frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

How Has Roman Concrete Lasted for Millennia? 1,900-Year-Old Latrine Offers Clues

https://www.smithsonianmag.com/smart-news/how-has-roman-concrete-lasted-for-millennia-a-1900-year-old-latrine-offers-new-clues-about-the-materials-impressive-durability-180989115/
55•divbzero•2h ago

Comments

skybrian•1h ago
Modern concrete has steel rebar, which is very useful, but eventually corrodes. Stainless steel rebar could be used if longevity mattered, but usually it doesn’t because the building will likely become functionally obsolete and need replacing before then.
warumdarum•1h ago
You could wouldnt even need that. https://de.wikipedia.org/wiki/Opferanode
eru•51m ago
Yes, but you have to replace the sacrificial anode every so often.

https://en.wikipedia.org/wiki/Galvanic_anode

Barbing•23m ago
Like the plate of food on the grass for the bees at the barbecue
masklinn•47m ago
> Stainless steel rebar could be used

There are also coated and non-metallic rebars.

terribleperson•24m ago
Coated rebar could be good, but the coating can be damaged anywhere from manufacture to installation.
simonebrunozzi•1h ago
A picture of one of these toilets would have been useful.
defrost•59m ago
https://www.science.org/cms/10.1126/sciadv.aeb0754/asset/924...

From: https://www.science.org/doi/10.1126/sciadv.aeb0754

demosthanos•55m ago
Related, Grady Hillhouse on the myth of Roman concrete.

> The miracle of modern chemistry has given us a wide variety of admixtures like superplasticizers to improve the characteristics of concrete beyond a Roman engineer’s wildest dreams. So why does it seem that our concrete doesn’t last nearly as long as it should? It’s a complicated question, but one answer is economics. There’s a famous quote that says “Anyone can design a bridge that stands. It takes an engineer to build one that barely stands.” Just like the sculptors job is to chip away all the parts of the marble that don’t look like the subject, a structural engineer’s job is to take away all the extraneous parts of a structure that aren’t necessary to meet the design requirements. And lifespan is just one of the many criteria engineers must consider when designing concrete structures. Most infrastructure is paid for by taxes, and the cost of building to Roman standards is rarely impossible, but often beyond what the public would consider reasonable.

https://practical.engineering/blog/2019/3/9/was-roman-concre...

A large part of why Roman concrete lasted longer than ours tends to is that we suffer from a shortage of narcissistic emperors with the means to wield entire economies towards their own immortality.

seydor•44m ago
> we suffer from a shortage of narcissistic emperors

not recently

defrost•42m ago
but not ones willing to put the effort into properly fixing a pond.
fragmede•30m ago
Who's pond? I'm willing to bet billionaires' estates have well tended ponds, contrary to public ponds. Or a reflecting pool.
Mistletoe•54m ago
I’ve often wondered why every good sidewalk I see has a WPA stamp on it from the 1930s and the modern ones are all crumbled and uneven.

https://share.gemini.google/5g0gxGyOmAPD

defrost•45m ago
.. and, also, ...

The folk of the 1930s were entirely capable of making poor quality concrete that barely lasted 30 years (source, my father, born 1935, still alive despite having mixed many a batch of concrete and having laboured).

The reason you don't see that walking about is that poor quality 1930s concrete was replace 50+ years ago.

TylerE•32m ago
Survivorship bias in action.
Barbing•16m ago
Gemini 3 Flash?

Edit: if you were an expert in this field and that link saved you from typing, and you mentioned you could confirm every word, that’d make sense - I think those Flash models were tested as being as reliable as a coin flip in some hallucination test scenarios, so linking it’s like… eh do I wanna read potentially-only-plausible history?

aizk•50m ago
Thinking of the ww2 plane with holes meme right now.
bombela•40m ago
> It turns out that another chemical reaction, known as carbonation, might also contribute to Roman concrete’s longevity.

Roman concrete was made lime cement (calcium dioxide); which cures via carbonation (hardens with carbon oxide). And adding pozzolan to lime makes it hydrolic (hardens with water). Is it surprising that it can still carbonate some? Modern concrete has steel which rust and crack concrete. You can use fiberglass rebar for longevity, or build without rebar even, but that is more costly and and less efficient.

litoE•28m ago
As I understand it, concrete has excellent resistance to compression but fails easily on traction, while steel bars are exactly the opposite. That is why you put rebar in concrete: the steel handles the traction loads and the concrete handles the compression. This works well because both materials have similar coefficients of thermal expansion, so as the temperature changes they both expand and contract at the same rate. I suppose you can engineer fiberglass to have the same thermal expansion coefficient and use it to replace steel (assuming it is just as strong on traction). But how would you "build without rebar even"? Wouldn't your beams start cracking at the bottom, where they are subject to traction?
aw1621107•26m ago
Just FYI, I think you're looking for "(in) tension" instead of "(on) traction".
mikepurvis•21m ago
To build without tension you have to build structures that basically look like Roman structures [1]: a bunch of tightly spaces arches so that the entire thing is in compression, with no meaningful tension anywhere.

But it turns out that's pretty inconvenient; we really like doing dozens of feet of span for highway overpasses, building floors, and everything else. So we put rebar in all the concrete and just acknowledge that that means it has an absolute maximum lifespan of a century or two, and will certainly not last for millennia the way pure concrete in pure compression can.

[1]: https://www.theartnewbie.com/blog/rome/roman-arch

aiauthoritydev•22m ago
The whole promise of engineering is not to build a bridge that stands but to build a bridge that barely stands. It is not a good idea to build a bridge that last 500 years. You likely destroyed valuable resources to build one. Build a bridge that lasts 100 years and save those resources. In 100 years the technology to build bridges improves so much that it is lot easier to build a new one. At least in most countries like India.
matheusmoreira•16m ago
Nobody is going to tear down old bridges and rebuild them at enormous costs just because technology changed. Some things can't or won't be redone and so it's worth it to build it to last and getting it right on the first try.
throwaway27448•11m ago
> In 100 years the technology to build bridges improves so much that it is lot easier to build a new one

This hasn't even been true for 200 years lol

croes•9m ago
Sometime knowledge gets lost and newer doesn’t automatically mean better
kQq9oHeAz6wLLS•29m ago
If you're talking about the algae, it came back after Obama's $34m renovation, too.

Turns out algae is hard to kill, especially when you feed the reflecting pool from a tidal basin.

userbinator•42m ago
Most infrastructure is paid for by taxes, and the cost of building to Roman standards is rarely impossible, but often beyond what the public would consider reasonable.

Would you pay 10x more to have something that lasts 100x or even 1000x longer? The upfront cost is higher, but the TCO is ultimately lower. IMHO it's ultimately a form of planned obsolescence. This becomes even more obvious when plenty of expense is spent just on "engineering" to deliberately reduce lifespan.

TylerE•32m ago
The thing is, we're actually pretty crappy at knowing what we'll need 50 years from now, much less 500. Doesn't make sense to overbuild for an unknown future, when hundred years from now us will likely be able to do a far better job anyway.
Aurornis•22m ago
No, for two reasons.

First, we can’t summon infinite money to pay for things. Paying 10X more per bridge means we can build 1/10th as many bridges or we have to start stealing from other budgets.

Second, we don’t know what the needs will be for the bridge in that location 100 or 1000 years from now. It could need to be torn down to be widened. Maybe we’re all riding around in electric vehicles that coordinate perfectly with each other and the bridge isn’t needed for cross traffic any more. We don’t know.

cma•12m ago
There's also stainless steel rebar
21asdffdsa12•4m ago
https://en.wikipedia.org/wiki/Galvanic_anode

How Has Roman Concrete Lasted for Millennia? 1,900-Year-Old Latrine Offers Clues

https://www.smithsonianmag.com/smart-news/how-has-roman-concrete-lasted-for-millennia-a-1900-year...
59•divbzero•2h ago•31 comments

Kimi K3: Open Frontier Intelligence

https://www.kimi.com/blog/kimi-k3
1425•vincent_s•15h ago•863 comments

An Engineer's Guide to USB Typе-С (2024)

https://www.ti.com/lit/eb/slyy228/slyy228.pdf?ts=1759892558029
90•gregsadetsky•6d ago•1 comments

Microsoft Comic Chat is now open source

https://opensource.microsoft.com/blog/2026/07/16/microsoft-comic-chat-is-now-open-source/
614•jervant•13h ago•138 comments

GrapheneOS recommended for domestic abuse victims

https://privacypros.com.au/privacy-hub/articles/dv-safe-phone-australia/
64•aussieguy1234•4h ago•33 comments

Pebble Mega Update – July 2026

https://repebble.com/blog/pebble-mega-update-july-2026
19•crazysaem•2h ago•0 comments

Decoy Font

https://www.mixfont.com/experiments/decoy-font
485•ray__•13h ago•115 comments

LM Studio Bionic: the AI agent for open models

https://lmstudio.ai/blog/introducing-lm-studio-bionic
204•minimaxir•9h ago•73 comments

$100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

https://www.tryai.dev/blog/ai-music-video-arena-claude-vs-gpt-5.6
207•hershyb_•9h ago•238 comments

The Little Book of Reinforcement Learning

https://github.com/alxndrTL/little-book-rl/
95•mustaphah•7h ago•11 comments

Solod: Go can be a better C

https://solod.dev
92•koeng•3d ago•27 comments

NotebookLM is now Gemini Notebook

https://blog.google/innovation-and-ai/products/gemini-notebook/notebooklm-gemini-notebook/
282•xnx•13h ago•143 comments

M 3.9 Experimental Explosion – 147 Km ENE of Ponce Inlet, Florida

https://earthquake.usgs.gov/earthquakes/eventpage/us7000t13l/executive
52•hnburnsy•5h ago•23 comments

Old Icons

https://leancrew.com/all-this/2026/07/old-icons/
27•zdw•5d ago•6 comments

UIUC AI Teaching Assistant

https://github.com/Center-for-AI-Innovation/ai-teaching-assistant-uiuc
9•teleforce•3h ago•0 comments

How Our Rust-to-Zig Rewrite Is Going

https://rtfeldman.com/rust-to-zig
469•jorangreef•18h ago•246 comments

'Likweli': A new monkey species discovered in the Congo Basin

https://news.yale.edu/2026/07/15/meet-likweli-new-monkey-species-discovered-congo-basin
67•gmays•8h ago•12 comments

Mathematics of Data Science

https://arxiv.org/abs/2607.11938
130•Anon84•9h ago•3 comments

Simulating everything, sort of: The promise and limits of world models

https://arstechnica.com/ai/2026/07/simulating-everything-sort-of-the-promise-and-limits-of-world-...
24•LorenDB•3d ago•1 comments

Detecting LLM-Generated Texts with “Classical” Machine Learning

https://blog.lyc8503.net/en/post/llm-classifier/
178•uneven9434•13h ago•122 comments

Immersive Linear Algebra Book with Interactive Figures (2015)

https://immersivemath.com/ila/
192•srean•14h ago•26 comments

Starlink from 1984

https://nemanjatrifunovic.substack.com/p/starlink-from-1984
6•ingve•5d ago•0 comments

Helium escaping from atmosphere of nearby rocky exoplanet in a habitable zone

https://www.science.org/doi/10.1126/science.aea9708
86•anyonecancode•9h ago•25 comments

The human-in-the-loop is tired

https://pydantic.dev/articles/the-human-in-the-loop-is-tired
132•haritha1313•5h ago•69 comments

Pseudpocalypse

https://dynomight.net/pseudpocalypse/
104•surprisetalk•2d ago•63 comments

Show HN: Clx – Compile Lua to Native Executables Through C++20

https://github.com/samyeyo/clx
105•_samt_•5d ago•11 comments

Ring-Zero: Scaling Zero RL to a Trillion Parameters for Emergent Reasoning

https://arxiv.org/abs/2607.12395
46•binyu•8h ago•15 comments

Abstracting Effects with Continuations

https://crowdhailer.me/2026-07-15/abstracting-effects-with-continuations/
55•crowdhailer•19h ago•2 comments

CD sales growth outpaced vinyl in the first half of 2026

https://consequence.net/2026/07/the-cd-revival-is-getting-hard-to-ignore/
83•speckx•12h ago•97 comments

How to Train a Gen AI Kick Drum Model on Your Old Linux Desktop with 6GB VRAM

https://www.zhinit.dev/blog/training-a-kick-drum-diffusion-model
116•zhinit•14h ago•56 comments