Fork me on GitHub

How do you get yourself psyched up to lay the foundations for a new project?


You get to research the best pieces of our ecosystem, which can be gratifying in itself If creating the boilerplate itself is too much of a hassle you can always start from a template and/or framework. If you keep things modular, eventually the architecture can become 'your own', without template cruft

Cora (she/her)01:02:03

I just channel this meme

😂 9
truestory 4
😰 1
Cora (she/her)01:02:37

more seriously, I try to use different libraries and techniques just to try it out and keep it interesting (but time box it if it's not a side project)

Cora (she/her)01:02:02

it's easy to be too conservative and just choose the most boring things, but also easy to be too aggressive and force techniques and libraries even if they don't fit the problem. so, balance 🙂


This is the life of a commercial project. Changes necessarily happen during the project. Requirements change. Architectures are discovered to not cover certain cases. A customer changes their mind. As a result, a system that we felt would be robust and elegant… isn't. That doesn't make it useless. Indeed, if it responded appropriately to changes, it will be the best it could be. But it just won't be what was hoped for back when the repo was fresh and hope filled the air 🙂


Non-commercial projects are a different matter. You can make them for yourself. This means that you can do whatever you want, and stick to it! But, that means it might not be useful


But they're elegant and fun 😊

💯 2

Some morse than others!

😆 8

That started as throw away code, and developed elegance as it progressed

simple_smile 1
Zach Mitchell, PhD05:02:33

Can't say I expected to see EatWhatYouKill in a stack trace 😂

at org.eclipse.jetty.util.thread.strategy.EatWhatYouKill.runTask(
	at org.eclipse.jetty.util.thread.strategy.EatWhatYouKill.doProduce(
	at org.eclipse.jetty.util.thread.strategy.EatWhatYouKill.tryProduce(
	at org.eclipse.jetty.util.thread.strategy.EatWhatYouKill.produce(
	at org.eclipse.jetty.util.thread.QueuedThreadPool.runJob(
	at org.eclipse.jetty.util.thread.QueuedThreadPool$

🦌 2
real_gun 2
😂 1
atat 2

Yeah, that still makes me chuckle every time I see it 🙂


I guess I can breathe a sigh of relief that the jig is up before they managed to insert themselves as a "mobility provider" into too many municipal transit networks. I'd have hated to see what little public transit we have in the USA privatized and hollowed out by means of a shiny app.

👍 1

I have enjoyed reading Jarret Walker's commentary on Uber over the years, especially his classic post on ridesharing as "elite projection":

👍 1
Martynas M07:02:21

I'm not sure if what they say is that it's not a solution that works for all possible scenarios or a solution that doesn't work at all...


Are clojure.spec releases tied to clojure releases? Or is it theoretically possible we'll be seeing a stable release of spec before 1.12?

Alex Miller (Clojure team)21:02:17

we've already released a spec update that 1.11 will depend on for example

Alex Miller (Clojure team)21:02:49

or are you asking about spec 2


a spec which doesn't contain alpha in the name

Alex Miller (Clojure team)21:02:36

well, we could just remove the alpha :)

🙏 3
Alex Miller (Clojure team)21:02:41

but to answer your question, I don't know what will happen by the time of 1.12 (given that we have not yet even released 1.11)