isntitvacant

isntitvacant, 1 hour ago

i frequently reflect on fond memories of firing up pilotwings 64 to take to the skies as a literal slice of kiwi fruit. it’s a core memory, really

isntitvacant, 5 days ago

@timelordiroh @BathysphereHat this seems to happen a lot with intensifiers — “awfully good”, “bad/dead/sweet-ass”, etc — so “hot minute” only has contextual value (which is kind of cool, since it accomplishes something you couldn’t do with a plain intensifier like “very”)

isntitvacant, 18 days ago

some low quality facts (in the sense that this book has taken me MONTHS and I'm still not done, but this is from memory):

• the "nam" part of "vietnam" is from mandarin "nan"/"南", literally "south"
• the order of japanese kana is inherited from sanskrit ordering (also the entire "s" row used to be palatalized, instead of just "shi")
• the "salumu" of "As-salamu alaykum" and "shalom" share a root, "peace"
• suspected origin of "italy" is greek "oitalios", "land of young cattle"

isntitvacant, 18 days ago

• saxons and francs were named for their favored weapons (knives & spears)
• celtic had a rich tradition in the north of spain and france
• pharaoh is metonymy (lit "great house"); as is egypt ("ha(t)-ka-ptah"; "temple of the soul of ptah" located in memphis
• (embarrassing for me to say, but) I didn't realize how long-lived nahuatl was after contact with spain.
• we live in a dwarf fortress sim: andalusia -> Vandalicia; "land of the vandals"

isntitvacant, 18 days ago

the larger theme the author explores is that, with some (very grim) exceptions, language is seldom effectively spread through force – invaders tend to pick up the language of the people they invade, if they stick around.

new ideas – in technology, trade, religion – appear to be how languages most typically spread; strong written or ritualized oral traditions reinforce the staying power of the language – at least insofar as I follow what this book is saying.

isntitvacant, 18 days ago

okay, one last tiny thing I learned from this book because I just realized it's after midnight and I should go to bed:

Sugar has its roots in Sanskrit ("sharkara"), BUT: when alexander's army first encountered sugarcane in India, they didn't have a way to describe it other than "honey without the help of bees"

isntitvacant, 18 days ago

(and bringing this full circle with Wasm: this has all got me thinking about how language communities work within programming – how programming language communities influence each other & the difficulties that emerge when they attempt to interoperate with one another; how far distant is an IDL from how we used aramaic or latin as a bridge for cross-language diplomacy?)

isntitvacant, 18 days ago

(... are demos where hosts written in one language communicate with guests written in another language really that reflective of how people would actually collaborate using this technology?)

isntitvacant, 17 days ago

@sunfish it strikes me that, in non-browser contexts, there’s much less of a premium on being able to communicate cross-language — it was already possible (via IPC, network RPC, etc.)

Wasm changes technical costs but doesn’t change the social costs that currently push operators of medium-to-large systems towards “approving” a set of languages for internal use.

Lightweight virtualization and runtime linking — as a tool for allowing multiple teams to ship a system — seem more compelling?

isntitvacant, 17 days ago

@sunfish I’m framing this from the perspective of a hypothetical 50-ish person sized startup: how does wasm change how my team builds systems? How does it change the skills I look for in hiring?

I come back to “it lets them work in small teams that ship work incrementally without having to lean on microservice RPC”; maybe reducing infra spend.

Wasm might even mean that I can hire for fewer languages, even, in that it increases the number of platforms I can target w/a single language.

isntitvacant, 1 month ago

@yosh it took me a (redacted quantity of time), but i got core module linking working in my tiny wasm interpreter written in rust

isntitvacant, 1 month ago

A startup that sets out to build a registry is either looking to get acquired, needs it to support some other product, or needs a plan to move some class of work away from github (or slack or teams) to their platform.

In the second case, a registry is an UberFeature comprised of a lot of sub-features, with an enormous support cost. What is so different about their product that they need a full-fledged registry?

isntitvacant, 1 month ago

In the third case, what are you going to do to differentiate your collaborative software eng platform from github and gitlab? (“have a single language focus” does not seem like a winning strategy.)

which brings us back to the first case: who do want to sell the company to? do you care? you’re asking your users to come along for the ride with you over the long term, so I’d argue “yes, you should”, but.

isntitvacant, 1 month ago

@richardstephens Yeah – I left them out because they're not quite hypergrowth registries; but they do serve as a clearing house for different registry protocols. (They might acquire a startup supporting a popular-enough registry, but it wouldn't be the exit investors are looking for)

I'm interested in your stance on artifact registries, also: it's hard to wage a war on Big Cached Data! Are you picturing a future where artifacts are built on-site+on-demand, or where artifacts are easily audited?

isntitvacant, 1 month ago

@aeva I put a 418 “I’m a Teapot” response in the cdn code for the npm registry with the intent that only misconfigured clients would see it but instead it took down registry dot yarnpkg dot com (because I forgot some proxies include the implicit :443 port in their host header)

isntitvacant, 1 month ago

@aeva “snuck” might be stretching it though, this was at the end of a two week marathon port from fastly to cloudflare workers precipitated by things happening at the company, so both my pair and I at least looked at the code and said, blearily, “yeah that looks ok”

(and to be fair cloudflare did say “hey you need to check host headers”)

isntitvacant, 1 month ago

@b0rk in the vein of “unreferenced”, “unnamed”.