are true – it happens, and it’s dangerous to climb up on the roof to clean the snow. There is such a thing as a heated satellite dish, but they are more expensive than regular dishes so they are rarely installed.

However, what people can’t seem to understand is that, unless you are trying to get a clear view over trees, there is NO advantage to mounting a dish high on a roof. Unlike a TV antenna, relative height does not influence signal strength (not to any measurable degree – think of the distance to the satellite, and realize that by going higher, all you’re doing is getting the dish an infinitesimal amount of distance closer). A much better location is any protected place where the dish has a clear view of the satellites – for many people that would be on the southernmost side of the house, perhaps a foot or two under the eaves. But for many others, the best place for the dish is on a metal pole (a 10 foot length of galvanized pipe of the proper diameter often works well) placed out in the yard somewhere. It’s definitely best to pour a bag or two of cement around the base of the pole (preferably after flattening the part that will be embedded in the cement slightly so it can’t turn – one or two solid whacks with a sledge hammer will usually deform it enough to prevent rotation. The advantage to having the dish on a pole, or on the side of the house is that if it gets full of snow, all you have to do is go out and give it a swipe or two with a broom (carefully, of course) and you are back in business.

But for some reason, many installers almost insist on putting the dish on the roof, which is a really bad idea unless you simply must put it there to clear the treeline (or a nearby building). Not only can the dish fill with snow, which you won’t be able to clean out easily, but the cable may be susceptible to damage from ice on the roof. Furthermore, trust me, unless you either a) know and trust the cable installer, or b) don’t really care if there are a few holes in your roof, you really don’t want a dish installer on your roof if you can possibly avoid it. Because, unfortunately, some installers do seem to take the attitude that “It’s not my house, I don’t have to live here, and by the time they discover that the holes are leaking I’ll probably be working someplace else.”

At least if they put the dish on the side of your house, you can inspect the workmanship, and the same is true if they put it on a pole. But professional installers hate the pole option because it generally involves at least two trips – one to dig the hole, set and level the pole, and pour the concrete, and another to install and aim the dish. Plus, they hate digging the trench to get the cable from the pole to the house. And one other thing is that if they happen to get the location of the pole wrong (where the view of one of the satellites is obstructed), then they have to dig up pole and concrete and try to move it to a better location.

So the trick is, if you ever get a dish installed, insist on an installer that will do a pole install. Make it clear to the scheduler that if the installer isn’t willing to do a pole install, you’ll send ‘em packing. There may be a slight additional charge for this, but you will be thankful on the (actually relatively rare) occasion that the dish fills with snow (usually only a certain kind of wet, sticky snow will actually stick).

Oh, and by the way, if you ever find yourself starting to believe the ads disparaging the dish, call up one of your friends that subscribes to cable and ask them if their cable ever goes out. The ads would have you believe that the dish is always going out while cable is ultra-reliable, but if you can find someone who has subscribed to both technologies, you may get a slightly different view, especially if they live any distance from the cable headend.

Technorati Tags: net neutrality, michigantelephone

The ever-popular Nerd Vittles site has released their Unabridged Installation and Upgrade Guide for Asterisk@Home 2.7. Before you begin, be sure to scroll down to the comments, particularly if you have a Zaptel card in your system.

For those of you not familiar with Asterisk@Home, here’s a short description, from the introductory paragraphs:

Want a rock-solid PBX at a rock-bottom price: free! Asterisk@Home 2.7 has hit the street with some new goodies and a new release of Asterisk. So here we go again. With this update, you get version 1.2.5 of Asterisk as well as the latest and greatest version of Linux, CentOS 4.2, the latest Festival Speech Engine (1.96), the latest version of the Asterisk Management Portal (1.10.010), the Flash Operator Panel (version 0.24), Open A2Billing, Digium card auto-configuration, loads of AGI scripts including weather forecasts and wakeup calls, xPL support, the latest SugarCRM Contact Management System with the Cisco XML Services interface and Click-to-Dial support, Microsoft File Sharing and Networking support through Samba (3.0.10), plus dozens and dozens of free utility software applications for Asterisk compliments of Nerd Vittles. And, yes, Asterisk@Home 2.7 still fits on a single CD! By popular request, we’ve also added something new in this tutorial: some tips and tricks to assist those of you upgrading from a previous version of Asterisk@Home. No, you can’t just copy your old config files over and expect things to work. But, in the Nerd Vittles’ tradition, we’ll show you how to take most of the torture out of the upgrade process. For those new to Nerd Vittles, be aware that we make slipstream changes to articles as users discover things we’ve missed in our first pass. So check for Comments before you begin or subscribe to our Comments RSS Feed. And add yourself to our Frappr map.

The installation process is pretty straightforward. You need a dedicated machine for this install although, for Windows users, here’s another option that doesn’t require a dedicated machine. The VMware edition from vmwarez.com runs in a window on your Windows XP or 2000 desktop and has been updated to version 2.7. You can download it here. For now, however, we’re assuming you’re doing the dedicated machine install: pure Linux on a clean machine. So begin by downloading the 2.7 ISO image from here, burn a CD (click here if you need a refresher course), use an old clunker PC or a $200 WalMart special, insert the CD you made, plug your machine into the Internet and turn it on. Then watch while Asterisk@Home loads CentOS/4.2 and all the Asterisk and Linux goodies imaginable: Apache, SendMail, Asterisk Mail, SugarCRM, MySQL, PHP, phpMyAdmin, SSH, Bluetooth, the Asterisk Management Portal, the Flash Operator Panel, Call Detail Reporting, and on and on. We’ve covered how to use most of the Linux products in our Mac HOW-TO’s (see sidebar), and they work exactly the same way with Asterisk@Home so keep reading. And, yes, this install will reformat (aka ERASE) your hard disk before it begins, but it now warns you first.

Note: Several edits have been made in this post since it was originally posted.

I’m going to add a few hints for those of you who, like me, come from a Windows background and may not be all that familiar with Linux. Please understand that I am NOT an expert on this stuff, having all of about a month’s familiarity with the Asterisk@Home package, but there were some things that drove me nuts at the beginning and they do not have to do the same for you.

The first thing is, after you complete the installation, do the upgrades, install Webmin, etc. you will find that you need to make changes to some configuration files. Nerd Vittles wants you to do it using a Linux text editor called nano, which I’m sure is a fine editor if you’re used to it, but it can drive you a bit crazy at first if you are used to working in Windows. So bear in mind that there are a couple of things you can do at this point. One is to get Samba (support for Windows-style networking) up and running, and in this version of Asterisk@Home they actually give you a way to do that without having to resort to doing it through Webmin (which, by the way, is still considerably easier than trying to configure Samba to work on your own). Once Samba is configured, you can then treat your Asterisk@Home box as a shared resource on your local network, and you can edit files using a text editor on your Windows box, but if you do that, there are a couple of pitfalls to watch out for.

The first is a thing called permissions – this is a form of security in Linux, which is a derivation of Unix, which was originally designed back in the days of time-sharing computers, when many users and many unrelated processes might be sharing the same system. Permissions are designed as a way to keep files restricted to only those people and processes that are supposed to have access to them. It would take far too much space here to go into greater detail about permissions but suffice it to say that when the permissions aren’t set correctly on a file, things break. If you plan on learning anything about Linux, permissions should probably be one of the first subjects you study.

Normally, if you use your Windows-based text editor to load a file in from your Asterisk box over the local network, make some changes, and save it back, the permissions won’t change (but beware of changed line endings – we’ll get to that in a moment). However, if you create a new file, or change the filename before saving it, or delete the original file before saving the changes, that may (probably will) change either the permissions, or the owner and user, and suddenly the file may become inaccessible to the software. Yeah, I really hate this, while folks who love Linux seem to think it’s one of the best features of that operating system. Get bitten by permissions a few times (when you are scratching your head wondering why Asterisk seems to be ignoring a file that you know is there) and you will probably wish the software ran under Windows. I don’t think most Linux users have any idea how foreign the concept of permissions is to a typical Windows user.

So, changing a file on your Windows box may not always be such a great idea. Fortunately, there are still easier ways of doing things than sitting right at tyour Asterisk box and using nano. The trick is to set up a SSH client on your Windows box such as PuTTY, assuming you are not in a country where where encryption is outlawed (if you are, there are probably components of Asterisk@Home itself that may be illegal in your country). Install PuTTY and use it to connect to your Asterisk@Home box, using an appropriate user name and password (root will let you do anything on the system, but it’s also the most dangerous way to operate if you don’t know what you’re doing). Before you connect for the first time, you may want to change a couple of configuration settings in PuTTY, so that your numeric keypad works as expected.

In PuTTY configuration, open up the Terminal settings and click on Keyboard. Then make sure you have the following settings checked:

The Backspace key: Control-? (127)
The Home and End keys: Standard
The Function keys and keypad: Xterm R6
Application keypad settings: Both should be Normal
Enable extra keyboard features: Check Control-Alt is different from AltGr

Click on Window and look at the scrollback buffer size – you may want to expand this a bit beyond the default (if you don’t see the need now, you can always come back and do it later). Then click on Translation and look to see which option is used for “Handling of line drawing characters” – if you get display weirdness in some situations you can try changing this around (I use “Use Unicode line drawing code points” but I will not guarantee that’s the best setting). Then click on Selection and note the mouse options – again you may want to change these as you get more familiar with the program, or if you already have a personal preference.

You really shouldn’t have to change anything else unless you’re operating under special circumstances. Once you have PuTTY configured, click on Session, put the IP address of your Asterisk box in the Host name field, give this configuration a name (in the Saved Sessions text box – I suggest using the name Asterisk), then click the button to save this configuration.

Now click the Open button and you should get a login prompt. Login using one of your user names and passwords that you set during installation. Once you login and are at a command prompt, it’s just as if you are sitting in front of your Asterisk@Home box.

Now, whether you are connecting through PuTTY, or actually at your Asterisk@Home box keyboard, you can make changes to the text files from within Linux itself. You could use nano, but there’s another way that many from a Windows background will consider easier, especially if you’ve ever used a dual-pane file manager like Norton/Windows/Total Commander (or a similar program). From the Linux command prompt, type mc -a (mc stands for Midnight Commander) and you will find yourself in front of a (hopefully) familiar environment. The -a option makes it render the line drawing characters as something at least a little more presentable when using PuTTY; you probably don’t need (or want) to use that option if you’re actually at your Asterisk box.

In Midnight Commander, as in the Windows variants, you can highlight a file, then press or click on F4 to bring up an editor, which is probably going to work a lot more as you’d expect it to work than nano if you come from a Windows/DOS background.

Remember how I talked about permissions above, and how they can give you fits? Well, you can use Midnight Commander to manage them, also. Highlight a file, then click on File (in the top menu bar), then I suggest using Advanced chown which shows both the permissions, and the owner and group settings, and allows you to easily change them.

If you want to learn more about Midnight Commander, there’s a FAQ here.

Should you decide to edit configuration files on your Windows box, be aware that Windows normally saves text files with a carriage return and a linefeed at the end of each line, while the Linux/Unix convention is to use a linefeed only (this is the other pitfall I was referring to). If you save a configuration file with carriage returns in it, you may get all sort of unpredictable (and bad) results. So, be sure to use a text editor that will let you save files in the format Linux prefers, and then remember to save the files that way (you may have to use the “Save As” command rather than just doing a simple “Save”). Just to confuse matters, I’ll mention that some other types of systems (Macs, I think) use a carriage return only to end a line. If your text editor doesn’t give you an option to save files in the Linux/Unix format, it might not be able to do it. That’s why it’s probably better to use nano or the Midnight Commander editor to make changes in the configuration files.

Now, here’s one other tip. If your Asterisk@Home box is behind a NAT (generally, behind a router, as it probably will be if you are using it, well, at home) and you are getting one way audio on some SIP connections, then you may need to add some lines to the top of sip.conf, under [general]:

nat=yes
externip = your fixed internet address, e.g. youraccount.dyndns.org
fromdomain = your fixed internet address (again)
localnet = 192.168.x.0/255.255.255.0

(Replace the x with the octet number used in your local network).

One other place you might need to put your fixed IP address is in the file etc/sysconfig/network after the HOSTNAME= part. If this isn’t done, sendmail may not work correctly and you may not be able to get your voicemails e-mailed to you. Additionally, you may need to go into /etc/hosts and make sure that there is NO line linking your fixed internet address to 127.0.0.1

I again remind you that I’m no expert in this stuff; what I’ve written above is what I have found works. If you know of a better way to do anything I’ve shown here, by all means please feel free to leave a comment (hints on configuring PuTTY and/or Midnight Commander for a better display are especially welcome).

Technorati Tags: Asterisk, Asterisk@home, PBX, VoiP, telephone, telephony, MichiganTelephone