We have been working around the clock to figure out why this happened. Of course, one assumption we made was to assume that GoogleVoice is not the issue.

Well we were wrong. We now believe that GoogleVoice has a security hole, and is the root cause behind why we sent texts to the wrong person’s.

There is an extremely simple to test to prove this:

1. Open a firefox tab and login to google voice with an account… account A.
2. Open another firefox tab and login to googlevoice. Log out of google voice and log into account B.
3. Go back to the tab where account A is open.
4. Hit “text” on the left hand side menu for account A.
5. Watch as you now see the texts from account B in your GoogleVoice session for account A.

We at Peek are webscraping GoogleVoice’s page. We do exactly the programmatic equivalent of the above. And thats why we sent GoogleVoice messages to the wrong user’s, when we asked for the messages for account A, we got the messages for Account B.

In the world of Spring 2.0, I was really torn about depency injection. My thought was that it actually substantially decreased code readibility. Reading Spring 2.0, felt like reading Hamlet… a big piece of Java code in one hand, with the xml config Coles Notes in the other hand, staring back and forth to really get what was going on. It was painful. However, the level of OO and factoring was much stronger. The pure coding part of maintenance was much easier, once you were up to speed.

Spring 2.5 actually felt like a step backwards. Annotations were introduced, but not quite with the robustness and depth I wanted. There several weird situations where I had to “break” the Spring MVC model – my two biggest complaints were Ajax and accessing the application through things other than the web (for instance a mobile device). So I then ended up with code that was half-autowired/half-configured using weird ways to get at the application contexts.

I was starting to get frustrated. You see, I regularily work on about 5 code bases, 4 are java:

-Two of the java projects are classic J2EE/POJO – the code is uglier, factoring & oo not as good… but I never lose days figuring out the underlying technology. But making changes to the code base can be god-awful, and it can be very tough designing code for testability in this world.

-The other two are Spring – the code is pretty, factored beautifully, Martin Fowler would be proud. But I guesstimate I lose a full day a month in the weeds of Spring trying to figure out “how” to do something.

Losing days to framework’s hurts, its far more frustrating than losing days to bugs or new libraries. Framework’s are supposed to help not hurt!

So, begrudgingly I moved up to Spring 3.0. So far… wow! Much simpler, its starting to feel like what I wanted from a framework. The explosion in annotations feels good. Basically, the increased reliance on convention over configuration and annotations are making me a believer in dependency injection again. I can code Spring 3.0 and barely ever touch a config file. The code base feels drastically simpler.

Next up, I’ve been meaning to try Spring Roo for a while. Anybody tried it out there?

In a world of philosophical theory, I basically see the articles drawn down to this contrast: “BlackBerry vs the State” and “Google vs the Individual”.

BlackBerry is actually taking a relatively heroic stance on individual rights. They are basically saying:
-the common person should have access to the most advanced encryption standard out there
-The common person’s rights to private communications trumps that of the state’s right to read your private communications
(sort of… they are saying “give us warrants and we’ll give you their communications fresh from our server’s in Canada where you have no jurisdiction, you can’t just read the data pipe as you see fit”)

Its almost a quasi 2nd amendment-ish view of life – why should we trust the government with our communications? The government doesn’t have rights to my vocal conversations with people, why should they have rights to my electronic conversations with people.

Google on the other hand is saying “you have no right to your data/communications, it belongs to the public”. All data should be publically available for scrutiny – by state, by corporation, by other individuals, etc.

These are massively different philosophical viewpoints! There needs to be a lot of public debate and standards put in place. I felt like this was addressed a few years ago with lots of countries rolling out new “privacy frameworks”, but it still seems that wasn’t sufficient.

The funny thing is that we may not need a government & standards body answer. The opensource community and the market can kind of answer this for us.

If I were a Google competitor (cough, cough Microsoft), I would make browser’s and OS’s that are anonymous and private by default. I’d scare the world with Google’s “data grab” and hit them where it hurts. If the Internet goes “anon” then Google will be in deep trouble. How can you have relevant, contextual advertisements if you know nothing about that person?

In fact, I think its high-time that Mozilla & Linux came with anonymity & privacy rights ON by default. Tor should be built into Ubuntu/Firefox/Thunderbird and used by default.

Android should be similar, http://guardianproject.info/ could be deployed by tons of ODMs as they launch new devices. They can give user’s a true option of anonymity & privacy. They don’t have to expose themselves on the Internet unless they do it deliberately.

Lastly, I want to talk about social networks. Privacy/anonymity seem irrelevant if all you are going to do is post on sites like Facebook, Twitter, etc… where you build up an online identity. Well, a few answers to that problem:

1. People have tackled the problem of anonymous “social networking/sharing” – drop.io & pastebin are two example services where this has been done.

2. Data can be temporal and not stored. Why should twitter store anything other than my past few days of tweets?

3. Social networks can opt-out of google searches. Why should you be able to find my Facebook profile when you search “Dan Morel” on google. I don’t want my private social network with my friends and family to be searchable by the public using Google.

4. The data can be encrypted end to end, only people in my network should be able to decode my updates. When I “add a friend” this could include some sort of behind the scenes private/public key swap. Why should anything I do or write be available to anybody other than user’s of my social network data?

Anonymity and privacy are not difficult problems to solve. Facebook & Twitter don’t have to participate in Google’s attempt to grab all of the world’s data and monopolize the monetization of it.

While the internet is regulated to net neutrality right now, the mobile internet is not at all.

Let me explain how a classic web startup on the plain old internet builds their technology:
1. Get an idea
2. Build your web app
3. Put it on the internet (use Amazon AWS, buy a server, use a hosting provider, etc, etc, etc, etc)
4. When you want to “expand internationally”, write your app in more languages

Ta da, you are launched. You have launched something that can be used by anybody in the world, on any computer. There were 0 delays due to “certification processes” and 0 certification costs and 0 business development so people can access your service. It is there.

In the world of mobile internet this is not the case. There are basically two paths to build an app.

Deal with the Device Makers:
1. Get an idea
2. Build an app
3. Nobody can access it
4. If its for a smartphone with mature OS/SDK ala Apple or google, apply for approval/launch. Some period of weeks later they launch it.
5. If its for something else… say a feature phone or LG, Nokia etc, its a long BD process.. months or years of proving yourself to get even a meeting. Think Seven or other mobile email providers and their business model. Then a certification process at the end…sigh.
6. That device will likely only work and be sold on one or two networks, so you can probably only reach 20-30% of your target market. And you might have to repeat the process for each country you want to launch in.

Or, you can instead deal with the carriers directly:
1. Get an idea
2. Build it
3. Nobody can buy it or use it
4. Go beg one of Verizon, AT&T, or T-Mobile to let you put their device on the network or your software on their phones.
5. They will probably say no
6. If its hardware – get PTCRB and FCC approval at a cost of probably $30k
7. If its software – go through a juggernaut product certification process of 4-6 months and possibly some cost
8. You will get bounced multiple times in the process when Google or somebody else “bigger” has something to get certified
9. Finally you launch, huzzah.
10. Every time you want to launch in a new country you have to go through the same process

Why is the world of mobile internet so different from the world of the internet?

1. Bandwidth is PRECIOUS. Verizon and everybody else in teleco knows that the mobile bandwidth curve will not explode like that of the fixed internet. Demand outpaces supply (talking to you here New York City). The mobile internet will continue to be expensive to access – $400 iPhone like devices paying $75/month for data access. Verizon et al want to keep it that way.

2. Its roots are teleco. Teleco-heads had to make products that always worked. Thats why the highest grade of service availability is ‘carrier-grade”. The culture and processes around mobile are around stability & availability… i.e. there are massive & long certification processes and they don’t like dealing with unproven start-ups.

3. Similar to above – control of sessions. Teleco technically always had to connect and control calls. They are used to being the controller. The internet never had this need. Power was pushed to the edges, everything was decentralized.

There really is no reason that the mobile internet cannot be like the fixed internet. There is no real reason that a wide gap should exist. Building services that use the mobile internet should be simple, the steps should feel like building a service for the fixed internet. And bold regulation/vision is needed to make this happen.

Google basically threw in the white flag and said… “nah, its easier to just partner with Verizon”. Which is very frustrating since Google and Apple have helped open the world of mobile to “apps” and new services. They need to continue that fight, not join hands with Verizon and quash the future of open mobile access.

We want to make a point loud and clear to Google, what you are doing in the Google-Verizon deal is plain wrong. Anything different from net neutrality for mobile networks is going to result in a lame duck mobile internet controlled and limited by two-three big powerful corporations. That is not how the Internet became what it is today and not how Google got to where it is now.

The Google-Verizon plan means building an internet where somebody like Verizon can simply “turn off” or “delay” services, like we are going to do tomorrow with our Google email fetching service.

We encourage other companies who offer services around Google to do similar and help send a message.

My big criticism. Anytime you talk about major mobile trends and don’t mention Nokia & GSM, its like talking about retail trends without talking about WalMart. Apple is predicting about 37M iPhones being sold this year, Nokia is expecting to sell 400M or so phones. Net total, Apple has sold roughly 50M iPhones. 50 million. I want to make sure that number is clear for when I juxtapose it against some numbers in a second.

Here is my list of some big, big, big things that happened in the past 5 years globally:

-on July 10th (two weeks ago), according to Ericsson, the 5 billionth GSM subscriber was added making it basically the most successful technology of all time, there are only 6.8B people in the world… and there are 5B GSM mobile subscribers!!! That is jaw-dropping.

-3 billion people now text… wow… who cares about social networking… 3B texters!!! I didn’t even think the global literacy rate was that high

-3G hit its 500M-th customer, so 10% of the world have “broadband in their pocket”… yawn. If you are building data services and not thinking about the 4.5 billion potential customers who run on GPRS… you are thinking pretty narrow-minded.

-Nokia makes the Nokia 1280, a $15 phone!!! $15! Why did you pay $400 for your iPhone again when 97% of what you do is texting and phone calls????

-BlackBerry sold roughly 100M units in the same timeframe iPhone sold 50M (they had a 10M unit headstart roughly)

-The combo of Texting + BlackBery + mobile email = mobile messaging explosion

-There are now about mobile social network users reaches 250M globally!! I actually agree with mashable that this is pretty damn big, and part of the above mobile messaging explosion.

The future, lower priced devices and lower rates. iPhone and Android get stuck at a few hundred million devices until they figure out how to make devices that work on 2G and cost far less. Somebody else figures that out and takes them out. Maybe/hopefully that somebody is Peek.

Your job is not entrepreneurial. I hate hearing the line – “Google is so entrepreneurial, its like a startup!”.

Real entrepreneurs have to worry about money and resources. If you work at Google, you do not.
Real entrepreneurs have to worry about awareness and marketing and doing it affordably and on a budget. Tech blogs broadcast every word from the mouth of Google.
Real entrepreneurs have to worry about culture, they can’t just hire a bunch of hyper-rational introverts with no friends, who are massively arrogant about their math skills…and cookie cutter follower zombies in the world of technology.
Real entrepreneurs have to buy their own lunch. They don’t get fat off of morning Googley donuts.
Real entrepreneurs have to worry about profitability and making customers happy, they can’t just invade the privacy of all their customers and launch buzz on them.

Honestly, for the life of me, I can’t fathom why a developer would go take a job at Google. Its like being an awesome carpenter and saying “I want to go work at Ikea” so I can churn out follower products like an open-source mobile OS 5 years after Symbian or web-based email 10 years after Hotmail or a webOS 3 years after Palm did it. Way to gooooo, you are such wonderful innovators, you are applying the Microsoft method of “do it several years after everybody else”. I’m sure they haven’t been burned by that, and I’m sure you won’t either.

Anyways, all this to say, if you want to try to do things differently, if you want actual, real entrepreneurial experience and a real start-up job… we need a killer dev at Peek. I dare you to get up off your comfy Googly job and email me dan at getpeek. You will work on some of the coolest tech in the business. We believe the fat world of HTML & javascript isn’t right for mobile, we want to make OSs, protocols, devices and apps that can be used across the whole world, not just by upper-middle class westerners who have broad 3G deployments and money to buy $400 devices.

We are looking for an experienced, strong developer to play a significant role in the Peek dev team.

The role is vast:
-mobile coding (app + embedded level)… this is done in C/C++
-Our Peekway servers that send, receive and store emails, social network, texts, etc. This is done in Java/J2EE (sometimes Spring)
-You will get to actually operate your code, and fix it when it breaks. We run on Amazon AWS on Linux, Apache, My SQL, JBOSS infrastructure
-All of our backend services… care interfaces, billing and so on… you’ll get your hands on that as well
-Crazy stand-alone apps like PeekSocial

We practice agile and devops (you will deploy/rollback/troubleshoot bugs in your own code), we deploy new features fast and furiously. You will have both massive discretion on your own projects, bugs, features that you implement as well as massive input into the future of what you develop on.

We are looking for a very, very, very good developer and person. What we need:

-we need a developer who has coded big, hard projects and made them successful
-languages aren’t crucial, we presume if you are awesome you can remember how to call malloc() and free() in C (if thinking about memory mgmt scares you, this isn’t the role for you) and we presume you can learn how to Autowire in Spring
-You need to have experience with scale… i.e. when you argue architecture, there should be experience backing those words, not just repeating Martin Fowler’s arguments (whoops, I just made fun of myself)
-You get things done… you are both an artist and a duct-tape programmer, you are passionate about both hitting dates and making everything follow a design pattern… you don’t do things like make ugly GUIs because “I’m not a GUI guy”
-You like the warm and fuzzy feeling when writing unit tests, they are like a cool glass of ice tea on a summer day to you
-You like “surprising” people with the new hack you thought of last night and wrote this morning
-When you troubleshoot you use science, metrics & fact. Not crazy logic. Especially not crazy logic about crazy race conditions.
-You are not an angry asshole, you like human beings and communicating with them in a non-aggressive/violent way… you like working in a team of really smart/energetic people and telling them about the human normal things you did on the weekend
-You care and have passion, if customers complain about something not working… it hurts deep inside. If there is an outage because of some code you broke, you think about pulling out your samurai sword and emailing a friend to be your second.

Peek launched nation-wide in the US late 2008. Since then we’ve launched successfully across Europe and in India, and picked up many awards along the way including Times Gadget of the Year and Wired Product of the Year. Our team are true start-up veterans – Dr. Amol Sarva played a key role as Finance Director in starting-up Virgin Mobile USA and our chairman, Dave Sprosty, was CEO of SpeakEasy, who were acquired by BestBuy where we became COO, and later CEO of BestBuy Mobile. Oh yeah and me, Dan Morel, I helped start a company called Redknee that went from 0 to IPO completely organically.

And we have cool purpose. We want to make the internet & data available to everybody globally. We want people in the rocky regions of Pakistan to be able to pull out $50 devices that let them communicate and get all the data they need. This means that we need to make devices that are affordable & use as little data as possible… but are still freakishly capable!!! Its a big, big challenge.

You will get to develop on all aspects of Peek – device, server, etc. We are Java in our world of servers and C/C++on the device. We call C++ and Java – Caca++.

-16 hour days
-probably get alarms 2-3x per night
-alarms and issues are really hard to solve, no real manual available, mostly trial and error
-on the job training only… no real prep or training available
-tons of advice and criticism but no real help
-massive physical and emotional stress… people’s lives are in your hands
-lots of your job is cleaning up other people’s messes
-if you do your job great, you can change the world… if you do your job really badly, you can cost society millions