no dogma podcast

Summary Laurent Bossavit talks about the myths like the 10x developer that have grown in the software industry. Details Who he is, what he does. His book - "The Leprechauns of Software Engineering", why he wrote it. The 10x developer, literary archeology. The telephone game, examples in the software world, cost of when defects are discovered. Industry does not have interest exposing faults, why is the word "belief" used in software, is software an engineering discipline, opinions over measurements, how did we end up with manifestos. What should we measure when judging software quality, why measuring bugs are like eating from the garbage can. How to make things better. How to get Laurent's book. Laurent's book recommendations.

Summary Amy Sterling and Chris Jordan of EyeWire talk about mapping the neurons and synapses of the brain. Details Who they are, what is EyeWire, how it started. About the brain, 80 billion neurons in a human brain, 100 trillion synapses. It used to take 1000 hours to map a neuron now it takes 80 hours. 250,000 users from around the world. Combined effort of players and AI. EyeWire is focusing on 1 cubic mm of a brain which has a 100,000 neurons and billion synapses. Where EyeWire's data goes after mapping. Why are they building EyeWire, to learn why we are the way we are, we don't know how many types of cell are in the brain. The EyeWire tech stack. Building a community, media engagement, internships, competitions. Moving towards open source for parts of EyeWire. How they make money, or not! Future work, IARPA. Joining EyeWire or other citizen science projects, World VR forum, Games for Change.

Summary David Mead of Start With Why talks about improving communication skills, leadership and handling conflict. Details Simon Sinek and start with why, David's own background. Start with why, golden circle. What we do, how we do it, why we do it; without why it is much harder to differentiate ourselves. Most companies start with what, examples of companies that start with why. People like to be around people like them and believe what they believe. Imperfect companies can have a nobel higher goal. How can engineers improve their communication; it's a skill that can be learned; give people tasks and roles that inspire. Better communication across the whole organization, sharing the big picture. Simple tips to improve communication skills, set goals that are attainable with low risk. How to handle conflict, don't take a position against something, stand for something. Conflict as a useful tool to resolve issues. What to do if conflict has become the norm, get back to the why; what to do when the why is not enough, "we can't fix people, we can provide the environment where they can be inspired to change themselves". Don't promote because of skill; when leading a team you are responsible for the team not the job, "management is about getting stuff done, leadership is about people". Always keep an eye on the bigger picture.

Summary Patrick Smacchia creator of NDepend explains how this tool can improve the quality of your code. Details Who he is, what he does. Why he is interested in code quality. NDepend is 10 years old. Transitioning from free to commercial. What is static analysis. Comparing NDepend to other tools, Roslyn analyzer. Finding spaghetti code, all rules are linq queries. Measuring technical debt, estimating the cost of fixing the code vs leaving it alone. Call graphs, dependency matrix, tree map, code coverage. Visual Studio Team Services plugin, quality gates, comparing code coverage per release, testability and maintainability. NDepend can analyze dlls, it looks at the intermediate language. Patrick loves the book "CLR via C#", Bryan talks about the time Jeffrey Richter stared him down. Future work. Getting a free trial. How to really pronounce Smacchia.

Summary Bill Wagner discusses the new Microsoft documentation service, a new way of learning about Microsoft's development offerings. Details Who he is, what he does, upcoming talks in Portland, Vermont, Boston and Sydney. The new .NET documentation project, why they are doing it, reorganizing the docs to help solve problems. New docs give more context, e.g. thread safety, advice on usage. Picking what to write about. Open to user contributions. Who keeps the docs up to date, internal pull requests; third party tools and platforms. Documentation for developers with non .NET backgrounds. Monitoring traffic to docs. Docs as a compliment to stack overflow. Bryan complains about lack of full samples - Bill talks about very a large example eShopOnContainers. Walkthroughs. How to request new docs. Why some of C# and .NET is not open source, process of open sourcing. Updates to Bill's books, invite to help with docs.

Summary Eric Schles talks about a set of tools he is building to identify and filter fake news stories. Details Who he is, a story on human trafficking. Importance of identifying "fake news". News Literacy Project, how Eric got involved. Manually categorizing news stories. Building software to the job, metrics to identify "fake news", stitch fix, word distance map, comparing to the manual process. Eric loves python, scraping. Other applications, machine generating long form content, "machines writing books". Providing an API. Scaling to handle large volumes; python 3.6; Asyncio and Kafka. Bias in the software; yellow journalism in 1890s and 1920s. How to help.

Summary Dylan Reisenberger talks about Polly, a resilience and transient-fault-handling library for .NET. Commonly used for retries, circuit breaking and fallback when calling remote services. Details Who he is. Quick overview of Polly, why do I need Polly - the network is not reliable. History of the Polly project. How popular it is. What a resilience framework is. Retries in Polly; backoff; doing other things during the retry. Policies, what they are. Handling exceptions and result codes. Circuit breaker; what it is; why use them. Using policies together, wrapping. Stability patterns, bulkhead isolation. Queues. How to execute a web request with Polly. Using Polly for things other than web requests. Re-authorization of requests. No .NET alternatives. Future work, caching, policy registry, metrics, reactive extensions. How to help.

Summary Ben Day, Pluralsight author and consultant talks about dev ops in the Microsoft world and how to introduce it in your organization. Details Dev ops will solve everything. definition is hard to pin down. Three questions, 1) how long from checkin to deployment, 2) what are the steps to get code deployed, 3) how much time is spent on production support issues. Why do we need dev ops. Who takes on the role of dev ops. What Microsoft offers. All the way from local dev to release. Do dev teams get dev ops members. People don't like change. Dev ops "levels of awesomeness". Seeing it really work. Continuous release with Microsoft, Ben's Pluralsight course, how quickly can we move code from dev to production.

Summary Rachel Roumeliotis of O'Reilly Media spoke to me about technology and development trends for 2017. Details Who she is and what she does. Upcoming conferences, OSCON and Fluent. Rachel and I discuss tech trends for 2017: open source, the big players, can every company do it? Code is not the only value, customer lock-in. "All businesses are software businesses", how common is that perception, is dev over valued sometimes. "Infrastructure changes", very hard to keep up, big companies telling small companies that they are doing things wrong. "The year of AI", again; AI silos, no overarching system. Keeping the customer in mind when working with tech.

Summary Michael Biercuk, director of the Quantum Control Laboratory at the University of Sydney talks to me about quantum computing and the future it will lead to. Details Who he is, what he does. Quick overview of quantum computing. How traditional computers work, transistors, charge etc. Moore's law, transistor size, nanometer size, tunneling. When quantum effects start to cause problems. What problems can only quantum computing solve; quantum supremacy. Can quantum computing crack ssl certs; decoherence is the big problem and how to delay it; finding a catalyst for the Haber process. Why is quantum computing faster. Programming a quantum computer. Bits, qbits and 1 & 0 at same time ; if...else with qbits. Current state of the art, academic, industrial and small commercial/startup. What will unlocking quantum computing do for us; computing is advancing every field; if we get to 300 qbits! Michael thinks harnessing quantum computing will transform society.