Automated tests for software
Nov. 8th, 2022 01:40 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
I have known for many years that I enjoy devising software solutions to difficult problems, then implementing and demonstrating them. A more recent discovery for me is that I enjoy writing automated tests for the software I write. I wonder if this is partly because I like writing code when I have a good understanding of what is going on and, having implemented code and got it appearing to work, mostly I already know how to write tests for it and can then just get on with that productively.
Agile purists may ask, weren't the tests already written? Not all of us get to work in such a shielded environment. When a deadline is pressing and another developer is made more productive by my piece already working, even buggily, then I'll get off the critical path as soon as I can. When I am proposing projects to wealthy, patient customers then I build the test-writing into every task on the chart, but I have not been in that enviable position for years now.
Agile purists may ask, weren't the tests already written? Not all of us get to work in such a shielded environment. When a deadline is pressing and another developer is made more productive by my piece already working, even buggily, then I'll get off the critical path as soon as I can. When I am proposing projects to wealthy, patient customers then I build the test-writing into every task on the chart, but I have not been in that enviable position for years now.
no subject
Date: 2022-11-08 04:51 pm (UTC)no subject
Date: 2022-11-09 09:10 am (UTC)The main disadvantage I've seen is where the talking to (prospective) users is very separate from the development effort, the developers seeing only the stories that are now or next on the roadmap. Having an understanding of the technical detail that's coming down the pipe in much later sprints can heavily inform how earlier stories ought to be implemented and that foundation can be expensive to adjust once built on.
no subject
Date: 2022-11-09 03:02 pm (UTC)My projects were variously big. I think perhaps my largest project was rewriting a dBase III position budgeting/tracking system for the budget department of a state agency. They needed some mods, I printed the code and it turned out to be a gigantic loop! So I dumped the code and rewrote the whole thing to be modular. This was back in the late '80s. My co-worker wrote a really nice menu interface based on the Lotus 1-2-3 menu, and we implemented record locking and a full audit. It was a REALLY sweet system. I had regular meetings with the stakeholders, they were always in the loop, and they were over the moon when we deployed. I got a commendation from the head of the finance dept when we were done. I kinda wish I had the code, I'm wondering how much I would understand it! I haven't touched that language since then, I don't know if it really exists anymore with SQL and its offspring ruling the world now.
no subject
Date: 2022-11-10 08:38 am (UTC)no subject
Date: 2022-11-11 11:52 am (UTC)I, too, am discovering that there is pleasure in setting up a decent set up unit tests. Fortunately for me though, those tests are a required part of the workflow to demonstrate that the models I write are Doing The Thing.
no subject
Date: 2022-11-15 10:13 am (UTC)no subject
Date: 2022-12-01 08:47 pm (UTC)I never got in to Power BI stuff as it wasn't tech that the government entities that I worked for were interested in. They tended to already be invested in Peoplesoft-type things and had their consultants outside of IT, so I just sat back and did my server wrangling. And now I'm just coasting towards retirement and learning the stuff that I want to learn for my own interests and pleasures.
no subject
Date: 2022-12-01 11:04 pm (UTC)