Ask HN: What Code That We Write Today Will Still Be Used in Ten Thousand Years? | Hacker News Reader

Ask HN: What Code That We Write Today Will Still Be Used in Ten Thousand Years?

I own excellent English translations of Homer's Illiad and Odyssey. The poems were composed roughly three thousand years ago, then passed down orally through the generations before they were written down at all.

While attributed to Homer, it's unclear that there ever was such a person; quite likely, whoever passed on a poem, added his or her own material.

There are many translations available; I selected the one that makes the most sense to modern Americans. The cover of Illiad depicts the Normandy Invasion from inside a troop carrier; Oddyssey's has the Earthrise as seen from Apollo XIII.

These two poems are the very foundations of Western Civilation.

We all hope that there will be some kind of civilization ten thousand years from now. What are we doing, to contribute to its foundation?

View on HN

I have put out this idea before: in the far future, there will probably be a lot of ancient software in use that has been used and tested so much that it has very close to zero bugs or defects. The value of very old bug free software will be so great that people and organizations will not push for new features. In my future world this will largely be open source code.

A counter to my future world with very high quality but slowly changing software would be the advent of strong AI. I can't imagine strong AI that would not quickly, robustly, and continuously rewrite itself in a safe sandbox then rapid redeployment scenario.

"very old bug free software"

MULTICS came pretty close to that. It's commonly thought that UNIX was so-named to ridicule MULTICS. In reality, the people who developed UNIX contributed to MULTICS, it's just that they wanted something fun and easy to work on after having worded so hard and so long on MULTICS.

It was written in the mid-sixties. The last MULTICS box was decommissioned in Halifax, I think in 2005 or so.

Quite a lot of what we regard as new, in today's operating systems, was implemented in MULTICS in the sixties.

I worked doing systems programming on Prime minicomputers, a long long time ago.

Primes ran MULTICS.

None in use. Although I think that repositories for good software are like good literature and will stick around and be studied. I could see people using it to study how we evolved the solutions, how we thought through problems (likely mocking us and laughing some), but hopefully learning where the field came from so figuring out where they can go and take us.

Just look to other sciences (biology, chemistry etc), we look back 200+ years and find that solutions to seemingly impossible problems were known all along but that some factor we needed to make it successful just didn't exist yet (a Tony Stark moment). Today we take it for granted I think in some cases that in the past 15 years we can do things with GPU's, DSP's etc that just 30 or so years ago were thought to be near insurmountable problems to solve in our lifetime.

I'd be surprised if any of it is still used in 100 years except for the odd elective course on the history of programming languages and hardware.

The Iliad and the Odyssey survived because for most of the intervening centuries educated people (i) revered the relative sophistication of the Classical civilisations and (ii) relatively little else considered of value was written for most of the intervening years. It would be quite sad if people of the next few millennia feel the same way about the output of the early 21st century Anglosphere, particularly if its unsurpassed technical achievements rather than cultural output that's the object of their reverence.

You may want to look at http://www.longnow.org and http://en.wikipedia.org/wiki/Long_Now_Foundation. One of their projects (the brainchild of Danny Hillis) is to build a clock that will run for 10,000+ years.

That's why I picked 10,000 years specifically.

Computer Algebra and Computational Mathematics will still be meaningful (although the companies that host them will no longer exist).

They all thought I was quite the strange guy for posing this question to the Portland Startup Weekend. More typical was the young lady who wanted to write an app that would help her make new friends.

Reply on news.ycombinator.com