Coffee Space – Coffee Space

RSS Mainstream

I found myself thinking about RSS again after Niki Usher published “This is the year of the RSS reader. (Really!)”.

I did then posted about this on dead social:

I started discussing a wish-list of features for RSS, but I thought it would be good to spell these out better here.

Hashing

Firstly, I think there should be some way to embed a hash of the content to ensure what is delivered is correct. I think the easiest way to implement this would be to add to the HTTP HEAD command, so that one of the fields is a strong hash of the content.

This would allow the following:

Redistribution - A CDN could deliver the message body, with it being authenticated by the original server as having the same expected hash. In the HEAD could also be a set of mirror locations to download the content.
Update Checking - Rather than pull the entire feed, an RSS reader could simply poll the head to check for changes. This would massively save on bandwidth for feeds that contain an entire site, for example.
Computation reduction - An RSS feed reader could download the content over HTTP, saving both the server and client computation (at the sacrifice of privacy).

All of this also applies to normal web pages. It just makes sense to make a hash available via the HEAD command.

Comments

One of the things I have not seen solved elegantly is a way to add comments to RSS feed items and how to post comments from an RSS feed reader. I have a few ideas on this:

Comments tag - There is some comments tag, i.e.: <comments>ABSOLUTE_URL</comments> that links to another RSS feed containing comments. In this way, each comment could also link to another <comments> tag, allowing threading! The only thing the RSS feed reader would need to check for is circular references. I would also impose some limits as to how many comments are loaded before some user interaction is required to load further comments.
Posting - One thing we will want to do is add comments to other’s comments threads. I think the best way to do this would be to author content on an external RSS feed and then publish the link to this. The site would then attempt to load from this URL and reply with the success of doing so, making a mirror and linking back. For Countries under GDPR, the site would need to occasionally poll the original RSS feed and check whether the comment has been updated or marked with some deletion flag. It would then be their responsibility to respect this.

Note: Each site would also likely want to perform some form of sanitization of the content they are about to mirror. Dealing with just text somewhat simplified the problem, removing all HTML tags simplified things further. From there it would be a case of searching keywords and manual moderation.

The greatest part about this type of implementation is that it would allow for distributed social media to be built around the concept.

Anyway, this is potentially a way to bring the Dead Social V2 to life…