Quick, how long does it take to fish a fish?

If you’re like me, you don’t know.

And the reason is obvious: The ocean is big. It’s deep. You can’t see the fish.

You know there’s fish, fish are real, because you’ve seen fish before. You might even have caught a fish or two.

You probably eat fish.

But how long it takes from you pull out your rod and get the hook, line and sinker into the water, till there’s an actual, real-life honest-to-God fish on the hook …​

No clue!

And that’s okay.

Why is that okay?

Now, consider this conversation that I’ve overheard

Customer: "How long will it take to find and fix this bug?"

Developer: "How long does it take to fish a fish?"

As a developer, it’s a great metaphor to use for washing your hands of all responsibility.

The customer, after all, sees software development as the modern day equivalent of the black arts.

It’s a mysterious process driven by quirky people talking in strange languages.

To the customer, there’s always some too-complex-to-understand reason for why the code wasn’t finished one schedule …​

And a too-complex-to-understand reason for why the whole thing has to be rewritten, in another language, without any new functionality being added.

And a too-complex-to-understand reason for why the previous programmers, who don’t work here anymore, has left a code base of such a quality that no new programmer is willing to change anything.

And so, when the customer who’s used to being in the tight grip of the unassailable, necessary but oh so annoying programmer, is force-fed this metaphor, it comes almost as a relief:

Okay, I can understand that …​ I don’t know how long it takes to fish a fish, either. Okay.


If you’re a fisherman, you damned better know how long it takes to fish a fish.


Because your livelihood depends on it.

Because you know a lot about fish.

And you do.

You know about the ocean. You know where the fish goes at certain times of year.

You know what bait to use, you know how deep the line should go. You know the signs to look for, you watch the birds diving into the water, the ever changing weather.

You have technology to help you map the sea.

You’ve done this for years, and you know all this stuff because you’ve had mentors, you’ve studied, you’ve tried and failed, and always been learning.

A professional programmer should not wash their hands of the responsibility by using an ill-fitted metaphor.

If you’re a professional programmer, be like the professional fisherman.

You have knowledge, tools and experience.

What kind of a bug is it?

Does it appear every time the user does something, or only sometimes?

Does it happen for all users, or only some users?

If it only happens for some users, what’s the commonality between the users?

Are they doing the exact same thing?

Have we watched them do it?

Do they have the same hardware?

Can we reproduce the bug?

When we can reproduce the bug, we can isolate it. When we can isolate it, we can fix it.

The good thing about being a programmer, is that, at least in principle, everything that happens is logical, by its very nature. Therefore, it is understandable, by its very nature.

I’m not saying it’s easy. It can be damned hard. But as a professional programmer, it us up to you to take on the responsibility.

Your experience should let you estimate how long it will take to fix the bug

Estimate the steps. Refine the estimates as you learn more. Explain what you need in order to get started:

"To start fixing this bug, first we need to be able to reproduce it. I cannot see it happening in the logs, so we need to find a user for whom this happens. Can you get me access to one such user, or two, or three?"

"If I’m able to reproduce the bug together with the user, I can instrument to see whether it’s a bug in the client software, or in the server software. If it’s on the client, I need help from Thomas. He’s not here today, but he’ll be in tomorrow. When I’ve looked at it with him, I can get you a better estimate …​"

Take responsibility to isolate the damned thing. Express your thinking, how you weigh the cues, how you eliminate some possibilities. Explain your reasoning. And estimate as you go along.

Know your fish.

Rewards will follow.