What are we trying to build?

We’re all here because we’re dissatisfied with Stack Exchange. But are we all dissatisfied about the same things?

Are we trying to be Stack Exchange, except not governed by the same people? Are we trying to be a repository of knowledge, and if so we agree sufficiently about what that means?

By the way, I invite people to read recent thoughts from SO founder Jeff Atwood: https://blog.codinghorror.com/what-does-stack-overflow-want-to-be-when-it-grows-up

18 Likes

If I’m not mistaken, the general feeling is that Stack Overflow / Stack Exchange has worked well, on the technical/platform level. The major problem with it, from most people’s POV (among members of this project, anyway), has been how the community has lost their voice over the platform over time.

So, we are aiming to rebuild what works about Stack Exchange - that is, most of it [1] [2] (IMO). The most significant change should be true community governance [3] [4] [5] [6] [7] [8].

Then comes the various ideas and proposals that people have been bringing up over a couple of channels. Some are very relevant improvements, but these are mainly of incremental nature (or should be, the way I personally see it).

TODO: paste relevant transcripts here; unless of course we manage to import the content from Discord to this Forum (I’ll be working with @bertieb on this).

7 Likes

I agree with Marc’s reply, and would like to add: while we ultimately want to build a rich, complete platform, I think we also need to identify (1) an MVP and (2) a couple communities that are annoyed enough at SE Inc. to be willing to be beta testers. I started a wiki page for functional requirements for the MVP (please enhance!): https://github.com/expert-engine/docs/wiki/Functional-Requirements:-MVP . I’m not trying to take charge, just writing down some stuff I think we have some amount of consensus on.

9 Likes

I’m not convinced that we should build a clone of Stack Exchange, for both fundamental and marketing reasons.

Fundamentally, as a repository of knowledge, Stack Exchange has some limitations which are due to the voting mechanism. If we’re going to build an alternative platform, I’d like to explore alternative mechanisms.

Voting with ownership begets reputation which begets competition.

If users own their posts and posts have votes, it’s natural that people will want to add these votes, whether the site does it for you or not. And then people will compete for the biggest number.

This creates a strong trend of optimizing for the ego of the participants rather than for the benefit of readers. Witness the anti-edit trend. Joel Spolsky’s vision for Stack Overflow was that once a question had multiple answers, someone would edit the best answer to add information from other answers. But in practice, try to do that and the author of the edited answer will complain that the edit denatures the answer while the author of the other answers complain that their answer has been stolen. Witness also the delete wars, where content that shouldn’t be there can’t be removed because someone would _gasp lose reputation!!!111

Voting does not always indicate correctness.

Wikipedia manages to be mostly reliable by requiring citations. The major limitation of Wikipedia is that this limits the scope to notable topics for which citations exist. The secondary limitation of Wikipedia is that on topics where it is difficult to evaluate the reliability of citations, the information can be unreliable (for example anything related to the history of certain regions of the world).

Stack Exchange lifts the notability requirement by introducing voting. But voting has limitations.

  • Well-written incorrect information gets more upvotes than badly written correct information.
  • Whoever posts first is more likely to get upvotes than whoever comes later.
  • Popular answers (what people want to hear) get more upvotes than unpopular, but more useful answers (what actually works or is actually good advice).

I don’t have an alternative mechanism to propose, but I would like to explore alternatives. For me that’s one of the main reasons to start a new platform.

With voting comes sockpuppetry.

Why is it a problem if someone has multiple accounts? There isn’t much benefit to having multiple accounts except when voting is involved. On SE we spend a significant amount of effort to bust sockpuppetteers, and that’s mostly because of voting rings.

If I want Stack Exchange, I know where to find it.

That’s the marketing reason. For the ~99% of SE users who don’t know what meta is, why would they bother going anywhere else? SE comes out on top on Google.

“Like SE except we respect Monica” won’t get much traction. “Like SE except better in some way that benefits YOU” would. What can we offer to Joe Shmoe?

21 Likes

For me the biggest question around this is -

if we could put what we want to achieve into 1 or 2 sentences - what would those sentences be?

3 Likes

Mine is the Wikipedia of the non-notable. A repository of knowledge which does not have to rely on citations for trustworthiness.

5 Likes

@gilles Not necessarily. On DIY, the best answers often rely on quoting from the National Electric Code or other reliable sources.

1 Like

Of course answers can rely on citations! That’s not specific to any particular site. But apart from the special case of Skeptics, the sites do not rely solely on citations. It’s ok to say “I tried this and it works”. It’s ok to say “in my experience …”. The difference with Wikipedia is that we don’t have to limit ourselves to knowledge that can be cited. This considerably broadens the body of knowledge that we can share.

I’ll amend my wording to make this clearer.

3 Likes

Mi Yodeya relies on citations – they’re not required, but they’re strongly encouraged. The additional value that the site brings is organizing information and providing good entry points – for any question of text or halacha or philosophy there are many relevant sources, and good answers summarize the key points from the most-relevant ones, point to others, and cite/link everything for further review. And all that’s wrapped up in the natural curiosity that led somebody to ask the question in the first place, whether it’s a practical matter of prayer, a whimsical question like whether T-Rex would be kosher for the time-traveler, or anything in between.

(I’ve now seen your update, but I’ll leave this here as an additional anecdatum. We also welcome practical-experience answers where those apply.)

1 Like

I think if we are really going to succeed in this project we really need to get intentional with EXACTLY what we want to do. And it needs to start even before requirements, figuring out in a couple of sentences our goal/mission then in a few bullet points the main objectives we have to achieve said goal.

Then our MVP requirements should branch from those bullet points. That way the work that we are all working on doesn’t lose sight of our goals.

At the same time as creating the bullet points of objectives we really need to define our core user(s)

this should give us the two main principles behind a product:

  • Who we want to be using our product(s)
  • Why they will want to use our product(s)

If we cant answer these two questions succinctly and if a large group of the core contributors can’t then we won’t succeed.

I’m here because I love the idea of contributing to creating a community of people, helping people on their journey, learning from others but also not dealing with all the politics and bs that SO has become known for.

I dont like how ‘closed’ SO has become for new people, how unclear/difficult it is to ask questions. I DONT like how PC insano it has become but im not here for that reason. Im here because I want to build a better community that thrives around asking questions. Not just to make SO2.0. When we ask questions in an environment that promotes creativity, openness and discussion I think it breeds ingenuity, community and human-ness. Its what we do best. Any time humanity has pushed itself forward its been through questions.

For me this is what Codidact could become. A place for questions. A place for answers. A place for community. Often I find the discussions had around a question to be the most valuable in SE - even though they are an afterthought IMO

It was actually @cellio story that drew me here in the first place. To me that is what SE has really missed the mark on this whole thing - diversity of opinion and people - coming together.

7 Likes

It might be a question that has to be asked after the question “What are we trying to build?” has been answered. But I think that the questions are interconnected:

How should the site evolve?

I’ve seen that people wanted the site to be “complete”, but am not sure what that means. StackExchange basically started with StackOverflow - “A site about computer stuff”. Over time, the community grew so large and the content became so broad that now there are dozens of more specialized sites (Software Engineering, Code Review, Graphics, Machine Learning, … and of course, all the non-computer stuff).

This should not be considered to be any sort of “proposal”, but only as a question: Is the goal to start with 170 specific, small sites that resemble the SE network sites, or are there plans to initially have sites with broader topics, and then have a process of fragmenting and specializing these when the need arises?

Although I’m not an expert in “community building”, I think that having an answer here might already be crucial to “bootstrapping” a new community: When the site starts with 170 sub-sites, it could be difficult to achieve the critical mass in any of them. To some extent, this problem can be alleviated by “seeding” the sites with the most prominent Q/As from their SE counterparts, as discussed in other threads. But this will not replace a community of people who are willing to actively contribute…

(If somebody thinks that this should be discussed separately, this post could be moved to another, more appropriate thread - similar to migrating a question from a “broader” site to a more specialized one :slight_smile: )

1 Like

Personally I think the focus at first should be around what do we want the community to look like. Maybe we create functionality to spin off ‘rooms’ where we write a framework that allows unlimited question ‘subjects’ and doesnt follow the same structure as SE - or maybe we dont.

I just think we need to focus on one thing at a time - and focus on the most basic, fundamental (but most important thing) first.

You mentioned it at the top. How should the site evolve. Is more strategy and implementation detail and comes later. Lets not worry about running before we even know how to walk or where we are walking to.

I don’t think we want to start with clones of 170 SE communities. I do think we want to start with some existing SE communities, ones that actively want to try an alternative to SE with an eye toward moving. Over time we should expect both more migrations of existing communities and proposals for new communities, which we’ll have to figure out how to handle at some point.

I’d like to see us start out with just a handful of communities from SE, ones that are willing to be beta testers and help us shape our platform. Those communities are likely looking at alternatives too, like BioStars and question2answer (what Physics Overflow uses), so we need to have a better value proposition.

5 Likes

This strikes a chord with me. My primary SE experience is in a relatively small community - Mi Yodeya. There, community really is an important aspect of what we’ve built. We’ve built friendships and projects together in the course of learning together. As you said, this aspect is an afterthought on SE, partly because SE is primarily optimized for SO, which is too massive to exist as a single community.

The site is optimized to invite questions and answers, specifically, with back-and-forth discussion half-heartedly enabled in comments, actively discouraged by policy, and pushed to a separate but linked chat platform. Now, I think that the clear emphasis on the primary content being on-topic and in conformance with the defined Q&A format is the core idea that makes SE distinct and valuable, and I’m not sure any platform could be valuable to me in the same way without something like that core feature. But on the other hand, I feel that the community we’ve built using the Q&A-optimized platform that SE offers could be even stronger on a platform that also optimizes for community-building, per se.

So, I would suggest that including socializing and free discussion as first-class features somehow could be an important differentiator from SE, particularly with respect to smaller communities around relatively niche subjects. This could include fully-integrated chatrooms, a chatroom automatically created and linked to each primary post, and/or a section of the site set up for free-form discussion and community-building, outside the strict Q&A format.

7 Likes

Wow - that reply made my heart sing!!!

This is exactly the kind of conversation we need. I completely resonate with this. Even in this tiny thread ive got 20 ideas racing for how we can solve these challenges and im sure im not the only one!

@Marco13 Can we do something to try and facilitate some more directed discussion around these points that might lead directly into some vision/purpose statements?

I still think its pretty foggy on what we are trying to do at Codidacts core.

But this is how great thing are made. A feeling that something is missing or can be made better.

By the community, for the community.

5 Likes

@cellio maybe you could compile a list of the most active existing communities that we would possibly want to seed/setup?

Almost like a sales list of people we could approach when we get get a bit further down the line with this?

We’ve started to discuss seeding in MVP: Data import. That discussion is primarily about the how. The who warrants a separate discussion.

1 Like

Thanks @mattjbrent, that’s an interesting perspective.

I do not agree with it. I don’t think the world needs another discussion platform. If you want discussion, what’s wrong with Reddit? And human-ness is more of a thing that happens (and must be let happen) than a goal: if you want a hug, get off your computer.

What drove me to Stack Exchange (initially Stack Overflow) in the first place was in part the lack of discussion. (There’s chat but it stays out of the way.) It was all about building a place where you can find answers right below the question, and not on page 27 of a thread in a message that’s incomprehensible out of context. Being part of a community came later — it started when I joined a beta site where I was one of the top users and got involved in curation.

But I don’t think our points of view are incompatible! It’s perfectly fine if discussion exists. As long as the answers can be found below the question, as long as I can ignore the discussions, I’m happy. A vast majority of Wikipedia visitors don’t look at talk pages. A vast majority of Stack Exchange visitors don’t look at chat and don’t pay much attention to comments.

Who we want to be using our product(s)

Everyone who can read. Eventually. (Catering to illiterate people here is beyond my imagination horizon.)

Why they will want to use our product(s)

To find knowledge that they can’t find elsewhere. And more for a small minority, but for the vast majority, it’s the knowledge.

11 Likes

Discussion: enabled but non-invasive. Comments should be collapsible, arguably threaded, or on a separate “panel” in the page (I don’t know the right terminology, but I mean something like those in-page “tabs” that change the view without forcing a full reload). By default they’re out of the way but you can easily see that they exist and can easily view them.

11 Likes

We’ve been discussing comments at MVP Proposal: Requests for Clarifications and Feedback

3 Likes