Simplify your code: Functional core, imperative shell

https://testing.googleblog.com/2025/10/simplify-your-code-functional-core.html

by reqo • 9 months ago

411 213 9 months ago

testing.googleblog.com

hinkley5• 9 months ago

Bertrand Meyer suggested another way to consider this that ends up in a similar place.

For concerns of code complexity and verification, code that asks a question and code that acts on the answers should be separated. Asking can be done as pure code, and if done as such, only ever needs unit tests. The doing is the imperative part, and it requires much slower tests that are much more expensive to evolve with your changing requirements and system design.

The one place this advice falls down is security - having functions that do things without verifying preconditions are exploitable, and they are easy to accidentally expose to third party code through the addition of subsequent features, even if initially they are unreachable. Sun biffed this way a couple of times with Java.

But for non crosscutting concerns this advice can also be a step toward FC/IS, both in structuring the code and acclimating devs to the paradigm. Because you can start extracting pure code sections in place.

Jtsummers6• 9 months ago

Command-Query Separation is the term for that. However, I find this statement odd:

having functions that do things without verifying preconditions are exploitable

Why would you do this? The separation between commands and queries does not mean that executing a command must succeed. It can still fail. Put queries inside the commands (but do not return the query results, that's the job of the query itself) and branch based on the results. After executing a command which may fail, you can follow it with a query to see if it succeeded and, if not, why not.

https://en.wikipedia.org/wiki/Command%E2%80%93query_separati...

jonahx• 9 months ago

Why would you do this?

Performance and re-use are two possible reasons.

You may have a command sub-routine that is used by multiple higher-level commands, or even called multiple times within by a higher-level command. If the validation lives in the subroutine, that validation will be called multiple times, even when it only needs to be called once.

So you are forced to choose either efficiency or the security of colocating validation, which makes it impossible to call the sub-routine with unvalidated input.

hinkley5• 9 months ago

The example in the wiki page is far more rudimentary than the ones I encountered when I was shown this concept. Trivial, in fact.

CQS will rely on composition to do any If A Then B work, rather than entangling the two. Nothing forces composition except information hiding. So if you get your interface wrong someone can skip over a query that is meant to short circuit the command. The constraint system in Eiffel I don’t think is up to providing that sort of protection on its own (and the examples I was given very much assumed not). Elixir’s might end up better, but not by a transformative degree. And it remains to be seen how legible that code will be seen as by posterity.

layer82• 9 months ago

In asynchronous environments, you may not be able to repeat the same query with the same result (unless you control a cache of results, which has its own issues). If some condition is determined by the command’s implementation that subsequent code is interested in (a condition that isn’t preventing the command from succeeding), it’s generally more robust for the command to return that information to the caller, who then can make use of it. But now the command is also a query.

jakewins3• 9 months ago

I think CQRS is something different than what’s being described here. “Query” code in CQRS can still “do stuff”: call an external database, grab locks, audit trail recording etc.

What’s being described here is something lower level, that you keep as much code as you can as a side-effect-free “pure functional core”. That pattern is useful both for the “command” and “query” side of a CQRS system, and is not the same thing as CQRS

hinkley5• 9 months ago

Found an example that is closer to how I learned about CQS:

https://hemath.dev/blog/command-query-separation/

Down at the bottom it gets into composition to make utility functions that compose several operations. Any OO system has to be careful not to expose methods that should have been private, so that’s not specific to CQS. It’s just that the opportunities to get it wrong increase and the consequences are higher.

rcleveng• 9 months ago

If your language supports generators, this works a lot better than making copies of the entire dataset too.

KlayLay• 9 months ago

You don't need your programming language to implement generators for you. You can implement them yourself.

akshayshah• 9 months ago

Sometimes, sure - but sometimes, passing around a fat wrapper around a DB cursor is worse, and the code would be better off paginating and materializing each page of data in memory. As usual, it depends.

hackthemack4• 9 months ago

I never liked encountering code that chains functions calls together like this

email.bulkSend(generateExpiryEmails(getExpiredUsers(db.getUsers(), Date.now())));

Many times, it has confused my co-workers when an error creeps in in regards to where is the error happening and why? Of course, this could just be because I have always worked with low effort co-workers, hard to say.

I have to wonder if programming should have kept pascals distinction between functions that only return one thing and procedures that go off and manipulate other things and do not give a return value.

https://docs.pascal65.org/en/latest/langref/funcproc/

POiNTx3• 9 months ago

In Elixir this would be written as:

  db.getUsers()
  |> getExpiredUsers(Date.now())
  |> generateExpiryEmails()
  |> email.bulkSend()

I think Elixir hits the nail on the head when it comes to finding the right balance between functional and imperative style code.

montebicyclelo• 9 months ago

    bulk_send(
        generate_expiry_email(user) 
        for user in db.getUsers() 
        if is_expired(user, date.now())
    )

(...Just another flavour of syntax to look at)

time4tea2• 9 months ago

Not a single person in this thread commented on the use of Date.now() and similar - surely clock.now() - you never ever want to use global time in any code, how could you test it?

clock in this case is a thing that was supplied to the class or function. It could just be a function: () -> Instant.

(Setting a global mock clock is too evil, so don't suggest that!)

fedlarm• 9 months ago

You could write the logic in a more straight forward, but less composable way, so that all the logic resides in one pure function. This way you can also keep the code to only loop over the users once.

email.sendBulk(generateExpiryEmails(db.getUsers(), Date.now()));

tadfisher• 9 months ago

That's pretty hardcore, like you want to restrict the runtime substitution of function calls with their result values? Even Haskell doesn't go that far.

Generally you'd distinguish which function call introduces the error with the function call stack, which would include the location of each function's call-site, so maybe the "low-effort" label is accurate. But I could see a benefit in immediately knowing which functions are "pure" and "impure" in terms of manipulating non-local state. I don't think it changes any runtime behavior whatsoever, really, unless your runtime schedules function calls on an async queue and relies on the order in code for some reason.

My verdict is, "IDK", but worth investigating!

hackthemack4• 9 months ago

It has been so long since I worked on the code that had chaining functions and caused problems that I am not sure I can do justice to describing the problems.

I vaguely remember the problem was one function returned a very structured array dealing with regex matches. But there was something wrong with the regex where once in a blue moon, it returned something odd.

So, the chained functions did not error. It just did something weird.

Whenever weird problems would pop up, it was always passed to me. And when I looked at it, I said, well...

I am going to rewrite this chain into steps and debug each return. Then run through many different scenarios and that was how I figured out the regex was not quite correct.

mrkeen5• 9 months ago

you want to restrict the runtime substitution of function calls with their result values?

I don't get how you got there from parent comment.

Pascal just went with a needless syntax split of (side-effectful) methods and (side-effectful) functions.

sfn422• 9 months ago

I would have written each statement on its own line:

var users = db.getUsers();

var expiredUsers = getExpiredUsers(users, Date.now());

var expiryEmails = generateExpiryEmails(expiredUsers);

email.bulkSend(expiryEmails);

This is not only much easier to read, it's also easier to follow in a stack trace and it's easier to debug. IMO it's just flat out better unless you're code golfing.

I'd also combine the first two steps by creating a DB query that just gets expired users directly rather than fetching all users and filtering them in memory:

expiredUsers = db.getExpiredUsers(Date.now());

Now I'm probably mostly getting zero or a few users rather than thousands or millions.

hackthemack4• 9 months ago

Yeah. I did not mention what I would do, but what you wrote is pretty much what I prefer. I guess nobody likes it these days because it is old procedural style.

ajusa2• 9 months ago

(author here)

This is actually closer to the way the first draft of this article was written. Unfortunately, some readability was lost to make it fit on a single page. 100% agree that a statement like this is harder to reason about and should be broken up into multiple statements or chained to be on multiple lines.

codazoda• 9 months ago

Glad to see this. This style seems like it’s out of vogue now, but I find it much, much easier to reason about.

ahoka• 9 months ago

But then you are creating references with larger then needed "reachability".

netdevphoenix5• 9 months ago

Took me a bit of scrolling to find this. I believe most of the other folks are functional devs or something. The 5 functions on a single line wouldn't pass the code review in most .net/java shops.

The rule I was raised with was: you write the code once and someone in the future (even your future self) reads it 100 times.

You win nothing by having it all smashed together like sardines in a tin. Make it work, make it efficient and make it readable.

HiPhish2• 9 months ago

email.bulkSend(generateExpiryEmails(getExpiredUsers(db.getUsers(), Date.now())));

What makes it hard to reason about is that your code is one-dimensional, you have functions like `getExpiredUsers` and `generateExpiryEmails` which could be expressed as composition of more general functions. Here is how I would have written it in JavaScript:

    const emails = db.getUsers()
        .filter(user => user.isExpired(Date.now()))  // Some property every user has
        .map(generateExpiryEmail);  // Maps a single user to a message

    email.bulkSend(emails);

The idea is that you have small but general functions, methods and properties and then use higher-order functions and methods to compose them on the fly. This makes the code two-dimensional. The outer dimension (`filter` and `map`) tells the reader what is done (take all users, pick out only some, then turn each one into something else) while the outer dimension tells you how it is done. Note that there is no function `getExpiredUsers` that receives all users, instead there is a simple and more general `isExpired` method which is combined with `filter` to get the same result.

In a functional language with pipes it could be written in an arguably even more elegant design:

    db.getUsers() |> filter(User.isExpired(Date.now()) |> map(generateExpiryEmail) |> email.bulkSend

I also like Python's generator expressions which can express `map` and `filter` as a single expression:

    email.bulk_send(generate_expiry_email(user) for user in db.get_users() if user.is_expired(Date.now())

hackthemack4• 9 months ago

I guess I just never encounter code like this in the big enterprise code bases I have had to weed through.

Question. If you want to do one email for expired users and another for non expired users and another email for users that somehow have a date problem in their data....

Do you just do the const emails =

three different times?

In my coding world it looks a lot like doing a SELECT * ON users WHERE isExpired < Date.now

but in some cases you just grab it all, loop through it all, and do little switches to do different things based on different isExpired.

solid_fuel• 9 months ago

I may have gotten nerd sniped here, but I believe all of these examples so far have some subtle errors. Using elixir syntax, I would think something like this covers most of the cases:

    expiry_date = DateTime.now!("Etc/UTC")

    query = 
          from u in User,
          where: 
            u.expiry_date > ^expiry_date 
            and u.expiry_email_sent == false,
          select: u

    MyAppRepo.all(query)
    |> Enum.map(u, &generate_expiry_emails(&1, expiry_date))
    |> Email.bulkSend()  # Returns {:ok, %User{}} or {:err, _reason}
    |> Enum.filter(fn 
      {:ok, _} -> true
      _ -> false
    end)
    |> Enum.map(fn {:ok, user} ->
      User.changeset(user, %{expiry_email_sent: true})
      |> Repo.update()
    end)

Mainly a lot of these examples do the expiry filtering on the application side instead of the database side, and most would send expiry emails multiple times which may or may not be desired behavior, but definitely isn't the best behavior if you automatically rerun this job when it fails.

----

Edit: I actually see a few problems with this, too, since Email.bulkSend probably shouldn't know about which user each email is for. I always see a small impedance mismatch with this sort of pipeline, since if we sent the emails individually it would be easy to wrap it in a small function that passes the user through on failure.

If I were going to build a user contacting system like this I would probably want a separate table tracking emails sent, and I think that the email generation could be made pure, the function which actually sends email should probably update a record including a unique email_type id and a date last sent, providing an interface like: `send_email(user_query, email_id, email_template_function)`

lmm3• 9 months ago

I have to wonder if programming should have kept pascals distinction between functions that only return one thing and procedures that go off and manipulate other things and do not give a return value.

What you want is to use a language that has higher-kinded types and monads so that functions can have both effects (even multiple distinct kinds of effects) and return values, but the distinction between the two is clear, and when composing effectful functions you have to be explicit about how they compose. (You can still say "run these three possibly-erroring functions in a pipeline and return either the successful result or an error from whichever one failed", but you have to make a deliberate choice to).

Warwolt• 9 months ago

Making a distinction between pure and effectful functions doesnt require any kind of effect system though.

Having a language where "func" defines a pure function and "proc" defines a procedure that can performed arbitrary side effects (as in any imperative language really) would still be really useful, I think

sandeepkd• 9 months ago

On the same page here, read it multiple times to see if I can convince my mind, this is bit off in terms of reading the code as its being executed. There are high chances of people making mistakes over the time with such patterns. As usual there is always a trade off involved, readability is the one taking hit here.

tags2k• 9 months ago

Since everyone's giving !opinions, in my C# DDD world you'd ideally be able to:

  _unitOfWork.Begin();

  var users = await _usersRepo.Load(u => u.LastLogin <= whateverDate);
  users.CheckForExpiry();

  _unitOfWork.Commit();

That then writes the "send expiry email" commands from the aggregate, to an outbox, which a worker then picks up to send. Simple, transactional domain logic.

Antibabelic• 9 months ago

Ada is a great modern language that preserves the distinction between functions and procedures that you mention.

MarkMarine3• 9 months ago

These chains become easy to read and understand with a small language feature like the pipe operator (elixir) or threading macro (clojure) that takes the output of one line and injects it into the left or rightmost function parameter. For example: (Elixir) "go " |> String.duplicate(3) # "go go go " |> String.upcase() # "GO GO GO " |> String.replace_suffix(" ", "!") # "GO GO GO!"

(Clojure) ;; Nested function calls (map double (filter even? '(1 2 3 4)))

;; Using the thread-last macro (->> '(1 2 3 4) (filter even?) ; The list is passed as the last argument (map double)) ; The result of filter is passed as the last argument ;=> (4.0 8.0)

Things like this have been added to python via a library (Pipe) [1] and there is a proposal to add this to JavaScript [2]

1: https://pypi.org/project/pipe/ 2: https://github.com/tc39/proposal-pipeline-operator

netdevphoenix5• 9 months ago

If you get an exception, you might not know where it comes from unless you get a stack trace. Code looks nice but not practical imo

shortrounddev22• 9 months ago

It also invites exceptions as error handling instead of a monadic (result) pattern. I usually do something more like

    Result<Users> userRes = getExpiredUsers(db);
    if(isError(userRes)) {
        return userRes.error;
    }

    /* This probably wouldn't actually need to return a Result IRL */
    Result<Email> emailRes = generateExpireyEmails(userRes.value);
    if(isError(emailRes)) {
        return emailRes.error;
    }

    Result<SendResult> sendRes = sendEmails(emailRes.value);
    if(isError(sendRes)) {
        return sendRes.error;
    }
    
    return sendRes; // successful value, or just return a Unit type.

This is in my "functional C++" style, but you can write pipe helpers which sort of do the same thing:

    Result<SendResult> result = pipe(getExpiredUsers(db))
        .then(generateExpireyEmails)
        .then(sendEmails)
        .result();

    if(isError(result)) {
        return result.error;
    }

If an error result is returned by any of the functions, it terminates immediately and returns the error there. You can write this in most languages, even imperative/oop languages. In java, they have a built in class called Optional with options to treat null returns as empty:

    Optional.ofNullable(getExpiredUsers(db))
        .map(EmailService::generateExpireyEmails)
        .map(EmailService::sendEmails)
        .orElse(null);

or something close to that, I haven't used java in a couple years.

C++ also added a std::expected type in C++23:

    auto result = some_expected()
        .and_then(another_expected)
        .and_then(third_expected)
        .transform(/* ... some function here, I'm not familiar with the syntax*/);

taeric9• 9 months ago

This works right up to the point where you try to make the code to support opening transactions functional. :D

Some things are flat out imperative in nature. Open/close/acquire/release all come to mind. Yes, the RAI pattern is nice. But it seems to imply the opposite? Functional shell over an imperative core. Indeed, the general idea of imperative assembly comes to mind as the ultimate "core" for most software.

Edit: I certainly think having some sort of affordance in place to indicate if you are in different sections is nice.

agentultra• 9 months ago

whispers in monads

It can be done "functionally" but doesn't necessarily have to be done in an FP paradigm to use this pattern.

There are other strategies to push resource handling to the edges of the program: pools, allocators, etc.

taeric9• 9 months ago

Right, but even in those, you typically have the more imperative operations as the lower levels, no? Especially when you have things where the life cycle of what you are starting is longer than the life cycle of the code that you use to do it?

Consider your basic point of sale terminal. They get a payment token from your provider using the chip, but they don't resolve the transaction with your card/chip still inserted. I don't know any monad trick that would let that general flow appear in a static piece of the code?

garethrowlands4• 9 months ago

Indeed, the general idea of imperative assembly comes to mind as the ultimate "core" for most software.

That's not what functional core, imperative shell means though. It's a given that CPUs aren't functional. The advice is for people programming in languages that have expressions - ruby, in the case of the original talk. The functional paradigm mostly assumes automatic memory management.

taeric9• 9 months ago

Right, I was just using that as "at the extreme" and how it largely exists to allow you to put a functional feel on top of the imperative below it.

I'm sympathetic to the idea, as you can see it in most instruction manuals that people are likely to consume. The vast majority of which (all of them?) are imperative in nature. There is something about the "for the humans" layer being imperative. Step by step, if you will.

I don't know that it fully works, though. I do think you are well served being consistent in how you layer something. Where all code at a given layer should probably stick to the same styles. But knowing which should be the outer and which the inner? I'm not clear that we have to pick, here. Feel free to have more than two layers. :D

mrkeen5• 9 months ago

This works right up to the point where you try to make the code to support opening transactions

Indeed. It's all well and good to impart some kind of flavour into your code and call it functional, but transactions do not give a crap about style.

A transaction needs to be able to 'back out' to fulfill 'all-or-nothing' semantics. Side effects are what make this impossible.

garethrowlands4• 9 months ago

Surely transactions are a pretty good example of where functional core / imperative shell is a good guide. You really don't want to be doing arbitrary side effects inside your transaction because those can't be backed out. Check out STM in Haskell for a good example.

zkmon• 9 months ago

I think it's just your way of looking at things.

What if a FCF (functional core function) calls another FCF which calls another FCF? Or do we do we rule out such calls?

Object Orientation is only a skin-deep thing and it boils down to functions with call stack. The functions, in turn, boil down to a sequenced list of statements with IF and GOTO here and there. All that boils boils down to machine instructions.

So, at function level, it's all a tree of calls all the way down. Not just two layers of crust and core.

skydhash• 9 months ago

Functional core usually means pure functional functions, aka the return value is know if the arguments is known, no side effects required. All the side effects is then pushed up the imperative shell.

You’ll find usually that side effect in imperative actions is usually tied to the dependencies (database, storage, ui, network connections). It can be quite easy to isolate those dependencies then.

It’s ok to have several layers of core. But usually, it’s quite easy to have the actual dependency tree with interfaces and have the implementation as leaves for each node. But the actual benefits is very easy testing and validation. Also fast feedback due to only unit tests is needed for your business logic.

bitwize• 9 months ago

I invented this pattern when I was working on a small ecommerce system (written in Scheme, yay!) in the early 2000s. It just became much easier to do all the pricing calculations, which were subject to market conditions and customer choices, if I broke it up into steps and verified each step as a side-effect-free, data-in-data-out function.

Of course by "invented" I mean that far smarter people than me probably invented it far earlier, kinda like how I "invented" intrusive linked lists in my mid-teens to manage the set of sprites for a game. The idea came from my head as the most natural solution to the problem. But it did happen well before the programming blogosphere started making the pattern popular.

socketcluster2• 9 months ago

Even large companies are still grasping at straws when it comes to good code. Meanwhile there are articles I wrote years ago which explain clearly from first principles why the correct philosophy is "Generic core, specific shell."

I actually remember early in my career working for a small engineering/manufacturing prototyping firm which did its own software, there was a senior developer there who didn't speak very good English but he kept insisting that the "Business layer" should be on top. How right he was. I couldn't imagine how much wisdom and experience was packed in such simple, malformed sentences. Nothing else matters really. Functional vs imperative is a very minor point IMO, mostly a distraction.

benoitg• 9 months ago

I’d love to know more, do you have any links to your articles?

CharlesW• 9 months ago

"Specific on the surface, generic underneath" (Medium paywalled): https://medium.com/tech-renaissance/generic-internals-specif...

foofoo12• 9 months ago

Even large companies are still grasping at straws when it comes to good code

Probably many reasons for this, but what I've seen often is that once the code base has been degraded, it's a slippery slope downhill after that.

Adding functionality often requires more hacks. The alternative is to fix the mess, but that's not part of the task at hand.

stitched2gethr• 9 months ago

I've seen it many times. And then every task takes longer than the last one, which is what pushes teams to start rewrites. "There's never enough time to do it right, but always time to do it again."

motorest2• 9 months ago

Probably many reasons for this, but what I've seen often is that once the code base has been degraded, it's a slippery slope downhill after that.

Another factor, and perhaps the key factor, is that contrary to OP's extraordinary claim there is no such thing as objectively good code, or one single and true way of writing good code.

The crispest definition of "good code" is that it's not obviously bad code from a specific point of view. But points of view are also subjective.

Take for example domain-driven design. There are a myriad of books claiming it's an effective way to generate "good code". However, DDD has a strong object-oriented core, to the extent it's nearly a purist OO approach. But here we are, seeing claims that the core must be functional.

If OP's strong opinion on "good code" is so clear and obvious, why are there such critical disagreements at such a fundamental levels? Is everyone in the world wrong, and OP is the poor martyr that is cursed with being the only soul in the whole world who even knows what "good code" is?

Let's face it: the reason there is no such thing as "good code" is that opinionated people making claims such as OP's are actually passing off "good code" claims as proxy's for their own subjective and unverified personal taste. In a room full of developers, if you throw a rock at a random direction you're bound to hit one or two of these messiahs, and neither of them agrees on what good code is.

Hearing people like OP comment on "good code" is like hearing people comment on how their regional cuisine is the true definition of "good food".

frank_nitti2• 9 months ago

These are great and succinct, yours and your teammate’s.

I still find myself debating this internally, but one objective metric is how smoothly my longer PTOs go:

The only times I haven’t received a single emergency call were when I left teammates a a large and extremely specific set of shell scripts and/or executables that do exactly one thing. No configs, no args/opts (or ridiculously minimal), each named something like run-config-a-for-client-x-with-dataset-3.ps1 that took care of everything for one task I knew they’d need. Just double click this file when you get the new dataset, or clone/rename it and tweak line #8 if you need to run it for a new client, that kind of thing.

Looking inside the scripts/programs looks like the opposite of all of the DRY or any similar principles I’ve been taught (save for KISS and others similarly simplistic)

But the result speaks for itself. The further I go down that excessively basic path, the more people can get work done without me online, and I get to enjoy PTO. Anytime i make a slick flexible utility with pretty code and docs, I get the “any chance you could hop on?” text. Put the slick stuff in the core libraries and keep the executables dumb

timpieces• 9 months ago

Yes I feel that when to apply certain techniques is frequently under-discussed. But I can't blame people for err-ing on the side of 'do everything properly' - as this makes life more pleasant in teams. Although I think if you squint, the principle still applies to your example. The further you get from the 'core' of your platform/application/business/what-have-you, the less abstract you need to be.

chamomeal• 9 months ago

That is pretty convincing advice!! Maybe I’ll try writing more specific scripts.

zdc1• 9 months ago

I see a similar problem in infra-land where people expose too many config variables for too many things, creating more cruft. Knowing what to hardcode and what to expose as a var is something a lot of devs don't seem to understand; and don't realise they don't understand.

veqq• 9 months ago

The more specific, the more brittle. The more general, the more stable. Concerns evolve/decay at different speeds, so do not couple across shearing layers. Notice how grammar/phonology (structure) changes slowly while vocabulary (functions, services) changes faster.

...

Coupling across layers invites trouble (e.g. encoding business logic with “intuitive” names reflecting transient understanding). When requirements shift (features, regulations), library maintainers introduce breaking changes or new processor architectures appear, our stable foundations, complected with faster-moving parts, still crack!

https://alexalejandre.com/programming/coupling-language-and-...

js85• 9 months ago

"Generic core, specific shell."

Your advice is the opposite of "functional core, imperative shell". The FCIS principle has IS which is generic, to be simple, because it's usually hard to test (it deals with resources and external dependencies). So by being simple, it's more unit testable.

On the other hand, FC is where the business logic lives, which can be complex and specific. The reason why you want that "functional" (really just another name for "composable from small blocks") is because it can be tested for validity without external dependencies.

So the IS shields you from technicalities of external dependencies, like what kind of quirks your DB has, or are we sending data over network or writing to the file, or does the user inputs comands in spanish or english, do you display the green square or blue triangle to indicate the report is ready, etc.

On the other hand, FC deals with the actual business logic (what you want to do), which can be both generic and specific. These are just different types of building blocks (we call them functions) living in the FC.

FCIS is exemplified by user-shell interaction. The user (FC) dictates the commands and interprets the output, according to her "business needs". While the shell (IS) simply runs the commands, without any questions of their purpose. It's not the job of IS to verify or handle user errors caused by wrong commands.

But the user doesn't do stuff on her own; you could take her to a pub and she would tell you the same sequence of commands when facing the same situation. In that sense, the user is "functional" - independent on the actual state of the computer system, like the return value of a mathematical function is only dependent on the arguments.

Another example is MVC, where M is the FC and VC is the IS. Although it's not always exactly like that, for variety of reasons.

You can think of IS as a translator to a different language, understood by "the other systems", while the FC is there to supply what is actually being communicated.

Twisol7• 9 months ago

I disagree that these two pieces of advice are opposed. I think they are orthogonal at worst, and in agreement at best.

"Functional core, imperative shell" (FCIS) is a matter of implementing individual software components that need to engage with side-effects --- that is, they have some impact on some external resources. Rather than threading representations of the external resources throughout the implementation, FCIS tells us to expel those concerns to the boundary. This makes the bulk of the component easier to reason about, being concerned with pure values and mere descriptions of effects, and minimizes the amount of code that must deal with actual effects (i.e. turning descriptions of effects into actual effects). It's a matter of comprehensibility and testability, which I'll clumsily categorize as "verification": "Does it do what it's supposed to do?"

"Generic core, specific shell" (GCSS) is a matter of addressing needs in context. The problems we need solved will shift over time; rather than throwing away a solution and re-solving the new problem from scratch, we'd prefer to only change the parts that need changing. GCSS tells us we shouldn't simply solve the one and only problem in front of us; we should use our eyes and ears and human brains to understand the context in which that problem exists. We should produce a generic core that can be applied to a family of related problems, and adapt that to our specific problem at any specific time using a, yes, specific shell. It's a matter of adaptability and solving the right problem, which I'll clumsily categorize as "validation": "Is what it's supposed to do what we actually need it to do?"

Ideally, GCSS is applied recursively: a specific shell may adapt an only slightly more generic core, which then decomposes into a smaller handful of problems that are themselves implemented with GCSS. When business needs change in a way that the outermost "generic core" can't cover, odds are still good that some (or all) of its components can still be applied in solving the new top-level problem. FCIS isn't really amenable to the same recursion.

Both verification and validation activities are necessary. One is a matter of internal consistency within the component; the other is a matter of external consistency relative to the context the component is being used in. FCIS and GCSS advise on how to address each concern in turn.

zelphirkalt3• 9 months ago

A functional core can actually be very generic. There is nothing in the functional paradigm, that makes functions less generic. They are as generic or specific as you write them.

lelanthran• 9 months ago

the correct philosophy is "Generic core, specific shell."
Nothing else matters really. Functional vs imperative is a very minor point IMO, mostly a distraction.

I'm torn on this. This really is the faster way to higher quality.

OTOH, if more developers knew this, I wouldn't be so much more faster when I create my systems for clients. I'd just be a "normal 1x dev".

I like implementing features, sans AI-assistance, in my LoB applications faster than devs with Claude code doing so on their $FRAMEWORK.

zelphirkalt3• 9 months ago

Non-functional core tends to become a buggy mess, with workarounds in the shell to account for those bugs, and then one needs to know about the nature of the core to correctly use the shell and so on. Functional core will lend itself very well to unit tests. Writing them will almost be trivial, when functional core is done right. Imperative shell is then less of an issue, because the blast radius of bugs is reduced to one usage of the core. The imperative shell should be kept as small as reasonably possible of course.

spoiler2• 9 months ago

In my head, and the way its describe the generic and specific are swapped. The core handles a specific, pure problem. The core is generic (do you get the data via http, databases, filesystem, etc) then becomes irrelevant to the core problem

zelphirkalt3• 9 months ago

I would even go so far to say, that large companies even struggle with this more than small ones. The amount of people needing to know how to build things properly is larger than in a small company, where one knowledgeable engineer might already be sufficient. Too many cooks spoiling the soup(/broth?). And lots of people are cooking these days.

quietbritishjim• 9 months ago

Meanwhile there are articles I wrote years ago which explain clearly from first principles why the correct philosophy is ...

I think this is a very common mistake. You've spent years, maybe decades, writing code and now you want to magically transfer all that experience in a few succinct articles. But no advice that you give about "the correct philosophy" is going to instantly transfer enough knowledge to make all large companies write good code, if only they followed it. Instead, I'm sure it's valuable advice, but more along the lines of a fragment within a single day of learning for a diligent developer.

A company I worked recently had a more extreme version of this mistake. It had software written in the 1980s based on a development process by Michael Jackson (no, not that one!), a software researcher that had spent his whole career trying to come up with silly processes that were meant to fix software development once and for all; he wrote whole books about it. I remember reading a recent interview with him where he mourns that developers today are interested in new programming languages but not development methodologies. (The code base I worked on was fine by the way, given that it was 40 years old, but not really because of this Jackson stuff.)

I'm reminded of the Joel on Software article [1] where he compares talented (naturally or through experience) developers as being like really talented expert chefs, and those following some methodology as being like people working at McDonald's.

[1] https://www.joelonsoftware.com/2001/01/18/big-macs-vs-the-na...

Twisol7• 9 months ago

But no advice that you give about "the correct philosophy" is going to instantly transfer enough knowledge to make all large companies write good code, if only they followed it.

Good old "Programming as Theory Building". It's almost impossible to achieve this kind of transfer without already having the requisite lived experience.

[0] https://ratfactor.com/papers/naur1_theory_building.pdf

ericmcer2• 9 months ago

Isn't this saying business layer should not be on the top?

Business layers should be accessible via an explicit interface/shape that is agnostic to the layers above it. So if the org decides to move from mailchimp to some other email provider the business logic can remain untouched and you just need to write some code mapping the new provider to the business logic's interface.

Maybe our visualizations are mixed up, but I always viewed things like cloud providers, libraries etc. as potentially short lived whereas the core logic could stick around forever.

semiinfinitely• 9 months ago

this looks like a post from 2007 im shocked at the date

diamondtin2• 9 months ago

I saw Gary posted his blog link on twitter, and I really like his article. I really didn't expect it to surface up at this moment (2025), and it's referred from a google blog. :shrug:

vietvu• 9 months ago

Me too. Aren't we already doing this? This is the basic I have been taught first.

semiquaver• 9 months ago

Yeah, it’s based on an old post: https://www.destroyallsoftware.com/screencasts/catalog/funct...

mrkeen5• 9 months ago

And "I call it my billion-dollar mistake. It was the invention of the null reference in 1965" is from 2009.

Hopefully by 2045 these ideas will have gotten a little more traction.

johnrob• 9 months ago

Functions can have complexity or side effects, but not both.

anttiharju• 9 months ago

All pure functions have complexity?

jackbravo• 9 months ago

Reminds me of this clean architecture talk with Python explains this very well: https://www.youtube.com/watch?v=DJtef410XaM

diamondtin2• 9 months ago

destory all software

postepowanieadm• 9 months ago

Something like that was popular in perl world: functional core, oop external interface.

dominicrose• 9 months ago

When you're not relying on a compiler you just have to right good code. And it's easier if the code never has to change, only grow. I know a confident Perl programming who rarely changes his mind about anything. When he codes something he keeps it.

I always feel like I have to "maintain" code so I usually get bored after 3k lines of code, but truth is code doesn't have to be maintained if we like it the way it is, which obviously includes all the functionality that comes with it.

SafeDusk2• 9 months ago

One of the core design principles at https://github.com/aperoc/toolkami

BergAndCo• 9 months ago

Spam

SafeDusk2• 9 months ago

Not sure man, I specifically stated this in my README way before this post: https://github.com/aperoc/toolkami/blob/main/README.md#comma....

I mean it's not much, but the concept just resonates with me and I want to share it. Sad I can't share even simple opinion nowadays ...

chairhairair• 9 months ago

The for-loop is just better.

wslh• 9 months ago

I don't really like the example (and it's from Google) because, beyond the general concept, it seems like the trigger for sending emails is calling bulkSend with Date.now() instead of the user actually triggering an email when it's really expired: user.subscriptionEndDate change to < Date.now().

itsthecourier• 9 months ago

that's nice, so should I get all the db users and then filter them in app?

lmm3• 9 months ago

Probably. Or better yet move the code to run where the data is so you're not moving the data around.

CharlieDigital2• 9 months ago

I wrote our AI agents code with a functional core + imperative shell and I have to agree: this approach yields much faster cycle times because you can run pure unit tests and it makes testing a lot easier.

We have tens of thousands of lines of code for the platform and millions of workflow runs through them with no production errors coming from the core agent runtime which manages workflow state, variables, rehydration (suspend + resume). All of the errors and fragility are at the imperative shell (usually integrations).

Some of the examples in this thread I think get it wrong.

    db.getUsers() |> filter(User.isExpired(Date.now()) |> map(generateExpiryEmail) |> email.bulkSend

This is already wrong because the call already starts with I/O; flip it and it makes a lot more sense.

What you really want is (in TS, as an example):

    bulkSend(
      userFn: () => user[],
      filterFn: (user: User) => bool,
      expiryEmailProducerFn: (user: User) => Email,
      senderFn: (email: Email) => string
    )

The effect of this is that the inner logic of `bulkSend` is completely decoupled from I/O and external logic. Now there's no need for mocking or integration tests because it is possible to use pure unit tests by simply swapping out the functions. I can easily unit test `bulkSend` because I don't need to mock anything or know about the inner behavior.

I chose this approach because writing integration tests with LLM calls would make the testing run too slowly (and costly!) so most of the interaction with the LLM is simply a function passed into our core where there's a lot of logic of parsing and moving variables and state around. You can see here that you no longer need mocks and no longer need to spy on calls because in the unit test, you can pass in whatever function you need and you can simply observe if the function was called correctly without a spy.

It is easier than most folks think to adopt -- even in imperative languages -- by simply getting comfortable working with functions at the interfaces of your core API. Wherever you have I/O or a parameter that would be obtained from I/O (database call), replace it with a function that returns the data instead. Now you can write a pure unit test by just passing in a function in the test.

I am very surprised how many of the devs on the team never write code that passes a function down.

nickpsecurity2• 9 months ago

Great examples. We were taught to pass variables, scalar or compound, into API's. Most of us were never taught to pass functions.

Even Python examples in trainings that look functional might not be. They put the function calls in as arguments. The beginner thinks the function returns some data, that would be in a variable, and they are implicitly passing that variable. Might as well, for readability, do the function call first to pass a well-named variable instead.

That was my experience. That plus minimizing side effects in functions. I've yet to really learn functional programming where I'd think to pass a function in an API. What are the best articles or books for us to learn that in general or in Python?

globular-toast3• 9 months ago

It's called dependency injection and there's loads written about it. It's a really powerful technique which is also used for dependency inversion, key for decoupling components. I really like how it enables simple tests without any mocking.

The book Architecture Patterns in Python by Percival and Gregory is one of the few books that talks about this stuff using Python. It's available online and been posted on HN a few times before.

QuadmasterXLII• 9 months ago

I would argue that the real key is to have a distinct core and shell, and to hold the core to a much higher standard of quality than the shell. In this article, being "functional" is just serving as a proxy for code quality.

ccortes• 9 months ago

In this article, being "functional" is just serving as a proxy for code quality.

It is not, it is being very specific about what it means and what it is referring to

droningparrot• 9 months ago

Haskell practically encourages this style of programming. Any function that touches IO needs to wrap outputs with an appropriate monad. It becomes easier to push all IO out to the edges of your program and keep your core purely functional with no monads

kaashif• 9 months ago

I wish that's what people did, some codebases I've seen are messes of monad transformer stacks the likes of which you've never seen.

I mean, what if you want to do IO and have mutable data structures inside a do block? I'm afraid I'm going to have to prescribe you a monad transformer. Be careful of the side effects.

kmeisthax• 9 months ago

Maybe JavaScript's colored functions[0] were trying to tell us something

[0] https://journal.stuffwithstuff.com/2015/02/01/what-color-is-...

lucifer153• 9 months ago

This is same idea with onion architect in "Grokking Simplicity: Taming Complex Software with Functional Thinking Book by Eric Normand"

metalrain• 9 months ago

I like the idea but the example doesn't make much sense.

In what application would you load all users into memory from database and then filter them with TypeScript functions? And that is the problem with the otherwise sound idea "Functional core, imperative shell". The shell penetrates the core.

Maybe some filters don't match the way database is laid out, what if you have a lot of users, how do you deal with email batching and error handing?

So you have to write the functional core with the side effect context in mind, for example using query builder or DSL that matches the database conventions. Then weave it with the intricacies of your email sender logic, maybe you want iterator over the right size batches of emails to send at once, can it send multiple batches in parallel?

edf13• 9 months ago

In what application would you load all users into memory from database and then filter them with TypeScript functions?

You’d be surprised! I have worked on a legacy PHP service which did something very similar

bad_username• 9 months ago

I am surprised by this example, for the same reason.

Generally, performance is a top cause of abstraction leaks and the emergence of less-than-beautiful code. On an infinitely powerful machine it would be easy and advisable to program using neat abstracrions, using purely "the language of" the business. Our machines are not infinitely powerful, and that is especially evident when larger data sets are involved. That's where, to achieve useful performance, you have to increasingly speak "the language of" the machine. This is inevitable, and the big part of the programmer's skill is to be able to speak both "languages", to know when to speak which one, and produce readable code regardless.

Database programming is a prime example. There's a reason, for example, why ORMs are very messy and constitute such excellent footguns: they try to gap this bridge, but inevitably fail in important ways. And having and ORM in the example would, most likely, violate the "functional core" principle from the article.

So it looks like the author accidentally presented a very good counterexample to their own idea. I like the idea though, and I would love to know how to resolve the issue.

vivzkestrel• 9 months ago

db.getUsers() I am sorry what? Who in their right mind loads all users from the database and then filters out the expired subscription ones. Shouldn't the database query do this?

ryangibb• 9 months ago

The MirageOS project [0] is a great collection of functionality pure OCaml libraries that are useful outside of unikernels. I've used the DNS library with an effectful layer for various nameserver experiments [1].

[0] https://mirage.io/

[1] https://ryan.freumh.org/eon.html

pjmlp• 9 months ago

All nice ideas, that unfortunately don't get appreciated on the age of offshoring and vibe coding.

Have to ship it non matter what.

svat• 9 months ago

Another good blog post that is IMO in the same vein: https://lambdaisland.com/blog/2022-03-10-mechanism-vs-policy (“Improve your code by separating mechanism from policy”). This blends harmoniously with “functional core, imperative shell”—the "mechanism" code is the "functional core", and the "policy" code is the "imperative shell"—and also a little bit with John Ousterhout's idea in A Philosophy of Software Design of "deep modules" (in this context, don't put policy stuff, i.e. arbitrary decisions, inside the module).

urxvtcd2• 9 months ago

I have written a small system in Elixir adhering to FCIS. Not used to the approach, I was pretty slow and sometimes it felt like jumping through hoops set by myself, lol, but I loved it, the code was very clean, testable, and refactorable. Highly recommend it as an exercise, it was surprising just how much state and IO can be pushed out.

procaryote3• 9 months ago

I like the general idea, but unless you're assuming some very clever language or even more clever ORM that fixes things implicitly, wouldn't

    email.bulkSend(generateReminderEmails(getExpiredUsers(db.getUsers(), fiveDaysFromNow)));

get all users and then filter out the few that will expire in 5 days, on a code level? That doesn't sound like it would scale

bribri• 9 months ago

I agree. They could have picked a better example. Just db.getUsers() alone should set off alarm bells as soon as you see it.

globular-toast3• 9 months ago

Irrelevant. The "bad" code does it too. It's talking about something specific and not "let's fix all the problems with this code".

procaryote3• 9 months ago

If you write one method poorly so you select too much and filter in a for loop, you just have a bad method you can fix.

If you pick and recommend a pattern where filtering should happen separately from retrieving, all your code will be bad

Give a man a fish / teach a man to fish, but bad.

spoiler2• 9 months ago

I think it's just a contrived example. They probably wanted to show more than a single thing composing in a very short post given it's from their Toilet series.

Replace it with `getUsers(filters)` or even a specialised function, and it starts making more sense.

ajusa2• 9 months ago

(author here)

It's exactly this - I do regret using "db" a bit now after reading all of the comments here, as it's taken away focus from the main point. But yes, the post had to fit on a single page, and I needed to pick something that most engineers would be familiar with.

dawidloubser• 9 months ago

Often the answer to "some very clever language" is a lazy functional language like Haskell, where it's quite common to express problems naturally on infinite or near-infinite lists (of numbers, users, whatever...) in this way - and the language's lazy evaluation semantics effectively turning it into an efficient streaming data pipeline of sorts.

But even if not, the example from the article is just hypothetical. `db.getUsers()` could be something that just retrieves rather efficient `[UserEmail, ExpiryTime]` pairs, and then you'd have to have a pretty enormous user base for it to not scale (a couple of million string/date pairs in memory should be no problem).

procaryote3• 9 months ago

Modern computers are fast enough that a lot of bad code won't outright break. That doesn't necessarily make it good code

I fixed a performance issue at some point where a missing index meant a scan of millions of rows every login. It worked, and could log in 3 people per second or so. It was still terrible code.

rockyj• 9 months ago

Interestingly, I have been harping on this for a while. Recently wrote a blog on how to separate business logic from infrastructure code and tie them together by composing functions together - https://rockyj-blogs.web.app/2025/10/25/result-monad.html

I also see that lately "code quality" is the least concern of most (even software product) companies, just ask AI to write code in a single file / module / class - then launch feature and fix if you have to. I could see that in a few years things will be extremely messy (but who can say).

kitd• 9 months ago

I never really got into Haskell in a big way, but one of the things I liked about the Haskell Wikibook [1] was how they presented Haskell code as being either in pure form or "do" form, and how the latter orchestrates the former, much as presented here. To a beginner like me not interested in monads etc, this was a very simple and explicit way of approaching coding in Haskell.

[1] https://en.wikibooks.org/wiki/Haskell

novoreorx• 9 months ago

While I largely agree with the philosophy, the example provided is not very practical. The code snippet `getExpiredUsers(db.getUsers(), Date.now())` is unlikely to occur in real-life scenarios. No one would retrieve all users and then filter them within the program. Instead, it should be `db.getExpiredUsers(Date.now())`.

We should never be too extreme on anything, otherwise it would turn good into bad.

doix• 9 months ago

No one would retrieve all users and then filter them within the program.

No one _should_ do that, but that's a common enough problem (that usually doesn't get found until code is running in production). I suspect with the rise of vibe coding, it's going to happen more and more.

regularfry• 9 months ago

Sometimes it's forced by using the wrong database in the first place, or the wrong data structure. It can be less pain to do a bit of post-processing in the application layer than to unpick either of those.

_flux• 9 months ago

With a good library you could do just that, by having the functions return only queries and then expand them to the actual values (by interacting with the DB) after applying the filtering to it?

soulofmischief2• 9 months ago

So would you then have to do `getActualUsers(db.getUsers())` or `query(db.getUsers())`?

Still smells like in such a case the developer avoids the complications of abstraction or OOP by making the user deal with it. That's bad API design due to putting ideology before practicality or ergonomics.

fulafel• 9 months ago

Depends on the programming environment: if it's a O(n) operation anyway, meaning you don't the times indexed, and your computation is colocated with the data and the db interface is using lazy sequences...

(Also, real-life systems of course do things inefficiently all the time)

foobarian• 9 months ago

This sounds to me like the old hexagonal architecture [1]

[1] https://en.wikipedia.org/wiki/Hexagonal_architecture_(softwa...

Twisol7• 9 months ago

Yep! I remember the phrase "functional core, imperative shell" being popularized by Gary Bernhardt in ~2012 [0][1]; in his talk Boundaries [0] (around 31:00), he even mentions "hexagonal architecture" by name.

[0] https://www.destroyallsoftware.com/talks/boundaries

[1] https://www.destroyallsoftware.com/screencasts/catalog/funct...

ericmcer2• 9 months ago

Famously how Doom was written and maybe part of why it was ported to so many platforms.

Hex is kind of a PITA for ground up projects, but if you are doing something where you know multi-platform/cloud/device whatever is important it is cool.

fsmv• 9 months ago

Google writes articles like these every week and hangs them in the bathroom. It's meant to be a quick one page tip thing. That's why the example isn't super realistic, it has to be short.

There's a link with more info at the top. I'm not sure why this one in particular made it to the front page of HN.

smusamashah• 9 months ago

How does it fit with Tell Don't Ask https://martinfowler.com/bliki/TellDontAsk.html

Or is it that the example in the article is a bit poor?

crymer11• 9 months ago

From your linked article:

But personally, I don't use tell-dont-ask. I do look to co-locate data and behavior, which often leads to similar results. One thing I find troubling about tell-dont-ask is that I've seen it encourage people to become GetterEradicators, seeking to get rid of all query methods.

Simplify your code: Functional core, imperative shell

Related Stories