Conversation
Notices
-
@macgirvin @pskosinski IANA protocol engineer, so please take what I say here as attempting to learn, not questioning your credentials ...
-
OK, so I just had another go at using #Diaspora. I'm still getting pretty much no interaction, despite posting a bunch of public posts ...
-
The interface is still confusing and requires too many clicks and page reloads to do simple, basic tasks, and it's sloooooowwwww
-
... and that's without the #XMPP chat turned on. I mean #Qvitter isn't flawless either but it's worth the pain coz people talk to me here
-
@rysiek ae, that's still pretty much their position
https://schub.io/blog/2018/02/01/activitypub-one-protocol-to-rule-them-all.html
-
@deadsuperhero @rysiek that's great, but ...
https://schub.io/blog/2018/02/01/activitypub-one-protocol-to-rule-them-all.html
-
@rysiek @deadsuperhero as more people find out #Hubzilla / #Friendica do a better job of what D* does, they are in danger of falling behind
-
@rysiek @deadsuperhero eg I'm clicking the 'mark all notifications as read' button. Not only should I not have to do that, it's not working!
-
@deadsuperhero @rysiek I guess it's akin to Tent wondering why all the established networks didn't implement their roll-your-own protocol ;)
-
@deadsuperhero @rysiek part of the problem could be I'm using a deblobbed Ubuntu 14.04, but the Firefox fork !trisquel supply is up-to-date
-
@deadsuperhero @rysiek I'm referred to the relevance of Metcalfe's Law to the discussion. Tent were too small to matter to Identi.ca et al
-
@rysiek @deadsuperhero if #Tent had pulled in a massive user base it would have been worth the technical hassle of supporting their protocol
-
@rysiek @deadsuperhero similarly, if the #fediverse grows massively, due to #MediaGoblin and others implementing #ActivityPub ...
-
@rysiek @deadsuperhero ... then the D* devs might start to see the technical debt of implementing ActivityPub as worth paying
-
@deadsuperhero @rysiek I do agree about the email style identifier. IMHO adoption of #OpenID has been limited by the URL style identifier
-
@rysiek let's not overstate the case. As @deadsuperhero points out D* protocol is *much* better documented now and a reference module exists
-
@rysiek @deadsuperhero D* have worked with MikeMcG and others so federation works with Hubzilla/ Friendica, Socialhome, and GangGo
-
@pskosinski @rysiek @deadsuperhero if it's merely a numbers game, how do the D* user numbers compare to #FarceBook?
-
@pskosinski since Hubzilla can federate with most of the other federated networks, it doesn't really matter how many use Hubzilla itself ...
-
@pskosinski which package has the better #UX? Hubzilla is far from perfect when it come to #UX, but has more features than D* (eg calendar)
-
@rysiek @deadsuperhero but I disagree about the following vs. 'sharing with' logic. This is one of the things I find very confusing about D*
-
@banjofox like I said to @rysiek let's not overstate the case. There are real compatibility headaches for D*. It's not just a case of #NIH.
-
@rysiek @banjofox the approach of supporting as many protocols as possible through a federation module/ library seems like a winner to me
-
@rysiek @banjofox the evolution from OStatus/ Pump.io to AP has moved us closer to unified federation. But there's more work to be done
-
@rysiek @banjofox also, I suspect that trying to make a single social protocol cover every possible use case risks ending up with xkcd-927
-
@artim how do I find and connect with you on D*?
-
@pskosinski yeah, Mike had a change of heart on #ActivityPub, see:
https://medium.com/we-distribute/got-zot-mike-macgirvin-45287601ff19
-
@pskosinski I agree Hubzilla/Zot needs a complete #UX makeover, like what Mastodon was originally founded to do for GNUSocial/OStatus
-
@pskosinski but I think federation between networks using Diaspora's protocol (D*, #Hubzilla, #Friendica, #SociaHome, #GangGo) has improved
-
@pskosinski ... but as I said the #UX of Diaspora needs work just as much or more than Hubzilla, just like most of the #FreeNetworks apps
-
@pskosinski @macgirvin ... but aren't these problems of the protocol implementations, not failings of the protocols themselves?