clacke
Re: libranet.de/display/0b6b25a8-3… (by myself)
Someone wrote a blog post on this, the heirloom program or the 100-year program or something and drew the SML conclusion, but I can't find it now.
It was written over a year ago. It wasn't the CollapseOS guy, nor was it anyone at Merveilles Town.
#PostCollapseComputing #HeirloomComputing #SmallComputing #PermaComputing
clacke
Unread arbitrary search results that aren't it but look interesting enough in themselves:
- SummerFest 2021: Festival of Artisanal Software
- Konrad Hinsen: IEEE: Dealing With Software Collapse
- Roy Schestowitz: Post Collapse Computing Part 4: The Road Ahead - Space and Meaning
- Maarten Steenhagen: What could post-collapse computing be?
- Steven Melendez: People are building technology that could survive the apocalypse (probably about CollapseOS) (ironic that the article URL collapsed)
- Tobias Bernard: Post Collapse Computing Part 4: The Road Ahead
Some of the authors are on fedi but I won't spam them with a mention in this slushbucket list.
clacke
Now I'm just finding neauoire, viznut and xj9 articles 🤣
No, it wasn't one of them who wrote it.
clacke
In the quoted post I also mentioned R5RS and possibly R7RS as centennial languages.
Ada would get a mention if the bootstrap was sorted out. Or is that clean now?
clacke
@mdhughes Chez, Racket and Guile more or less do R6RS, but there are a dozen R5RS and already more R7RS-small than there are R6RS.
Libraries aren't defined in R5RS, but the number of ways it has been solved are: (1) Fewer than the number of implementations, (2) Not difficult to implement.
I don't expect a centennial program to have many dependencies anyway, maybe no transitive dependencies. I don't expect the dependencies to be moving targets either, so the centennial program might just vendor them.
No unicode is bad, but even though the standard says you don't have to support it, most R5RS de facto support it, so I expect you can also write to R7RS-ok-not-THAT-small and get away with it.
Digital Mark λ 📚 🕹 💾 🥃
@clacke R5RS doesn't have library/modules, so it's unlikely to be long-lived. R7RS-small has bad decisions like abandoning Unicode, and R7RS-large is a bloated piggy, it'll never be done or implementable.
The two most solid impls of Scheme (Chez & Racket) are R6RS (Racket also has its stuff on top), and probably will be for forseeable future/maybe forever.
Maybe R8RS will sort that out next decade. Until then I just simplify my R6RS code if I need it to be more portable.
#scheme
clacke
@mdhughes I'm not deep in Scheme lore and haven't followed the standardization process. Here's my take as a mere lazy user of Racket and R5RS+:
As I understand it, R5RS was most importantly an R4RS with a codified macro mechanism. R7RS-small was most importantly R5RS with a codified library mechanism.
If there is ever an R8RS I hope it mostly just codifies the de facto R7RS-small+ with unicode and whatever else "everyone" had to solve based on real-world concerns and doesn't try to advance the language beyond that.
clacke
Digital Mark λ 📚 🕹 💾 🥃
@clacke Well, R4RS/IEEE had non-hygienic defmacro in most impls.
R6RS added better IO types, hashtables, exceptions, a saner replacement for defmacro, which let records generate full names and some more niceties, AND had full Unicode support like a 21st C language.
Then the impls who weren't smart enough to do all that said "we'll make our own R7RS (small) without even hookers or blackjack". But now R7RS-large is just adding more and more crap to catch back up.
clacke
Some insights about what makes a 100-year language and what it even means for a language to be "dead" or "alive":
clacke
Found it! I realized that the right trick was to limit the search to lobste.rs rather than trying to find the magic phrase that would make the right article float to the top.
clacke
Haelwenn /элвэн/ :triskell:
iooioio
@clacke
Kinda reminds me of this:
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.