We need to look through all the ActivityPub codebases #1
Labels
No Label
4opens
bug
Discussion
duplicate
enhancement
geekproblem
help wanted
invalid
question
wontfix
No Milestone
No Assignees
9 Participants
Notifications
Due Date
No due date set.
Blocks
#5 Pleroma is a runner for a indymedia site codebase.
indymedia/indymedia-reboot
Reference: indymedia/indymedia-reboot#1
Loading…
Reference in New Issue
There is no content yet.
Delete Branch "%!s(<nil>)"
Deleting a branch is permanent. Although the deleted branch may exist for a short time before cleaning up, in most cases it CANNOT be undone. Continue?
To find the best for modifying to build a news service.
This looks likey https://pleroma.social/
https://gitlab.com/explore/projects?tag=activitypub
SpellCheck:
#ActivityPub
Source: https://www.w3.org/TR/activitypub/
We need to look through all the activertypub installsto We need to look through all the ActivityPub codebasesupdated :)
Ones we have a text outline of the code we can use @hamishcampbell can send this out to geek contacts as a first step.
Pleroma is probably good enough, and can handle a lot of traffic. It's codebase is Elixir, so I'm not very familiar with that. Also it's user interface is javascript heavy, but then so too is Mastodon's.
If you're going to use Pleroma then I'd recommend containerizing it and treating it as a blob, otherwise maintaining it over time becomes a problem. The configuration of Pleroma has evolved in a very undisciplined way which can make migrations from one version to the next tricky.
I understand its split into two "The project consists of several components: Pleroma is the server implementation, and comes bundled with PleromaFE, the default frontend. Other useful utilities are also provided, such as an ActivityPub relay. "
So we would need a new front end the PleromeFE side? the are a number of frontends which would be the best to use, pros/cons to do the news activertypub site?
What do you think the work load on that would be.
ActivityPub relays should be treated with caution, because it is possible to end up relaying content which is illegal or counterproductive. It's a very similar problem to email open relays. In general only relay posts if you absolutely have to. afaik Pleroma doesn't have any sort of whitelisting when it comes to deciding what to relay.
again I'll mention https://gohugo.io/ :)
or https://writefreely.org/
the latter already can post using activitypub.
as long as you don't need fancy javascript (basically for interactivity) then both options more than suitable in my eyes.
but I openly loathe javascript ... so am slightly biased
Yes I'd also avoid javascript heavy systems if you don't need that. It also makes things more friendly for screen readers. If you're just generating blog-like news items then writefreely might be better.
Been blogging on WriteFreely every day this month, for some things I think its great but it has a lot of federation issues. I think we would be better off starting with Pleroma and building a new front end. It has most of those issues figured out more then any other part of the fediverse. It might be useful to check out MoodleNet's fork of pleroma, they forked over a year ago to build an educational platform ontop, maybe it is already cleaned up enough that it would be a better starting point. Mayel (who was in charge of that process) would probably be a great person to answer some questions about using Pleroma as a base either way.
(also thank you @hamishcampbell for inviting me here, I have been close to indymedia my whole life, and have been thinking about how a project like this might happen for years. I was a close friend of Brad Will if that means anything to anyone here)
It seems activityPub is having some flaws.. which don’t provide to the final user the privacy and control.. that ZOT could give… but this is a long discussion happening in the self-hosting fediverse community..
Here’s a cool introduction to Zot.. https://zotlabs.org/help/about/about#Technical_Introduction
Anyway hubzilla might be the software to change to be able to do this… it supports loads of stuff..
ofcourse, this concern comes after:
https://likeable.space/notice/9qwGoknsLbUGV0M5lQ
but i think this happens mostly because mastodon is draw in such a way that everything is public has in twitter..
Hubzilla is nice for geeks but not so nice for outreach to normal people. As we are doing a project to balance the power of geeks and empower normal people #activityPub is better for a #4opens project. For the security/offline p2p minded we are planing to bridge to scuttlebutt useing the "pub" have one of the core devs putting a proposal up for this here over the next weeks.
We plan to reboot the "open" early indymedia project with activityPub and also reboot the latter closed indymedia project with scuttlebutt. As they both have value.
https://unite.openworlds.info/Open-Media-Network/Open-Media-Network/wiki/4opens
i move this talk to here: #2
Hubzilla and Zap are ok, but I agree with the above assessment. The security of actually existing Zot implementations are not all that much better than ActivityPub ones. Also whatever you may think of W3C at least ActivityPub is a published open standard (subject to all the usual caveats about the half-baked condition of the current specification).
The Zot apps have needed a designer to improve the UX for years, so I'm not holding my breath for that to change any time soon.
ajeremias referenced this issue2020-01-17 22:07:44 +00:00
OK focues - the are other activertypub projects lots of them can we look at a few more. what we use might come down to the programers we find and the languages they like to program rather than the “best” codebase. lets keep it open when we have good info, and a code outline, lets move it to a wiki page and start outreaching for programers :)
Yes. Ease of hackability combined with good enough documentation often beats everything else. Systems written in relatively uncommon languages, like Elixir or Racket, may be harder to customize even if they have technical advantages.
For indymedia you should perhaps look at: https://github.com/dessalines/lemmy
It's AP reddit clone.
As for Zot apps. I do personally think they are much more advanced then AP ever will be, and I find it pity that they are dismissed based on UI. UI is exactly what Zot misses but if everyone will avoid it because of that it will be hard to make it happen. Together with the project I'm involved in we decided to go against the current (on mastodon boom) and went into Zot. This year we want to focus on help with Themes and general UI too so having more people and more exposure to Zot would probably be beneficial. Zot supports standards like carddav, caldav, webdav so you sync data from it with your device. It does groups and events, chats, message editing, and even e2ee crypto. Even DM's are not stored plain text in the database.
Edit
As for UI since hubzilla supports Themes you can make it look as you want: eg. https://hub.disroot.org/channel/disroot
Also due to modular nature of Zot, you can decide to enable only narrow feature set to fit your needs. Hubzilla (for example) is a platform you build your project on, so if you do not need certain features (wikis, websites or notes for example) you don't enable them.
@muppeth intresting to look at lemmy but "This is a very early beta version, and a lot of features are currently broken or in active development, such as federation." this takes it out of running for now for the easey news project we are attempting.
I like hubzilla and would be happy bridging formats, data is just data, but format wars are #geekproblem that we are clearly rejecting #OMN
Activertypub works fine for what we are building and is a full gold #4opens project. lets focus.
OK the are a load more activertypub implementations lets give them all a fair view.
We have this one from https://epicyon.net/ which is basic and would work for tiny installs. its is also part of this project that am looking at https://freedombone.net/ I have a test install up and running over tor will post the link when it settles down. Not a mainstream candedate due to scaling issues but the developer is frendly and he might be intrested in doing a news based version.
Some background https://the-federation.info/protocol/activitypub
The are still other implementations we should look at https://the-federation.info/#projects
What about Wordpress with the ActivityPub plugin? I havn't tried it, but it might be better suited for a news site. Also there are many people who are familiar with installing Wordpress.
Yep its over kill as we have the whole legacey of WP in there for the gerneral project. But its a good option for some use cases, we should look at it. do you have any WP skills, could test it on my WP blog http://hamishcampbell.com if you up for upgradeing the PHP and makeing the SSL work ;)
Would be nice to get it fedreatated.
its awesome. and too easy... you install it on a wordpress this plugin: https://wordpress.org/plugins/activitypub/ and then you are a node in the network..
@username@yourwebapge.org
https://social.coletivos.org/web/statuses/103336322566455260
Not #activitypub yet and not the same project, but intresting https://dev.lemmy.ml/
More codebaseds to look at https://github.com/autogestion/pubgate
A good list of codebases to look through https://git.feneas.org/feneas/fediverse/-/wikis/watchlist-for-activitypub-apps
How deep do we need to go to build this https://github.com/go-fed/activity
This one https://federated.id/
CPub is a general ActivityPub server built upon Semantic Web ideas. It could "understand" any text information in good tradition of RDF-based Indy-engines like Samizdat.
How do you think this would fit -how easey to build an IMC?
Open-Media-Network/Open-Media-Network#26
CPub might be a robust infrastructure to store linked data. Usage of Robust Immutable Storage should allow to store all posts, comments, etc. in IPFS let say. So communities could have all the generated content even after disappearing of particular sites. That's about the backend part.
Frontend should be nice looking mobile first fancy pancy. And as CPub will be compatible with Mastodon API like Pleroma does, any PleromaFE could be used as a basis for IMC FE.
But. CPub is still in progress.
The IMC project is an affinity group - so we are planing to work through consensus and diversity of strategy to move the project past where it was ripped apart by internal stresses after 10 years of running as a successfully worldwide radical grassroots media project.
To do this we have a #nothingnew policy beyond moving the project into modern standards #activertypub we need to reboot the project with work flows intact. Thus, we are working to this #UX Open-Media-Network/Open-Media-Network#26
and the original workflows and process.
Of course this can and will be updated as needed, but we have a "chicken and egg" issue that we need a working affinity group/s to reach consensus on where we go. The #indymediaback project is a way of bootstrapping this code/process in #4opens way.
To keep “diversity of strategy” in place we are using the #OMN framework.
Thanks for the info @hamishcampbell. I'm not familiar with the #nothingnew policy. Could you please point me out what does this policy mean?
#nothingnews is an attempt to over come the churning of alt tech and the #dotcons as been doing this stuff for more than 20 years and have seen hundreds of me to/NGO projects grow and fail, rise repeat for ever into a flow of pointlessness. its a BLOCK that we have lived with for way to long.
In resent years the first project to get me motivated was Mastodon and #activertypub as both were based on #nothingew they were evaluations of existing projects and standards - twitter/tweetdeck and atom/pubsubhub.
The #fevivers has fallowed this in that it is direct copy's of existing useful #dotcons - peertube is youtube, pixalfeed is Instagram etc. This on balance has been a positive, though not with out its problems. Has clearly overcome the last 10 years of pointless in #geekproblem alt-tech for now.
#indymediaback is an example, like mastodon etc. of rebooting an historical usefulness social Alt-tech project from before it was ripped apart and then building it using modern standards from the experience of moving past the last 10 years failer
Now academic alt-tech is a whole another subject :)
@rustra What is your part/agender is all this :)
Thanks for this clear explanation. I share this vision in context of the Indymedia project. When I've joined the CPub I thought that it could fit the needs of new modern general server for Indy as well.
Think the thing that courses the tech churning is a lack of respect for joined up social thinking in the #geekproblem world view. This tiny minority think they can invent from their limited social experience and ignore wider social context and history of radical movements. This is why the work that come from this always fails humaine agenders and thus feeds the dominant pointlessness. Its not just the sprit of the age, which it is, its the sprit of a tiny minority of people that is now grown into a world wide problem from/for the #dotcons and the alt-tech crew.
The are simple ways out of this mess and the #OMN and #indymediaback are examples of this path that we need to take.
Do you think you could build a #OMN based #indymedia on #CPub ?
I think intentions behind CPub suits well for OMN: linked (open) data, storage with content-addressing. And no need to wait it will be ready as Pleroma backend can be used already for OMN.
I'm contributing to CPub for it. And you've mentioned the IMC affinity group with consensus decision making. If you mean this not sure that I could actively participate in it.
This one look a bit corp but would likly work as a good backend for any of our projects https://github.com/xwiki/xwiki-platform
This one might be intresting https://forestwiki.com/wiki for #indymedia and or #OGB its under active dev and they might have some AP and RSS code runing in a bodge state.