Author Archives: jeremydmiller

Remembering Why Agile was a Big Deal

Earlier this year I recorded a podcast for Software Engineering Radio with Jeff Doolittle on Agile versus Waterfall Software Development where we discussed the vital differences between Agile development and traditional waterfall, and why those differences still matter. This post is the long-awaited companion piece I couldn’t manage to finish before the podcast posted. I might write a follow up post on some software engineering practices like continuous design later so that I can strictly focus here on softer process related ideas.

I started my software development career writing Shadow IT applications and automation for my engineering group. No process, no real practices either, and lots of code. As you’d probably guess, I later chafed very badly at the formal waterfall processes and organizational model of my first “real” software development job for plenty of reasons I’ll detail later in this post.

During that first real IT job, I started to read and learn about alternative iterative development processes like the Rational Unified Process (RUP), but I was mostly inspired by the brand new, shiny Extreme Programming (XP) method. After tilting the windmill a bit at my waterfall shop to try to move us away from the waterfall to RUP (Agile would have been a bridge too far), I jumped to a consulting company that was an influential, early adopter of XP. I’ve worked almost exclusively with Agile processes and inside more or less Agile organizational models ever since — until recently. In the past couple years I’ve been involved with a long running project in a classic waterfall shop which has absolutely reinforced my belief in the philosophical approaches to software development that came out of Agile Software (or Lean Programming). I also think some of those ideas are somewhat lost in contemporary Scrum’s monomaniacal focus on project management, so here’s a long blog post talking about what I think really was vital in the shift from waterfall to agile development.

First, what I believe in

A consistent theme in all of these topics is trying to minimize the amount of context switching throughout a projectand anybody’s average day. I think that Agile processes made a huge improvement in that regard over the older waterfall models, and that by itself is enough to justify staking waterfall through the heart permanently in my book.

Self-contained, multi-disciplinary teams

My strong preference and a constant recommendation to our clients is to favor self-contained, multi-disciplinary teams centered around specific projects or products. What I mean here is that the project is done by a team who is completely dedicated to working on just that project. That team should ideally have every possible skillset it needs to execute the project so that there’s reduced need to coordinate with external teams, so it’s whatever mix you need of developers, testers, analysts all working together on a common goal.

In the later sections on what I think is wrong with the waterfall model, I bring up the fallacy of local optimization. In terms of a software project, we need to focus on optimizing the entire process of delivering working software, not just on what it takes to code a feature, or write tests, or quickly checking off a process quality gate. A self-contained team is hopefully all focused on delivering just one project or sprint, so there should be more opportunity to optimize the delivery. This generally means things like developers purposely thinking about how to support the testers on their team or using practices like Executable Specifications for requirements that shortens the development and testing time overall, even if it’s more work for the original analysts upfront.

A lot of the overhead of software projects is communication between project team members. To be effective, you need to improve the quality of that communication so that team members have a shared understanding of the project. I also think you’re a lot less brittle as a project if you have fewer people to communication with. In a classic waterfall shop, you may need to be involving a lot of folks from external projects who are simultaneously working on several other initiatives and have a different schedule than your project’s schedule. That tends to force communication into either occasional meetings (which impact productivity on its own) or through asynchronous communication via emails or intermediate documentation like design specifications.

Let’s step back and think about the various types of communication you use with your colleagues, and what actually works. Take a look at this diagram from Scott Ambler’s Communication on Agile Software Teams essay (originally adapted from some influential writings by Alistair Cockburn that I couldn’t quite find this morning):



In a self-contained, multi-disciplinary team, you’re much more likely to be using the more effective forms of communication in the upper, right hand of the graph. In a waterfall model where different disciplines (testers, developers, analysts) may be working in different teams and on different projects at any one time, the communication is mostly happening at the less effective, lower left of this diagram.

I think that a self-contained team suffers much less from context switching, but I’ll cover that in the section on delivering serially.

Another huge advantage to self-contained teams is the flexibility in scheduling and the ability to adapt to changing project circumstances and whatever you’re learning as you go. This is the idea of Reversibility from Extreme Programming:

“If you can easily change your decisions, this means it’s less important to get them right – which makes your life much simpler. ” — Martin Fowler

In a self-contained team, your reversibility is naturally higher, and you’re more likely able to adapt and incorporate learning throughout the project. If you’re dependent upon external teams or can’t easily change approved specification documents, you have much lower reversibility.

If you’re interested in Reversibility, I gave a technically focused talk on that at Agile Vancouver several years ago

I think everything in this section still applies to teams that are focused on a product or family of products.

Looking over my history, I’ve written about this topic quite a bit over the years:

  1. On Software Teams
  2. Call me a Utopian, but I want my teams flat and my team members broad
  3. Self Organizing Teams are Superior to Command n’ Control Teams
  4. Once Upon a Team
  5. The Anti Team
  6. On Process and Practices
  7. Want productivity? Try some team continuity (and a side of empowerment too) – I miss this team.
  8. The Will to be Good
  9. Learning Lessons — Can You Make Mistakes at Work?
  10. Indelible Proof of a Healthy Team

Deliver serially over working in parallel

A huge shift in thinking behind Agile Software Development is simply the idea that the only thing that matters is delivering working software. Not design specifications, not intermediate documents, not process checkpoints passed, but actual working software deployed to production.

In practice, this means that Agile teams seek to deliver completely working features or “vertical slices” of functionality at one time. In this model a team strives for the continuous delivery model of constantly making little releases of working software.

Contrast this idea to the old “software as construction” metaphor from my waterfall youth where we generally developed by:

  1. Get the business requirements
  2. Do a high level architecture document
  3. Maybe do a lower level design specification (or do a prototype first and pretend you wrote the spec first)
  4. Design the database model for the new system
  5. Code the data layer
  6. Code the business layer
  7. Code any user interface
  8. Rework 4-6 because you inevitably missed something or realized something new as you went
  9. Declare the system “code complete”
  10. Start formal testing of then entire system
  11. Fix lots of bugs on 4-6
  12. User acceptance testing (hopefully)
  13. Release to production

The obvious problems in this cycle is that you deliver no actual value until the very, very end of the project. You’re also struggling quite a bit in the later parts of the project because you’re needing to re-visit work that was done much earlier and you often struggle with the context switching that entails.

In contrast, delivering in vertical slices allows you to:

  • Minimize context switching because you’re finishing work much closer to when it’s started. With a multi-disciplinary team, every body is focused on a small subset of features at any one time which tends to improve the communication between disciplines.
  • Actually deliver something much earlier to production and start accruing some business payoff. Moreover, you should be working in the order of business priority, so the most important things are worked on and completed first. Which also serves to fail softer compared to a waterfall project cycle.
  • Fail softer by delivering part of a project on time if even if you’re not able to complete all the planned features by the theoretical end date — as apposed to failing completely to deliver anything on time in a waterfall model.

In the Extreme Programming days we talked a lot about the concept of done, done, done as opposed to being theoretically “code complete.”


Rev’ing up feedback loops

After coming back to waterfall development the past couple years, the most frustrating thing to me is how slow the feedback cycles are between doing or deciding something and knowing whether or not anything you did was really correct. It also turns out that having a roomful of people staring at a design specification document in a formal review doesn’t do a great job at spotting a lot of problems that present themselves later in the project when actual code is being written.

Any iterative process helps tighten feedback cycles and enables you to fix issues earlier. What Agile brought to the table was an emphasis on better, faster, more fine-grained feedback cycles through project automation and practices like continuous integration and test driven development.

More on the engineering practices in later posts. Maybe. It literally took me 5 years to go from an initial draft to publishing this post so don’t hold your breathe.


What I think is wrong with classic waterfall development

Potentially a lot. Your mileage may vary from mine (my experiences with formal waterfall processes has been very negative) and I’m sure some of you are succeeding just fine with waterfall processes of one sort or another.

At its most extreme, I’ve observed these traits in shops with formal waterfall methods, all of which I think work against successful delivery and why I think these traits are problematic.

Over-specialization of personnel

I’m not even just talking about developers vs testers. I mean having some developers or groups who govern the central batch scheduling infrastructure, others that own integrations, a front end team maybe, and obviously the centralized database team. Having folks specialized in their roles like this means that it takes more people involved in a project in order to have all the necessary skillset, and that means having a lot more effort to communicate and collaborate between people who aren’t in the same teams or even in the same organizations. That’s a lot of potential project overhead, and it makes your project less flexible as you’re bound by the constraints of your external dependencies.

The other problem with over-specialization is the fallacy of local optimization problem, because many folks only have purview over part of a project and don’t necessarily see or understand the whole project.

Formal, intermediate documents

I’m not here to start yet another argument over how much technical documentation is enough. What I will argue about is a misplaced focus on formal, intermediate documents as a quality or process gate. Especially if those intermediate documents are really meant to serve as the primary communication between architects, analysts, and developers. One, because that’s a deeply inefficient way to communicate. Two, because those documents are always wrong because they’re written too early. Three because it’s just a lot of overhead to go through authoring those documents to get through a formal process gate that could be better spent on getting real feedback about the intended system or project.

Slow feedback cycles

Easily the thing I hate the most about “true” waterfall projects is the length of time between getting adequate feedback between the early design and requirements documents and an actually working system getting tested or going through some user acceptance testing from real users. This is an especially pernicious issue if you hew to the idea that formal testing can only start after all designed features are complete.

The killer problem in larger waterfall projects over my career is that you’re always trying to remember how some piece of code you wrote 3-6 months ago works when problems finally surface from real testing or usage.



I’d absolutely choose some sort of disciplined Agile process with solid engineering practices over any kind of formal waterfall process any day of the week. I think waterfall processes do a wretched job managing project risks by the slow, ineffective feedback cycles and waste a lot of energy on unevenly useful intermediate documentation.

Agile hasn’t always been pretty for me though, see The Surprisingly Valuable and Lasting Lessons I Learned from a Horrible Project about an absolutely miserable XP project.

Ironically, the most successful project I’ve ever worked on from a business perspective was technically a waterfall project (a certain 20-something first time technical lead basically ignored the official process), but process wasn’t really an issue because:

  • We had a very good relationship with the business partners including near constant feedback about what we were building. That project is still the best collaboration I’ve ever experienced with the actual business experts
  • There was a very obvious problem to solve for the business that was likely to pay off quickly
  • Our senior management did a tremendous job being a “shit umbrella” to keep the rest of the organization from interfering with us
  • It was a short project

And projects like that just don’t come around very often, so I wouldn’t read much into the process being the deciding factor in its success.








Marten v4.0 Planning Document (Part 1)

As I wrote about a couple weeks ago in a post called Kicking off Marten V4 Development, the Marten team is actively working on the long delayed v4.0 release with planned improvements for performance, the Linq support, and a massive planned jump in capability for the event store functionality. This post is a the result of a long comment thread and many discussions between the Marten community.

We don’t yet have a great consensus about the exact direction that the event store improvements are going to go, so I’m strictly focusing on improvements to the Marten internals and the Document Database support. I’ll follow up with a part 2 on the event store as that starts to gel.

If you’re interested in Marten, here’s some links:


Pulling it Off

We’ve got the typical problems of needing to address incoming pull requests and bug issues in master while probably needing to have a long lived branch for v4.

As an initial plan, let’s:

  1. Start with the unit testing improvements as a way to speed up the build before we dive into much more of this? This is in progress with about a 25% reduction in test throughput time so far in this pull request
  2. Do a bug sweep v3.12 release to address as many of the tactical problems as we can before branching to v4
  3. Possibly do a series of v4, then v5 releases to do this in smaller chunks? We’ve mostly said do the event store as v4, then Linq improvements as v5 — Nope, full speed ahead with a large v4 release in order to do as many breaking changes as possible in one release
  4. Extract the generic database manipulation code to its own library to clean up Marten, and speed up our builds to make the development work be more efficient.
  5. Do the Event Store v4 work in a separate project built as an add on from the very beginning, but leave the existing event store in place. That would enable us to do a lot of work and mostly be able to keep that work in master so we don’t have long-lived branch problems. Break open the event store improvement work because that’s where most of the interest is for this release.

Miscellaneous Ideas

  • Look at some kind of object pooling for the DocumentSession / QuerySession objects?
  • Ditch the document by document type schema configuration where Document A can be in one schema, and Document “B” is in another schema. Do that, and I think we open the door for multi-tenancy by schema.
  • Eliminate ManagedConnection altogether. I think it results in unnecessary object allocations and it’s causing more harm that help as it’s been extended over time. After studying that more today, it’s just too damn embedded. At least try to kill off the Execute methods that take in a Lambda. See this GitHub issue.
  • Can we consider ditching < .Net Core or .Net v5 for v4? The probable answer is “no,” so let’s just take this off the table.
  • Do a hunt for classes in Marten marked public that should be internal. Here’s the GitHub issue.
  • Make the exceptions a bit more consistent

Dynamic Code Generation

If you look at the pull request for Document Metadata and the code in Marten.Schema.Arguments you can see that our dynamic Expression to Lambda compilation code is getting extremely messy, hard to reason with, and difficult to extend.

Idea: Introduce a dependency on LamarCodeGeneration and LamarCompiler. LamarCodeGeneration has a strong model for dynamically generating C# code at runtime. LamarCompiler adds runtime Roslyn support to compile assemblies on the fly and utilities to attach/create these classes. We could stick with Expression to Lambda compilation, but that can’t really handle any kind of asynchronous code without some severe pain and it’s far more difficult to reason about (Jeremy’s note: I’m uniquely qualified to make this statement unfortunately).

What gets dynamically generated today:

  • Bulk importer handling for a single entity
  • Loading entities and tracking entities in the identity map or version tracking

What could be generated in the future:

  • Document metadata properties — but sad trombone, that might have to stay with Expressions if the setters are internal/private :/
  • Much more of the ISelector implementations, especially since there’s going to be more variability when we do the document metadata
  • Finer-grained manipulation of the IIdentityMap

Jeremy’s note: After doing some detailed analysis through the codebase and the spots that would be impacted by the change to dynamic code generation, I’m convinced that this will lead to significant performance improvements by eliminating many existing runtime conditional checks and casts

Track this work here.

Unit Testing Approach

This is in progress, and going well.

If we introduce the runtime code generation back into Marten, that’s unfortunately a non-trivial “cold start” testing issue. To soften that, I suggest we get a lot more aggressive with reusable xUnit.Net class fixtures between tests to reuse generated code between tests, cut way down on the sheer number of database calls by not having to frequently check the schema configuration, and other DocumentStore overhead.

A couple other points about this:

  • We need to create more unique document types so we’re not having to use different configurations for the same document type. This would enable more reuse inside the testing runtime
  • Be aggressive with separate schemas for different configurations
  • We could possibly turn on parallel test running to speed up the test cycles

Document Metadata

  • From the feedback on GitHub, it sounds like the desire to extend the existing metadata to tracking data like correlation identifiers, transaction ids, user ids, etc. To make this data easy to query on, I would prefer that this data be separate columns in the underlying storage
  • Use the configuration and tests from pull request for Document Metadata, but use the Lamar-backed dynamic code generation from the previous section to pull this off.
  • I strongly suggest using a new dynamic codegen model for the ISelector objects that would be responsible for putting Marten’s own document metadata like IsDeleted or TenantId or Version onto the resolved objects (but that falls apart if we have to use private setters)
  • I think we could expand the document metadata to allow for user defined properties like “user id” or “transaction id” much the same way we’ll do for the EventStore metadata. We’d need to think about how we extend the document tables and how metadata is attached to a document session

My thought is to designate one (or maybe a few?) .Net type as the “metadata type” for your application like maybe this one:

    public class MyMetadata
        public Guid CorrelationId { get; set; }
        public string UserId { get; set; }

Maybe that gets added to the StoreOptions something like:

var store = DocumentStore.For(x => {
    // other stuff

    // This would direct Marten to add extra columns to
    // the documents and events for the metadata properties
    // on the MyMetadata type.

    // This would probably be a fluent interface to optionally fine tune
    // the storage and applicability -- i.e., to all documents, to events, etc.

Then at runtime, you’d do something like:


Either through docs or through the new, official .Net Core integration, we have patterns to have that automatically set upon new DocumentSession objects being created from the IoC to make the tracking be seemless.

Extract Generic Database Helpers to its own Library

  • Pull everything to do with Schema object generation, difference detection, and DDL generation to a separate library (IFeatureSchemaISchemaObject, etc.). Mostly to clean out the main library, but also because this code could easily be reused outside of Marten. Separating it out might make it easier to test and extend that functionality, which is something that occasionally gets requested. There’s also the possibility of further breaking that into abstractions and implementations for the long run of getting us ready for Sql Server or other database engine support. The tests for this functionality are slow, and rarely change. It would be advantageous to get this out of the main Marten library and testing project.
  • Pull the ADO.Net helper code like CommandBuilder and the extension methods into a small helper library somewhere else (I’m nominating the Baseline repository). This code is copied around to other projects as it is, and it’s another way of getting stuff out of the main library and the test suite.

Track this work in this GitHub issue.

F# Improvements

We’ll have a virtual F# subcommittee to be watching this work for F#-friendliness:

HostBuilder Integration

We’ll bring Joona-Pekka Kokko’s ASP.Net Core integration library into the main repository and make that the officially blessed and documented recipe for integrating Marten into .Net Core applications based on the HostBuilder in .Net Core. I suppose we could also multi-target IWebHostBuilder for ASP.Net Core 2.*.

That HostBuilder integration could be extended to:

  • Optionally set up the Async Daemon in an IHostedService — more on this in the Event Store section
  • Optionally register some kind of IDocumentSessionBuilder that could be used to customize session construction?
  • Have some way to have container resolved IDocumentSessionListener objects attached to IDocumentSession. This is to have an easy recipe for folks who want events broadcast through messaging infrastructure in CQRS architectures

See the GitHub issue for this.

Command Line Support

The Marten.CommandLine package already uses Oakton for command line parsing. For easier integration in .Net Core applications, we could shift that to using the Oakton.AspNetCore package so the command line support can be added to any Core 2.* or .Net Core 3.* project by installing the Nuget. This might simplify the usage because you’d no longer need a separate project for the command line support.

There are some long standing stories about extending the command line support for the event store projection rebuilding. I think that would be more effective if it switches over to Oakton.AspNetCore.

See the GitHub issue


This is also covered by the Linq Overhaul issue.

  • Bring back the work in the linq branch for the revamped IField model within the Linq provider. This would be advantageous for performance, cleans up some conditional code in the Linq internals, could make the Linq support be aware of Json serialization customizations like [JsonProperty], and probably helps us deal more with F# types like discriminated unions.
  • Completely rewrite the Include() functionality. Use Postgresql Common Table Expression and UNION queries to fetch both the parent and any related documents in one query without needing to do any kind of JOIN s that complicate the selectors. There’d be a column for document type the code could use to switch. The dynamic code generation would help here. This could finally knock out the long wished for Include() on child collections feature. This work would nuke the InnerJoin stuff in the ISelector implementations, and that would hugely simplify a lot of code.
  • Finer grained code generation would let us optimize the interactions with IdentityMap. For purely query sessions, you could completely skip any kind of interaction with IdentityMap instead of wasting cycles on nullo objects. You could pull out a specific IdentityMap<TEntity, TKey> out of the larger identity map just before calling selectors to avoid some repetitive “find the right inner dictionary” on each document resolved.
  • Maybe introduce a new concept of ILinqDialect where the Expression parsing would just detect what logical thing it finds (like !BoolProperty), and turns around and calls this ILinqDialect to get at a WhereFragment or whatever. This way we could ready ourselves to support an alternative json/sql dialect around JSONPath for Postgresql v12+ and later for Sql Server vNext. I think this would fit into the theme of making the Linq support more modular. It should make the Linq support easier to unit test as we go. Before we do anything with this, let’s take a deep look into the EF Core internals and see how they handle this issue
  • Consider replacing the SelectMany() implementation with Common Table Expression sql statements. That might do a lot to simplify the internal mechanics. Could definitely get us to an n-deep model.
  • Do the Json streaming story because it should be compelling, especially as part of the readside of a CQRS architecture using Marten’s event store functionality.
  • Possibly use a PLV8-based JsonPath polyfill so we could use sql/json immediately in the Linq support. More research necessary.

Partial Updates

Use native postgresql partial JSON updates wherever possible. Let’s do a perf test on that first though.

A Small Case Study in Test Automation (and other things)

I’m trying to walk a line here in this post between avoiding specifics about a client project for obvious reasons, but providing enough detail to make this post worthwhile for that client. One of our client’s development managers is interested in speeding up their testing, and I’m hoping to use this post to lay out some ideas and approaches to improve the testing procedures in this system.

I’ve been part of an integration project for the past couple years that validates, routes, and processes financial transactions coming from an external partner of our client’s all the way to a very large 3rd party hosted in our client’s environment. We’re in the middle of some significant changes in the integration to that 3rd party application that is going to trigger a round of regression testing of the entire system — and that’s where this post comes in. Testing this application has been very challenging and extremely time consuming. Any opportunity to make regression testing be quicker and more effective is going to make everyone’s jobs easier.

It’s not just that the testing itself is slower than desired. Because the testing is slow and not easily repeatable, the development team can’t really do much technical improvement through refactoring as they learn more about the system behavior and how the code structure is working out over time. That’s been a definite negative for code and architectural quality.

Before I get into the details of the existing system, know that what I’m showing and discussing here is a bit of an idealized version of how I wish we had architected the system and what we’ve recommended to the client for the longer term. The real system is a bit messier and significantly harder to test than what I’m presenting here — but there’s a lesson for you, testability should be a first class architectural goal in many cases (and Conway’s Law is legitimately something to work around).

From a 10,000 foot level, here’s the entire system:

TestAutomationScenario-High Level

The workflow is:

  1. A couple times a day, a new flat file containing new transactions will be dropped into a file share
  2. The File Reader console application is executed to find this file, parse it into little transaction messages, and publish those messages to Rabbit MQ. There’s a little bit of database tracking going on for reporting and just general activity tracking.
  3. Rabbit MQ publishes the transaction messages to the subscribing Transaction Processor application (an ASP.Net Core application with an active subscriber for these incoming messages).
  4. The Transaction Processor handles each transaction message by:
    1. Pulling in a helluva lot of information from the 3rd Party Application and other information from a Configuration DB related to the account number in the transaction message
    2. Using the information from the previous step to validate whether the transaction can be processed normally, or has to go into a queue for manual resolution
    3. For the valid transactions, use the information from step #1 to decide how the money in the incoming transaction will be applied (routing to sub-transactions)
    4. Send the routed sub-transactions from the previous step to the 3rd Party Application through its externally facing API.

While there are some unit tests and intermediate level integration tests today on some of the subsystems, the overall official testing effort to date has relied strictly on end to end, manual testing of the entire system. Some of the emphasis on black box, end to end testing is due to our client’s mandatory regulatory auditing requirements and that can’t completely go away. However, there’s worlds of opportunity and new willingness to explore other alternatives like white box testing techniques or new processes for testing as a complement to the formal audit-style testing, so let’s jump into some ideas for making things work more efficiently.

Some Necessary Shifts in Testing Philosophy

First off, there’s an important shift from trying to prove that the system is working perfectly with strictly black box testing to thinking about testing as a feedback mechanism to identify and remove problems in the code so that the code can be deployed to production. If you look at testing as more of a feedback cycle, you can utilize the testing pyramid idea to maximize feedback about how your system functions with more efficient testing techniques.

Secondly, I think you have to have collaboration between testers, developers, and architects to make white box testing more effective. Part of that is increasing the testability of the system architecture, and another part of that is trying to avoid duplication in effort between tests written by developers and other tests performed by the testers. Moreover, if developers are actively engaged in writing tests — and they should be in my world view — it’s very helpful to have the testers involved in the content of those developer-written tests. In other words, I think that having strict separation between testers and development can be very inefficient. I know there are folks who strongly believe that strict independence for the testers from the developers is necessary, but I think that does more harm than good.

For more information on whitebox vs. blackbox testing and improving test feedback, also see:

If you’ll buy any of the two previous paragraphs, or you’re at least open-minded to continue, let’s see how some of this Test Pyramid thinking would play out in our big integration system.

At a high level, I would want the testing strategy to focus on:

  • Some kind of Behavioral Driven Development approach for all the business rules for validating and routing the transactions.
  • Mid-level integration tests on all the code that acts as a gateway or service proxy to the 3rd Party Application. This would include both the code that sends commands to the 3rd Party Application and the code that queries or reads the 3rd Party Application.
  • Mid-level integration tests on the File Reader that probably stubs the outgoing Rabbit MQ and just measures how the File Reader parses the incoming files, writes tracking information, and what messages it publishes.
  • A handful of fully end to end tests through the entire system to prove out all the integration points — but by and large you use finer grained tests to test out business rules and the integration with the 3rd Party Application.


The Transaction Processor

Most of the meat of the bigger transaction processing project is within what I’m calling the Transaction Processor shown below in a little more detail:

TestAutomationScenario-Transaction Processor

There’s a couple big responsibilities here:

  • Querying data from the 3rd Party Application with its heinously unusable, custom Xml query language to use inside the business rules
  • Look up some configuration parameters about accounts from a second Configuration DB
  • Carry out validation rules against incoming transactions
  • Route the incoming transactions into sub-transactions based on business rules
  • Post the sub-transactions to the 3rd Party Application with its, shall we say, interesting XML API.

Channeling some Domain-Driven Design thinking here, let’s go straight into the business rules for validation and routing. The business rules required a lot of input parameters, there were a lot of permutations to build and test, and the developers new to the problem domain had plenty of misunderstandings early on about the desired behavior.

From an architectural standpoint, I think it is extremely important to completely isolate these business rules from the 3rd Party Application, the configuration databases, and even the incoming flat file format because:

  • It was very difficult to set up test scenario inputs in the 3rd Party Application
  • There’s a tremendous number of test cases because of the permutations on account state and transaction parameters involved, so there would be a large benefit to tests being quick to author and execute
  • This logic is key to the business and has already evolved significantly since this project started. It’s imperative that this logic be safe to change over time, and that happens most effectively when it’s cheap to write new tests and quick to execute the existing test coverage.
  • I probably shouldn’t say this too loudly, but I think this client should reconsider coupling their ecosystem to the 3rd Party Application

To that end, the business rules should only depend on a domain model that’s internal to the Transaction Processor. We’ll use the A-Frame Architecture idea from Jim Shore’s Testing Without Mocks paper to isolate the business rule behavior from the infrastructure. The domain model objects that implement all the business logic will have no dependency whatsoever on the external dependencies. Instead, we’ll effectively write our own mapping layer to take the data returned from the 3rd Party Application and the Configuration DB and build all the state the domain model needs, then hand that to the business logic code in the domain model.

From the perspective of testing, there’s a lot of opportunity to get the business rules wrong. Rather than depend solely on design or requirements documents, I strongly recommend using Behavior Driven Development (BDD) techniques here to author executable specifications that are readable and reviewed (if not written) by the business domain experts and testers. What I largely recommend here is that the developers mostly write the test harness code, but business domain experts and more likely testers will own the content and meaning of the BDD specifications. Working in this manner, we should be able to treat the BDD specifications as the official tests for the business rule behavior even though this doesn’t run the entire process.

So that handles the business rules, now on to the rest of the Transaction Processor. The “controller” code in the diagram is playing a coordination role to mediate between the business rules and the code that interacts directly with the 3rd Party Application’s external API endpoints. I’d mostly use unit tests and maybe even *gasp* interaction testing with mock objects to test out the workflow and error handling of this code.

The service gateway code that interacts with the 3rd Party Application was extremely problematic in both development and testing. In retrospect, I wish we’d hammered at this code in isolation much more before even bothering trying to run end to end tests. The big issue we never pushed through (yet) was how to establish known system state in the 3rd Party Application so that we could write reliable automated tests around just the service gateway code in the Transaction Processor. I think it would be worthwhile for domain experts and/or testers to be involved in this step as well to verify the expected results are really happening in the 3rd Party Application.

Lastly, I’d opt to do some bigger tests for just the Transaction Processor where you directly enqueue the transaction messages in Rabbit MQ and test the entire Transaction Processor stack all the way down to the external dependencies. The point of these tests are to prove out the integrations and configuration. You don’t try to recreate all the business rules functionality tests covered by the smaller, faster unit tests.


“Some” End to End Tests

There are absolutely some issues that can only be tested through true, end to end tests. Integrations, configuration, environments, and security are examples. We’ll still write and perform some end to end tests, but we won’t try to recreate the business functionality tests covered.

No matter what though, the tests need to be as easily repeatable as possible so there’s still going to be a level of automation to speed things along. Here’s my thoughts on what that might look like:

  • The flat file format was originally used by mainframe applications, so as you can imagine, it’s not remotely user friendly to edit or read. I’d suggest using some custom code that can transform a much simpler format to the mainframe-friendly format so the testers can write new test cases more efficiently and everyone else can actually read and understand the test inputs
  • The undeniable, cardinal rule of automated testing is that you have to have known inputs and expected outcomes. In this system, that means being able to set up the 3rd Party Application in a known state for each end to end testing scenario. The failure to do that (not a technical impossibility, but it’s a long story) is my single biggest regret from this project. See My Opinions on Data Setup for Functional Tests for more on what I recommend for test input data.
  • Automating the testing of asynchronous workflows like this system can be very challenging. The biggest issue is making an automated test harness understand when the work is really done across multiple systems so it can proceed to the “assert” part of the standard “arrange, act, assert” test workflow. I’ve had some success with this in the past by making the test harness listen to the various application logging or some kind of visible side effect like data being written to a database to “know” when the work is complete.
  • Tests do fail from time to time, so I’d actually try to have the end to end test harness able to gather up the relevant logs for all the systems active in the test. That’s even more valuable if you can somehow manage to correlate the logging activity with only the active test run.
  • Finally, the big expensive end to end tests my client has to follow for official certification and auditing? Yeah, you have to do that, but my very strong recommendation and where I think they’re starting to head is to use finer-grained and more efficient testing techniques to remove problems first. Then come back and do the laboriously slow audit tests when you can justifiably expect success with few iterations.




There’s a couple big points I wanted to drive home in this post:

  • Embrace the test pyramid idea, and try to get over any aversion to white box testing because of its advantages for efficiency
  • Treat testing as a feedback mechanism more than a certification process
  • Tests of all type need to be repeatable to be effective feedback. Manual testing, and especially manual testing where it’s time consuming to set up the necessary system state first, is not very repeatable
  • I think you need to embrace the Agile idea of blurring the lines between roles. Developers and architects need to be involved in the automated testing for a better chance of success. Testers may need to get their hands dirty directly in the code or at least exploit their knowledge of the coding internals in order to make the testing more efficient
  • Developers, testers, and architects need to collaborate to be truly successful in testing. Waterfall style testing where all testing happens at the end is just not the way to be successful
  • Try to avoid duplicating effort between developer written tests and the tester activity, which might be just yet another way of saying the testers and developers need to be collaborating as the project goes on
  • Feedback cycles of all kinds are valuable for quality software

Using Postgresql Advisory Locks for Leader Election

If you’re running an application with a substantial workload, or just want some kind of high availability, you’re probably running that application across multiple servers (heretofore called “nodes” because who knows where they’re physically running these days). That’s great and all, but it’s not too uncommon that you’ll need to make some kind of process run on only one of those nodes at any one time.

As an example, the Marten event store functionality as a feature to support asynchronous projection builders called the “async daemon” (because I thought that sounded cool at the time). The async daemon is very stateful, and can only function while running on one node at a time — but it doesn’t have any existing infrastructure to help you manage that. What we know we need to do for the upcoming Marten v4.0 release is to provide “leader election” to make sure the async daemon is actively building projections on only one node and can be activated or fail over to another node as needed to guarantee that exactly one node is active at all times.

From Wikipedia, Leader Election “is the process of designating a single process as the organizer of some task distributed among several computers.” There’s plenty of existing art to do this, but it’s not for the feint of heart. In the past, I tried to do this with FubuMVC using a custom implementation of the Bully Algorithm. Microsoft’s microservices pattern guidance has some .Net centric approaches to leader election. Microsoft’s new Dapr tool is supposed to support leader election some day.

From my previous experience, building out and especially testing custom election infrastructure was very difficult. As a far easier approach, I’ve used Advisory Locks in Postgresql in Jasper (I’m also using the Sql Server equivalents as well) as what I think of as a “poor man’s leader election.”

An advisory lock in Postgresql is an arbitrary, application-managed lock on a named resource. Postgresql simply tracks these locks as a distributed lock such that only one active client can hold the lock at any one time. These locks can be held either at:

  1. The connection level, such that the lock, once obtained, is held as long as the database connection is open.
  2. The transaction level, such that a lock obtained within the course of one Postgresql transaction is held until the transaction is committed, rolled back, or the connection is lost.

As an example, Jasper‘s “Durability Agent” is a constantly running process in Jasper applications that tries to read and process any persisted messages persisted in a Postgresql or Sql Server database. Since you certainly don’t want a unique message to be processed by more than one node, the durability uses advisory locks to try to temporarily take sole ownership of replaying persisted messages with a workflow similar to this sequence diagram:

Transaction Scoped Advisory Lock Usage

That’s working well so far for Jasper, but in Marten v4.0, we want to use the connection scoped advisory lock for leader election of a long running process for the async daemon.

Sample Usage for Leader Election

Before you look at any of these code samples, just know that this is over-simplified to show the concept, isn’t in production, and would require a copious amount of error handling and logging to be production worthy.

For Marten v4.0, we’ll use the per-connection usage to ensure that the new version of the async daemon will only be running on one node (or at least the actual “leader” process that distributes and assigns work across other nodes if we do it well). The async daemon process itself is probably going to be a .Net Core IHostedService that runs in the background.

As just a demonstrator, I’ve pushed up a little project called AdvisoryLockSpike to GitHub just to show the conceptual usage. First let’s say that the actual worker bee process of the async daemon implements this interface:

public enum ProcessState

public interface IActiveProcess : IDisposable
    Task<ProcessState> State();
    // The way I've done this before, the
    // running code does all its work using
    // the currently open connection or at
    // least checks the connection to "know"
    // that it still has the leadership role
    Task Start(NpgsqlConnection conn);

Next, we need something around that to actually deal with the mechanics of trying to obtain the global lock and starting or stopping the active process. Since that’s a background process within an application, I’m going to use the built in BackgroundService in .Net Core with this little class:

public class LeaderHostedService<T> : BackgroundService
    where T : IActiveProcess
    private readonly LeaderSettings<T> _settings;
    private readonly T _process;
    private NpgsqlConnection _connection;

    public LeaderHostedService(LeaderSettings<T> settings, T process)
        _settings = settings;
        _process = process;

    protected override async Task ExecuteAsync(CancellationToken stoppingToken)
        // Don't try to start right off the bat
        await Task.Delay(_settings.FirstPollingTime, stoppingToken);
        _connection = new NpgsqlConnection(_settings.ConnectionString);
        await _connection.OpenAsync(stoppingToken);
        while (!stoppingToken.IsCancellationRequested)
            var state = await _process.State();
            if (state != ProcessState.Active)
                // If you can take the global lock, start
                // the process
                if (await _connection.TryGetGlobalLock(_settings.LockId, cancellation: stoppingToken))
                    await _process.Start(_connection);

            // Start polling again
            await Task.Delay(_settings.OwnershipPollingTime, stoppingToken);

        if (_connection.State != ConnectionState.Closed)
            await _connection.DisposeAsync();


To fill in the blanks, the TryGetGlobalLock() method is an extension method helper to call the underlying pg_try_advisory_lock function in Postgresql to try to obtain a global advisory lock for the configured lock id. That extension method is shown below:

// Try to get a global lock with connection scoping
public static async Task<bool> TryGetGlobalLock(this DbConnection conn, int lockId, CancellationToken cancellation = default(CancellationToken))
    var c = await conn.CreateCommand("SELECT pg_try_advisory_lock(:id);")
        .With("id", lockId)

    return (bool) c;

Raw ADO.Net is so verbose and unusable out of the box that I’ve built up a set of extension methods to streamline its usage that you might observe above if you notice that that isn’t quite out of the box ADO.Net.

I’m generally a fan of strong typed configuration, and .Net Core makes that easy now, so I’ll use this class to represent the configuration:

public class LeaderSettings<T> where T : IActiveProcess
    public TimeSpan OwnershipPollingTime { get; set; } = 5.Seconds();
    // It's a random number here so that if you spin
    // up multiple nodes at the same time, they won't
    // all collide trying to grab ownership at the exact
    // same time
    public TimeSpan FirstPollingTime { get; set; } 
        = new Random().Next(100, 3000).Milliseconds();
    // This would be something meaningful
    public int LockId { get; set; }
    public string ConnectionString { get; set; }

In this approach, the background services will be constantly polling to try to take over as the async daemon if the async daemon is not active somewhere else. If the current async daemon node fails, the connection will drop and the global advisory lock is released and ready for another node to take over. We’ll see how this goes, but the early feedback from my own usage on Jasper and other Marten contributors other projects is positive. With this approach, we hope to enable teams to use the async daemon on multi-node deployments of their application with just Marten out of the box and without having to have any kind of sophisticated infrastructure for leader election.

Kicking off Marten v4 Development


If you’re not familiar with Marten, it’s a library that allows .Net developers to treat Postgresql as a Document Database and also as an Event Store. For a quick background on Marten, here’s my presentation on Marten from Dotnetconf 2018.

The Marten community is gearing up to work on the long awaited v4.0 release, which I think is going to be the most ambitious release of Marten since the original v1.0 release in September of 2016. Marten is blessed with a strong user community, and we’ve got a big backlog of issues, feature requests, and suggestions for improvements that have been building up for quite some time.

First, some links for anyone who wants to be a part of that conversation:

Now would be a perfect time to add any kind of feedback or requests for Marten. We do have an unofficial “F# advisory board” this time around, but the more the merrier on that side. There’ll be plenty of opportunities for folks to contribute, whether that’s in coding, experimenting with early alphas, or just being heard in the discussions online.

Now, on to the release plans:

Event Sourcing Improvements

The biggest area of focus in Marten v4.0 is likely going to be the event store functionality. At the risk of being (rightly) mocked, I’d sum up our goals here as “Make the Marten Event Store be Web Scale!”

There’s a lot of content on the aforementioned Marten v4.0 discussion, and also on an older GitHub discussion on Event Store improvements for v4.0. The highlights (I think) are:

  • Event metadata (a long running request)
  • Partitioning Support
  • Improved projection support, including the option to project event data to flat database tables
  • A lot of improvements (near rewrite) to the existing support for asynchronous projections including multi-tenancy support, running across multiple application nodes, performance optimizations, and projection rebuild improvements
  • New support for asynchronous projection builders using messaging as an alternative to the polling async daemon. The first cut of this is very likely to be based around Jasper (I’m using this as a way to push Jasper forward too of course) and either Rabbit MQ or Azure Service Bus. If things go well, I’d hope that that expands to at least Jasper + Kafka and maybe parallel support for NServiceBus and MassTransit.
  • Snapshotting
  • The ability to rebuild projections with zero downtime

Linq Support Improvements

I can’t overstate how complex building and maintaining a comprehensive Linq provider can be just out of the sheer number of usage permutations.

Marten v4.0 is going to include a pretty large overhaul of the Linq support. Along the way, we’re hopefully going to:

  • Rewrite the functionality to include related documents altogether and finally support Include() on collections that’s been a big user request for 3+ years.
  • Improve the efficiency of the generated SQL
  • More variable data type behavior within the Linq support. You can read more about the mechanics on this issue. This also includes being quite a bit smarter about the Json serialization and how that varies for different .Net data types. As an example, I think these changes will allow Marten’s Linq support to finally deal with things like F# Discriminator Union types.
  • Improve the internals by making it more modular and a little more performant in how it handles marshaling data from the database to C# objects.
  • Improve Marten’s ability to query within child collections
  • More flexibility in event stream identifier strateggies


Other Improvements

  • Document Metadata improvements, both exposing Marten’s metadata and user customizable metadata for things like user id or correlation identifiers
  • An option to stream JSON directly to .Net Stream objects like HTTP responses without going through serialization first or even a JSON string.
  • Optimizing the Partial update support to use Postgresql native operators when possible
  • Formal support for .Net Core integration through the generic HostBuilder mechanism



Right now it’s all about the discussions and approach, but that’s going pretty well right now. The first thing I’m personally going to jump into is a round of automated test performance just to make everything go faster later.

The hope is that a subsequent V5 release might finally take Marten into supporting other database engines besides Postgresql, with Sql Server being the very obvious next step. We might take some preliminary steps with V4 to make the multi-database engine support be more feasible later with fewer breaking API changes, but don’t hold me to that.

I don’t have any idea about timing yet, sorry.

Jasper’s Efficient and Flexible Roslyn-Powered Execution Pipeline

You’ll need to pull the very latest code from the sample application linked to in this post because of course I found some bugs by dog-fooding while writing this:/

This is an update to a post from a couple years ago called Jasper’s Roslyn-Powered “Special Sauce.” You can also find more information about Lamar’s code generation and runtime compilation support from my Dynamic Runtime Code with Roslyn talk at London NDC 2019.

The more or less accepted understanding of a “framework” as opposed to a “library” is that a framework calls your application code (the Hollywood Principle). To be useful, frameworks should be dealing with cross cutting infrastructure concerns like security, data marshaling, or error handling. At some point, frameworks have to have some way to call your application code.

Since generics were introduced way back in .Net 3.5, many .Net frameworks have used what I call the “IHandler of T” approach where you are almost inevitably asked to implement a common layer supertype interface like this:

public interface IHandler
    Task Handle(T message, IContext context);

From a framework author’s perspective, this is easy to implement, and most modern IoC frameworks have reasonably strong support for generic types (like Lamar generic types support for example). Off the top of my head, NServiceBus, MassTransit, and MediatR are examples of this approach (my recollection is that NServiceBus did this first and I distinctly remember Udi Dahan describing this years ago at an ALT.Net event).

The other general approach I’ve seen — and what Jasper itself uses — is to have a framework call your code through some kind of dynamic code that adapts the signature of your code to the shape or interface that the framework needs. ASP.Net MVC Core Controller methods are an example of this approach. Here’s a sample controller method to demonstrate this:

public class WithResponseController : ControllerBase
    private readonly ICommandBus _bus;

    public WithResponseController(ICommandBus bus)
        _bus = bus;
    // MVC Core calls this method, and uses the signature
    // and attributes like the [FromBody] to "know" how
    // to call this code at runtime
    public Task Create([FromBody] CreateItemCommand command)
        // Using Jasper as a Mediator, and receive the
        // expected response from Jasper
        return _bus.Invoke(command);

To add some more complexity to designing frameworks, there’s also the issue of how you can combine the basic handlers with some sort of Russian Doll middleware approach to handle cross cutting concerns. There’s a couple ways to handle this:

  • If a framework uses the “IHandler of T” approach, folks often try to use the decorator pattern to wrap the inner “IHandler”. This approach can lead to folks crashing and burning by getting way too complex with generic constraints. It can also lead to some pretty severe levels of object allocations and garbage collection thrashing from the sheer number of objects being created and discarded. To add more fuel to the fire, this approach can easily lead to absurdly large stack trace exceptions that can be very intimidating to parse for the average developer.
  • ASP.Net Core’s middleware approach through functional composition. This can also lead to some dramatically bad stack traces and similar issues with object allocation.
  • Do some runtime code generation to piece together the middleware. I believe that NServiceBus does this internally with its version of Behaviors.

Lastly, most .Net web, service bus, or command execution frameworks use some sort of scoped IoC container (nested container in Lamar or StructureMap parlance) per request or command execution to deal with object scoping and cleanup or disposal in their execution pipelines.

Jasper’s Unique Approach

Jasper was originally conceived and planned as a replacement to the older FubuMVC project (but it’s gone off in different directions later of course). As a framework, some of FubuMVC’s design goals were to:

  1. Minimize the amount of framework artifacts (interfaces, base classes, attributes, et al) in your application code — and it generally succeeded at that
  2. Provide a very strong model for composable middleware (what we called Behaviors) in the runtime pipeline — which also succeeded in terms of capability, but at runtime it was extremely inefficient, object allocations were off the charts, the IoC integration was complex, and the exception stack traces when things went wrong were epically big.

With Jasper, the initial goals were to recreate the “cleanliness” of the application code and the flexibility of FubuMVC’s “Russian Doll” approach, but do so in a way that was much more efficient at runtime. And because this actually matters quite a bit, make sure that any exceptions thrown out of application code while running within Jasper have minimal exception stack traces for easier troubleshooting.

Roslyn introduced support for compiling C# code at runtime. Some time in 2015 I was in the office with some of the old FubuMVC core contributors and drew up on the whiteboard a new approach where we’d generate C# code at application bootstrapping time to weave in both the calls to the application code and designated middleware.

Later on, I expanded that vision to try to also encompass the object construction and cleanup functionality of an IoC container in the same generated code. The result of that initial envisioning has become the combination of Jasper and Lamar, where Jasper actually uses Lamar’s registration model to try to generate and inline the functionality that would normally be done at runtime through an IoC container. The theory here being that the fasted IoC container is no IoC container.

Alright, to make this concrete, let’s see how this plays out in real usage. In my last post, Introducing Jasper as an In Process Command Bus for .Net, I demonstrated a small message handler in a sample Jasper application with this code (I’ve stripped out the original comments to make it smaller):

public class ItemHandler
    public static ItemCreated Handle(
        CreateItemCommand command,
        ItemsDbContext db)
        var item = new Item
            Name = command.Name


        return new ItemCreated
            Id = item.Id

The code above is meant to:

  1. Create a new Item entity based on the incoming CreateItemCommand
  2. Persist that new Item with Entity Framework Core
  3. Publish a new ItemCreated event message to be handled somewhere else (how that happens isn’t terribly important for this blog post)

And lastly, the [Transactional] attribute tells Jasper to apply its transactional middleware and outbox support to the message handler, such that a single database commit will save both the new Item entity and do the “store” part of a “store and forward” operation to send out the cascading ItemCreated event.

Internally, Jasper is going to generate a new class that inherits from this base class (slightly simplified):

public abstract class MessageHandler
    public abstract Task Handle(
        IMessageContext context, 
        CancellationToken cancellation

For that ItemHandler class shown up above in the sample application, Jasper is generating and compiling through Roslyn this (butt ugly, but remember that it’s generated) class:

    public class InMemoryMediator_Items_CreateItemCommand : Jasper.Runtime.Handlers.MessageHandler
        private readonly Microsoft.EntityFrameworkCore.DbContextOptions _dbContextOptions;

        public InMemoryMediator_Items_CreateItemCommand(Microsoft.EntityFrameworkCore.DbContextOptions dbContextOptions)
            _dbContextOptions = dbContextOptions;

        public override async Task Handle(Jasper.IMessageContext context, System.Threading.CancellationToken cancellation)
            var createItemCommand = (InMemoryMediator.Items.CreateItemCommand)context.Envelope.Message;
            using (var itemsDbContext = new InMemoryMediator.Items.ItemsDbContext(_dbContextOptions))
                // Enroll the DbContext & IMessagingContext in the outgoing Jasper outbox transaction
                await Jasper.Persistence.EntityFrameworkCore.JasperEnvelopeEntityFrameworkCoreExtensions.EnlistInTransaction(context, itemsDbContext);
                var itemCreated = InMemoryMediator.Items.ItemHandler.Handle(createItemCommand, itemsDbContext);
                // Outgoing, cascaded message
                await context.EnqueueCascading(itemCreated);
                // Added by EF Core Transaction Middleware
                var result_of_SaveChangesAsync = await itemsDbContext.SaveChangesAsync(cancellation);



If you want, you can see the raw code at any time by executing the dotnet run -- codegen command from the root of the sample project.

So here’s what I’m claiming is the advantage of Jasper’s approach:

  • Allows your application code to be “clean” of framework artifacts and much more decoupled from Jasper than you can achieve with many other application frameworks like MVC Core
  • Using the diagnostic commands to dump out the generated source code in the application, Jasper can tell you exactly how it’s handling a message at runtime
  • By just generating Poor Man’s Dependency Injection code to build up the EF Core dependency and also to deal with disposing it later, the generated code eliminates any need to use the IoC container at runtime. And even for the very fastest IoC container in the world — and Lamar isn’t a slouch on the performance side of things — pure dependency injection is faster. Do mote that Jasper + Lamar can’t do this for every possible message handler and will have to revert to a scoped container per message with service location in some circumstances, usually because of a scoped or transient Lambda service registration or internal types.
  • The generated code minimizes the number of object allocations compared to a typical .Net framework that depends on adapter types
  • Jasper will allow you to use either synchronous or asynchronous handlers as appropriate for your use case, so no wasted keystrokes typing out return Task.CompletedTask; littering up your code.
  • The stack traces are pretty clean and you’ll see effectively nothing related to Jasper in logged exceptions except for the outermost frame where the generated MessageHandler.Handle() method is called.
  • Some other .Net frameworks try to do what Jasper does conceptually by trying to generate Expression trees and compile those down to Func objects. I’ve certainly done that in other tools (Lamar, StructureMap, Marten), but that’s rarified air that most developers can’t deal with. It’s also batshit insane for async heavy code like you’ll inevitably hit with anything that involves IO these days. My theory, yet to be proven, is that by relying on C#, Jasper’s middleware approach will be much more approachable to whatever community Jasper attracts down the line.


For some follow up reading if you’re interested:

Introducing Jasper as an In Process Command Bus for .Net

A couple weeks ago I wrote a blog post called If you want your OSS project to be successful… about trying to succeed with open source development efforts. One of the things I said was “don’t go dark” when you’re working on an OSS project. Not only did I go “dark” on Jasper for quite awhile, I finally rolled out its 1.0 release during the worst global pandemic in a century. So all told, Jasper is by no means an exemplary project model for anyone to follow who’s trying to succeed with an OSS project.

This sample application is also explained and demonstrated in the documentation page Jasper as a Mediator.

Jasper is a new open source tool that can be used as an in process “command bus” inside of .Net Core 3 applications. Used locally, Jasper can provide a superset of the “mediator” functionality popularized by MediatR that many folks like using within ASP.Net MVC Core applications to simplify controller code by offloading most of the processing to separate command handlers. Jasper certainly supports that functionality, but also adds rich options for asynchronous processing commands with built in resiliency mechanisms.

Part of the reason why Jasper went cold was waiting for .Net Core 3.0 to be released. With the advent of .Net Core 3.0, Jasper was somewhat re-wired to support the new generic HostBuilder for bootstrapping and configuration. With this model of bootstrapping, Jasper can easily be integrated into any kind of .Net Core application (MVC Core application, web api, windows service, console app, “worker” app) that uses the HostBuilder.

Let’s jump into seeing how Jasper could be integrated into a .Net Core Web API system. All the sample code I’m showing here is on GitHub in the “InMemoryMediator” project. InMemoryMediator uses EF Core with Sql Server as its backing persistence. Additionally, this sample shows off Jasper’s support for the “Outbox” pattern for reliable messaging without having to resort to distributed transactions.

To get started, generated a project with the dotnet new webapi template. From there, I added some extra Nuget dependencies:

  1. Microsoft.EntityFrameworkCore.SqlServer — because we’re going to use EF Core with Sql Server as the backing persistence for this service
  2. Jasper — this is the core library, and all that you would need to use Jasper as an in process command bus
  3. Jasper.Persistence.EntityFrameworkCore — extension library to add Jasper’s “Outbox” and transactional support to EF Core
  4. Jasper.Persistence.SqlServer — extension library to add persistence for the “Outbox” support
  5. Swashbuckle.AspNetCore — just to add Swagger support

Your First Jasper Handler

Before we get into bootstrapping, let’s just start with how to build a Jasper command handler and how that would integrate with an MVC Core Controller. Keeping to a very simple problem domain, let’s say that we’re capturing, creating, and tracking new Item entities like this:

public class Item
    public string Name { get; set; }
    public Guid Id { get; set; }

So let’s build a simple Jasper command handler that would process a CreateItemCommand message, persist a new Item entity, and then raise an ItemCreated event message that would be handled by Jasper as well, but asynchronously somewhere off to the side in a different through. Lastly, we want things to be reliable, so we’re going to introduce Jasper’s integration of Entity Framework Core for “Outbox” support for the event messages being raised at the same time we create new Item entities.

First though, to put things in context, we’re trying to get to the point where our controller classes mostly just delegate to Jasper through its ICommandBus interface and look like this:

public class UseJasperAsMediatorController : ControllerBase
    private readonly ICommandBus _bus;

    public UseJasperAsMediatorController(ICommandBus bus)
        _bus = bus;

    public Task Create([FromBody] CreateItemCommand command)
        // Using Jasper as a Mediator
        return _bus.Invoke(command);

You can find a lot more information about what Jasper can do as a local command bus in the project documentation.

When using Jasper as a mediator, the controller methods become strictly about the mechanics of reading and writing data to and from the HTTP protocol. The real functionality is now in the Jasper command handler for the CreateItemCommand message, as coded with this Jasper Handler class:

public class ItemHandler
    // This attribute applies Jasper's transactional
    // middleware
    public static ItemCreated Handle(
        // This would be the message
        CreateItemCommand command,

        // Any other arguments are assumed
        // to be service dependencies
        ItemsDbContext db)
        // Create a new Item entity
        var item = new Item
            Name = command.Name

        // Add the item to the current
        // DbContext unit of work

        // This event being returned
        // by the handler will be automatically sent
        // out as a "cascading" message
        return new ItemCreated
            Id = item.Id

You’ll probably notice that there’s no interface and mandatory base class usage in the code up above. Similar to MVC Core, Jasper will auto-discover the handler classes and message handling methods from your code through type scanning. Unlike MVC Core and every other service bus kind of tool .Net I’m aware of, Jasper only depends on naming conventions rather than base classes or interfaces.

The only bit of framework “stuff” at all in the code above is the [Transactional] attribute that decorates the handler class. That adds Jasper’s own middleware for transaction and outbox support around the message handling to just that message. At runtime, when Jasper handles the CreateItemCommand in that handler code up above, it:

  • Sets up an “outbox” transaction with the EF Core ItemsDbContextservice being passed into the Handle() method as a parameter
  • Take the ItemCreated message that “cascades” from the handler method and persists that message with ItemsDbContext so that both the outgoing message and the new Item entity are persisted in the same Sql Server transaction
  • Commits the EF Core unit of work by calling ItemsDbContext.SaveChangesAsync()
  • Assuming that the transaction succeeds, Jasper kicks the new ItemCreated message into its internal sending loop to speed it on its way. That outgoing event message could be handled locally in in-memory queues or sent out via external transports like Rabbit MQ or Azure Service Bus

If you’re interested in what the code above would look like without any of Jasper’s middleware or cascading message conventions, see the section near the bottom of this post called “Do it All Explicitly Controller”.

So that’s the MVC Controller and Jasper command handler, now let’s move on to integrating Jasper into the application.

Bootstrapping and Configuration

This is just an ASP.Net Core application, so you’ll probably be familiar with the generated Program.Main() entry point. To completely utilize Jasper’s extended command line support (really Oakton.AspNetCore), I’ll make some small edits to the out of the box generated file:

public class Program
    // Change the return type to Task to communicate
    // success/failure codes
    public static Task Main(string[] args)
        return CreateHostBuilder(args)

            // This replaces Build().Start() from the default
            // dotnet new templates

    public static IHostBuilder CreateHostBuilder(string[] args) =>

            // You can do the Jasper configuration inline with a 
            // Lambda, but here I've centralized the Jasper
            // configuration into a separate class

            .ConfigureWebHostDefaults(webBuilder =>

This isn’t mandatory, but there’s just enough Jasper configuration for this project with the outbox support that I opted to put the Jasper configuration in a new file called JasperConfig that inherits from JasperOptions:

public class JasperConfig : JasperOptions
    public override void Configure(IHostEnvironment hosting, IConfiguration config)
        if (hosting.IsDevelopment())
            // In development mode, we're just going to have the message persistence
            // schema objects dropped and rebuilt on app startup so you're
            // always starting from a clean slate
            Advanced.StorageProvisioning = StorageProvisioning.Rebuild;

        // Just the normal work to get the connection string out of
        // application configuration
        var connectionString = config.GetConnectionString("sqlserver");

        // Setting up Sql Server-backed message persistence
        // This requires a reference to Jasper.Persistence.SqlServer

        // Set up Entity Framework Core as the support
        // for Jasper's transactional middleware

        // Register the EF Core DbContext
        // You can register IoC services in this file in addition
        // to any kind of Startup.ConfigureServices() method,
        // but you probably only want to do it in one place or the 
        // other and not both.
            x => x.UseSqlServer(connectionString),

            // This is important! Using Singleton scoping
            // of the options allows Jasper + Lamar to significantly
            // optimize the runtime pipeline of the handlers that
            // use this DbContext type

Returning a Response to the HTTP Request

In the UseJasperAsMediatorController controller, we just passed the command into Jasper and let MVC return an HTTP status code 200 with no other context. If instead, we wanted to send down the ItemCreated message as a response to the HTTP caller, we could change the controller code to this:

public class WithResponseController : ControllerBase
    private readonly ICommandBus _bus;

    public WithResponseController(ICommandBus bus)
        _bus = bus;

    public Task<ItemCreated> Create([FromBody] CreateItemCommand command)
        // Using Jasper as a Mediator, and receive the
        // expected response from Jasper
        return _bus.Invoke<ItemCreated>(command);

“Do it All Explicitly Controller”

Just for a comparison, here’s the CreateItemCommand workflow implemented inline in a controller action with explicit code to handle the Jasper “Outbox” support:

// This controller does all the transactional work and business
// logic all by itself
public class DoItAllMyselfItemController : ControllerBase
    private readonly IMessageContext _messaging;
    private readonly ItemsDbContext _db;

    public DoItAllMyselfItemController(IMessageContext messaging, ItemsDbContext db)
        _messaging = messaging;
        _db = db;

    public async Task Create([FromBody] CreateItemCommand command)
        // Start the "Outbox" transaction
        await _messaging.EnlistInTransaction(_db);

        // Create a new Item entity
        var item = new Item
            Name = command.Name

        // Add the item to the current
        // DbContext unit of work

        // Publish an event to anyone
        // who cares that a new Item has
        // been created
        var @event = new ItemCreated
            Id = item.Id

        // Because the message context is enlisted in an
        // "outbox" transaction, these outgoing messages are
        // held until the ongoing transaction completes
        await _messaging.Send(@event);

        // Commit the unit of work. This will persist
        // both the Item entity we created above, and
        // also a Jasper Envelope for the outgoing
        // ItemCreated message
        await _db.SaveChangesAsync();

        // After the DbContext transaction succeeds, kick out
        // the persisted messages in the context "outbox"
        await _messaging.SendAllQueuedOutgoingMessages();

As a huge lesson learned from Jasper’s predecessor project, it’s always possible to easily bypass any kind of Jasper conventional “magic” and write explicit code as necessary.

There’s a lot more to say about Jasper and you can find a *lot* more information on its documentation website. I’ll be back sometime soon with more examples of Jasper, with probably some focus on functionality that goes beyond other mediator tools.

In the next post, I’ll talk about Jasper’s runtime execution pipeline and how it’s very different than other .Net tools with similar functionality (hint, it involves a boatload less generics magic than anything else).


Using Alba for Integration Testing ASP.Net Core Web Services

There’s a video of David Giard and I talking about Alba at CodeMash a couple years ago on YouTube if you’re interested.

Alba is a helper library for doing automated testing against ASP.Net Core web services. Alba started out as a subsystem of the defunct FubuMVC project for HTTP contract testing, but was later extracted into its own project distributed via Nuget, ported to ASP.Net Core, and finally re-wired to incorporate TestServer internally.

There are certainly other advantages, but I think the biggest selling point of adopting ASP.Net Core is the ability to run HTTP requests through your system completely in memory without any other kind of external web server setup. In my experience, this has made automated integration testing of ASP.Net Core applications far simpler compared to older versions of ASP.Net.

We could use FubuMVC’s OWIN support with or without Katana to perform the same kind of automated testing that I’m demonstrating here years ago, but hardly anyone ever used that and what I’m showing here is significantly easier to use.

Why not just use TestServer you ask? You certainly can, but Alba provides a lot of helper functionality around exercising HTTP web services that will make your tests much more readable and remove a lot of the repetitive coding you’d have to do with just using TestServer.

To demonstrate Alba, I published a small solution called AlbaIntegrationTesting to GitHub this morning.

Starting with a small web services application generated with the dotnet new webapi template, I added a small endpoint that we’ll later write specifications against:

public class Answers
    public int Product { get; set; }
    public int Sum { get; set; }

public class AdditionController : ControllerBase
    public Answers DoMath(int one, int two)
        return new Answers
            Product = one * two,
            Sum = one + two

Next, I added a new test project using xUnit.Net with the dotnet new xunit template. With the new skeleton testing project I:

  • Added a reference to the latest Alba Nuget
  • My preference for testing assertions in .Net is Shouldly, so I added a Nuget reference for that as well

Since bootstrapping an ASP.Net Core system is non-trivial in terms of performance, I utilized xUnit.Net’s support for sharing context between tests so that the application only has to be spun up once in the test suite. The first step of that is to create a new “fixture” class that holds the ASP.Net Core system in memory like so:

public class AppFixture : IDisposable
    public AppFixture()
        // Use the application configuration the way that it is in the real application
        // project
        var builder = Program.CreateHostBuilder(new string[0])
            // You may need to do this for any static 
            // content or files in the main application including
            // appsettings.json files
            // DirectoryFinder is an Alba helper
            // Override the hosting environment to "Testing"

        // This is the Alba scenario wrapper around
        // TestServer and an active .Net Core IHost
        System = new SystemUnderTest(builder);

        // There's also a BeforeEachAsync() signature
        System.BeforeEach(httpContext =>
            // Take any kind of setup action before
            // each simulated HTTP request
            // In this case, I'm setting a fake JWT token on each request
            // as a demonstration
            httpContext.Request.Headers["Authorization"] = $"Bearer {generateToken()}";

        System.AfterEach(httpContext =>
            // Take any kind of teardown action after
            // each simulated HTTP request


    private string generateToken()
        // In a current project, we implement this method
        // to create a valid JWT token with the claims that
        // the web services require
        return "Fake";

    public SystemUnderTest System { get; }

    public void Dispose()

This new AppFixture class will only be built once by xUnit.Net and shared between test unit classes using the IClassFixture<AppFixture> interface.

Do note that you can express some actions in Alba to take immediately before or after executing an HTTP request through your system for typical setup or teardown operations. In one of my current projects, we exploit this capability to add a pre-canned JWT to the request headers that’s required by our system. In our case, that’s allowing us to test the security integration and the multi-tenancy support that depends on the JWT claims at the same time we’re exercising controller actions and the associated database access underneath it. If you’re familiar with the test pyramid idea, these tests are the middle layers of our testing pyramid.

To simplify the xUnit.Net usage throughout the testing suite, I like to introduce a base class that inevitable accumulates utility methods for running Alba or common database setup and teardown functions. I tend to call this class IntegrationContext and here’s a sample:

public abstract class IntegrationContext : IClassFixture<AppFixture>
    protected IntegrationContext(AppFixture fixture)
        Fixture = fixture;

    public AppFixture Fixture { get; }

    /// <summary>
    /// Runs Alba HTTP scenarios through your ASP.Net Core system
    /// </summary>
    /// <param name="configure"></param>
    /// <returns></returns>
    protected Task<IScenarioResult> Scenario(Action<Scenario> configure)
        return Fixture.System.Scenario(configure);

    // The Alba system
    protected SystemUnderTest System => Fixture.System;

    // Just a convenience because you use it pretty often
    // in tests to get at application services
    protected IServiceProvider Services => Fixture.System.Services;


Now with the AppFixture and IntegrationContext in place, let’s write some specifications against the AdditionController endpoint shown earlier in this post:

public class DoMathSpecs : IntegrationContext
    public DoMathSpecs(AppFixture fixture) : base(fixture)

    // This specification uses the shorthand helpers in Alba
    // that's useful when you really only care about the data
    // going in or out of the HTTP endpoint
    public async Task do_some_math_adds_and_multiples_shorthand()
        var answers = await System.GetAsJson<Answers>("/math/3/4");
    // This specification shows the longhand way of executing an
    // Alba scenario and using some of its declarative assertions
    // about the expected HTTP response
    public async Task do_some_math_adds_and_multiples_longhand()
        var result = await Scenario(x =>
            x.ContentTypeShouldBe("application/json; charset=utf-8");

        var answers = result.ResponseBody.ReadAsJson<Answers>();

Alba can do a lot more to work with the HTTP requests and responses, but I hope this gives you a quick introduction to using Alba for integration testing.

Using Oakton for Development-Time Commands in .Net Core Applications

All of the sample code in this blog post is on GitHub in the OaktonDevelopmentCommands project.

Last year I released the Oakton.AspNetCore library that provides an expanded command line experience for ASP.Net Core applications (and environment tests too!) that adds additional command line flags for the basic dotnet run behavior. Oakton.AspNetCore also gives you the ability to embed completely different named command directly into your application — either from your own application or Oakton.AspNetCore can pull in commands from Nuget libraries.

To make this concrete, I started a new sample project on GitHub with the dotnet new webapi template. To add the new command line experience, I added a Nuget reference to Oakton.AspNetCore and modified the Program.Main() method to this:

// I changed the return type to Task
public static Task Main(string[] args)
    return CreateHostBuilder(args)
        // This extension method makes Oakton the active
        // command line parser and executor

To show the ability to add commands from an external library, I also swapped out the built in DI container with Lamar, and added a reference to the Lamar.Diagnostics Nuget to expose Lamar’s diagnostics reports via the command line.

Just to show the Lamar integration, I added just the one line of code to the host builder configuration you can see below:

public static IHostBuilder CreateHostBuilder(string[] args) =>
        .UseLamar() // Overriding the IoC container to use Lamar
        .ConfigureWebHostDefaults(webBuilder =>

Now, to switch the actual command line to see the results of all of that, you can see Oakton’s built in command line help by typing the command dotnet run --?, which gives this output:

    Available commands:
         check-env -> Execute all environment checks against the application
          describe -> Writes out a description of your running application to either the console or a file
    lamar-scanning -> Runs Lamar's type scanning diagnostics
    lamar-services -> List all the registered Lamar services
    lamar-validate -> Runs all the Lamar container validations
               run -> Runs the configured AspNetCore application
          sayhello -> I'm a simple command that just starts up the app and says hello

A couple notes about the output above:

  • With the dotnet cli mechanics, the basic usage is dotnet [command] [optional flags to dotnet itself] -- [arguments to your application]. The double dashes marks the boundaries between arguments and flags meant for dotnet itself and arguments or flags for your application. So as an example, to run your application compiled as “Release” and using the “Testing” environment name for your .Net Core application, the command would be dotnet run --configuration Release -- --environment Testing.
  • The check-env, describe, and run commands come in from the base Oakton.AspNetCore library. The run command is the default, so dotnet run actually delegates to the run command.
  • All the lamar-******* commands come from Lamar.Diagnostics because Oakton.AspNetCore can happily find and apply commands from other assemblies in the project.
  • We’ll build out sayhello later in this post

Assuming that you are a Lamar (or a StructureMap) user, the first step to unravel most IoC configuration problems is the old WhatDoIHave() method to see what the service registrations are. To get at that data faster, you can use the dotnet run -- lamar-services command just to dump out the WhatDoIHave() output to the console or a text file.

However, adding Lamar.Diagnostics to your application adds some dependencies you may not want to be deployed. With a little help from the .Net Core SDK project system, we can just tell .Net to only use Lamar.Diagnostics at development time by using the <PrivateAssets> tag in the csproj file like this:

<PackageReference Include="Lamar.Diagnostics" Version="1.1.5">


I don’t know how to do that without breaking into the csproj file, but once you do, the Lamar.Diagnostics assembly will not be deployed if you’re using dotnet publish to bundle up your application for deployment.

The “AspNetCore” naming is unfortunately a misnomer, because as of .Net Core 3.* the Oakton extension works for any project configured and bootstrapped with the new generic HostBuilder (purely console applications, worker applications, or any kind of web application).

Now, to add a custom command to the application directly into the application without polluting the deployed application, we’re just using some conditional compilation as shown in this super simple example:

// This is necessary to tell Oakton
// to search this assembly for Oakton commands

namespace OaktonDevelopmentCommands
    // The conditional compilation here just keeps this command from
    // being present in the Release build of the application
    // This is also an OaktonAsyncCommand if you need to 
    // call async APIs
    [Description("I'm a simple command that just starts up the app and says hello")]
    public class SayHelloCommand : OaktonCommand<NetCoreInput>
        public override bool Execute(NetCoreInput input)
            // Super cheesy, just starting up the application
            // and shutting it right down
            using (var host = input.BuildHost())
                // You do have access to the host's underlying
                // IoC provider, and hence to any application service
                // including the compiled IConfiguration as well
                Console.WriteLine("Hey, I can start up the application!");

            // Gotta return true to let Oakton know that the command succeeded
            // This is important if you're using commands that need to report
            // success or failure to the command line.
            return true;

It’s hoke-y, but the command up above will only be compiled into your application if you are compiling as “Debug” — as you generally do when you’re working locally or running tests. When you deploy as a “Release” build, this command won’t be part of the compiled executable. As silly as it looks, this is being very useful in one of my client projects right now.

If you want your OSS project to be successful…

Don’t take any of this too seriously because I wrote this really fast as I was procrastinating instead of working with some ugly legacy code today.

First off, I don’t know why the hell you’d pay attention to me because I’m only middling successful at OSS in terms of the effort I’ve put in over the years to whatever the benefits are. That being said, I do know quite a bit about what not to do, and that’s valuable.

First off, have the right idea and build something that solves some kind of common problem. It helps tremendously if you’re going into some kind of problem area without a lot of existing solutions. My aging StructureMap project was the very first production ready IoC tool for .Net. I seriously doubt it would have been terribly successful if it hadn’t been for that fact. Likewise, Marten has been successful because the idea of Postgresql as a Document Database just makes sense because of Postgresql’s unusually strong JSON support. I can say with some authority that the project concept was appealing because there were 3-4 other nascent OSS projects to do exactly what Marten became a couple years ago.

If you’re trying to go build a better mousetrap and walk into a problem domain with existing solutions, you just need to have some kind of compelling reason for folks to switch over to your tool. My example there would be Serilog. NLog or Log4Net are fine for what they are, but Serilog’s structured logging approach is different and provided some value beyond what the older logging alternatives did.

Try not to compete against some kind of tool from Microsoft itself. That’s just a losing proposition 95% of the time. And maybe realize upfront that the price of a significant success in .Net OSS means that Microsoft will eventually write their own version of whatever it is you’ve done.

Oh, and don’t do OSS at least in the .Net world if you’re trying to increase your professional visibility. I think it’s the other way around, you increase your visibility through talks, blog posts, and articles first. Then your OSS work will likely be more successful with more community visibility and credibility. Other folks might disagree with that, but that’s how I see it and what I’ve experienced myself both in good and bad ways.

If at all possible, dog-food your OSS project on a real work project. I can’t overstate how valuable that is to see how your tool really functions and to ensure it actually solves problems. The feedback cycle between finding a problem and getting it fixed is substantially faster when you’re dog-fooding your OSS project in a codebase where you have control versus getting GitHub issues from other people in code you’re not privy to. Moreover, it’s incredibly useful to see your colleagues using your OSS tool to find out how other folks reason about your tool’s API, try to apply it, and find out where you have usability issues. Lastly, you’re much more likely to have a good understanding of how to solve a technical problem that you actually have in your own projects.

All that being said about dog-fooding however, I’ve frequently used OSS work to teach myself new techniques, design ideas, and technologies.

Be as openly transparent about the project as you can early on and try to attract other contributors. Don’t go dark on a project or strive for perfection before releasing your project or starting to talk about it in public. I partially blame “going dark” prior to the v1.0 release for FubuMVC being such a colossal OSS failure for me. In 2011 I had a pretty packed room at CodeMash for a half day workshop on the very early FubuMVC, but I quit blogging or talking about it much for a couple years after that. When the FubuMVC team got a chance to do present again at CodeMash 2013 for the big v1.0 rollout, there were about a dozen people in the room and I was absolutely crushed.

Or just as valuable, try to get yourself some early adopters to get feedback and more ideas about the direction of the project. Not in terms of downloads and usage per se, but absolutely in terms of the technical achievement and community, Marten is by far and away the most successful OSS project I’ve had anything to do with over the years. A lot of this I would attribute to just having the right concept for a project, but much of that I believe is in how much effort I put into Marten very early on in publicizing it and blogging the project progress early. You know that saying that “with enough eyeballs, all bugs are shallow?” I have no idea if that’s actually true, but I can absolutely state that plenty of early user feedback and involvement has a magical ability to improve the usability of your OSS project.

Contributors come in all different types, and you want as many as you can attract. An eventual core team of contributors and project leaders is invaluable. Pull requests to implement functionality or fix bugs is certainly valuable. Someone who merely watches your project and occasionally points out usability problems or unclear documentation is absolutely helpful. People who take enough time to write useful GitHub issues contribute to projects getting better. Heck, folks who make itty bitty pull requests to improve the wording of the documentation help quite a bit, especially when there’s a bunch of those.

This deserves its own full fledged conversation, but make sure there’s enough README documentation and build automation to get new contributors up and running fast in the codebase. If you’re living in the .Net space, you make sure that folks can just open up Visual Studio.Net (or better IDE tools like JetBrains Rider;) and go. If there is any other kind of environment setup, get that scripted out where a simple “build” or “./” command of some sort or another build out whatever they need to run your code and especially the tests. Docker and docker-compose is turning out to be hugely helpful for this. This might force you to give up your favorite build automation tooling in favor of something mainstream (sorry Rake, I still love you). And don’t be like MS teams used to be and always use all kinds of weird Visual Studio.Net project extensions that aren’t on most folks’s box.

Documentation is unfortunately a big deal. It’s not just a matter of having documentation though, it’s also vital to make that documentation easy to edit and keep up to date because your project will change over time. Years ago I was frequently criticized online for StructureMap having essentially no documentation. As part of the huge (at the time) StructureMap 2.5 release I went and wrote up a ton of documentation with code samples in a big static HTML website. The 2.5 release had some annoying fluent interface APIs that nobody liked (including me), so I started introducing some streamlined API usage in 2.6 that still exists in the latest StructureMap (and Lamar) — which was great, except now the documentation was all out of date and people really laid into me online about that for years.

That horrendous experience with the StructureMap documentation led to me now having some personal rules for how I approach the technical documentation on ongoing OSS projects:

  1. Make the project documentation website quick to re-publish because it’s going to change over time. Your best hope to keeping documentation up to date is to make it as painless as possible to update and publish documentation updates
  2. Just give up and author your documentation in markdown one way or another because most developers at this point understand it
  3. Embed code samples in the documentation in some sort of way where they can be kept up to date
  4. As silly as it may sound, use some kind of “Edit me on GitHub” link in each page in your documentation website that let’s random people quickly whip up little pull requests to improve the documentation. You have no idea how helpful that’s been to me over the past 3-4 years.
  5. Make it easy to update and preview the documentation website locally. That helps tremendously for other folks making little contributions.

There are other solutions for the living documentation idea, but all of the projects I’m involved with use the Storyteller “stdocs” tooling (which I of course wrote over years of dog-fooding:)) to manage “living documentation” websites.

Try to be responsive to user problems and questions. This is a “do as I say, not as I do” kind of thing because I can frequently be slow to acknowledge GitHub issues or questions at times. Just letting them know that “I saw this” can buy you some good will. Try to be cool in the face of folks being assholes to you online. Remember that interactions and folks generally come off worse online than they would in real life with body language cues, and also remember that you’re generally dealing with people when they’re already frustrated.

Assuming that your project is on GitHub, I highly recommend Gitter as a way to communicate with users. I find the two way communication a lot faster to help unwind user problems compared to working asynchronously in GitHub issues.