I’d like to create a topic under the FEP category but cannot. The create topic button is dimmed when I visit there.
Rather than give me access though, I’d recommend opening it up so anyone above a certain level can create topics there.
Trust level 1 and above should be able to create topics now. @how can you review and change if you like to specify differently? I did not remove the activitypub.hosts group privilege.
Is there anything wrong with this process?
@aschrijver Thanks, that worked.
@how Great question!
Given how the initial build-out happened but then fizzled, I think there are several problems to address. So for me the answer is yes because it doesn’t specify a place to talk about the process, how it’s going, what the issues are.
The FEP category seems to me the best place, one drawback being some spam or a lot of OT stuff which we can address if it comes. So basically I wanted to first solve the problem of where to talk about FEPs and now that this category is more permissive, I think that’s done.
I’d invite anyone who sees a problem with the FEP process to create a topic describing the problem they see. This effort has a ton of potential to achieve its laudable goals but it’s going to take some diligence to reduce friction and build some momentum.
Yes, I’d also be in favor of having the #activitypub:fep category be more accessible, so anyone can start a “I wanna create a FEP on [this and that], wdyt?”. The formal parts are then in the FEP repo, with links to discussion(s) / post(s) here, if needed.
I disagree with this approach.
The #activitypub:fep category was a controlled space where people would take time, that required focus to read and think about things to solve from casual conversations evolving elsewhere in the forum. What you’re doing here is sabotage.
I’m sorry you see it this way. I’m just trying to restart a process that seemed to have hung. You can disagree with the approach but I don’t appreciate that view of my intention.
Which category would you propose to discuss a pre spec proposal?
You do not appreciate my view of your intention, as I do not appreciate your interference into a process that is documented and that you choose to ignore. You’re very welcome to help “restarting a process that seemed to have hung”, but you should do it following the rules, not taking power. These rules were made to accompany the FEP process and keeping the relevant discussion space slow paced. There are trust levels for a reason – you need to understand how things work. Without @pukkamustard’s and @cjs’ involvement, I don’t see any resolution since they are the ones who are in charge with the FEP process.
Anywhere a discussion makes sense, it should happen. But if a discussion does not happen, probably there’s no need for specification. See for example how the discussion about group has been going on. Only when there’s a need that is shared across implementations can we start thinking about a spec. If the need is not clarified first, how can you generate a spec? This is how XEP, IETF, and other standardization processes work: you discuss first, and from the discussion comes the spec proposal, which is discussed, criticized, refined, etc.
I guess what @weex took for a hung process is a combination of too many specs and not enough time to discuss them. Maybe the FEP process is too short.
Again, the #about topic in the #activitypub:fep category states that the spec topic comes after the FEP in the git repository. The #activitypub category or any of the #software categories are suitable for starting discussions that may lead to specifications. Please just don’t clutter the FEP category with pre-drafts.
See how the Move FEP under a new Standards category discussion moves soooo slowly. It requires time and attention to make things move in a group. That’s a fact of life. We need time to process… I hope that topic will resolve soon though, because indeed Processing the FEP backlog is useful and it would be great to have FEPs resumed.
1 Like
I feel the issue is that most discussion takes place in other settings and on different channels, mostly ad-hoc between individuals who looked into each other’s codebases to figure out how to interoperate on something. The casual discussions on SocialHub are progressing very slowly with long time intervals between posts, and they often peter out without any real decision made. This understandably leads to devs just implementing and focus on standardization later (which may or may not happen, and may be just as ad-hoc).
And if someone decides to take the effort and write a FEP, with the small amount of federated devs we have, they are likely to be the only one to which the enhancement is important at that time. Gathering proper feedback becomes an involved process in time and effort. Or takes place insufficiently or not at all. If something is truly new, then the early adopter automatically becomes the decider.
All this is not necessarily a problem. We can decide that the way this works is okay and fine. But it is imho good to brainstorm about process improvement to streamline all this. @weex is doing this in the open in #community category, and anyone including @pukkamustard and @cjs is free to respond. Agree that we should use forum structure as intended and only make changes with informed consent by those responsible and involved.
Note that I too gave FEP feedback in March, got some feedback, responded and since then the topic has languished (until recently with renewed intention to pick up the FEP process). The notion that things weren’t running smoothly yet, was also shared by me. But is also still young and remains largely untested in practice (for AP community), so that may be as expected.
The larger problem (maybe… it can also be just my impression) is that this community now constitutes mostly of ‘just a forum’. I don’t feel there exists a real Community of Action that is at work here to evolve the fediverse. Just people publishing interesting posts and having discussions.
1 Like
This wasn’t a conscious choice. I was going off the information I found in doing a reasonable amount of research. The thread you linked is not in the FEP FEP and while pinned, I didn’t realize it had rules in it which where essential to the FEP process. Doing a broader search I found the entirety of the FEP process is actually spread among four or five places which becomes a problem if we assume an adversarial environment. It would be a big burden on any editor or admin if they were tasked with enforcing rules written in so many places, especially when some of those places can be edited without notice.
I’m not sure what power we’re talking about. I’m for the principle of least-privilege and would prefer the FEPs were moving right along without my involvement. They aren’t and I’m stepping up to try and get the ball rolling. Power it seems is the root of many problems. Once things are working, I’d love nothing more than to hand off the torch and get on with other goals.
This seems a premature optimization since I can’t see when FEP discussion has ever gotten faster than what a single person could keep up with. Successful processes scale and meet challenges without affecting scaling.
Both pukkamustard and cjs have been involved to some extent in this restart and have shown support even for moving the repo but their time and energy to devote to FEP is limited. You’ve mentioned them, but they have not chimed in. Do you have a timeout on that request? No functional process stays blocked indefinitely.
I’d encourage you to stop reading into my actions. I’m trying to help. I’m not perfect and I don’t know everything that’s ever happened on this forum, but I see value in more than one FEP being finalized so trying to get there.
1 Like