Hacker News new | past | comments | ask | show | jobs | submit login

RSS is the only way to read news :D Reddit, HN - everything goes to my RSS ;)



Then why not use RSS, or Atom, or CSV, or HTML, for that matter?

That OPML "outline" has no tree structure at all, except for the single top level container item:

    <outline title="HN Personal Blogs" text="HN Personal Blogs">
What is the point of using OPML for a flat structure?

Outlines are all well and fine, but a list of links to blogs is hardly an outline.

And isn't JSON the simplest most ubiquitous accessible standard format for outlines and even arrays of links these days, if not just raw-dogging XML?

More thoughts and opinions and history of OPML, RSS, Dave Winer, More, Frontier, Manila, Radio UserLand, XML-RPC, SOAP, Xanadu, spreadsheets, CSV, JSON, XML, etc:

https://news.ycombinator.com/item?id=21170434

https://news.ycombinator.com/item?id=16224154

https://news.ycombinator.com/item?id=20780928

https://news.ycombinator.com/item?id=35674125

https://news.ycombinator.com/item?id=36544675


I think it's just that OPML has traditionally been used to import/export lists of RSS feeds, and therefore most readers are capable of importing an OPML file whereas other formats would not be as widely supported.


Why do RSS feeds need to be imported/exported as anything other than RSS feeds?

According to this thread, four out of five RSS apps on Linux Mint failed to import the OPML file, and the one that did was buggy:

https://news.ycombinator.com/item?id=36627431

Maybe I'm thick, but it bewilders me that someone wouldn't simply use RSS or Atom directly, instead of abusing OPML to mis-represent it (which is itself just as terribly designed as RSS, by the same person, Dave Winer -- don't even get me started).

Is this some kind of a masochistic corrupted data ingestion challenge from TikTok that I'm not aware of, akin to eating Tide Pods or washing your clothes in Campbell's Soup?

Does anyone even ever use OPML for outlines any more? Or is it like YAML's backflip, where its original name meant "Yet Another Markup Language" until somebody pointed out to their great surprise that it WASN'T actually a markup language, so they retroactively recursively renamed it "YAML Ain't Markup Language" on opposite day?

So does OPML no longer stand for "Outline Processor Markup Language" any more, but now stands for something else like "OPML Presents Mere Lists"?

What's wrong with exporting and importing RSS as RSS? Has its meaning also changed, flipping places with OPML, so it now means "RSS Serves Structures", so you use RSS for nested trees and OPML for flat lists now?

I suppose you could recursively encode outlines as nested RSS feeds inside the text content other RSS feeds as <![CDATA[ at each level, or since those can't be nested, exponentially doubling the number of escaped entities at each level.

I Wanna Be <![CDATA[ -- Sung to the tune of “I Wanna Be Sedated”, with apologies to The Ramones:

https://donhopkins.medium.com/i-wanna-be-cdata-3406e14d4f21


This is the way. I just swapped as many news, social (reddit/etc) and the like over to newsblur via RSS and I'm VERY happy. This OPML will go a long way to filling out my categories :-) Thanks a bunch for your efforts!


RSS traditionally kinda sucks for discovery purposes though. (Unless you count these HN threads as the discovery part)

I do wonder what's a good way of doing discovery in software, FOSS-style. Using open source client side and free APIs if needed.


What about subscribing to OPML files hosted on URLs ? That way one can follow OPML list "hosts" (like the one on OPs repository) and discover new RSS feeds. I'm quite surprised that this feature do not exists in RSS reader I tried.


Good. I'm tired of apps suggesting stuff to me.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: