If we look at what #Mastodon did to #ActivityPub, we start to realize that #XMPP not having a client and server from a single vendor that is available on all platforms is indeed a strength, not a weakness.
-
@lutindiscret @thevril @daniel and at the same time hamstrung it to a very limited and misshapen version of the actual specification that everyone needs to implement if they want to be compatible with the mastodon servers.
Basically they inadvertently executed an embrace, extend and extinguish operation.
-
@lutindiscret @thevril @daniel and at the same time hamstrung it to a very limited and misshapen version of the actual specification that everyone needs to implement if they want to be compatible with the mastodon servers.
Basically they inadvertently executed an embrace, extend and extinguish operation.
@mariusor @lutindiscret @thevril @daniel
As I understand things, Mastodon came first and ActivityPub was an effort to ratify the open protocol. But ActivityPub is sufficiently ill-defined (not very, just sufficiently) that any implementation will always be opinionated a little bit. If you want to interact with servers, you have to spend a lot of effort debugging the interface to each server. That's my experience at least, so I don't really blame Mastodon for what they have done.
-
@mariusor @lutindiscret @thevril @daniel
As I understand things, Mastodon came first and ActivityPub was an effort to ratify the open protocol. But ActivityPub is sufficiently ill-defined (not very, just sufficiently) that any implementation will always be opinionated a little bit. If you want to interact with servers, you have to spend a lot of effort debugging the interface to each server. That's my experience at least, so I don't really blame Mastodon for what they have done.
@khleedril yes, I'm aware of the chronology. Yet since they've adopted it, there was barely an effort from Mastodon to improve their spec conformance - and I'm thinking about basic things here, despite becoming an incorporated entity with developers on their payroll. I believe my statement still stands.
-
@lutindiscret @thevril @daniel and at the same time hamstrung it to a very limited and misshapen version of the actual specification that everyone needs to implement if they want to be compatible with the mastodon servers.
Basically they inadvertently executed an embrace, extend and extinguish operation.
@mariusor @lutindiscret @thevril @daniel Do you think there’s still room to improve things? What could be done here - I didn’t dig into the details so I’m not sure if it’s a totally lost cause (in which case, ignore me!)?
-
@mariusor @lutindiscret @thevril @daniel Do you think there’s still room to improve things? What could be done here - I didn’t dig into the details so I’m not sure if it’s a totally lost cause (in which case, ignore me!)?
@lornajane I don't know, hopefully not. My bitterness mostly stems from the fact that I'm a "small fediverse developer", meaning that I built fediverse servers that are not in the mainstream.
My software is making use of some features in the ActivityPub and Activity vocabulary that are spec compliant but they get rejected by Mastodon servers.
When I pointed out these problem(s) on the Mastodon bug tracker they weren't acknowledged, it was argued that they're bad features, etc. Around some of them I could work around (like requiring webfinger resource resolvers), but around others I couldn't and I won't.
-
@daniel But XMPP doesn't have anything on desktop that compares to Conversations. It helps to have a unified solution to set the minimum bar higher.
@kolev @daniel I love #XMPP but not having a single cross-platform fully-featured client makes onboarding normies much more difficult, hindering adoption greatly.
Nothing wrong with being the #Ubuntu or #LinuxMint of decentralised #FOSS messengers... people who need the Archlinuxes and BSDs will find them in time.
I have managed to onboard lots of people to cross-platform messengers like #DeltaChat, #ElementMessenger and #SignalApp, but not a single person to XMPP. Cross platform apps also make switching OSes eaiser, e.g. migrating from Windows to Linux, or Android to LinuxMobile.
-
@kolev @daniel I love #XMPP but not having a single cross-platform fully-featured client makes onboarding normies much more difficult, hindering adoption greatly.
Nothing wrong with being the #Ubuntu or #LinuxMint of decentralised #FOSS messengers... people who need the Archlinuxes and BSDs will find them in time.
I have managed to onboard lots of people to cross-platform messengers like #DeltaChat, #ElementMessenger and #SignalApp, but not a single person to XMPP. Cross platform apps also make switching OSes eaiser, e.g. migrating from Windows to Linux, or Android to LinuxMobile.
@opensourceopenmind @kolev @daniel What #XMPP clients would you recommend (doesn't have to cross-plattform) for
?
-
@phocque @opensourceopenmind @kolev @daniel
gajim or dino for GNU/Linux
monocles for Android
Monal for iOS and MacOS -
@phocque @opensourceopenmind @kolev @daniel
gajim or dino for GNU/Linux
monocles for Android
Monal for iOS and MacOS@greatsword @phocque @opensourceopenmind @kolev Be careful recommending #monocles. They’ve stopped back porting critical security updates from #Conversations_im some time ago. Just go with the original. Monocles is just the Blabber disaster waiting to happen again.
-
@lornajane I don't know, hopefully not. My bitterness mostly stems from the fact that I'm a "small fediverse developer", meaning that I built fediverse servers that are not in the mainstream.
My software is making use of some features in the ActivityPub and Activity vocabulary that are spec compliant but they get rejected by Mastodon servers.
When I pointed out these problem(s) on the Mastodon bug tracker they weren't acknowledged, it was argued that they're bad features, etc. Around some of them I could work around (like requiring webfinger resource resolvers), but around others I couldn't and I won't.
@mariusor @lornajane @lutindiscret @thevril @daniel
I am not too optimistic. I see a current fediverse that can make mainstream social media decentralized, by providing a bare level of interoperability, and a straightjacket of quirks to adopt to make it work. This fediverse either stays niche, or sees a hostile corporate takeover.
A "back to the original promise of ActivityPub" may be in order, which 1) lays a robust foundation to build on, and 2) is evolved under custodianship of the commons.
-
@mariusor @lornajane @lutindiscret @thevril @daniel
I am not too optimistic. I see a current fediverse that can make mainstream social media decentralized, by providing a bare level of interoperability, and a straightjacket of quirks to adopt to make it work. This fediverse either stays niche, or sees a hostile corporate takeover.
A "back to the original promise of ActivityPub" may be in order, which 1) lays a robust foundation to build on, and 2) is evolved under custodianship of the commons.
@mariusor @lornajane @lutindiscret @thevril @daniel
Social coding commons is a movement which envisions a 'peopleverse' as the future of the social web. By facilitating the emergence of self-sustainable commons organized enough to evolve that. And cocreate a commons based value economy that thrives in this social networking environment.
On the forum there's a section to discuss social protocols..
Proto social
Collective dedicated to a service-oriented social web.
Discuss Social Coding (discuss.coding.social)
And more related topics in other categories.
-
@mariusor @lornajane @lutindiscret @thevril @daniel
Social coding commons is a movement which envisions a 'peopleverse' as the future of the social web. By facilitating the emergence of self-sustainable commons organized enough to evolve that. And cocreate a commons based value economy that thrives in this social networking environment.
On the forum there's a section to discuss social protocols..
Proto social
Collective dedicated to a service-oriented social web.
Discuss Social Coding (discuss.coding.social)
And more related topics in other categories.
@smallcircles in my opinion there's no further need to "discuss" protocol as in we've already committed to ActivityPub. What we can do is to be the best implementation of that so we don't fuck it up for the other guys. Or we can just decide ActivityPub is not it, and clear the space for someone with more drive in this regard.
-
@smallcircles in my opinion there's no further need to "discuss" protocol as in we've already committed to ActivityPub. What we can do is to be the best implementation of that so we don't fuck it up for the other guys. Or we can just decide ActivityPub is not it, and clear the space for someone with more drive in this regard.
@mariusor the problem is what we do is
Design standard; implement; try making good UI; think UX; wonder how to get traction/adoption
What we should do is:
Wonder how to get adoption/traction; UX; UI implements (once, multiple implementation are pointless actually); iterate until critical mass, network effect is here and GAFAM site is dead; standard.
Decentralization is a mean, not a goal. We thing and act wrong
-
@mariusor the problem is what we do is
Design standard; implement; try making good UI; think UX; wonder how to get traction/adoption
What we should do is:
Wonder how to get adoption/traction; UX; UI implements (once, multiple implementation are pointless actually); iterate until critical mass, network effect is here and GAFAM site is dead; standard.
Decentralization is a mean, not a goal. We thing and act wrong
@lutindiscret if your main goal is
> get adoption/traction
you already failed as indie social media.
If all your efforts are based on this foundation everything you do will be tainted by the same incentives that mainstream social media has.
I'm half joking but, what we need to do is to build platforms that nobody wants to use, only then we will have achieved the zen of social media.
More pragmatically it means to build social media that is easy and cheap to run, easy to gather Dunbar's number range of people around it, and easy to link to other similar communities. Anything else will force you into the wrong direction.
