Huh; from the title, I assumed this was going to be a service that acted like a blog editor (with accounts, draft posts, rich-text WYSIWYG editing, etc.), but where there was no index for the resulting posts by author, instead just giving you "floating" pages accessible only by their URL, like Gists/Pastebins are (but nicely formatted, like regular blog-posts are.)
I've always wanted such a thing; it'd be perfect to combine with a microblogging site like Twitter or Tumblr, where instead of posting long things to the microblogging site, you post links to your "floating" essays amongst your short posts. Your microblog would then be the canonical "blog"—the canonical index of these articles—rather than people assuming they could subscribe to your presence at whatever article editing+hosting service you chose to use (which might, after all, change with each article.)
It sounds like you want a CMS service (content management system) that's not tied to a blog like taxonomy. If you can't find one, there are tools made to build that sort of thing.
All three happen to be PHP based, not sure why most of the CMS solutions lean towards it. If that doesn't work for you, search terms like "simple", "cms", "your preferred language" might help yield answers that are a good fit.
You're sort of right, insofar as a CMS doesn't enforce a blog's reverse-chronological sorting... but that's not quite what I meant; a CMS is still fundamentally about having a collection of things by the same editorial staff put together—e.g. on the same domain.
What I was picturing is to throw out the level where you could infer "same editorial staff", so instead you just get the shared hosting service (http://example.com/) and then directly under that, articles (http://example.com/foo). Like a pastebin: there's no source attribution to a given post, even implicitly.
The lack of internal attribution is important, because any internal attribution mechanism—e.g. mentioning the author in a <meta> tag, or putting their name in the URL—can be used to build an index of posts by author (just googling "site:http://example.com/author" gets you half-way there), which people might then treat as canonical instead of relying on the microblog.
So, I'm not against self-hosting a CMS, but my requirement that articles have no article-internal attribution—not even implicitly based on where they're hosted—means that self-hosting is kind of moot, unless I know a bunch of other people I can invite to also post on my self-hosted CMS until it turns into its own pastebin.
(Also, I have other, more fanciful requirements, like the ability of articles to effectively have their own embedded code and databases, turning them into their own little one-page web-app, like NYT's fancier long-form data-science articles are. CMSes don't really go there; they expect code you want to use in an article to be added to the site as a plugin, rather than being treated as part of an article and thus isolated+not-run when on any other article, deleted when the article is deleted, etc. The only thing that could pull this off right now is hosting each article as its own Heroku app; and even then it's not a perfect fit. You could theoretically write a blogging engine that posted articles by creating Heroku apps, though... hmmmm, now there's an idea.)
"If you can't find one, there are tools made to build that sort of thing"
Meaning, perhaps if you couldn't find a trustworthy 3rd party that would host it, it might be worth $5 a month to host it yourself, and make it available to others.
Edit: On your other points:
There's nothing stopping you from buying a generic sounding domain like "pastethings.io". Not sure why that doesn't separate it far enough.
As for "code in a post", not sure what you mean. Javascript that runs in the browser? Formatted source code for viewing? Server side code that actually runs, but embedded in a post?
The nice part about that idea is the flexibility you get over where you share specific content, and how each piece can be platform-agnostic. You could essentially augment Twitter or Tumblr and create posts for each with a single interface.
So I've actually been building what you described [0], though it's focused on publishing text without formatting right now. But if you ever give it a try, I'd love to hear any thoughts you have on it!
Once IPFS gets into browsers you could use it like this. Just publish an HTML document into IPFS and share the link on the microblog. The page is a stand-alone object, hosted by anyone that views it, with a canonical, content-derived IRI.
Part of what I envision when I say "blogging engine on the back" is that you'd be able to update your posts. I'm picturing long-form 'evergreen' content here—big, fancy articles with lots of pictures and graphs and maybe even interactive code-samples, which you'd want to maintain over time to keep up with the state of things.
One annoying thing about the way blogs currently work is that the main reverse-chronological 'feed' stream that people subscribe to when they follow "your blog" is composed directly of articles you create, so people rarely post this kind of content—precisely because, when they update it, they'd either have to "pollute" their articles index (which people also rely on being searchable) with a little post saying they updated the original article and linking back to it; or they'd have to repost the updated article as a whole, duplicating it in the articles index and giving people coming from Google two separate pages to land on.
Separating blogging into article-posting and microblogging fixes this, because your microblog basically becomes an 'event feed'—you can have one post saying "Posted http://example.com/foo" and another one later saying "Updated http://example.com/foo". Your articles are no longer in the same namespace as news about your articles, so now you're free to have a one-to-many relationship between the articles and the news-posts about them.
The thing I worry about with sites like that is that I never know whether they're going to be here tomorrow. There was another site, http://wrttn.in, which seemed quite similar, but is now defunct. If I hadn't saved some of the good essays from that site, they'd be gone forever.
Interesting. So it's like Wordpress in that it's easy to install and has a web interface except it generates a static site instead of using a DB. I can see the appeal in that.
I'm sure there's a special feature that makes this valuable to some, but how is it different from creating your own static pages (using Pandoc or a similar system) and uploading them to a web server?
Then mention feathers in their "more" section http://blogless.datenbrei.de/more-about-blogless/
"Is it some kind of Medium.com, Known or feathe.rs? Yes, in some way, but single-user and without commercial interest."
Definition of blog: "a regularly updated website or web page, typically one run by an individual or small group, that is written in an informal or conversational style."
This is a static site generator written in PHP meant to allow authors to upload content-filled pages. You can call a dog a sheep but it's still a dog. Not demeaning the project just the "we're not a blog" copy.
Ok, we replaced "without a blog" and tried to use representative language from the article. If someone suggests a better (more accurate and neutral) title, we can change it again.
I've always wanted such a thing; it'd be perfect to combine with a microblogging site like Twitter or Tumblr, where instead of posting long things to the microblogging site, you post links to your "floating" essays amongst your short posts. Your microblog would then be the canonical "blog"—the canonical index of these articles—rather than people assuming they could subscribe to your presence at whatever article editing+hosting service you chose to use (which might, after all, change with each article.)
Too bad this isn't that.