Well I guess I am going to have to be that guy:

Does EVERY new feature released by github require an HN announcement? I mean I can see reasons for the big, useful stuff used by many people, but this is such a non-event I am having trouble understanding why it is here.

It's lighthearted fun. Some guy at Github said, what the hell. And within a day of a new language being announced[1], it's already "supported" by github. Is this any kind of a big deal? no. But I know I cracked a smile.

[1] http://news.ycombinator.com/item?id=5272634

EDIT: was not aware that the "Tom" in TOML is the CEO of github. So nevermind.

<jealousy> And now whenever we see a config file that looks like that style (which is common) it'll be considered TOML. Oh, the spoils of nerd fame ;) </jealousy>

I'll second this. My friend made a pretty cool and useful language (moonscript.org) and after it got popular he submitted a pull request for syntax highlighting. It took months and lots of people commenting for it to get merged in. So to see this BS front-paged grinds my gears a bit.

Did you flag it? This site is a democracy (of sorts) so what does well is what everyone collectively votes for. If enough people flag items they think are offtopic and items disappear, that's also democratic.

This is not "released by GitHub". This is just a guy (who happens to have founded GitHub) who wrote something cool.

That's exactly what HN is (supposed to) be about.

The "Wrote something cool" is all well and good and I agree that is the kind of thing that HN is about.

However that story was already covered yesterday: http://news.ycombinator.com/item?id=5272634

This story is nothing more than "Hey, Github now supports syntax highlighting of the brand-new nobody-is-really-using-yet config format written by its founder which was mentioned yesterday" which seems like a non-event not worthy of taking up space.

It's just someone posting it and if people like it, they upvote it. You seem to be part of the minority..

Without commenting on the merit of this story I'd like to point out that 25 (or 500) upvotes does not mean that the majority of people like the story or that the people who do not like the story are in the minority. Unlike comments there are no downvotes for submissions.

But one can flag a submission if need be.

Nice. Can we figure all this out all these things before we get too far with this? https://github.com/mojombo/toml/issues/42

Interestingly enough, TOML is somewhat similar to valid CoffeeScript hash/array syntax, apart from the auto-vivifying hashes bit (and native dates, natch). Here's the top of that highlighted document:

    # This is not a TOML document. Boom.
    
    title: "TOML Example"
    
    owner:
      name: "Tom Preston-Werner"
      organization: "GitHub"
      bio: "GitHub Cofounder & CEO\nLikes tater tots and beer."
      dob: "1979-05-27T07:32:00Z" # First class dates? 'Fraid not.
    
    database: 
      server: "192.168.1.1"
      ports: [ 8001, 8001, 8002 ]
      connection_max: 5000
      enabled: true
    
    servers:
    
      # Indentation matters. Tabs or spaces. CS don't care.
      alpha:
        ip: "10.0.0.1"
        dc: "eqdc10"
    
      beta:
        ip: "10.0.0.2"
        dc: "eqdc10"    

Proof: http://goo.gl/3xV4P

That looks like valid YAML to me.

Definitely top priority when they're still running Rails 2.3.

Why does it matter if they are on 2.3 or whatever if they can keep pushing the features? I don't care which Rails they are using ( provided they are secure, of course)

Now, why this is on HN, is a different question that is worthy of being asked.

Gotta love developer-driven companies.

Just because Tom created it! A few days old format definitely not used by many I guess

Very entertaining looking at all the libs that have popped up to support TOML.

Has anyone looked at the Go version? For all the Go examples I've looked at (mostly networking) they are mostly short and easy to follow. I was shocked at how much code the author wrote.

https://github.com/pelletier/go-toml/tree/master/src/toml

    (*subtree)[keys[len(keys)-1]] = value

This is the first time I see this kind of syntax in Go, what does (*subtree) do?

I believe it means: Dereference the pointer `subtree` and assign `value` to the key `keys[len(keys)-1]` in the map yielded by the dereference.

Can anyone tell me the motivation behind TOML? Why isn't JSON enough?

Discussed in yesterday's thread http://news.ycombinator.com/item?id=5272634

The intention is to be easier for humans to read and write. Strict JSON requires a clutter of double-quotes, curly braces, parentheses and commas which have no place in a human-editable format.

first class dates, seems like its ready to replace xml and much more streaming/real time friendly.

It'd be nice if they'd add Rust to the list.

The good news is that anyone can do it :D

Check out https://github.com/github/linguist. It's the library GH uses to detect and highlight languages; and _many_ people external from GH have contributed to it. It's just a matter of someone adding a lexer for Rust and submitting a pull request :D

I _almost_ added support to linguist once, but something stopped me. I can't quite remember what it was... I _think_ it was that only edge pygments supported Rust.

Writing syntax highlighter for TOML takes ~10mins. Rust syntax isn't so simple.

there's probably more TOML than Rust on GitHub already, though

I highly doubt that. At the very least: Servo.

Also: rustc

And core/std libs.