2024: “We should rewrite everything in Rust!”2018: “We should rewrite everything in Go!”2010: “We should rewrite everything in C#!”2000: “We should rewrite everything in Java!”1990: “We should rewrite everything in C++”But hey, I’m sure this time the n...

mos_8502 :verified:

2024: “We should rewrite everything in Rust!”

2018: “We should rewrite everything in Go!”

2010: “We should rewrite everything in C#!”

2000: “We should rewrite everything in Java!”

1990: “We should rewrite everything in C++”

But hey, I’m sure this time the new shiny language that’s fashionable this year will be the one that changes everything and never gets replaced.

Edit: All of those languages are fine in and of themselves. Use the tools you like. But don't act like you're doing anything new. There has been nothing new in a very long time.

Irenes (many)

@mos_8502 btw if you hadn't guessed, it was you who reminded us of our thread, thanks

mos_8502 :verified:

@irenes I figured, but I didn't want to assume.

I want to be clear, I'm not, in any way, telling people to "not use language X" or to "use language X instead". 100% cool by me to use whatever tools make you happy. I'm just sick of the endless cycle of "this time we'll do it right and never have to change again!".

Brian Swetland

@mos_8502 @irenes Are there people who seriously think "we'll never have to change again"? One of my takeaways from some of the big security oopses of the last decade or two is "we don't know what we don't know."

There's value in learning from past mistakes and trying to make the next thing a bit more resistant to breaking in the same ways that previous things broke. But that doesn't mean there won't be exciting new ways for things to break discovered in the future!

Sound like job security!

Irenes (many)

@swetland @mos_8502 *sigh* reflecting back on our past we don't think we ever seriously believed that there would be a single final programming language, but we sure did buy into the purity narrative anyway

Brian Swetland

@irenes @mos_8502 My big takeaway from some of this recent chatter is I should clearly be spending more of my time on my long running, slow moving, hobby language project.

My goal: a small systems language that tries to combine the joy that C has brought me over the past 3 decades with some of the syntactic and safety innovations that Go, Rust, etc have introduced, while trying to avoid the horror of modern "big compiler" UB landmines.

Because, of course, my solution is "yet another language!"

Irenes (many)

@swetland @mos_8502 oh! well it sounds really fun, for whatever that's worth, and we do think that fun is an under-rated reason to do things. a lot of stuff we have great admiration for was done for fun, not for any higher goal - though it ended up serving other goals too.

Dave Anderson

@irenes Co-sign doing things for the sheer fun of it. We get to imprint ideas onto fancy sand, and the fancy sand does stuff with the ideas! It should be impossible to forget how wild that is, and yet, I need to remind myself regularly.

Brian Swetland

@irenes @mos_8502 Fun is a major goal to be sure. I've written a number of virtual machines, interpreters, linkers, assemblers, etc over the years but never a real self-hosting to-the-metal compiler. Sure it's been done before by many people, but *I* haven't done it yet (well at least not to a level of completion I'm satisfied with).

Irenes (many)

@swetland @mos_8502 it's among our ambitions as well

we'd like ours to be based on Futamura projections and partial evaluation, because we don't choose easy projects... ><

Irenes (many)

@danderson for sure!