Hacker News Reader

Hacker News

Comment by XelNika | Hacker News Reader

Parent Full thread

XelNika

It doesn't say 2007 anywhere, but the page is clearly at least that old.

https://web.archive.org/web/20071115173112/https://www.sqlit...

I also found material from 2004 with the term.

https://www.tcl.tk/community/tcl2004/Presentations/D.Richard...

> Regardless it's vague and irrelevant material

It's not vague at all, the term makes sense to distinguish "in-process" from client/server models. The page also includes an explanation on the very next line.

> Adding an original timestamp to the page would've been 100x more helpful

Not everyone reading about a database product is a developer. A timestamp would not help anyone unfamiliar with the history of the serverless term. The alternative is to rewrite all instances of "serverless" in the documentation which is a waste of time. Modern "serverless" is a stupid buzzword and this page clearly serves as a protest.

View on HN

Complete history of the document in question is here: https://www.sqlite.org/docsrc/finfo?name=pages/serverless.in...

It was, indeed, written in 2007, but based on ideas that predate that.

I would love to know who uses "serverless" instead of "in-process". Why add a new term at all?

And if the definition has since been muddied, then all the more reason to avoid using it instead of creating even more niche definitions.

> I would love to know who uses "serverless" instead of "in-process". Why add a new term at all?

"In-process" is meaningless to non-IT people, they don't even know what a process is. The SQLite dev probably created the term for marketing purposes, i.e. the exact same reason cloud providers adopted it 10 years later.

> And if the definition has since been muddied, then all the more reason to avoid using it instead of creating even more niche definitions.

I disagree. The term in relation to SQLite is clearly defined and predates the modern version, there's really no need to go back and change it. You're also disregarding the statement made by the SQLite dev by keeping this page and updating it with a clarification.

SQLite is meaningless to non-IT people.

Those people wouldn't end up on a random documentation page for SQLite then, making this argument superfluous.

> I would love to know who uses "serverless" instead of "in-process". Why add a new term at all?

This discounts which term came first. Back in 2007 it was just fine to talk about this as being serverless, the marketing term gained popularity years later. They even talk about the more recent definition on the page, I really don't get why people in this sub thread get triggered by some random documentation page written over a decade ago. There's no need to further change or delete that page because this discussion is lacking any practical relevance.

I didn't make the argument, the other poster did.

It doesn't matter what definition came first. It was bad back then when embedded and in-process already existed. Now it's even worse.

Non-IT people make IT decisions all the time. If you can sell your IT product to C-levels, they'll force IT to use it. Look at how companies misuse things like blockchain, ML and AI just because of the hype around those words.

Yea, that's now how SQLite has ever been sold. All the other comments on this post have since shown just how useless the "serverless" label is for this.

Reply on news.ycombinator.com