Dialtone is a Promise
Dialtone is a promise.
Please guys, I use Skypeout, I want to rely on it but I can't. If you'd have given more information during this time prehaps I would still be with you, but alas my trust of you is broken.
— martin.porcheron
commenting on a Skype blog
It's the promise of connection. To humanity, to family, to government and social services.
It's a promise your phone will ring when someone calls.
Jesse Robbins 
writes of Telecoms Sans Frontières (Telecommunications without Borders) restoring phone networks in post-earthquake Peru. Keeping that promise, keeping people connected.
Here in the United States, massive deregulation dismantled our payphone networks. When we have our own earthquakes (I live two kilometers from the Hayward fault), storms and other disasters, payphones are often the only landline phones we can reach, the only phones that may work. Payphones are vanishing from the American landscape because of the false assumptions that mobile networks are as reliable as landline networks and that every person has a mobile phone.
Telephony is even more central to developing communities where billions of people have never made their first telephone call.
And we use Skype. Billions of minutes called. Not to mention Skype IM, one of the most popular text chats on Earth.
So the dialtone promise matters.
Skype broke this promise last week.
Skype's motto for years was "It just works."
Mostly.
Your landline phone company defined the dialtone promise. They make that promise by owning and operating all the parts.
Skype's network is not like a local telco. Skype depends on the kindness and infrastructure of strangers. On PCs, their processors and memory, their Internet connectivity, their electricity, their ISPs.
Can Skype keep the dialtone promise? Or must Skype define a new "Internet dialtone" that acknowledges the new reality?
Skype hedges their dialtone promise with cautions.
"Skype is not a replacement for your ordinary telephone and Can't be used for emergency calling."
It's fine print.
- Do people read it?
- Do they believe it?
- Is this enough to set expectations appropriately?
- What should users learn about Skype and this service category from the outage?
Brands are promises too.
What has this breach, this 48 hour breach, this one thousandth of Skype's life, what has this breach done to Skype's brand?
skype
Comments
Dialtone - a promise?:
Dialtone is a localized loop of copper wire where voltage is applied and a tone is generated by inducing noise. This concept was used to signal end users that a connection to the local telephone switch was active. However soothing, there are no promises about what happens on the other end of the local switch, where the trunks to the Public Telephone Network (PSTN) are located. Mother's Day calling behaviors always proved this to me when I picked up the phone to call on that day. Dialtone was present but the PSTN had been engineered with insufficient bandwidth to process all of those calls.
A similar condition exists when the internet model is used for real time protocols. These voice and video streams must be delivered end to end (from my house to yours) in order for any service to fucntion over this shared resource.
It seems to me as though the challenge is to preserve necessary bandwidth for real time protocols as well as improve both local connections as well as wide area internets. The local connection needs to properly designed as well -- power backup (UPS is required) and redundant designs for core equipment (such as having two wireless APs) can go a long way to doing the right things within your control.
It's a calculated risk to use a shared internet for services which are either assumed or are implied as reliable. Network engineering can reduce the instance of failure, but we're always learning about how to make a better network. I'm hopeful Skype is following a similar paradigm.
Posted by: PhilC | August 22, 2007 06:24 AM