Comment by pydry
10 hours ago
DRY is misunderstood. It's definitely a fundamental aspect of code quality it's just one of about 4 and maximizing it to the exclusion of the others is where things go wrong. Usually it comes at the expense of loose coupling (which is equally fundamental).
The goal ought to be to aim for a local minima of all of these qualities.
Some people just want to toss DRY away entirely though or be uselessly vague about when to apply it ("use it when it makes sense") and thats not really much better than being a DRY fundamentalist.
DRY is misnamed. I prefer stating it as SPOT — Single Point Of Truth. Another way to state it is this: If, when one instance changes in the future, the other instance should change identically, then make it a single instance. That’s really the only DRY criterion.
I like this a lot more, because it captures whether two things are necessarily the same or just happen to be currently the same.
A common "failure" of DRY is coupling together two things that only happened to bear similarity while they were both new, and then being unable to pick them apart properly later.
> then being unable to pick them apart properly later.
Which is often caused by the "midlayer mistake" https://lwn.net/Articles/336262/
I said this elsewhere in the comments, but I think there's sort of a fundamental tension that shows up sometimes between DRY and "a function/class should only do one thing." E.g., there might be two places in your code that do almost identical things, so there's a temptation to say "I know! I'll make a common function, I'll just need to add a flag/extra argument..." and if you keep doing that you end up with messy "DRY" functions with tons of conditional logic that tries to do too much.
Yeah there are ways to avoid this and you need to strike balances, but sometimes you have to be careful and resist the temptation to DRY everything up 'cuz you might just make it brittler (pun intended).
That’s how I understand it as well. It’s not about an abstract ideal of duplication but about making your life easier and your software less buggy. If you have to manually change something in 5 different places, there’s a good chance you’ll forget one of those places at some point and introduce a bug.
Yes, and there are many different kinds of truth, so when two arise together—we can call this connascence—we can categorize how these instances overlap: Connascence of Name, Connascence of Algorithm, etc.
That's how I understood it. If you add a new thing (constant, route, feature flag, property, DB table) and it immediately needs to be added in 4 different places (4 seems to be the standard in my current project) before you can use it, that's not DRY.
> If you add a new thing (constant, route, feature flag, property, DB table) and it immediately needs to be added in 4 different places (4 seems to be the standard in my current project) before you can use it, that's not DRY.
The tricky part is that sometimes "a new thing" is really "four new things" disguised as one. A database table is a great example because it's a failure mode I've seen many times. A developer has to do it once and they have to add what they perceive as the same thing four times: the database table itself, the internal DB->code translation e.g. ORM mapping, the API definition, and maybe a CRUD UI widget. The developer thinks, "oh, this isn't DRY" and looks to tools like Alembic and PostGREST or Postgraphile to handle this end-to-end; now you only need to write to one place when adding a database table, great!
It works great at first, then more complex requirements come down: the database gets some virtual generated columns which shouldn't be exposed in code, the API shouldn't return certain fields, the UI needs to work off denormalized views. Suddenly what appeared to be the same thing four times is now four different things, except there's a framework in place which treats these four things as one, and the challenge is now decoupling them.
Thankfully most good modern frameworks have escape valves for when your requirements get more complicated, but a lot of older ones[0] really locked you in and it became a nightmare to deal with.
[0] really old versions of Entity Framework being the best/worst example.
1 reply →
Renaming it doesnt change the nature of the problem.
There should often be two points of truth because having one would increase the coupling cost more than the benefits that would be derived from deduplication.