News for hackers. Stuff that matters.

jasonswett

Re:Agent Skill for TDD on Saturday June 06, 2026 @08:29AM (#48424397)
The reason it works is because there's a difference between the model knowing something and the agent doing something. Claude will happily write giant untested functions even though it "knows" that short functions are easier to understand and then testing enables safe refactoring etc. The model also "knows" many conflicting "facts", such as the fact that testing is smart and that testing is a waste of time. It can't act on both beliefs at the same time. That's why nudging it toward your own preferred behaviors works.
Re:Agent Skill for TDD on Friday June 05, 2026 @08:55PM (#48420297)
I HATE this. I call it speculative coding. Claude often calls it "defensive" programming. It's easily my #1 LLM pet peeve. I have yet to figure out a reliable way to make this stop happening.
Re:Agent Skill for TDD on Friday June 05, 2026 @08:53PM (#48420281)
OP here. I don't know, in my experience Claude took "clean the kitchen before we make dinner" to heart in an astonishingly productive way. I haven't tried many other analogies though.
Re:Agent Skill for TDD on Friday June 05, 2026 @08:49PM (#48420254)
Thanks, I'm glad you like it!
Re:Agent Skill for TDD on Friday June 05, 2026 @08:49PM (#48420250)
My experience has been that yes, LLMs already know about TDD, OOP, etc., but they won't necessarily BEHAVE according to what they know unless you tell them. And of course, they "know" a lot of things that conflict with each other.
Re:Agent Skill for TDD on Friday June 05, 2026 @08:47PM (#48420241)
Good point! Will add a date.
Totally agree, although I'd maybe replace "knowledge" with "knowledge or behavior".
> Just because a piece of duplication costs something doesn’t automatically mean that the de-duplicated version costs less. It doesn’t happen very often, but sometimes a de-duplication unavoidably results in code that’s so generalized that it’s virtually impossible to understand. In these cases the duplicated version may be the lesser of two evils.
> One of the biggest benefits of duplication that the author fails to identify the locality of logic

There are a lot of things in the post that I fail to identify. There are also some things I wrote there that I no longer completely agree with.

Here's a newer post in which I go deeper, and in which I do address the locality of logic.

https://www.codewithjason.com/duplication/

Re: on Monday April 17, 2023 @12:34PM (#35603142)
I like reading articles about programming but I don't really want to get more emails. I also prefer reading offline over reading on the screen.

I figured there were probably other people like me so I decided to start a snail mail programming newsletter. You can find it here: https://www.codewithjason.com/snail-mail-newsletter/

So far 92 people have signed up from 19 different countries. If you choose to sign up, I look forward to send you letters in the mail (and hopefully getting letters back!).

Re:Monty Hall Problem on Thursday June 09, 2022 @02:11PM (#31684756)
The rule is "after the player has guessed a door, open ALL the other doors except the door that was guessed plus one other door".

I think your confusion is because if there are 100 doors then the rule is self-evident. But when there are only 3 doors, the rule of "open all doors except the door that was guessed plus one other" is indistinguishable from "open just one door".

The key is knowing that the rule is the "open all..." rule.

Re:When I do TDD and when I don’t on Thursday February 17, 2022 @03:37PM (#30378342)
I personally don't see any reason why the type of test matters when doing TDD. As I see it you can do TDD with integration tests just as much as you can do TDD with unit tests. The only difference to me is that the work involved might take on a bit of a different character. But that doesn't make it not TDD to me.
Agreed. Although I think the reason why most freelancers sell staff aug work as opposed to what you might call "strategic project work" is that it's about 100X easier (at least in my experience) to find staff aug work.
When I used to freelance I divided my marketing activities into two categories: "hunting" and "farming".

Hunting tactics include things like responding to job board ads, reaching out to people you know to see if they can help you find a gig, or going to a conference to try to find a new client. Like real hunting, hunting for clients can be fast, but it’s only successful a fraction of the time, and you don’t always get something good.

Farming tactics include things like writing blog posts, writing books, speaking at local meetups, speaking at conferences, guesting on podcasts, hosting podcasts, and other things that demonstrate your technical competence to a (if done successfully) large number of people. Farming can yield much better results than hunting, but there can be quite a long interval between the time you plant the seed and the time you get to reap the harvest. And just like in real farming, not all the seeds you plant will sprout into plants that bear fruit.

In my experience, farming tends to yield better clients than hunting, although you can't control when the clients are going to come.

I hate to say it but I think your safest bet is for you and your colleagues to each go your separate way (for now) and find contract gigs individually. In my experience it's really hard to find a client who wants to hire multiple people at once. The easiest kind of client to find is a software organization that needs a programmer and is willing to use a contractor rather than a full-time employee.

And BTW, the sad secret I learned about freelancing over the course of ~8 years doing it is that 99% of freelancers are just staff-aug contractors. They're not making great money and they don't have a special lifestyle. The only difference between most freelancers and regular employees is that the freelancers file 1099s instead of W-2s. Not trying to be a downer about it but I think it's helpful to acknowledge the reality of it.

Lastly, I've been asked many times how to get freelance clients, so I wrote a post about it here: https://www.jasonswett.net/my-advice-to-brand-new-freelance-...

Hope that helps.

Re:List of oldest continuously inhabited cities on Wednesday April 22, 2020 @11:03AM (#22946290)
What about London?
I think my estimate was a little on the high side. I've tempered my opinion since then (and I've edited that part out of the post). I don't think 9 out of 10 developers are incompetent.

Older »