If I understood correctly, the queue implementation in the blog post holds a transaction while an operation is in progress.

I see the advice to make it as short as possible, but why can’t we update the status column to, say, “processing” and avoid potentially long transactions at all?

This works well for jobs that are long-ish. You need another process to sweep for orphaned jobs and requeue or fail them. Add a timestamp of when it got picked up to keep track

Not a novel idea, but it always resonates with me when I hear that the goal of every software engineer is to manage complexity.

That said, I still find it hard to formulate for myself how to tell the difference between accidental and essential complexity.

It’s a fantastic performance booster for a lot of mundane tasks like writing and revising design docs, tests, debugging (using it like a super smart and active rubber duck), and system design discussions.

I also use it as a final check on all my manually written code before sending it for code review.

With all that said, I have this weird feeling that my ability to quickly understand and write code is no longer noticeable, nor necessary.

Everyone now ships tons of code and even if I do the same without any LLM, the default perception will be that it has been generated.

I am not depressed about it yet, but it will surely take a while to embrace the new reality in its entirety

For debugging, it’s also great trawling through logs and stack traces.

Makes a late night oncall page way easier when the bot will tell you exactly what broke

As a person with ADHD, I feel personally attacked.

I guess you understand this and are making a joke, but that "attack" would appear to be intentional (and motivating).

I find that I don't have major issues doing a thing once I get started on it. The main problem is choosing from among many things that I could reasonably consider "the thing", and then feeling confident enough in that choice to start.

What about doing the thing intently for a week and then realizing later you haven't touched the project in 6 months?

This sounds not too dissimilar to the release the POC to prod mentality.

There are times where you obviously need to do the thing to understand the thing to see the process of doing the thing. This allows for breaking the process down into better steps. Just writing code to do things you think is doing thing but prove not to do the thing when actually doing the thing is common.

I'm talking about having completely different, unstarted, overall projects in mind.

that sounds like paralysis by analysis. just pick a project and go.

Same. I'm tempted to print this post out and hang it for inspiration. But I guess that would also not be doing the thing.

I have bad ADHD and printed the strangestloop.io blog post out and put it on the wall by my work desk in Oct 2023 according to the printout timestamp. I still haven't done the thing in some meaningful areas, and the print has honestly kind of been dispiriting. I'm going to take this post as the prompt to take it down.

I'm going to consider this with the same weight I would if my future grey-bearded self popped out of a portal to say it, thank you. I've had a sticky note on my monitor for a few years that just says "SHIP SHIP SHIP SHIP SHIP SHIP"; it might be time for that to go before it becomes much more depressing.

sometimes i find that being okay with not doing the thing is exactly the thing i need to do to be okay with getting around to doing the thing

Coming up with excuses is not doing the thing.

Who claimed it was?

As Martin recently posted, now the production should take about 4 more months until we can buy the paperback version.

That said, if you have an O'Reilly subscription, you can already enjoy the new edition.

+1

I'd also like to understand whether there are still any cases when MongoDB is the right choice

Also, take a look at windmill.dev. It’s a beast

I think windmill got the entities right (script, flow, app). The last I tried it was a bit more complex to setup than n8n.

We also had to abandon it, due to setup complexity coupled with capacity constraints.

We’ve similarly had a lot of success with self-hosting Windmill.

12 out of 13 chapters are already available on the O'Reilly website in the "raw and unedited" format.

The book is scheduled for publication in February 2026, but if you have an O'Reilly subscription, you can already access the new content.

That's exactly the "simple vs easy" difference that Rich Hickey emphasized in his famous talk.

Even in this thread, there was a comment (now deleted) saying that a staff engineer at GitHub is unlikely to know what real scale is.