Are we trying to make a more friendly/welcoming site?

Sorry if this question has already been answered, I haven’t look through all the fourm posts, just a good chunk of them.

The FAQ does not mention a goal of being a more welcoming/friendly place however:

I see a lot of posts on here mentioning stackexchange as being ‘unwelcoming’ & that being a reason for codidact to exist. I’ve even seen posts linking here from stackexchange with a message along the lines of ‘if you want something more beginner friendly’… But I havn’t seen any from the admins or official statements suggesting that is a primary goal.

My personal opinion is that stackexchange is already welcoming to new comers - so I’m just curious about what decisions have been made so far?

7 Likes

I think it would be silly not to. SE has a horrible reputation for unfriendliness and being mean to newbies and so by being more welcoming, that will help draw traffic here.

The other flip side of the coin is how SE was lacking the tools to moderate poor quality content properly leading to burnout and lashing out at new members. When upper management tried to fix that, they decided to blame the power users for the poor reputation for unfriendliness and that whole thing went over like a ton of bricks.

I think that fixing both those things would be a major selling point for us.

6 Likes

Which comes from the people that can’t be bothered to read a few instructions and get familiar with a site before barging in and demanding their homework be done for them.

This whole “unfriendly to newbies” myth is a false red herring. SE is unfriendly to those can’t follow directions or don’t care enough to bother because the world owes them an answer. That’s a Good Thing. These are not the users we want.

The only real problem is that they complain, and management that thinks more users is always better tries to “fix” the problem. That’s a management problem, not a user treatment problem.

9 Likes

I think there are two types of “bad question” new users: Those who don’t know the rules and who would be receptive to change if they did know them, and those who don’t care anyways (the “give me the code plz” people).

How effective was SO’s mentor system? For the few in the first camp, would that be a good system to add here to help weed out those who care for the rules and stop crap being publicly visible?

3 Likes

If somebody would call Codidact a “more beginner friendly” alternative, I would call BS.

Of course user interaction design, “making it more usable” - also specifically for beginners - is part of developing this thing, and efforts in that direction have already been made.

But in general, I’d say the goal is a “more community friendly” version.

7 Likes

We are not trying to make a site that excludes people, based on anything - identity, ability, proficiency in English, etc etc. We are trying to make a site that’s suitable for a community, which necessarily comprises people from all backgrounds and abilities. We will not be unwelcoming, and we will not pick and choose who’s allowed to ask or answer or otherwise participate here.

11 Likes

@ArtOfCode Your comment (which I think everyone agrees with) also applies to Stackexchange in it’s current form.

It’s a good selling point in general, but it’s not something that would make us different from Stackexchange.

My question is if we’re trying to solve an unfriendliness problem seen on Stackexchange, specifically it’s reputation for being mean to newbies And if so, how

2 Likes

This is a little unclear, and could be misunderstood during development or (especially) later, by disgruntled users looking for a defense against moderation. Do you mean, for example, “we won’t use arbitrary attributes to pick and choose”, “we won’t prejudge, no matter how plausible it seems”, “we will use only consistent, open, fair principles to judge whether someone can continue participating”, or “we will literally never restrict participation for any reason”?

The first of those seems like a no-brainer. The second one seems like a good idea, but almost automatic (except, perhaps, for discriminating on English ability); it would be difficult to do anything else without asking for and somehow verifying a lot of information up-front. Neither of these are a significant improvement on SE, with the possible exception of somehow improving on the state of English-only requirements. The third one seems useful and meaningful, although a bit challenging, especially given the value of secret anti-abuse algorithms. The fourth one sounds like you’re ruling out post bans and suspensions for low-quality posts or misbehavior, which I assume is a non-starter.

Or is there some other distinct point on that spectrum that you were intending?

2 Likes

In addition: It is up to each community to define its scope, which includes the expected level of experience or expertise and some idea of applicability beyond the asker. Some communities will be more expert-oriented and some more beginner-encouraging; that happens in any community.

Regardless of those bounds, we on our instance expect community members to be polite in their dealings with other users. It’s not rude to close an off-topic question; it is rude to say “that’s a dumb question” or “go read a textbook” or the like. We encourage communities to be as helpful and encouraging as they can, recognizing that this represents a range of activity that varies based on site maturity, site volume, and individual users.

If you can help guide a lost newcomer, great! If all you can do is gently point that person to the help, fine! If you don’t have time to do that, just leave it for someone else – you don’t need to comment on everything you see. If something needs a vote of any type, vote. Etc. We don’t expect coddling; we appreciate mentoring when someone gives that gift; we require respectful behavior at all times.

People have different definitions of “friendly” and “welcoming”, which has been part of the problem both on SE and in our discussions here. Focus on the behavior not the labels.

11 Likes

I think that designing the general UX to make it as easy as possible to to write good questions is all that we can do for new visitors to the site.

There are serious people with good questions, and there are normal, non-technical people with good questions, and then there are people with good questions but who lack… something… necessary to present a good question and who can’t be bothered to follow even a few basic rules or guidelines and so cannot and, more importantly, will not present their questions well.

I like the changes to phrasing that SE made to their feedback mod actions recently that made things perhaps simpler, but also made it easier to show these new visitors what they need to do to make that question better. I think this is a good example of how to make feedback (which will obviously be necessary) more helpful in a positive way, focusing on what can change and improve about the question to make it a better question.

But, the other things, throwing all those dedicated high rep people who do the lions share of the mod and new user interface work under the bus just hurts. I’m not sure how to quantity what makes a good experience and community for those high rep people, besides that the organization running the site just not suck.

7 Likes

I think something that could help with the “If you can guide a lost newcomer, great!” sort of thing would be to introduce a way to positively reinforce that sort of helpful/friendly behavior - some way to tell an experienced user “Good job helping that newcomer!” Maybe the opposite of a flag for comments? A way to mark a comment, “Well done helping with that comment!”?

5 Likes

On Stack Overflow a lot of users were (are) frustrated with the never-ending flood of low-quality questions and were a bit rude as a result (although there was little to no blatant rudeness).

This could either have been a natural emotional response or a conscious effort to get users to go away, since nothing else seems to work and the staff clearly didn’t seem too concerned.

Instead of trying to understand why users are being rude and trying to address the root cause, Stack Overflow instead went with just telling them to stop, which likely just increased frustration.

There may also be a culture or personality element of users who don’t know they’re being perceived as rude, or they’re in denial about it.

Some rude comments are part of the reason why people consider Stack Overflow to be unwelcoming. Another part, perhaps even the biggest part, is the flood of downvotes and close votes questions get, and sometimes the lack of answers. Yes, these are bad questions, but askers don’t (necessarily) know this beforehand.

What should we learn from this?

  • Rudeness is bad. We shouldn’t tolerate it, but this would’ve been much less of a problem if they actually did something about concerns raised.

  • We should downvote and close things to maintain quality, but we should try to minimise the need for this. This is also related to the point below.

  • We need excellent measures in place to prevent getting overwhelmed by bad questions. This can come in many forms:

    • Make it clear what is expected of users before they ask their question.

    • Also make it clear what they can expect. Make it clear that what you’re telling them are rules, not just some things they can ignore if they want.

    • Promote the site as “exclusive” (both externally and on the site itself) to help set expectations and hopefully filter out the “wrong” users and keep the right ones.

    • Review questions before they go public. Either all of them or some automatically classified subset, using AI. At some point we may consider users to be experienced enough to ask a good question, although on Stack Overflow there are a number of low-effort questions by users with a few thousand reputation. This status should probably be based on questions only, not answers or other contributions (answering and asking are 2 different skillsets), and be lost if some number of your recent questions are closed. Not that we need to tell users whether or not they have this status.

    • Prioritisation of questions. Being able to see “hot” questions (ones that are highly upvoted in a short time) is one way to prioritise, but this doesn’t help much if you’re looking for questions to answer.

    • Make it harder to ask questions or register on the site. Perhaps there’s a delay of a few hours or perhaps it’s invitation-based or perhaps there are just a few hoops you need to jump through before being able to register or ask a question.

    • Present alternative sites with lower standards to potential askers. A lot of users may see Stack Overflow as the only place where they can get their question answered. While it’s possibly the best place, there are plenty of alternatives. And certainly Stack Overflow itself would become an alternative to Codidact. But it doesn’t help if users don’t know about it. Telling users about alternatives might not be the best strategy in terms of growing and retaining the user-base, but it could help with the ratio of good users to bad users.

4 Likes

No! That feeds right into the “elitist” mode, that is (I think mostly incorrectly) stated regarding SE.

This is a possible solution. And it can be done pretty easily, I think. We don’t need AI and all kinds of fancy stuff. Just make it a Privilege thing. Ask a Question at TL0 (or TL1 ? I don’t know the exact level really.) and you are told that it is going into a Review Queue. (I don’t like Review Queues the way SE does them, but here it makes sense). Once you have asked ‘n’ “good enough” questions (positive score or whatever is appropriate), then your questions don’t need to be reviewed any more.

We do this essentially the same as Edits for low-rep users on SE. And do the same thing for Answers (but for Answers make the privilege based only on Answers - some people never ask Questions, some people only ask Questions. There is overlap of skills (which is how “general rep ==> general privileges” is supposed to work), but in reality, while both Q & A should have:

  • Proper language (whether English or otherwise) - spelling, grammar, etc.
  • Understanding of community requirements (what types of questions are allowed)
  • Understanding of technical requirements (e.g., how to properly include code)
    there are definitely some differences and a good Asker may not be a good Answerer (at that time, people can improve) and vice versa.
2 Likes

I strongly disagree with quite a bit of the fundamental ideas of this - if you market yourself as a site where you get ‘answers to your questions, fast!’, then proceed to do anything other than answer the question, you’ve kind of (I.e. quite directly) lied to the new user. This is a problem already. Askers don’t care if it’s bad, they’re promised an answer regardless.

That is, the marketing needs to be clear that we want quality questions as well as answers. The system behind how SE works needs to change, in other words, which is part of why we’re here.

7 Likes

Not MVP, certainly, and getting AI and moderation good enough to do better than SO’s Triage (and it would need to be a lot better) is … non-trivial at best. Doing anything more than an extremely quick one- or two-user check (a la Olin’s idea of requiring an upvote for new users’ questions before opening up answering) will also slow down answers drastically, which is quite painful, and won’t catch that much.

I would support requiring askers to be registered (i.e., confirmed email), so that we know we can reach them with notifications if something is wrong with their questions. Anything beyond this is extremely dubious, since it cuts down on a lot of good questions as well as bad, and worse, does so to an extent that’s almost impossible to gauge. (Questions that aren’t asked because it’s too annoying are invisible. Questions that aren’t asked because the asker tried previously and mentally crossed Codidact off their list of options are even more invisible.)

Note that Beeminder has had surprising success, at least with its admittedly idiosyncratic userbase, in deliberately advertising all its competitors. They hypothesize that the mechanism for this is that advertising in a small niche primarily expands the niche, which benefits all companies within it, and then Beeminder also reaps the benefits of being obviously transparent and user-focused. Q&A isn’t really a niche that can be expanded much in that way (despite SE’s years of considerable effort to do so), so in a purely practical sense this isn’t likely to help us much beyond the obvious good public image and the occasional bad question that gets asked elsewhere.

1 Like

Is the “fast” part really part of the marketing? I prefer good answers a day later to bad answers in five minutes. Indeed, I dislike the “I want my answer now!” mindset. Also “fast” is an invitation to badly formulated posts (because it takes time to formulate well),

I would prefer a site that says “we provide good answers” to “we provide fast answers.” And conversely, if I ask the question, I do not like to be expected to be around all the time to quickly react to any clarification requests. Sure, I’m not going to fire the question up and immediately leave, but I don’t want to be expected to stay for the next half hour to see whether someone asks for clarification.

I decidedly do not want the site to promote fastness.

7 Likes

But how about good answers 5 days later instead of tomorrow?
Your perspective is that of a site creator, wanting a good website, right?

Now think about the user’s perspective:
You’re stuck with a problem, quite possibly at work. You can ask on SO, where you can expect answers in about half a day if your question isn’t too niche. It might not be a good answer, but it will probably point you in a new direction.

That’s what we have to compete with, and it’s a ruthless competition for efficiency.

That’s fine with me, too. Sure, it is nice if you get a helpful answer right away, but if I feel I need an immediate answer now then I’ll prefer other ways to get information.

I rarely asked questions where an immediate answer will make a difference, and even in the few cases I did, I didn’t really expect an immediate answer.

Maybe it is because I started internet communication on Usenet where it could easily take a day until someone else even saw the question.

4 Likes

I participated in the mentoring experiment. From memory I wanna say maybe one in 5 or 6 people who entered mentorship left it with a substantially better question.

6 Likes

Except the current CoC is already exclusionary to a huge chunk of the world’s population that happens to be religious. “Hate speech” to most irrational people already includes simple disagreement on different worldviews, no matter how delicately conveyed. The “hate speech” requirement is not conducive to civil discourse, already invalidates your “all backgrounds” statement, as well as the unwelcoming and pick and choose statements. Preventing hateful speech is already covered by the other clauses in the CoC.

That’s fine - it’s your community and you can define it, but you can’t pretend it’s not exclusionary already based on the current CoC. @cellio, of all people, I would expect to have this front and center from recent events. I don’t mean this to be a tangent to the CoC thread, just to answer the question at the top and point out that the site is already unwelcoming and discriminatory to a large chunk of the world.

8 Likes