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?

Excellent question. Reflection on this is important for the other excellent question @gilles raised about why other extant QA software won’t do; and when formulating an MVP.

Objections to SE/SO/etc could fall into several buckets, such as (with examples, not necessarily my views):

  • governance - “I don’t like those in charge, and what they’re doing”
  • social - “I don’t like prevailing social norms w.r.t. to new users / poor questions / etc”
  • procedural - “I don’t like how one moderator can decide on things / I don’t like the process for complaints being handled”
  • mechanical - “I think questions should get multiple chances at the reopen queue / I think anonymous votes should have some weight or give/take rep”
  • technical - “I think the API should be more open to make automated tasks (eg spam hunting) easier / I think the stack software should be open source so we can run our own $TOPIC community”

Many of these are interlinked or overlapping; and some we will want to modify more than others.

6 Likes

To me, what keeps me coming back to SE is its signal to noise ratio. I love learning things, and SE is a wonderful platform for all sorts of people to share information. I dislike other places cough reddit cough because of how unfocused, opinionated, and flat out useless most of their content is. I’m not in SE for the community, despite how pleasant it is. I’m there because people talk about things that matter, and get to the point.

People need to feel welcome to make contributions. No one wants to get out of their comfort zone if it will only hurt them, so being nice is what allows good content to exist in the first place.

SE has that. Or at least it’s community has it. Lately the company doesnt seem to care much.

Of the objections @bertieb listed, I identify with all of them save procedural, but primarily with governance. Additionally, I’m not sure where animated ads and illegal relicensing would fall under, but thats on my list too.

7 Likes

I identify with all except ‘technical’ ^^

That would fall under governance. Btw, on Discord, we seem to have reached an agreement in that: (i) We won’t run ads at all[1] [2] [3] [4] [5], and (ii) unjust/illegal relicensing is an extremely easy problem to avoid [6][7].

2 Likes

Discussion and the like

I think it should be possible to have meta content for every item (but not required), and hidden by default so it does not clutter the main content.

Meta content of different kinds:

  • Read-only (votes, views, info from profile page, close status, revision history, etc.),

  • Threaded discussion of the content (similar to comments)

  • Seeking consensus on changing/building content (and/or formalised voting on decisions about it)

  • Off topic

  • Say hello to user (subject to blocking)

  • “See also” (like the hot network list)

There could be views that collect the same kind of meta content, say all current/open discussions for tag wikis.


The site

The slogan for the site could be: The place for professionals and learners alike

The site could have several corners that are qualitatively very different, not based on the same model (as Stack Exchange mostly is). I am thinking mostly of Stack Overflow here, but it could probably also apply to other facts-based sites (mathematics, English, physics, etc.)

I think the differentiation should be that we will (also) have a corner that is designed for learning, especially for beginners. One of the artefacts from that corner should be a set of FAQs (e.g. in Usenet style) that both serves that very purpose and also serves as a very quick and efficient introduction to a beginner (or new to it) in a field.

There is a huge untapped market here and the expectation (whether justified or not) of a place for beginners creates a lot of tension on Stack Overflow and also creates the image of Stack Overflow as a “toxic” place (unfair or not). A lot of contributors have a wish of teaching others as their motivation for contribution, including Jon Skeet.

The artefact from the normal Q&A is curated questions/answers, maintained/kept up to date and of publication quality (on the level of Wikipedia or better)

The site should not be free for all: The barrier for participation should be the willingness to put in an effort (those with the minimum-effort attitude are not welcome).

Possible corners (only the first two for MVP):

  • Knowledge repository

    Q&A as we know it

  • Human search engine / learning

    • Especially for beginners seeking advice, pointers to resources,
      drafting a real question that can be put in the Q&A corner, homework help, etc.

    • Content hidden from search engines

    • A safe place, without voting

    • An artefact could be FAQs, in the style of Usenet’s (before AOL destroyed it)

  • Chat

    For very quick interaction (meta type stuff is covered elsewhere)

  • Blog-type content

    Longer

  • Shooting the breeze

  • Meta place

    Governance, feeds from the different sources, voting on big issues, moderator elections

  • Professional services market (yes, controversial)

    Support for micropayments.

  • User pages

    Including support for structured content (for instance, Trello-like)

(Meta: OK, I had to stop myself before it became too long. Take it as a set of ideas, rather than a well-developed whole.)

5 Likes

This intrigues me.

I’ve been mulling over something like this for quite a while - for those who know me best - and it’s kinda tough to really get a bead on what the “next” site would be.

So what will follow from me is an attempt at what I think I want to see in a future Q&A-oriented technical site.

Mission

We wish to become a reliable source of answers to complex questions across subject matters.

Justifications

  • Reliability implies curation. Curation is vital. Being able to clean up the faff and get rid of the noise is just as important as having experts actually answering in the first place.
  • Complex questions are important to codify early on, since the “easy” stuff is already easy enough to find. A new Q&A community’s value-add doesn’t come in with answering what’s already been solved; it’s all about tackling the harder stuff.
  • Cross-subject implies that the Q&A platform will eventually not be limited to technical sites, but to all manner of sites. This is an attempt to bridge the gap for those sites looking for community.

In my head, everything else becomes part of “the how”. I’m sure there are some other MVP posts I can read up on and comment to the effect of to cover that aspect.

7 Likes

The question of “what are we trying to build” is tightly coupled with “who are we trying to build it for.” And with that, what are the people that we are trying to build it for wanting to use it for?

Some reoccurring comments elseweb about SO have been have been that “the most interesting questions are closed” and that the reason that they ask questions is to get the opinions of experts that are in that community (many of these see their questions and the answers that they get as part of a chat or discussion type environment where they are lost to time rather than part of a larger repository of knowledge).

Many of the questions are blog post seeds (this gets into long form blog ownership questions - which can go down another rabbit hole of design and requirements).

As I see it, there are three different main audiences. There is the curated material, the opinion browsing, and the ephemeral “need help with this.” Part of the challenge is that SO has had is that all of these are lumped into one Q&A engine and the friction of different people, as part of different audiences, work to maintain the site in different and sometimes contradictory ways.

If working at recreating the SO model is the approach, it may be necessary to concentrate on while and, well, frankly discourage the other two somehow. SO tried to do this by making the Q&A engine itself hostile to those other uses, and that may be part of the interpretation that SO the community is hostile to them.

Another interesting site out there is everything2.com and its still active sister sites, perlmonks.com. Both of these exhibit very long lasting communities (E2 has its origins in '98 and perlmonks in '99). While perlmonks has a very narrow focus (perl) it does have its Q&A approach as seen in Categorized Questions and Answers.

There are many discussion sites. There are quite a few sites that do Q&A to one degree or another. There are different models for new users (MetaFilter has a one time $5 charge, https://lobste.rs does a “you have to get an invite from someone”). There are scores of nearly forgotten discourse sites that attempted to do Q&A. I recall a period of time when people kept trying to make the old Not Programming Related site (one such launch attempt can be seen at HN - Not Constructive, a place for the discussions not allowed on Stack Overflow)

I would like to see something different - a different way to capture and retain knowledge than what SO (or Reddit or Quora) present.

3 Likes

Loving the discussion and opinions im reading on here.

My personal summary of what i’ve read so far.

First and foremost, one of the most useful parts of a Q&A forum is getting to an answer (or useful related information) quickly.

There is a recurring theme of ‘areas’. Experienced professionals have a different requirement than someone who just graduated. Some people don’t want overt dynamic interaction between users, some do Theres a desire from a lot of existing SE users that enjoy the education aspect of SE and teaching others.

This idea that there could be different sections to an answer. Where a view or UI could be toggleable (customizable view depending on what you want to see) Where Q&A really becomes its own sub-topic thread that could host a chat, post comments Q&A etc. If we store meta info on various parts of this it would be quite feasible to make this view customization quite powerful. IE only show me the most upvoted comments. Only show answers that are accepted. This could be stored on a users profile and use a query-like syntax as consumable REST endpoints.

A way to organize information that makes it easy to digest a topic, sub-topic or sub-sub-topic. A possible way to flag content with certain descriptions that could allow users to toggle various content.

Opinion-based questions can still be a legitimate question - although does require careful moderation. Using content flags would allow a user to only view non-opinionated questions.

Allow a consensus of upvotes to count towards an ‘answer’ just because the person asking the question hasnt accepted it doesnt mean its not the accepted answer.

I still believe chat or comments can be an integral part of the question/answer without being frowned upon. I still find lots of pertinent information in comments that are sometimes more valuable to me than the answer itself. (often browser support or some related information/link)

More open moderation. Moderation will be one of the most important aspect of Codidact. Strong moderation tools will help keep the site organized, help keep experienced users coming back and help new users make a valid impact and contribution. moderation should enable all users to contribute rather than placing strong restrictions and reasonably large burdens on new users, moderation enables a new user to interact and get involved without compromising quality of the content.

We need to think about how we bring value to new users. How we improve interaction within the community with new users without compromising the quality of questions and answers. Often newer people will need to ‘talk’ more as opposed to an experienced user.

6 Likes

Seems like a perfect summary to me. I want to add a thought that maybe there is a way to have discussions about posts without cluttering a comment section - maybe make the comments expandable so there is more freedom to have a discussion in the comments. While I agree that comments are just to add thoughts to a post, I think interesting discussions can take place in them that shouldn’t be censored.

2 Likes

My take: Ask questions, get answers - you have a voice here.

It requires some context - but chances are, if you’re here, you understand it.

As we gain more visibility, there have been quite a few suggestions to deviate from the SE model in significant ways, which are definitely not without merit. But the truth is, that is not the objective I was aiming for when joining this particular project, personally - and being with the team from day one, I can say it wasn’t/isn’t the objective of the majority, if not all of the core contributors.

In other words: Building something that differs significantly from the Q&A model as we came to know from SE is not what I personally believe in, and is not what I signed up for.

Fellow @MakotoTheKnight said in a post from another thread:

And in fact, I disagree. The reason none of the previous software implementations succeeded in replacing Stack Exchange is definitely not that they tried to mimic it. On the contrary, it’s probably because they didn’t do it well enough (and not just on a technical level).

7 Likes

This resonates with me. My goal in coming here was to help stand up something very much like SE’s MVP, quickly, and then iterate on improving it. So yes, we would end up changing some things that we want to change, after careful discussion and study, but we’d start with something that basically works, get that up and running and used by some communities that want to move, and then do further development informed by (and in collaboration with) those communities.

I want Codidact to be awesome, including in ways that differentiate it from SE. We have the chance to fix things. But we have to agree on what those things (and their fixes) are, and then we have to do it, and if it takes a year or more to get there, we run the risk of being irrelevant when we do. Right now each dissatisfied community will do its own thing; what can we offer to help us all work together instead?

7 Likes

But if we build it, will they come? I don’t think “like Stack Exchange but with free governance” will win enough people over on its own. We need to have some concrete, direct value to offer as well.

3 Likes

I think “doesn’t treat volunteers like garbage” is already a fairly strong selling point, at least on some communities, but that might not be enough. On the other hand, the mod-resignation posts and ensuing meta discussions point to a few communities that might be disgruntled enough to move to anything that’s easy.

11 Likes

We need to agree on an MVP before thinking about how to sell it; its already hard enough to think about advertising a service that doesn’t exist yet. If the service isn’t even described then its practically impossible since nothing is set in stone. There are ways to do everything, I’ve already proposed a way to advertise Codidact on SE to try to bring the community over, but it depends on some technical aspects of the implementation.

If we import new questions as they are made on SE, and we answer the imported question, we can have a tickbox for the answerer to automatically “crosspost” the solution to SE with attribution to their answer on Codidact. Quote+link source answers are accepted on many stacks, and it can interest people if done correctly. Plus, newcomers get the added advantage that they don’t feel like they are leaving their community behind if they only answer questions on Codidact.

Whats the closest thing we have to an MVP right now? Is it the github wiki, or is there something else being worked on?

3 Likes

Ooh, I love the “crosspost this Codidact answer to SE with attribution” idea! Free publicity and it doesn’t harm SE!

7 Likes

Actually, no. The decision on what goes into the MVP vs what will come later is based on what we want to be usable from day 1, and this depends on what we want to sell from day 1. (Using “sell” in a non-monetary sense of course.)

Building a new platform is harder in the sense of more work than selling a new platform, but selling a new platform is harder in the sense of less chance of success than building a new platform. The world is full of failed start-ups that had a solid technical idea but either didn’t find a market or didn’t manage to penetrate that market.

Be careful not to overdo it, or Codidact references risk being banned as spam.

We’re working it out topic by topic on the mvp tag in this forum.

6 Likes

That’s a strong selling point to the 0.02% of users who post 1% of the content and do 10% of the moderation. To supplant Stack Exchange, we also need to have something that will interest the “silent majority”.

And be careful not overdoing the ”for people who are disgruntled by Stack Exchange“ aspect. That encompasses all the people who are frustrated because their crappy questions have been rightfully closed. (More of a problem on SO than on typical small communities.)

4 Likes

@gilles: Maybe the Pareto principle (minding the rough percentages) can do a large part of the job? And then, good SEO can do the rest?

A large proportion of SE traffic is from search engines. We’ll have all of their content. As long as it’s better moderated and we present the content better than they do, it should be just a matter of time until our platform becomes prevalent.

Beating SE on search rankings is going to be almost impossible. I don’t see it happening anywhere within the next 3 years. Search engines like google and duck duck go have SE-specific excerpt code, even. We’ll need to do a lot more than just have good HTML and wait.

6 Likes

Initially, we need to get enough of the hard-core users, who are more likely to be disgruntled, and then help them bring the next “tier” along. On one of my sites, most of the active core has stopped contributing and some have deleted their accounts; if we got a bunch of them to come over, then others would be more likely to follow because that’s where the content went. On other sites this won’t work as well; nobody’s going to be able to compete with SO that way.

Which is why I’ve been thinking primarily about the smaller, connected communities to start, the ones where users know each other and have more of a community feel and not just a “content library” feel, as important as that is too. We can’t survive on just that, but it feels like we can get started there. But I’m no expert; marketing and promotion are not my strong suits.

8 Likes

I’m not suggesting a departure from the Q&A model that is known. However, I’m not suggesting that this site “simply copy” what Stack Exchange has done and hope that grants success.

It seems like a lot of people here want this to be successful. Be prepared to defend that notion of success. I just remain mildly unconvinced that mirroring Stack Exchange will do that.

1 Like