RSS feed readers are BACK, babe
-
@blaine @cwebber I have a question for both of you while I have your ear.
It's about RSS 2.0!
One thing about ActivityPub that makes it hard to implement is that we don't have an explicit description of how a polling mode would work. I started one here: https://w3id.org/fep/b06c
I think it may make sense to describe how RSS 2.0 could fit in the picture, too. Having a defined mapping for RSS items into AS2, defining how IDs and uniqueness can be treated, and so on.
-
@evan @cwebber the discussion of protocols is often deeply embedded in the standards world, and protocols are defined by standards.
Really sorry - one of the documents that is referenced in the link you've shared was the last time I tried to engage, and the experience was *horrendous*. My good-faith contributions were flat-out rejected by an opinionated tech mini-lord, and your name is on the doc. That's not "protocol development" or consensus building.
-
@blaine @cwebber I have a question for both of you while I have your ear.
It's about RSS 2.0!
One thing about ActivityPub that makes it hard to implement is that we don't have an explicit description of how a polling mode would work. I started one here: https://w3id.org/fep/b06c
I think it may make sense to describe how RSS 2.0 could fit in the picture, too. Having a defined mapping for RSS items into AS2, defining how IDs and uniqueness can be treated, and so on.
@evan @cwebber I feel like one day we should have a beer again. My "same picture" take has been (for many, many years) that ActivityPub is two-way RSS with usable identity (even if the standard doesn't admit it). XMPP is the same, and the main innovation from atproto is actually-malleable schemas. Private data will bring it in-line with AP, while perf. optimizations will work the other direction. All asides, though: I care about what the people want, not what the computers want. ❤️
-
@evan @cwebber I feel like one day we should have a beer again. My "same picture" take has been (for many, many years) that ActivityPub is two-way RSS with usable identity (even if the standard doesn't admit it). XMPP is the same, and the main innovation from atproto is actually-malleable schemas. Private data will bring it in-line with AP, while perf. optimizations will work the other direction. All asides, though: I care about what the people want, not what the computers want. ❤️
@blaine @cwebber Your take on the topology is spot on, and there's a traceable lineage from polled RSS to pushed RSS to OStatus to ActivityPub. I think AP supports innovation through different schemas, but it really requires supporting the AP API. See https://www.w3.org/wiki/Activity_Streams/Primer/Extensions for examples.
-
undefined Piero Bosio ha condiviso questa discussione
-
RSS feed readers are BACK, babe!
If it gets too much to do polling all the time and that overwhelms the network, maybe we can add directed delivery, publish subscribe style
It could be like email
@cwebber Indeed, global like counts and trending hashtags are only useful for the advertising industry.
-
RSS feed readers are BACK, babe!
If it gets too much to do polling all the time and that overwhelms the network, maybe we can add directed delivery, publish subscribe style
It could be like email
@cwebber I actually do like the PDS part of bluesky, and I like the DID part except for the fact that it's fake. PDSes that can talk to each other via activitypub sounds highly desirable to me.
-
@evan @cwebber the discussion of protocols is often deeply embedded in the standards world, and protocols are defined by standards.
Really sorry - one of the documents that is referenced in the link you've shared was the last time I tried to engage, and the experience was *horrendous*. My good-faith contributions were flat-out rejected by an opinionated tech mini-lord, and your name is on the doc. That's not "protocol development" or consensus building.
@blaine @cwebber this is the AP Webfinger profile, right? I remember this.
I just re-read the issue comment history. It's weird that I didn't reply anywhere in here, given that I was tagged. I am sorry that your idea didn't get into the doc, even as a future enhancement. It seems like a useful feature and part of the point of Webfinger.
I opened some new issues on that repo recently. I'm going to reopen this one, if you don't mind. It seems like a really good topic for discussion.
-
@cwebber I actually do like the PDS part of bluesky, and I like the DID part except for the fact that it's fake. PDSes that can talk to each other via activitypub sounds highly desirable to me.
You should check wafrn!
It has blue sky and activitypub support.
-
You should check wafrn!
It has blue sky and activitypub support.
@irelephant @cwebber but the pages load sooo slooooow
-
@irelephant @cwebber but the pages load sooo slooooow
Really? Its pretty fast for me rn.
-
@blaine @cwebber this is the AP Webfinger profile, right? I remember this.
I just re-read the issue comment history. It's weird that I didn't reply anywhere in here, given that I was tagged. I am sorry that your idea didn't get into the doc, even as a future enhancement. It seems like a useful feature and part of the point of Webfinger.
I opened some new issues on that repo recently. I'm going to reopen this one, if you don't mind. It seems like a really good topic for discussion.
-
Really? Its pretty fast for me rn.
@irelephant @cwebber when i click the ⋯ next to this post, then click "open original page", and i start a stopwatch the moment i see the new tab open, i see:
- a picture of a bug and a sentence of text.
- the sentence of text is replaced with "did you know nachos aren't even mexican".
- a skeleton of the wafrn layout, and a throbber.
- eventually, the post.Final time on stopwatch: 2.26 seconds.
I do the same thing on a cosocial post: https://cosocial.ca/@evan/115220440152189152
Result: Under 0.5 seconds.
-
@irelephant @cwebber when i click the ⋯ next to this post, then click "open original page", and i start a stopwatch the moment i see the new tab open, i see:
- a picture of a bug and a sentence of text.
- the sentence of text is replaced with "did you know nachos aren't even mexican".
- a skeleton of the wafrn layout, and a throbber.
- eventually, the post.Final time on stopwatch: 2.26 seconds.
I do the same thing on a cosocial post: https://cosocial.ca/@evan/115220440152189152
Result: Under 0.5 seconds.
@irelephant @cwebber The cosocial page loaded so fast I actually couldn't successfully stopwatch it.
-
@blaine @cwebber Anyways, I'm sorry your experience there was horrendous, especially since I had encouraged you to participate. As one of the authors of that report, I should have been more careful to make sure your suggestion was included. I apologize for not doing a better job; I'll try to learn from the experience to do better in the future.
-
@irelephant @cwebber The cosocial page loaded so fast I actually couldn't successfully stopwatch it.
2 seconds for the initial load when there is a PWA update is not bad.
#I-consider-it-an-acceptable-time #reducing-that-would-make-the-server-go-from-18€-to-81 -
undefined julian ha condiviso questa discussione
-
2 seconds for the initial load when there is a PWA update is not bad.
#I-consider-it-an-acceptable-time #reducing-that-would-make-the-server-go-from-18€-to-81gabboman@app.wafrn.net mm using PWA as an excuse to hand wave away slow page loads is so 2010s :smirk_cat:
NodeBB is a PWA and loads plenty fast, although there is a server side rendering component there.
Feed RSS
Gli ultimi otto messaggi ricevuti dalla Federazione
-
gabboman@app.wafrn.net mm using PWA as an excuse to hand wave away slow page loads is so 2010s :smirk_cat:
NodeBB is a PWA and loads plenty fast, although there is a server side rendering component there.
-
2 seconds for the initial load when there is a PWA update is not bad.
#I-consider-it-an-acceptable-time #reducing-that-would-make-the-server-go-from-18€-to-81
-
@blaine @cwebber Anyways, I'm sorry your experience there was horrendous, especially since I had encouraged you to participate. As one of the authors of that report, I should have been more careful to make sure your suggestion was included. I apologize for not doing a better job; I'll try to learn from the experience to do better in the future.
-
@irelephant @cwebber The cosocial page loaded so fast I actually couldn't successfully stopwatch it.
-
@irelephant @cwebber when i click the ⋯ next to this post, then click "open original page", and i start a stopwatch the moment i see the new tab open, i see:
- a picture of a bug and a sentence of text.
- the sentence of text is replaced with "did you know nachos aren't even mexican".
- a skeleton of the wafrn layout, and a throbber.
- eventually, the post.Final time on stopwatch: 2.26 seconds.
I do the same thing on a cosocial post: https://cosocial.ca/@evan/115220440152189152
Result: Under 0.5 seconds.
-
-
Really? Its pretty fast for me rn.
-
@irelephant @cwebber but the pages load sooo slooooow