Because the world needs another standard

https://jsonfeed.org/

My personal favourite parts from the site:

JSON is […] less prone to bugs.

And:

It’s at version 1, which may be the only version ever needed.

perfection from the first take

a trait also attributed to Ed Wood, Jr.

Why? Because, if it is not broken, fiddle with it till it is!

the time has come to disrupt rss feeds gets millions of venture capital, spends on hookers and blow

you’re just jealous that you didn’t think of it first.

https://tt-rss.org/fox/tt-rss/blob/master/classes/handler/public.php#L185 :confounded:

But you didn’t do a quick write-up and publish it under a fancy domain name!

fox, I meant the venture capital, hookers and blow.

It’s spreading…

It would be easy enough for fox to add support for this to TT-RSS. I just really question the whole idea of it being “less malformed” and “more human readable”. The entire concept of feeds and feed readers automates all of this, so any “malformed” content is a problem with the generator. In other words, write a proper generator and, viola, you don’t have a malformed feed. Having malformed JSON feeds will be just as common place as RSS/Atom feeds because some people just can’t be bothered to code things properly.

And honestly, none of this crap is really “human readable”. I can read XML about as well as I can read JSON. Then again, it’s not meant to be human readable since, again, the whole concept is to automate the transfer of information.

In the end, my instinct tells me a few big sites will adopt this because “Hey, look at us,” and all the readers in the world will need to add support for it.

i’m not going to recognize this idiotic abortion of a thing nobody asked for by implementing support for it

In the end, my instinct tells me a few big sites will adopt this because “Hey, look at us,”

i’m sure hacker news and other cancerous shitholes of the startup culture are hard at work already

and all the readers in the world will need to add support for it.

this is why we have plugins

I just really question the whole idea of it being “less malformed”

this is the hilarious part, XML document can at least be validated using a DTD, JSON has nothing of the sort and its syntax and type documentation is basically "¯\(ツ)/¯ ".

i wouldn’t be surprised this garbage exists because some dimwit wanted to syndicate shit with his nodejs backend or whatever and XML was just too fucking hard for his wee little brain.

e: lmao, one of these two geniuses literally wants to kickstart a microblog. god is dead.

e2: what we really need is a protobuf-based syndication. you heard it here first. original idea, do not steal.

LOL… nodejs really seems like some guy who only does frontend dev wanting to be able to do backend without learning anything new.

e: People bash PHP, but I’ve been playing around with Ghost blog and frankly, I can run an entire LEMP stack with WordPress (or TT-RSS) with 128 MB of RAM. Ghost blog on nodejs uses 80 MB right out of the box, not including the web server or database.

let me tell you about a popular framework known as ruby on rails

  PID USER      PR  NI    VIRT    RES    SHR S %CPU %MEM     TIME+ COMMAND
21043 fox       20   0 1764736 302724  12392 S  0.0 14.7 201:13.83 ruby
21125 fox       20   0 1683736 287948  13284 S  0.0 14.0 844:38.96 ruby
21105 fox       20   0 1751324 278532  12604 S  0.0 13.5 846:33.66 ruby
21081 fox       20   0 1668384 255180  12440 S 18.9 12.4 853:50.59 ruby
21064 fox       20   0 1668376 225564  12020 S  0.0 11.0 845:23.95 ruby
21018 fox       20   0  441088 123912   4104 S  0.0  6.0  49:02.21 ruby

this forum has like 3 users at peak, i’m sure eating all this is entirely justified because of reasons

e: also, another overbloated rails abortion aka gitlab literally needs nodejs to precompile assets for reasons i unironically don’t want to fucking know because it probably would make me angry enough to uninstall it

[quote=“fox, post:11, topic:193, full:true”]>I just really question the whole idea of it being “less malformed”

this is the hilarious part, XML document can at least be validated using a DTD, JSON has nothing of the sort and its syntax and type documentation is basically "¯\(ツ)/¯ ".[/quote]
I’m absolutely not defending this JSON feed stuff, but it is possible to define validation schemas for JSON and then perform the validation. I’ve no idea if this is any better/worse than XML and DTDs though.

I think I would have trouble sleeping at night knowing that crap was running.

Thankfully never had to deal with Ruby.

I’m absolutely not defending this JSON feed stuff, but it is possible to define validation schemas for JSON and then perform the validation.

oh cool, looks like there’s an IETF draft which may or may not be supported by anything (i honestly have no idea). i haven’t found any mentions of it on the jsonfeed authors website, although admittedly i wasn’t looking very hard.

well guys i guess the era of XML is over promptly writes out a boolean falue as “false”, crashes client

https://tt-rss.org/fox/tt-rss/commits/jsonfeed-test-branch :effort:

i’m not personally running it but not gonna stop those eager to disrupt everything

Actually there is validation available. http://json-schema.org/ (which I see someone’s already alluded to.)

Just a shame that in their haste to write a badly written spec, https://jsonfeed.org/ didn’t actually provide one. The spec might have been better written if they had.

Presuming, of course, that someone proposing such nonsense in the first place is actually aware of it.

Just a shame that in their haste to write a badly written spec, https://jsonfeed.org/ didn’t actually provide one. The spec might have been better written if they had.

this is basically startup culture.txt. thinking things through is an obsolete paradigm. it was patently obvious with how they boldly declared the initial version of their spec was literally perfect.

personally i would be surprised if writing this spec took them more than 15 minutes because it’s basically something very close to RSS transplanted into a different transport format with some attributes arbitrarily renamed and “oh well you can embed random json objects everywhere, have fun” thrown in for extensibility.

maybe in a few years YML (comedy option: INI) becomes the hot new thing with the hacker news crowd and we’ll live to see someone register and shit out YmlFeed with similar amount of effort.