3 ms·

The problem with RSS is that most feeds treat all items the same so the same prominence is attached to a nuclear strike on New York as is attached to a child sc

by erfgh 9y ago

The problem with RSS is that most feeds treat all items the same so the same prominence is attached to a nuclear strike on New York as is attached to a child scraping its knee.

tvmalsv 9y ago
I'm ok with that. I can eyeball a bunch of RSS entries very quickly, and I think a "NYC NUKED" headline will get my attention just fine. I'll attach my own level of interest.
Gys 9y ago
To me that is another benefit of rss. I want to do my own 'manual' filtering, to decide myself what is important to me. I do not want some algorithm to feed me. Especially not if that algorithm is commercially biased: its main goal is not to inform me, the goal is to make money from me.
- JohnBooty 9y ago
  The problem I've found with aggregating multiple feeds into a single chronological stream is that some feeds totally overwhelm others because they publish more frequently. If Feed A publishes 10 articles a day and Feed B publishes 10 articles a year, the articles from Feed B get totally engulfed and pushed way, way, way down the list... effectively they become invisible.
  - dispo001 9y ago
    My aggregator limits the number of items by feed and by domain. The problem is less of an issue if you add many more feeds and filter harder. Those rarely publishing do still get drowned out but it gives you plenty of other things to read. Thanks to volume, at times, if the slow publisher posts something awesome (they often do) it is reposted by the more spammy domains. I do have other ideas to fix that issue but the ink is still wet on those.
dispo001 9y ago
While major news outlets pretty much always have separate feeds for different categories the real awesomeness of RSS was pretty much not understood: There is the <category> element! I aggregate many many thousands of feeds, I've never seen anyone use the category properly. I actually cant find any attempt at a specification that clearly describes it. (I might be confused by the versions and feeds using a mix of elements from different versions) You know things are bad when w3schools[1] is the only one providing a full description of the difference between non hierarchic and hierarchic. On the domain attribute they are as stupid as everyone else. I'm not even sure how the domain attribute should be used. I do instinctively know that putting domain="syndic8" is wrong as it is not a domain. That however isn't stopping everyone from using it as an example in their version of the spec. Anyway: The <category> element specifies one or more categories the channel or item belongs to separated by commas. The element can ALSO specify the hierarchic location in a category. The value of the <category> element is then a forward-slash separated string. So you could entirely have something totally like: <category>local/traffic acidents,injury,children</category> Where "local > traffic" is the hierarchy while "injury" and "children" are more like tags. On the rare occasion I see a <category> element in an RSS feed they 99% of the time use "General". Using separate feeds to accomplish the same(?) is not great but it works. Here is an OPML I bake myself for the guardian.[2] Their feeds do actually use categories complete with domains. Lots of them![3] Don't ask me how to consume this feedstock. To me the weirdness of the lack of category is primarily in every properly structured site having them. If you are pulling the title, pubdate and content from your database it shouldn't take much effort to also pull the category and have a <category>pokemon</category> element in the news item. Fun fact: While it is missing from the w3 validator spec[4] the harvard.edu version of the spec[5] does mention the historic awesome "ratings" element. A thing from back in the days when non-corporate technology was still a thing. Namely The PICS rating system[6]. This hidden treasure was pretty much hated to death by loud mouth corporate propagandists. (I'm putting it mildly) The public rating your article is the worse kind of scrutiny when working for a mainstream propaganda outlet portraying it self as if news. The status quo eventually nailed the coffin shut by promoting it as a parental-oversight adult content filter which became the only surviving implementation. Sadly the RSS spec only allows rating the channel rather than its items which spoils most of the theoretical awesome. :( Thus therefore as the implication thereof, at least in theory, RSS has the most awesome tagging, categorization, rating and review system ever envisioned (not-seen) on the internet. People just don't know how to technology. [1] - https://www.w3schools.com/XML/rss_tag_category_channel.asp https://www.w3schools.com/XML/rss_tag_category_channel.asp [2] - http://opml.go-here.nl/theguardian.opml http://opml.go-here.nl/theguardian.opml [3] - view-source:https://www.theguardian.com/business/vodafonegroup/rss https://www.theguardian.com/business/vodafonegroup/rss [4] - https://validator.w3.org/feed/docs/rss2.html#hrelementsOfLtitemgt https://validator.w3.org/feed/docs/rss2.html#hrelementsOfLti... [5] - https://cyber.harvard.edu/rss/rss.html#optionalChannelElements https://cyber.harvard.edu/rss/rss.html#optionalChannelElemen... [6] - https://www.w3.org/PICS/ https://www.w3.org/PICS/

by erfgh 9y ago

The problem with RSS is that most feeds treat all items the same so the same prominence is attached to a nuclear strike on New York as is attached to a child scraping its knee.

tvmalsv 9y ago

I'm ok with that. I can eyeball a bunch of RSS entries very quickly, and I think a "NYC NUKED" headline will get my attention just fine. I'll attach my own level of interest.

Gys 9y ago

To me that is another benefit of rss. I want to do my own 'manual' filtering, to decide myself what is important to me. I do not want some algorithm to feed me. Especially not if that algorithm is commercially biased: its main goal is not to inform me, the goal is to make money from me.

JohnBooty 9y ago
The problem I've found with aggregating multiple feeds into a single chronological stream is that some feeds totally overwhelm others because they publish more frequently. If Feed A publishes 10 articles a day and Feed B publishes 10 articles a year, the articles from Feed B get totally engulfed and pushed way, way, way down the list... effectively they become invisible.
- dispo001 9y ago
  My aggregator limits the number of items by feed and by domain. The problem is less of an issue if you add many more feeds and filter harder. Those rarely publishing do still get drowned out but it gives you plenty of other things to read. Thanks to volume, at times, if the slow publisher posts something awesome (they often do) it is reposted by the more spammy domains. I do have other ideas to fix that issue but the ink is still wet on those.

dispo001 9y ago

While major news outlets pretty much always have separate feeds for different categories the real awesomeness of RSS was pretty much not understood: There is the <category> element! I aggregate many many thousands of feeds, I've never seen anyone use the category properly. I actually cant find any attempt at a specification that clearly describes it. (I might be confused by the versions and feeds using a mix of elements from different versions) You know things are bad when w3schools[1] is the only one providing a full description of the difference between non hierarchic and hierarchic. On the domain attribute they are as stupid as everyone else. I'm not even sure how the domain attribute should be used. I do instinctively know that putting domain="syndic8" is wrong as it is not a domain. That however isn't stopping everyone from using it as an example in their version of the spec. Anyway: The <category> element specifies one or more categories the channel or item belongs to separated by commas. The element can ALSO specify the hierarchic location in a category. The value of the <category> element is then a forward-slash separated string. So you could entirely have something totally like: <category>local/traffic acidents,injury,children</category> Where "local > traffic" is the hierarchy while "injury" and "children" are more like tags. On the rare occasion I see a <category> element in an RSS feed they 99% of the time use "General". Using separate feeds to accomplish the same(?) is not great but it works. Here is an OPML I bake myself for the guardian.[2] Their feeds do actually use categories complete with domains. Lots of them![3] Don't ask me how to consume this feedstock. To me the weirdness of the lack of category is primarily in every properly structured site having them. If you are pulling the title, pubdate and content from your database it shouldn't take much effort to also pull the category and have a <category>pokemon</category> element in the news item. Fun fact: While it is missing from the w3 validator spec[4] the harvard.edu version of the spec[5] does mention the historic awesome "ratings" element. A thing from back in the days when non-corporate technology was still a thing. Namely The PICS rating system[6]. This hidden treasure was pretty much hated to death by loud mouth corporate propagandists. (I'm putting it mildly) The public rating your article is the worse kind of scrutiny when working for a mainstream propaganda outlet portraying it self as if news. The status quo eventually nailed the coffin shut by promoting it as a parental-oversight adult content filter which became the only surviving implementation. Sadly the RSS spec only allows rating the channel rather than its items which spoils most of the theoretical awesome. :( Thus therefore as the implication thereof, at least in theory, RSS has the most awesome tagging, categorization, rating and review system ever envisioned (not-seen) on the internet. People just don't know how to technology. [1] - https://www.w3schools.com/XML/rss_tag_category_channel.asp https://www.w3schools.com/XML/rss_tag_category_channel.asp [2] - http://opml.go-here.nl/theguardian.opml http://opml.go-here.nl/theguardian.opml [3] - view-source:https://www.theguardian.com/business/vodafonegroup/rss https://www.theguardian.com/business/vodafonegroup/rss [4] - https://validator.w3.org/feed/docs/rss2.html#hrelementsOfLtitemgt https://validator.w3.org/feed/docs/rss2.html#hrelementsOfLti... [5] - https://cyber.harvard.edu/rss/rss.html#optionalChannelElements https://cyber.harvard.edu/rss/rss.html#optionalChannelElemen... [6] - https://www.w3.org/PICS/ https://www.w3.org/PICS/