Download / Listen:

Herding Code 143 – Paul Stack on Continuous Delivery

• K. Scott asks Paul for a description of what Continuous Delivery is.
• Paul talks about the differences between Continuous Testing, Continuous Integration, Continuous Delivery and Continuous Deployment.
• Scott brings up the difficulties that databases bring about as it relates to deployments.
• Paul talks about how a distributed cache layer has helped alleviate those deployment problems for him.
• Jon asks how these processes lead software to have a better architecture.
• K. Scott asks how source control plays into the process.
• Paul talks about feature switching, which can be used to turn features on and off.
• Paul talks about the benefits of rapid feedback.
• There is a question from Twitter about databases and data warehousing and PowerShell equivalents of Chef and Puppet.
• Another question from related to the processes related to deploying with TFS.
• Scott talks about his experience prototyping Octopus Deploy.
• Kevin brings up the topic of rollbacks.
• Paul talks about A/B testing and canary testing.
• K. Scott asks about pushback on Continuous Delivery.
• Jon asks if it is possible to incrementally work towards Continuous Delivery.
• K. Scott asks about the pitfalls of Continuous Delivery.
• Jon asks if there are teams or situations that Continuous Delivery would not work.
• K. Scott asks about benchmarking of deployments.
• Another question from twitter related to Continuous Deployment of desktop software.
• Paul gives examples of technology making Continuous Delivery easier.
• Kevin asks if decisions made at the beginning of a project make it harder to implement Continuous Delivery.

Show Links:

• Paul Stack blog
• @stack72 
• SQL Compare
• Team City
• Feature switch library – NFeature
• Flickr feature toggles
• Chef
• Puppet
• Octopus deploy

Show notes by @rossfuhrman – thanks!

Download / Listen:

Herding Code 142 – Scott Guthrie on the ASP.NET MVC, Web API, and Razor Open Source Announcement

Show Notes:

• K Scott asks Scott Guthrie about the recent open source announcements about ASP.NET MVC, Web API, and Razor. Scott G. details what’s changed, highlighting both the new transparency due to working in public repositories and the changes to accept external code submissions.
• Scott G. explains that these are still supported products with dedicated Microsoft engineering investment.
• K Scott mentions a question from twitter – How far will this go? Will we see other products and projects following a similar model?
• Some more questions from twitter – What kind of feedback is Scott G. hoping to see? How will feedback be handled?
• Scott G. talks about how pull requests will be implemented. Developers will have to fill out a form and after they’re on record pulls will reviewed for various factors and then be integrated.
• Scott K. asks a about ownership of the code contributed to the project. Is OuterCurve involved?
• Jon asks a question from twitter: Was this the goal from the beginning? What was the biggest hurdle – legal or logistics. Scott G. mentions the community response to including jQuery in the ASP.NET MVC Project Template.
• Kevin asks Scott about patch contributing that has some performance issues and is the patch rejected or are the issues fixed. Scott G. thinks that minor issues in code might still be accepted or just asked to be fixed. A patch that does noting but "Format C Drive" will be rejected outright, other than that the process is pretty flexible.
• ASP.NET MVC 4 is not taking new features on as it’s currently in a Release Candidate mode.
• K Scott asks how Microsoft chose git as it’s source control. Microsoft sure has made a lot of OSS developers happy using git.
• Kevin points out that the Windows Azure SDK’s are on GitHub and asks why the ASP.NET components weren’t put there as well. Are there plans to move those projects from GitHub to CodePlex now? Scott G. says that CodePlex didn’t support git when the Azure SDK’s were released. With the announcement of git on CodePlex they’ve made decisions based on where they fit best, and ASP.NET content had historically been on CodePlex. There aren’t any plans to move from one to the other, and the beauty of DVCS is that they can be worked on in either place..
• Scott K. talks about how the team received feedback from blogs and mailing lists and now CodePlex discussions and asks about whether feature/roadmap discussions will be public or not.
• Jon asks about how release versions vs. nightly code builds will be supported from Microsoft. Scott G. explains that support will still attempt to help, but of course a released version is recommended for production scenarios.  He also reminds that product support will help with any .NET support scenario, including ASP.NET open source code.
• Jon asked a about how this will affect the Mono project, and Scott G. hopes that it does.
• Scott K. asks about how the release schedule might change now that the projects are open source.
• Scott G. talks about how, by going open source, customer feedback can be potentially received in real time which hopefully increases product quality.
• Kevin asks about community contributions of major new features. Scott G. says it’ll be a learning process, but they’re hoping to see some great ideas from the community. He discusses how Microsoft’s been incorporating open source libraries for a while, including JSON.NET, jQuery, Modernizr, etc., so now there’s flexibility to incorporate features both as core code and as external libraries.
• K Scott says Microsoft has been doing a great job incorporating community projects into their products rather than reinventing the wheel each time.
• Scott K. asks how Microsoft decides to create new projects or use existing solutions from the OSS community. Scott G. says it’s important to keep the MVC core concepts simple while allowing for advanced scenarios, and he and Scott K. discuss the balance between keeping concept count and clutter low while including support for popular scenarios.
• Scott G. mentions that he hopes the new open source view gets Microsoft feedback sooner so that changes can be made faster to final releases which will translate to better products.
• K Scott asks about what’s new in the world of Windows Azure. Scott G. clarifies that his new role includes ASP.NET and the web stack as well, and says there’s a lot of exciting stuff in the works for Azure. Scott G. says he’d like to come back on Herding Code to talk about it when it’s released.

Show Links:

• Scott Guthrie – @scottgu
• Scott’s blog post: ASP.NET MVC, Web API, Razor and Open Source

Show notes by @buildstarted – thanks!

Download / Listen:

Herding Code 140 – Lightning Round with Hadi Hariri

Show Notes:

• K Scott asks Hadi about EasyHttp.
• Hadi explains some of the problems and annoyances EasyHttp solves.
• Configuring the web request was a pain
• It was annoying to change the data structure of the classes to work with dynamic JSON
• EasyHttp supports all HTTP verbs (including PATCH and OPTIONS)
• It works really smoothly with dynamics
• Hadi talks about a library he used called JsonFx, which he prefers to JSON.NET
• Hadi says he decided to build out EasyHttp due to his work on YouTrackSharp
• That’s it! Scott runs off to tackle some more lightning round interviews.
• K Scott asks about EasyMVC
• Hadi talks why he created EasyMVC, explaining how the convention based routing works.
• EasyMVC also includes some filters which use conventions to handle content negotiation.
• K Scott asks what Hadi thinks about ASP.NET Web API. Hadi says he dislikes it, as it pushes developers towards having separate controllers for HTML and services.
• K Scott wraps up by asking Hadi about Kotlin.
• Kotlin is a new language in the JetBrains early access program.
• Kotlin attempts to improve on some of the shortcoming that JetBrains has seen working with Java.
• Kotlin is perhaps conceptually similar to Scala, but a lot easy to learn
• Object oriented with nullable types.
• Kotlin targets both the JVM and JavaScript
• It’s open source, and in early alpha phase.
• That’s it! Hope you liked K Scott’s lightning round interviews!

Show Links:

• Hadi Hariri – @hhariri
• EasyHttp project
• JsonFx project
• YouTrackSharp project
• EasyMVC project
• Blog post: With HTTP, your application is your API
• Blog post: The Kotlin Journey Part I : Getting things set up
• Kotlin project

Download / Listen:

Herding Code 140 – Lightning Round with Morten Kromberg

Show Notes:

• Morten describes the history and purpose of APL.
  • Did you know that APL stands for "A Programming Language"?
  • The first book about APL was written in 1962. It’s as old as COBOL and FORTRAN.
  • APL was first written as a mathematical notation, and was used in teaching for 4 years before an interpreter was even written.
• It’s a dynamic, interpreted language, and it encourages a functional style. Dyalog APL, which branched off about 30 years ago, is even more functional.
• People are still developing with APL, especially within the financial sector.
• Morten’s company makes APL interpreters.
• K Scott asks about the ability to work with APL from .NET. Morten says that, while it’s not a managed language, they have full interop so you can both create .NET classes in APL and consume them. You can use the GUI features in Dyalog APL, or you can interop with Windows / .NET GUI’s, and it can even be used as an ASP.NET scripting language.
• K Scott asks why such an interesting language that’s been around for so long isn’t well known. Morten speculates on some reasons and talks about why he thinks it’s seeing a resurgence.
• Morten and K Scott talk about some of the examples which really show off the language, such as a one line implementation of Conway’s Game Of Life.
• Morten talks about how APL sees matrices as a fundamental concept, expressing them at a level higher than objects.
• APL is a very agile language, as it encourages direct interaction from domain experts.
• Morten recommends http://tryapl.org, an interactive website where you can learn more about APL.
• That’s it! Scott runs off to tackle some more lightning round interviews.

Show Links:

• Morten Kromberg – @mkromberg
• Dyalog APL
• APL on Wikipedia
• Conway’s (very concise) Life
• TryAPL

Download / Listen:

Herding Code 139 – Lightning Round with Roy Osherove

Show Notes:

• K Scott asks Roy about his new book.
• Roy talks about the LeanPub approach.
• Roy describes the two parts of the book – elastic leadership and community contributions featuring notes from team leaders.
• K Scott says that our industry often throws developers into leadership positions, and the transition can be difficult. Roy agrees, and says these are the notes he’d wished he had when he was a new leader. Back then he thought he was doing a good job, and was having fun, but wasn’t really doing his job.
• Roy talks about how he got started, explaining how his passion is at the crossroads of where people and software meet. This book is coming out because it needs to come out – it’s a missing book.
• Different team phases require different leadership types, describing his elastic leadership approach that deals with the three phases he’s observed:
  • Chaos phase
  • Learning phase
  • Self organizing team
• What are some common mistakes?
  • Not recognizing which phase your team is in
  • Being afraid of talking to people about difficult things
  • Not understanding how to influence behavior
• That’s it! Scott runs off to tackle some more lightning round interviews.

Show Links:

• Roy Osherove – @royosherove
• 5whys blog
• http://leanpub.com/teamleader

Download / Listen:

Herding Code 138 – Douglas Crockford

Show Notes:

• K Scott asks Douglas what he meant when he said that the human brain wasn’t designed for this sort of work? What were we designed for? Douglas talks about how our minds are selected for hunting and gathering, but we have to work with what we’ve got, relating this to some tips for defensive programming.
• K Scott asks about JSLint. Douglas talks about how he thinks all languages should have code quality tools.
• The increment operator came over from C++ due to pointer operations. Bad idea?
• K Scott asks where he sees JavaScript going. Are things moving too fast, or too slow? Douglas talks about the difficulty of supporting multiple browsers and versions. EcmaScript 5didn’t break things, but EcmaScript 6 is making some bets to allow for that.
• EcmaScript 5 introduced "use strict" – how does that work? Douglas talks about the use of useless expressions.
• K Scott asks what developers should be doing today. Douglas says developers should be working in the intersection of EcmaScript 3 and the strict parts of EcmaScript 5 to be ready for EcmaScript 6.
• That’s it! Scott runs off to tackle some more lightning round interviews.

Show Links:

• Douglas Crockford
• JavaScript, The Good Parts
• JSLint
• EcmaScript 5
• Strict Mode Is Coming To Town (YUI Blog post by Douglas Crockford)
• ECMAScript 5: The New Parts

Download / Listen:

Herding Code 137: Mass Assignment, New New iPad, JavaScript libraries, Windows 8, Visual Studio, and Sad Trombones

Show Notes:

• K Scott asks everyone’s opinions on the GitHub / Ruby on Rails “mass assignment” debacle.
  • Everyone talks about mass assignment binding issues in MVC frameworks, including Rails and ASP.NET MVC – is this a security issue in the frameworks, or the web developer’s responsibility?
  • Jon says that it’s often tricky to debug negative cases, Kevin says that everything should be secure by default, and Scott K can go either way on it.
• K Scott asks us all who will buy the new new iPad.
  • Scott K says it struck him that they were limited in demonstrating it by the projection resolution.
  • Kevin wants one.
  • Jon thought it was interesting that Retina isn’t a hard DPI spec, it’s driven by the expected distance the device will be from your eyes – can he get a Retina effect by just sitting far from his desktop monitors?
  • There’s a discussion about the lack of a version number.
  • Jon wonders if that high quality of display will show up on other devices, or if Apple bought all the pixels. Oh, and patents.
  • K Scott asks Kevin if Samsung users laugh at him.
  • Jon says that’s no longer a issue now that software updates brought 4G to this iPhone… magic!
• Twitter question: What JavaScript libraries is everyone using?
  • Kevin talks about the Mocha JavaScript test framework.
  • Jon mentions Upshot from the ASP.NET Single Page Application framework.
  • K Scott talks about Sylvester and Zoomooz.
  • Scott K talks about tiny libraries like Zepto, Ender, and the Micro.js list.
  • Jon says he likes cdnjs.com for JavaScript library hosting.
• Scott K talks about the difficulty he had in shutting down Windows 8.
  • Jon says it’s all about search now… and what’s wrong with hitting the start button to power off?
  • Scott K says we’ve been trained for decades not to do that.
  • Jon says this is the biggest shift since Windows 95… there’s a lot of learning and unlearning to do.
• What about Visual Studio 11?
  • There’s some discussion about the color and design aspects. Should there be color? Metro?
  • Jon says at least it’s a lot faster, and he likes the quick search.
  • Scott K said it worked fine once he figured out what it was for… and maybe there should be fewer items in the menu to begin with.
  • Jon throws out a crazy idea – what about the ribbon interface for Visual Studio and kind of convinces Kevin.
  • Scott K says the memory usage is still way too high. Jon asks if that really matters. After some discussion, Scott K says something’s slow in there.
  • Oh, hey, the macro recorder’s gone now. Jon actually used that in Visual Studio recently.
  • Kevin asks if there’s really no way to customize the install anymore. The gang all agree they don’t want to install stuff like C++ and VSTO. Jon says the blog post says that few people actually customize the installation, but Scott K doesn’t believe it.
  • Scott K runs through some fun issues on the Visual Studio UserVoice.
  • Kevin put up a bajillion issues on Connect long ago, most are Closed – Won’t Fix.
• Jon talks about a post about 24 bit / 192 khz audio he read. Nobody seems to care all that much.
• Jon asks everyone what they’ve been up to.
  • K Scott’s travelling around and working on project that’s Ruby / Mongo on the backend and ASP.NET MVC on the front end. This freaks Kevin out.
  • Jon’s been working on ASP.NET MVC / Web API release stuff and hacking on Code52 project late at night.
  • Scott K spoke recently at NodePDX on is doing a bunch of crazy stuff at work around deployment.
  • Kevin remembers what a DSN is when nobody else does.
• Somebody sneaks in a Sad Trombone. Jon is unable to figure out who is playing tricks and motions to adjourn.

Show Links:

• GitHub posts: Public Key Security Vulnerability and Mitigation and Responsible Disclosure Policy
• Brad Wilson post: Input Validation vs. Model Validation in ASP.NET MVC
• Mocha – JavaScript test framework
• Sylvester – Vector and Matrix math for JavaScript
• Zoomooz – an easy-to-use jQuery plugin for making any web page element zoom
• Zepto – a minimalist JavaScript framework for modern web browsers*, with a jQuery-compatible syntax
• Ender – The no-library library
• microjs – tiny JavaScript libraries
• cdnjs.com – The missing CDN (for all the other JavaScript libraries)
• Windows 8 blog post – Evolving the Start menu
• Visual Studio 11 Beta Performance blog post series
• Visual Studio UserVoice
• Kevin Dente’s Visual Studio Connect Bug list
• 24/192 Music Downloads …and why they make no sense
• Scott K’s talk at NodePDX – info / video

Download / Listen:

Herding Code 136: Code52 with Paul Jenkins, Brendan Forster, and Andrew Tobin

Show Notes:

• Jon starts by asking how this whole idea got started.
• Brendan explains the timeline of a one week spin up cycle for a project.
• Jon asks if the projects just die at the end of a week, and if people are continuing to work on past projects. Paul and Andrew talk about the continuing work on all projects, including MarkPad .
• Jon says he thinks the concern of abandoned open source projects is overblown – collaboration is good, working
• Jon asks about the projects they’ve done so far, and the guys run through the list.
• Brendan explains how the different projects are selected.
• Jon asks about how technologies are selected. Brendan says that it’s all based on who shows up, and Andrew mentions the wide variety of technologies that have been covered so far.
• Paul says that the "bite sized projects" have been a good way for developers who are new to open source to get started. Jon asks how new developers can get ramped up, and Brendan lists some of the onboarding resources. Andrew explains that it’s hard to write much documentation when the projects are running for a week. Andrew mentions some of the Git documentation they’ve got written up.
• Scott K says he’s got a ton of personal projects that he doesn’t have time for. Everyone talks about the benefit of getting a team on a project for a week.
• Jon asks about some of the common frameworks that they use over and over. Paul talks about MahApps.
• Scott K asks about non-CLR project, like some crazy project in obscure languages. Brendan says the main focus is on building something useful in a week, which usually leads towards common languages. Brendan says that projects in other languages really need a "champion" to show up and lead the project for a week – and the door’s open.
• Jon asks about the community reaction overall. The guys talk about the wide attention, as well as some negative reactions from onlookers who want to see other technologies represented.
• Jon says he thinks there was some pent up need for energy in the .NET open source space, and asks if that was part of the reason for getting this started.
• Jon asks Paul about the "Mah" name he’s used for his open source projects.
• There’s a discussion of developing while hungry, including the Pretzel project name and the "Jon should make me a delicious cake" incident.
• Andrew talks about the GTFO project – GitHub Tools For Outlook
• Jon asks about the .NET developer community in Australia. Is it  as huge as it seems?
• Question from Twitter – @wolfbyte asks: "How do you balance the shifting of tools / ideas / processes against the goal of attracting people to open source development.
• Another question from @wolfbyte: "Are you guys tired yet?" (Yes!)
• Jon asks if there’s thought as to how to scale things going forward to all the work doesn’t fall on Paul/Brendan/Andrew for all projects. Is there a possibility to bring in guest leads for a week?
• There’s a discussion of HattersGonnaHat and the Konami code. Jon announces a new KDD movement: Konami Driven Development.
• Some discussion of Windows 8 development (update: the 3/5 – 3/12 project is Windows 8).
• Would it be possible to do more cross-platform work? Silverlight?
• What will the future projects be? Jon campaigns (in vain) for his Diff/Merge 2000 project proposal.
• Jon asks for more information about how they run all the behind-the-scenes communications and infrastructure. GitHub, App Harbor, JabbR – all are low friction and low / no cost.
• Jon asks about the amount of work involved in accepting pull requests.
• Brendan talks about the test coverage in code submissions. Paul says that having testing frameworks in place makes it a lot more likely that people will write the tests.
• Brendan wraps with a call to look at Code52.org and get started. Jon pledges to fly to Australia in luggage class, then offends everyone by confusing Australia with New Zealand.

Show Links:

• Code52 site (@code_52)
• Code52 JabbR room
• Andrew Tobin – blog, twitter
• Brendan Forster – blog, twitter
• Paul Jenkins – blog, twitter
• Code52 projects to date:
• MarkPad – A Markdown editor for
• jibbr – A Jabbr bot designed for collaborative
• Ideastrike – A collaborative idea voting site
• Samurai – A tiled, turn-based player-vs-plaryer game with an MVC 4 backend and the beginning of Windows Phone and iOS front ends
• Pretzel – A static site generator with Markdown and Liquid (and Razor on the way)
• metro.css – A CSS bootstrap package to simplify building web applications with a Metro look
• Carnac – A utility that displays keyboard shortcuts as you type them
• GT4O – An addin for managing Github tasks inside Outlook.
• i18n Packages for MVC4 – NuGet packages to simplify internationalization in ASP.NET MVC applications
• MyFinances for Windows 8 – A Javascript-based application for WinRT
• GitHub post – A New Coding Project Every Week at Code52
• #HattersGonnaHat
• Konami code
• MahTweets
• MahApps

Download / Listen:

Herding Code 135: Remco Mulder and Jeff Schumacher on Continuous Testing

Show Notes:

• Scott K kicks things off with a horrible old school BASIC joke.
• Remco explains how NCrunch got started because he was living under a rock.
• Jeff explains how Giles got started by finding AutoTest and seeing it not maintained (in reality it was).
• Remco talks about the number of test frameworks.
• Jon asks Remco about adding MSTest support – Remco groans about how difficult MSTest was to integrate with NCrunch.
• Jeff mentions that xUnit’s test runner is complete for backward compatibility.
• Jeff talks about Machine Specs and avoiding versioning hell with Impromptu Interface.
• Jon asks about each library’s approach to Continuous Testing and to explain it to a laymen.
• Rem explains what Continuous Testing is in relation to NCrunch.
• Scott comments on a "strange situation" of testing compiled code creates slow tests and asks about any shortcuts Rem or Jeff have taken.
• Jon comments how NCruch & Giles gives immediate feedback as he’s coding.
• Jeff mentions that Giles has the ability to filter tests so that you only see what you want to see.
• Remco explains how NCrunch attempts to determine impact to order how tests are run so the most pertinent tests run first.
• Jon gives his quick history of the testing feedback cycle: separate project phase, then repository check-in step with continuous integration, local test runers, and now tests running as we write the code. Are we done? Where can we go from here?
• Jeff comments about how Continuous Testing is like the red squiggly for code problems.
• Scott asks about alternative language support for NCrunch and Giles.
• Remco mentions that Salesforce.com has a Selenium "cluster" to continuously run UI tests.
• Remco talks about Visual Studio integration for NCrunch.
• Jeff talks about a branch that supports Mono.
• Jon talks about Roslyn, and Scott K reminds us that Mono was doing compiler as a service long ago – 2008?
• Remco jokes about clippy.
• Jeff talks about Continuous testing and Pair Programming.
• Jeff wanted to get to the point where he didn’t think about the tooling but only the tests and the code.
• Jeff talks about Visual Studio magazine and an article about Continuous Testing.
• Jon comments how easy NCrunch is able to setup and get going.
• Remco wants to make Continuous testing really really easy to get people to start using it.
• Jeff recommends looking at all the different continuous testing options, mentioning Greg Young’s Mighty Moose.

Show Links:

• Jeff Schumacher – blog, twitter
• Remco Mulder – blog, twitter
• NCrunch
• Giles
• Mighty Moose from Greg Young
• Continuous Testing: Think Different [Visual Studio Magazine online]
• Project Roslyn

Show notes by Ben Dornis. Thanks!

Download / Listen:

Herding Code 135: Remco Mulder and Jeff Schumacher on Continuous Testing

Show Notes:

• Scott K kicks things off with a horrible old school BASIC joke.
• Remco explains how NCrunch got started because he was living under a rock.
• Jeff explains how Giles got started by finding AutoTest and seeing it not maintained (in reality it was).
• Remco talks about the number of test frameworks.
• Jon asks Remco about adding MSTest support – Remco groans about how difficult MSTest was to integrate with NCrunch.
• Jeff mentions that xUnit’s test runner is complete for backward compatibility.
• Jeff talks about Machine Specs and avoiding versioning hell with Impromptu Interface.
• Jon asks about each library’s approach to Continuous Testing and to explain it to a laymen.
• Rem explains what Continuous Testing is in relation to NCrunch.
• Scott comments on a "strange situation" of testing compiled code creates slow tests and asks about any shortcuts Rem or Jeff have taken.
• Jon comments how NCruch & Giles gives immediate feedback as he’s coding.
• Jeff mentions that Giles has the ability to filter tests so that you only see what you want to see.
• Remco explains how NCrunch attempts to determine impact to order how tests are run so the most pertinent tests run first.
• Jon gives his quick history of the testing feedback cycle: separate project phase, then repository check-in step with continuous integration, local test runers, and now tests running as we write the code. Are we done? Where can we go from here?
• Jeff comments about how Continuous Testing is like the red squiggly for code problems.
• Scott asks about alternative language support for NCrunch and Giles.
• Remco mentions that Salesforce.com has a Selenium "cluster" to continuously run UI tests.
• Remco talks about Visual Studio integration for NCrunch.
• Jeff talks about a branch that supports Mono.
• Jon talks about Roslyn, and Scott K reminds us that Mono was doing compiler as a service long ago – 2008?
• Remco jokes about clippy.
• Jeff talks about Continuous testing and Pair Programming.
• Jeff wanted to get to the point where he didn’t think about the tooling but only the tests and the code.
• Jeff talks about Visual Studio magazine and an article about Continuous Testing.
• Jon comments how easy NCrunch is able to setup and get going.
• Remco wants to make Continuous testing really really easy to get people to start using it.
• Jeff recommends looking at all the different continuous testing options, mentioning Greg Young’s Mighty Moose.

Show Links:

• Jeff Schumacher – blog, twitter
• Remco Mulder – blog, twitter
• NCrunch
• Giles
• Mighty Moose from Greg Young
• Continuous Testing: Think Different [Visual Studio Magazine online]
• Project Roslyn

Show notes by Ben Dornis. Thanks!