I figure I'd take a moment to
interject some discussion into Skype Journal's very fair summary of how Yahoo voice
compares to Skype. I've quoted the Journal's Kevin Delany in italics and added my thoughts:
[Yahoo] have the North American market secured, while Skype is still struggling to enter Canada and fully bloom
in the USA.
I don't know where Kevin lives, but I've had more Skype conversations on Skype in the good
old U.S. of A. than on any other VoIP tool.
Yahoo Voice uses SIP, a standard protocol, as opposed to
Skype's closed one.
Yup, and for the umpteenth time, Skype, I would like you to add bleeping SIP
support to SkypeOut, for the love of Pete!
They have the power to advertise it on the #1 Most
Visited site in the world.
Aren't we forgetting about the eBay angle? eBay is the top e-commerce
site on the web, and one of the most heavily trafficked in all categories. If I had a link from eBay to one of my
blogs, I could retire tomorrow.
Some of their prices beat Skype's
This won't matter in
the long term. Price isn't the cornerstone of VoIP software appeal, or I should say it's no longer the
cornerstone.
Why Skype is still better:
Skype has the European market. Yahoo clearly
mentions on the bottom of their information page: "Intended for use by U.S. residents only.",
leaving Canada and the rest of the world in the dust.
This is true. But it's only a matter of time
before Yahoo starts rolling up European and Canadian PRIs.
They are fully P2P and encrypted. Nobody's
listening, while Yahoo will submit under the pressure of big brother to eavesdrop.
Skype will submit
under the pressure, eventually, too. The bottom line is profit, and if the inability to tap phone calls, whether
software driven or not, will get trumped by law enforcement any day of the week. These guys are in business to make
money, and, last time I checked, Skype isn't an open source product.
As much as they like to think it,
Yahoo does not own the IM world and doesn't have nearly the amount of IM users that Skype has.
Nor as
many as AIM. Good point, Kevin.
No ads, no bloat and nothing you don't want.
The Skype
API is bloat by definition: you have to have the PC desktop version of Skype installed and running just to use it.
Again, for the umpteenth time, Skype--separate the Skype API from the desktop Skype software, for the love of
Pete.
Calling is the central focus of Skype, while Voice is now a "feature" of
YIM.
I actually see this as a perceived disadvantage for Skype. Integration between text,
data-recall, web, and voice applications should be viewed as more and more seemless, not the other way around.
Voice is just an application running within the scope of a larger network, so saying something like "all voice,
only voice" is sort of like ignoring the core issue of importance at the heart of the VoIP revolution: software
and networks are flexible enough to do more than one thing at once.
It's fully
cross-platform.
Kevin nailed this one. Of course, I would've place this point at the top, not at the
bottom.Despite my critique, I actually think Skype Journal hits most of the right points.Read | Permalink | Email this | Linking Blogs | Comments
A few days ago, I took a phone interview for the local newspaper. The reporter, who was certainly no tech geek, seemed
very interested, and perhaps even excited, by the idea of Internet Telephony. He was happy to learn that, yes, Vonage
really can save you some serious dough on your monthly phone bill (so can Packet8 and a host of other providers).As I expected, he questioned me about the quality of VoIP services. He was under the impression that VoIP-based
phone service didn't (and couldn't) sound as good as traditional phone service. Not surprisingly, he was taken
aback when I revealed, at the conclusion of our twenty-minute call, that he'd been speaking to me via my VoIP phone
service.This got me thinking. We IP communications engineers, bloggers, and avant garde technology geeks
reside on the cutting edge, confident in our favored technologies, certain that our ideas about the greater
technological good will prevail. But are we missing the mark? Are we not sensitive enough to the true
challenges facing our little revolution? How is the world going to be transformed into a place where the is no
more PSTN, and the global voice infrastructure is--exclusively--IP based?To address these questions, I've
fashioned a list of the top ten issues that I see as counter-affecting the movement to an IP-based global
infrastructure.1. Disbelief about pricing and featuresConsumers, especially
fickle, middle-class consumers and those that are *ahem* older, seem to have a hard time believing that you can
actually lower your voice service cost and get a more feature-complete service offering by switching to Voice over
IP. We, as the thought leaders in this industry, can deal with these problems with better marketing, better
brand alignment (why doesn't eBay slap its name up there next to the Skype logo?), and better product evangelism.
2. Uncertainty about security Perception is reality. Even if the security of VoIP services
is indeed somewhat better than that of traditional land-line services, it takes a lot to convince a skeptical consumer
that something new can actually be more secure than the tried and true. There has been a sort of inferred
connection between Voice over IP services and identity theft which I, as a networking guy, don't really *get*.
It's easier to look over your shoulder at the supermarket checkout than it is to try to sniff your (encrypted)
signaling packets from a MacGyver-type connection to the local cable company's data pedestal. But, like I said,
perception is reality. 3. The question of E911 and government interference
This is more of a real problem than a purely perceptual one. Setting aside the lawsuits lodged against Vonage last year
due to failed 911 calls, there needs to be an "out of court" and "out of Congress" consensus on the
matter of E911. NENA has taken steps toward this concensus. But
industry-wide acceptance is still a ways off, and that's leaving Skype out of the equation. Regulating Skype into
E911 compliance will be a slippery affair, at best. So, the private sector needs a unified voice. That's about all
politicians will listen to. Otherwise, the mandate that gets legislated will
suck. 4. Uninformed-ness about VoIP technology use Getting back into
the realm of the perceived, there seems to be this idea that, in order to use Vonage or services like it, you need to
connect your phones to your computer, or somehow use your computer to facilitate calls--kind of like Skype or
Gizmo. The first person who expressed this misunderstanding to me was my mother. The second person was the
reporter I spoke to on the phone. And I know I've heard it from others, even informed folks like the distance learning
engineer at the local community college, whom I chatted with at a funeral a few weeks ago. Today's VoIP telephone
service marketeers need to do a better job of making the point that all you need is broadband, and the ATA does all the
work. 5. Difficulty in usage & hook up Basic problems that disenchant
mainstream consumers from VoIP telephone service will get worse as more and more people try Internet phone
services. The two basic problems I'm thinking of are: a. Some VoIP service providers have dial-plans
that force you to dial eleven digits even in situations when the call is in a local exchange and would ordinarily only
require seven digits. Sounds minor, but this is a big deal. Let's not forget how finicky consumers are! b. It's still too difficult to
connect all the phones in your home--or office--to an Internet phone provider. While some companies offer wireless
phone "systems" to address this issue (like Packet8 with their Uniden offering), the traditional Bells have
the upper-hand here. They have service technicians that can perform the hookups you need. I see this as a huge
challenge. As a networking guy and somebody who doesn't mind wires, electrical tape, and snippers, my whole house is
hooked to VoIP, but what about everybody else? 6. Paranoid media coverage
This has surfaced only recently. The mainstream media, and even the technology media (to some degree) has gotten
carried away with the idea of VoIP spam, what people are calling "SPIT'. Early on in my blogging for
Weblogs Inc., I even got sucked into the hype. But potential (not actualized) problems such as SPIT are often
flagged with such a sense of
urgency that consumers get the jitters, and media recommends the "wait and see" approach. This isn't good
for the revolution. Lesson to VoIP marketeers: communicate early and often with the mainstream media.
7. The issue of network neutrality There's just not enough hours in the day to explore all
angles of the network neutrality
issue. High-end content providers (like IPTV investors) want to divide the Internet into two or more service
classes--one for "fast" services like IP television, and one for "slow" services like the ordinary
web. Sounds good, right? Well, not exactly. You see, the people who will soon be purveying IPTV (you guessed it--the
Bells) will be able to use this two-tier architecture to marginalize content providers who they see as competitors.
They've already identified Google as one company they see as exercising too much control over content. Vonage and
other VoIP providers aren't far behind and the black list. I stand for network neutrality (that is, keeping all
applications, regardless of vendor or type of traffic) in a single tier of service, with best-effort delivery.
The telcos have floated the assertion that BitTorrent and other bandwidth hogs are in fact clogging the Internet, but
these ideas have outrightly dispelled as false by many people in the know. Today's bandwidth is plentiful, and
tomorrow's bandwidth can be plentiful. The whole idea of a two-tiered Internet seems to me to illustrate the Bell's
desire to control more of the Internet than they already do (which is the lion's share), as a means of boxing out
challenges to their revenue streams. Over time, I have come to believe that the answer to accomodating new,
bandwidth-heavy services is to create more bandwidth on the Internet. I've said it before, and I'll say it again:
bandwidth is not a resource for which there is a limited supply. All you need for more bandwidth is more fiber in the
ground. And this is arguably cheaper than trying to enforce multiple classes of service over the public Internet.
8. (Non-existant) Vendor interop This is a problem created and nurtured by the VoIP
industry itself. Skype users can't call Packet8 users (unless they dial out through the PSTN---and what does that do for
the revolution?? nothing!), and Cisco CallManager VoIP systems, at least until recently, didn't work with Avaya phones.
The end result? Limits on consumer use of available products and service. Unless you dominate the industry the
way Apple dominates the digital music industry, it just doesn't work. Interoperability between products and
service is paramount to the transformation of the global voice network. Kudos to SIPPhone for equipping their Gizmo Project software with the ability to interact with other SIP providers'
services. The rest of the industry should be following their example. 9. Bad enterprise attitudes
about infrastructure upgrades A few days ago, I was talking to a top infrastructure executive at
one of the largest manufacturing firms in the midwest. We got on the subject of transitioning his highly-distributed
global network to a unified, or converged, footprint. This means combining his voice and data applications into a
single logical schema, using common network technologies (most notable, IP), to do everything on his network. I
suggested that he consider switching to VoIP. His response? "The wiring in our building is managed by the phone
company, and we're not going to impede that relationship," or something very similar. Now, the parameters of
a voice convergence project may not actually be there in his firm, but his enterprise's attitude towards
progress in IP communications was disappointing. Unfortunately, with the profit-motive ever in the mind of today's
decision-making executives, many of whom are people who, yes, think that you need a PC to use Vonage, the negative
enterprise attitude towards VoIP is indeed the prevailing one. This has got to change. If the largest
private networks in the world can't converge, how can the largest public ones? 10. Genuine quality
of service issues Finally, we look at the challenge of quality of service. In an enterprise
setting, QoS is relatively easy to achieve, if the architects of the networked voice solution can convince the capital
spenders to purchase the right equipment in the right quantities. But, in residential scenarios, where service
consumers aren't "spending other people's money", QoS is not technically possible to guarantee. This is
because the companies that provide network connectivity to the last mile, such as cable operators and DSL carriers,
refuse to enforce QoS rules that would benefit their competitors in the VoIP industry--folks like Packet8, Vonage, and
Broadvoice. One Canadian telco has even gone as far as charging extra for QoS guarantees. Absurd. Even
still, there are certain things the larger independent VoIP providers could be doing to at least improve last-mile
QoS. For one, they should start hocking
broadband routers that implement queuing, a common practice for improving voice quality on a one-hop link such as
your cable or DSL connection. Ted Wallingford is an advanced networking
consultant. His web site is http://www.macvoip.com.Permalink | Email this | Linking Blogs | Comments
Today, I had the
pleasure of speaking with Kevin Ford, the savvy behind Parliant Software, the
Ottawa-based maker of Phone Valet, the excellent telephony automation package for the Mac. Kevin was very
accomodating, and I really enjoyed our talk. He spoke with the passion of a creative engineer who truly understands the
customers he serves.I learned some great nuggets about the reasoning behind some of the design decisions in
the last version of PV that I looked at--version 3. In my original review, I kind of panned PV's phone book as
being sort of a sorry replacement for Apple's Address Book. I took the liberty of directly comparing Ovolab's
Phlink to PV, since Phlink is heavily integrated with Address Book by default. I still prefer Phlink's approach, but
now I know a bit more about the thought process behind Phone Valet's address book, or lack thereof, as it were.Kevin told me that Parliant had never really intended to compete with or replace any particular contact
management app, but wanted to be a "good neighbor" and allow users to integrate whatever CRM systems they had
at their disposal. This make sense and will be taken into account in my next review of Kevin's software.I also had an opportunity to
talk with a VP at CommuniGate today. As you may know, CommuniGate is a vendor
of high-end, Unix-based software for high-volume messaging applications. Now, CommuniGate is in the process of ramping
a whole lot of SIP call-signaling capabilities into their platform, essentially transforming it into a true convergence
server for service providers. He mentioned that there are many great developments still to come, including Ajax-based
front-end goodies for subscriber-facing call-management apps. Cool, cool, cool. Well, it's getting late.
See you tomorrow, blogosphere.Permalink | Email this | Linking Blogs | Comments
As some of you might already know, via my incessant Cleveland banter and constant linking to local web sites, I am an
Ohioan. Not a native Ohioan, mind you, but a transplant from Detroit. It took every nerve in my body to root for the
Indians today as they squashed the Detroit Tigers, who have MLB's current home run leader. Whenever the Tigers lose, a
little piece of my heart loses along with them. But hey, I'm firmly cemented in Ohio now. Local sports. Local politics.
Local business. Local microbreweries. The Indians, the Browns, Broadvox, OneCleveland, and of course, the so-called rock
and roll hall of fame.As such, I was ecstatic when Brad Hill asked me to blog for Weblogs Inc.'s newest
family member, and the first of the network's local/state blogs: Blogging
Ohio. I hope you enjoy reading this blog as much as I enjoy living in the state of Ohio, whose rolling acres are
more filled with cattle farms and old steel mills than they are with VoIP startups and Silicon Valley exuberance.Welcome aboard, Blogging Ohio!Permalink | Email this | Linking Blogs | Comments
Whether it's misspelling the word iPod or simply complaining like a teched-out Andy Rooney, PC World's John C. Dvorak's
primary appeal seems to be to the casual, back-to-basics PC user who can't detect that his comments are often totally
out in left field. Case in point: Dvorak thinks Apple needs to open source Mac OS X. Apparently he doesn't know that OS
X's Unix kernel and much of its development toolset are already open source, but I digress.
If the Windows
test keeps going the way it's going, the results may indicate that Mac users are more likely to shift to Windows than
we used to think.
Dvorak clearly doesn't understand why a certain segment of the consumer population
continues to choose Mac OS X. They choose it for the precise reason that it isn't Windows (bloated, too-little-too-late
on features, and usually buggy through the second or third maintenance release) and that it isn't Linux (learning
curve, few commercial applications, and a rabid community of desktop Linux believers that often sound, to the average
observer, like zealots). If Mac users were impressed by how fast Windows ran on a certain piece of hardware, they
would save the 25% luxury tax they spend to own a Mac and just buy a cheap, fast Dell. But that hasn't happened
yet, has it?
But what will happen to Mac OS X? I suspect that the testing of Windows on a Mac might be
duplicated in reverse, with a similar test of the Mac OS X running on a conventional PC. Here again, we'd need to look
at the test-marketing results. In this scenario, the idea would again be to determine-by testing-whether or
not getting OS X onto PCs would help or hurt Apple as a company. The same three factors would be assessed: practicality
(is anyone interested?), functionality (does it work at all?), and political marketability. In the case of political
marketability, one additional variable enters the picture: Microsoft perceiving this as a threat to its business.
Dude, John, Microsoft is already in trouble. Look at their "innovations" over the last five years.
Look at their stock price, and look at their executive brain-drain. Steve Ballmer embarasses himself repeatedly at
events. The question isn't "does MS think Apple's Intel version of OS X is a threat to Windows?"
because Mac OS X is already a threat to Windows. Yet Microsoft has done a more than fair job containing that
threat by using its single-best asset: its installed based of companies that use Windows services.
The
one thing keeping OS X from gaining any share in the PC industry is that Microsoft owns the directory services market
in 90% of American corporations, and by controlling the framework for central authentication, Microsoft can keep Mac
users (and Linux users) out of the equation. If Apple did release Mac OS X for Intel beige boxes, the
threat posed to Microsoft wouldn't increase one bit. Because Mac OS X still wouldn't have top-billing as an
Active Directory client, and it still wouldn't have a feature-complete version of Outlook.
Since no
company, including massive IBM, has been able to compete with or unseat Microsoft from the desktop, Microsoft's stance
alone may prevent any universal acceptance of OS X on the desktop from ever happening. In fact, I assume that as this
is being written, Microsoft has coders in its skunk works tearing into OS X looking for deep flaws that it can exploit
and publicize. Don't think otherwise. It only makes sense that they'd do this.
Yeah, but they don't need
to. Publicizing a buffer overflow exploit in the Mac OS X kernel is something Microsoft doesn't need to do, because the
overwhelming strength of their exclusionary tactic against OS X is the hobbling of Windows-integrated services for
Apple's OS, as I said earlier. This one thing is what causes IT managers to deny OS X any consideration in the
enterprise, and this one thing is all Microsoft needs to keep doing to permanently hold OS X in its current position as
a niche OS.
Thus a cloud is rising over OS X and its future unless Apple makes its boldest move ever: turning OS X into
an open-source project. That would make OS X versus Linux become the most interesting battle within the computer scene.
With all the attention turned in that direction, there would be nothing Microsoft could do to stem a reversal of its
fortunes.
What? Can anybody help me understand why open-sourcing OS X would solidify Apple's position against
Microsoft? Or even against Linux? Dvorak has made the mistaken assumption, as many analysts do, that Unix
operating systems can compete with Win32 in the server room. You want to know why IBM is pushing Linux? Because
they have consultants they bill at $200 per hour to set up Linux boxes with IBM storage arrays, not because they have a
margin on the OS itself. IBM and Microsoft are involved in very different business models, and I'm afraid
Apple has a different business model entirely: sell a hardware platform and then sell residual software upgrades after
the fact, all at a profit. This isn't what IBM, the leading Linux consultant, or Microsoft, the leading software
company, are doing. So one-upping those guys with open source would require a drastic reconsidering of Apple's
current, profitable, business model. All of this seems born of the idea that a computer company can't sell
$10 billion a year and be considered a serious competitor. But to sell $10 billion a year profitably?? That's a hell of
a feat for any supply-chain oriented manufacturer. Just ask Delphi.Permalink | Email this | Linking Blogs | Comments
Skype announced it has surpassed the one hundred million user account mark. This is significant, but let's do some
mental math real quick. Assuming 1 in 10 people who activated a Skype account abandon it because Skype
isn't sticky enough for them, that's 10 million fewer accounts Skype can claim.Now, let's say 1 in 20 Skype
users has multiple accounts. Take me, for example. Having multiple accounts is the only way I can test certain features.
So one guy, two accounts. Ding Skype's number by another 5 million. So now we're down to 85 million
active, real, unique Skype users, right? Well, sort of. You see, in the earlier days of Skype, spam
sign-ups were fairly common, and the net effect is probably another 2 million users off the bottom line. Then,
you've got the Skype users of China who frankly don't count because they aren't free consumers and can't choose to use
Skype in the manner prescribed for all the free nations of the world. These Chinese users may be hobbled, but Skype
wants to give them full credit as regular users. So, assuming there are 6 million of them, I'll meet Skype in the
middle. Ding another 3 million. So now, we're down to a mere 80 million. Finally, take out all of the
automated endpoints and API hacks which use unique user names and I think you're probably down closer to 75
million. Not to mention, only 6 to 8 million users are purportedly online at a time. This means that the
active user base is only 10% or so of the total user base. That doesn't scream ubiquity. Splitting
hairs? Probably. They still have the best penetration rate of any desktop VoIP software on the planet.Permalink | Email this | Linking Blogs | Comments
Tonight I had the privilege of recording a podcast with some of the best minds in the telecom industry: Ken Camp, Andy
Abramson, Dameon Welch-Abernathy, and Alec Saunders. I was most impressed with the different points of view each of
these gentlemen brought to our round-table discussion. Ken's experience with AT&T served as a nice counterweight to
my relative lack of carrier-grade knowledge, while Alec's enterprising Voice 2.0 knowledge was highly transcendental.
Andy's market understandng was, as usual, very informed, and Dameon's hands-on experience hacking with some of the best
(and lesser-known) VoIP goodies was evident. I also found out that Dameon's name is pronounced just like
"Damon". Perhaps that's why he goes by Phoneboy. Ken did a great job moderating the discussion.
Though the technology didn't exactly cooperate (Dameon started on a WiFi VoIP connection and ended on a land-line and
Alec was having some jitterbugs on his connection) the exercise served as a great proof of concept for the notion of an
all-VoIP podcast roundtable. This is the kind of thing of lot of us have been talking about for quite a while, so it was
neat to actually do it. The tool we used was Gizmo Project, which has a built-in party-line style conference bridge as
well as built-in recording, very handy for podcasters. I know Ken is busy in the editing room, so the first VoIP
Think Tank podcast should be posted soon. I'll let you know when it is.Permalink | Email this | Linking Blogs | Comments
From the transcendental thought dept...I drive a compact car. I always have. When I was in high
school, gasoline cost me 84 cents a gallon. When I started driving a compact car, it cost me a mere 8 bucks to fill up
the tank with gas. I know that some of my peers in this industry, who are more, ahem, *seasoned*, can recall a time
when gas was even cheaper. My grandpa Hershel ran a service station in rural Ohio where I know he sold gas for as
little as a quarter a gallon. The basic precept of supply and demand in a free economy is an axiom. It
states that if supply meets demand, the price for a given commodity will remain stable. It states that if the supply of
the commodity increases relative to the deman, the price of that commodity drops, and if the demand for that commodity
increases relative to the supply--well, then the price goes up. The basic problem with gasoline--and with
network bandwidth--is that the supply of the commodity is in the hands of a select group of people who control
production. By slowing the production and delivery of commodities like oil and bandwidth, the price is forced upward.In the case of oil, the cartel which controls the lion's share of global oil production is OPEC, a group of
oil-producing nations which are run mostly by dictators. Iran is one such member, and it is run by a known anti-semite
and radical anti-American. Another member is Venezuela, whose communist dictor Hugo Chavez (who owns a controlling
intrerest in Citgo) also holds down a seat in the cartel. The parallels to the bandwidth market are very
clear. In the case of global Internet bandwidth, whose corporate roots are here in the United States, the cartel which
controls bandwidth "production" are the telecommunications and television delivery companies, folks like
AT&T, Comcast, Cablevision, Adelphia, and Verizon. All very powerful organizations with controlling interests in
the domination of core Internet bandwidth as well as last-mile delivery of digital services.And there
exists among this group of companies, at the boardroom level, a desire to see the Internet more or less divided into
two classes of service--one for "fast" applications like paid television programming (controlled by the Bells
of course) and another for "slow" applications like the public Web and marginal services offered by competing
companies like Vonage and Lingo. By placing these competitors in the "slow lane", as some telecom executives
have suggested, they incent them to pay bigger bucks for premium bandwidth, thus driving up the cost of the bandwidth
to everybody but its producers--the big players like AT&T and Verizon. Here's where the parallels get
really ugly: the producers of both commodities--global bandwidth and oil--have discovered a way in which to drive up
the price of a something that is essentially in unlimited supply, negating the precept of supply and demand. The earth
isn't running out of oil, and it will never run out of global Internet bandwidth. If I, as a regional Bell, can create
artificial value in my commodity by charging more for bigger chunks of it at a time (that's the Internet "fast
lane") I can force the price of competing with me up.This is exactly what has happened with OPEC. The
oil producers ramp down production, forcing the global markets to bid higher on oil, giving rise to an increase in
prices across all oil producers' sales channels, even if they aren't OPEC members. Why? Because those guys can get more
for the same thing. Remember, you only have to be a penny cheaper than your competition to sell a commodity
successfully. If the Bells ramp down production of quality bandwidth, the cost of competing with them will skyrocket,
content producers will pay more for bandwidth, the cost of content will increase, and the consumer will pay more for
something that is, underneath all the bureaucratic cover, still basically a commodity in unlimited supply. It's all about who controls that commodity. How much did you spend for gasoline today? Are you willing to
undergo a similar increase in the cost of bandwidth?Permalink | Email this | Linking Blogs | Comments
Working on a project, I have been goofing around with a pair of Linksys WIP300 WiFi SIP phones. This is my first hands-on experience with these units, so I thought I would pass along my thoughts. This is a pretty cool piece of gear, first off. It seems to work very well with Asterisk. The large color display is similar to that of a cell phone, and the built-in configuration menus are easy to navigate. The sound is excellent using G.711 codec. Haven't tried G.729 yet as I don't have a working G.729 on my Mac Mini, which is my Asterisk box, and I haven't tried direct phone-to-phone calling. With a basic Soho access point, I was able to walk a few hundred feet away before I had any difficulty hearing my wife on the other end of the call. The battery cover slips off too easily, and, though the phone provides a web interface, I haven't yet figured out how to modify the pre-set selection of wallpaper images and ring tones. I figured tftp would be supported, but haven't discovered a tftp server setting. And I can't figure out how to actually transfer files to the phone. I portscanned it with nmap but that didn't uncover any open ports. Oddly, the web port didn't even show up. So that's the skinny on the WIP300 so far.Permalink | Email this | Linking Blogs | Comments
