davenicolette

davenicolette, 7 days ago

@baldur The tone is fine.

davenicolette, 12 days ago

@jamesshore @jasongorman I've been arguing with myself about whether to post this comment, because of the deep respect I have for James Shore's work. But here goes. I have doubts about the reasons nullables are described as better than mocks in this article: https://www.jamesshore.com/v2/projects/nullables/how-are-nullables-different-from-mocks
(1/3)

davenicolette, 12 days ago

@jamesshore @jasongorman The first doubt is about something implicit in the article, not stated explicitly - that sociable tests should be written instead of isolated tests. I'm not convinced of that. I think both types of tests yield value.
The second doubt concerns the sample test - isolated and interaction-based. ISTM the solution to this is simply not to write such tests. They're aren't useful. Isolated tests that check the behavior of a unit of code are useful. (2/3)

davenicolette, 12 days ago

@jamesshore @jasongorman The third doubt concerns tests that break when you refactor. If we follow the guideline to write tests first, then it's literally impossible to cause a test to break by changing the production code - whether it's an intentional behavioral change or a refactoring.
I can see value in nullables, but not the kinds of value described in the article. In present company I'm definitely the dumbest person in the room, so please set me straight about this. Thanks. (3/3)

davenicolette, 11 days ago

@jamesshore @jasongorman Initial impression (haven't finished all the material yet): The sample command line test with an "off" switch sure looks/feels a lot like a hand-rolled test double. Just a first impression.

davenicolette, 11 days ago

@jamesshore @jasongorman Under Tradeoffs, I'm not sure "hand-written stub code" and "multiple test failures" are introduced by sociable tests; they're common with all kinds of tests and approaches to testing.
Also, "changes to production code" seem to be pretty similar to the changes we make to enable feature toggles; so this isn't a tradeoff either. No tradeoffs so far, really, unless I'm reading it wrong.

davenicolette, 11 days ago

@jamesshore @jasongorman Now I see what you mean by "interaction tests." The same as what I usually call "implementation-aware tests." Generally not such a good idea, by any name.

And by "state-based" you just mean "behavior-focused". OK.

I'm getting the gist now. A Narrow Test is a unit test. Overlapping Sociable Tests are integration tests. Mainly a matter of terminology so far. Still learning.

davenicolette, 11 days ago

@jasongorman @RonJeffries @jamesshore The rest of the article at https://www.jamesshore.com/v2/projects/nullables/testing-without-mocks contains great general advice about basic software design that all developers should learn. Haven't discovered anything particularly "new" yet, but everything so far is excellent on the merits as far as I can tell. Still learning.

davenicolette, 11 days ago

@jasongorman @RonJeffries @jamesshore The idea of modifying production code to accommodate test code bothers me a little because it might increase the threat surface of the code. The existence of methods that are visible to test code but ostensibly not to production code "feels" a little like the old practice of having a patch area in object programs to facilitate hot fixes in production. I guess it's no worse than other approaches with OO languages.

davenicolette, 11 days ago

@jasongorman @RonJeffries @jamesshore I think I'd like to do some sort of small project using this technique to see how it goes. Maybe one of the refactoring katas or something. Reading is fine but I think the best way to get a feel for something like this is to put it into practice. Thanks for publishing it!

davenicolette, 9 days ago

@jasongorman @RonJeffries @jamesshore I started exploring Nullables with Java. I plan to continue to try it with other languages. So far, I've seen few advantages over using mocks. That leads me to suspect I don't properly understand the approach. Please let me know where I've gone astray. https://github.com/neopragma/sociabletests

davenicolette, 8 days ago

@jamesshore @RonJeffries @jasongorman Thanks for the detailed feedback. I'll adjust the code accordingly. The stubs were separate originally and I factored them out to reduce duplication, as you surmised. The only difference of opinion is about names. The kata provides two files, but the input data source needn't be a file, so I would avoid using the word File in a class name. A minor thing, I guess.

(1/2)

davenicolette, 8 days ago

@jamesshore @RonJeffries @jasongorman I'll make the changes in the form of further progressions in development and modify the notes as well, so others can see me make the initial mistakes and how I corrected them based on your guidance. That might make the project more useful to people who want to get started with this.
(2/2)

davenicolette, 8 days ago

@jamesshore @RonJeffries @jasongorman Then I suggest the following: I'll use the naming convention you recommend, and at a future stage of the exploration (maybe when we're trying another language) we can introduce a different type of data store, and see how that affects the existing code at that point. That might help make the project useful, as well.

davenicolette, 14 days ago

@luckytran Reminds me of police and national guard responses to student protests against the Vietnam war in the 1960s-70s.

davenicolette, 15 days ago

@baldur The review is certainly well-written. I've found that when a movie is over-hyped, I end up disappointed even if it turns out to be pretty good. I would have enjoyed it more without the hype. FWIW I thought Civil War was a good action movie about journalists in a war zone. The main characters change and grow through their experiences, which is characteristic of a good story. I'd recommend seeing it, but don't set your expectations too high.

davenicolette, 20 days ago

@gdinwiddie With respect, you don't know what I'm proposing. You're reacting to a few words and filling in the blanks from your own perspective. That's understandable.

I would like you to understand what I'm proposing. As I said, I'll see how I can get my notes online. It's a tl;dr thing. Then you can criticize the actual idea rather than relying on assumptions. I would love that because it would help me improve it.

davenicolette, 20 days ago

@gdinwiddie In the meantime, let me address the other two questions in your last reply.

I don't think the Confederates are interested in "engaging." We're past that point, by their choice, not mine.

You ask why am I not moving somewhere myself, without first asking whether I already have. I'm where I want to be.

And yes, moving within the US is pretty easy.

davenicolette, 20 days ago

@gdinwiddie OK George, I threw some HTML documents up on Github Pages. I hope you'll take a few moments to read them and provide feedback - this time, feedback on the actual ideas. I appreciate your intelligence and perspective. I expect you to be strongly opposed to every single idea. Don't disappoint me on that score, my friend!
https://neopragma.github.io/state/

davenicolette, 19 days ago

@gdinwiddie Hi George. I know you have other things to do besides looking at this stuff, but I really would appreciate your feedback on the idea. I'm especially interested in following up on your statement that I'm not thinking through the details. I'm sure there are details I haven't thought of. Maybe you can help identify areas that require deeper consideration. Thanks in advance!

davenicolette, 19 days ago

@gdinwiddie That's interesting. The whole thing is a proposal for a new government. What wording or phrasing would make this clearer, in your opinion?

davenicolette, 19 days ago

@gdinwiddie Fair enough, but that dosn't help me make the message clearer. There's a Manifesto, a Constitution, a Legal Code, a series of background essays, and a Transition Plan for replacing the US with the American State. Which part of that doesn't look like a proposal to you? Which parts are not detailed enough?

Do phrases like, "The proposed model..." and "The purpose of the proposed political system..." not suggest there's some sort of proposal?

davenicolette, 19 days ago

@gdinwiddie So you read a very small part of the material, and now you say you didn't see things that are documented in the rest of the material. Well, thanks for glancing over it, anyway.

davenicolette, 18 days ago

@gdinwiddie Thank you for the thoughtful response. As you and I do the same general sort of work, I'm sure you're familiar with the idea of "remembering the future," in which we ask people who are stuck in a bad work situation to imagine what a better situation would look like and visualize how it might differ from the present. (1/2)