Increase exposure

One way of increasing our exposure is to use Codidact as a source when answering on other forums. As long as we are treating Codidact as any other source, there is absolutely nothing wrong with that. I did that with this answer on SO

Improve technical stuff

Sure, there are very few people who joins a forum because it has such amazing features. I'm not saying that this should be the primary selling point. But there are two types of improvements that I believe is quite important.

1. Features that are common on other forums that new user will miss when they come here. Sure, we don't need to implement everything, but we should definitely make sure that nothing is missing.

2. Existing features that just feel clunky and more of a prototype for a feature than a final implementation.

For example, the comment thread system feels very clunky. I love the idea, but it doesn't feel very well executed. It could be that I'm not used to it though.

We should have used the leverage of the Monica incident (too late now)

I think it was a mistake to spend so much time of deciding exactly how codidact should be before launching. Sure, there are benefits with that. I won't deny that.

But we did miss the possibility of using the dissatisfaction generated by the Monica incident to our advantage. This may not be very helpful now, but I think it's worth mentioning.

Focus on being a community

SO is not really a community. And I think that's a bit sad sometimes. I think that's one thing where we could get an edge. A suggestion is to make it a little bit of a mix between a Q/A forum and a social network. A few suggestions:

• Personal blog with commenting features and have a separate category for that besides Q&A, Code reviews and Meta.
• Discussion forums. Let's face it. This Q&A form is not especially good for discussions. There was a forum before Codidact launched for discussing how everything should be here. Maybe that could be brought back in some way?

posted almost 3 years ago

CC BY-SA 4.0

3y ago

klutt‭

1273 reputation 17 12 162 52

By having decent source code formatting that isn't completely inferior to other sites like Stack Overflow. We might want to post/view snippets longer than 13 lines without suffering some scroll wheel. See my feature request here.

My first impression when coming to any Codidact site requiring source code is "wow, this site wasn't designed with programming in mind at all". After which my knee-jerk reaction is: "eww yet another 'forum', exactly what I didn't need".

posted over 2 years ago

CC BY-SA 4.0

Lundin‭

8917 reputation 36 121 1025 165

In light of another fiesta on a "competitor site", I poked my head in here again. Here's my twocents.

When a ship starts going down, there's only three courses of action for every individual

1. Fix it
2. Abandon it
3. Stick with it

The key insight is that individuals chooses the option that sounds the best to them, personally. Given that the ship is going down rather fast at the moment, you'd think there's plenty of motivation to abandon ship. But it's not, here's the thing:

I can't get my daily Q&A fix here.

There are no questions, what am I going to answer? If I abandon ship, I'd be swimming.

This is not entirely a chicken and egg problem. There's no way you can answer a non-existent question. You can, however, ask a question even if you only get subpar answers. You have to get the questions in first. The answerers are easy to, shall we say, entice in the current climate.

Aside from that, there's just a lot of UI/UX pains that adds up. A lot of workflow habits are broken by UI differences, and I don't know whatever is justifying these differences are worth it. I suggest a very loose goal of breaking the least amount of workflow habits carried over from the (ahem) "competitor site".

Exhibit A: I can't resize this answer box. And I'm triggered.

posted over 1 year ago

CC BY-SA 4.0

1y ago

Passer By‭

161 reputation 0 1 16 6

Search engine optimization?

I thought this goes without saying, but apparently we aren't doing too well there for some reason.

The other day I was having a discussion with someone at SO regarding how hard it was to find specific information about anything on SO, even when using Google for the search. As an example I used a topic which I knew there was no good "canonical" for at SO, which had once resulted in me writing one on Codidact instead.

The topic was endianness, a common programming term and the question was simply "What is endianness?" or more specifically "What is CPU endianness?" I have written a Codidact post here with the title What is CPU endianness? and with a tag "endianness".

Using either that very title in Google, or just typing "what is endianess" results in just a few decent hits, with Wikipedia rightly being the #1 hit. Then follows a flood of misc tutorial sites of diverse or unknown reputation, as well as technical articles and the like. Lots of them probably didn't make any particular effort to end up in Google ranks. And yet Codidact is nowhere to be found.

The term "endianness" should be a good example to use for a search term, because it is completely unambiguous and only appears in a software engineering context.

I have to Google "What is CPU endianness?" with the exact title within citation marks to get Codidact appear at all, then as the #1 hit (the other hits being SO posts by me where I link to the Codidact post, as well as some SO scraper site).

This can't be right.

posted about 1 year ago

CC BY-SA 4.0

1y ago

Lundin‭

8917 reputation 36 121 1025 165

I just submitted a proposal to DuckDuckGo here for a new "bang" for their search syntax. If approved:

!coddsw search_term

will trigger the following URL:

https://software.codidact.com/posts/search?search=search_term

I did the same for EE Codidact (here the announcement).

I chose the prefix codd because it was not taken and no existing bang begins with that prefix, so that if someone wants to propose bangs for other Codidact sites, they could use the same prefix. For example coddphoto for Photography or coddphy for Physics.

Hope this helps a bit.

posted over 1 year ago

CC BY-SA 4.0

Lorenzo Donati‭

250 reputation 1 3 25 9

Just my two cents:

1. I found this community because of someone's username on Stack Overflow. That's probably a good start. However I then typed 'codidact' into Google. The first result was codidact.org, the third result was codidact.com. I would have expected it the other way around. Also codidact.com's homepage is the page for all communities, not the software site. So again I had to scroll down to find that. I'm not sure about codidact, but my impression of stack exchange is that's it's based on stack overflow as a base. Probably the highest user count is in stack overflow. So I would think that's the community you want to build the most, and thus it should have front page importance across your site. But that's up to you to decide.

2. The way I first discovered stack overflow was through google search results. Same with coderanch. Personally, when I have an issue, I don't go onto stack overflow and search for an answer. I go onto google. If the answer I want is on a stack overflow page, that's good. If it's somewhere else I don't care. I only care I can find my answer. So what I'm saying is, good visibility in google search results is a good way to get people to visit your site. They are going to care more about the quality of your answers that whether you're stack overflow or someone else.

Growing the community that can answer questions is another issue. Not sure what you can do about that. Probably all the disgruntled people from stack overflow who left or wanted to leave would be a good start.

posted almost 3 years ago

CC BY-SA 4.0

metal.carratt‭

189 reputation 0 1 18 0

As a (for now at least) casual user, I can report that a bad first impression is that there are way too many "500 server errors".

Within a few tens of minutes I ran into two, one for some profile action (clicking on some tab in a user profile?) and one for en edit action.

It was the same impression some months ago and, like for submitting comments to the official Stack Overflow blog, I now know to save everything offline in a text file in order not to risk losing content and/or to have the option to try again some time in the future (including when submitting this very post). That shouldn't be necessary.

If there are some known errors or limitations, I think users should at least be informed somehow (e.g., information about workarounds or things to avoid).

Perhaps take a look at those 500 server error error logs?

posted almost 3 years ago

CC BY-SA 4.0

3y ago

Peter Mortensen‭

101 reputation 0 1 10 6

Never Too Late

Due to, shall we say, recent AI-related hallucinations, pretty much everything that was possible PR-wise in 2019 is possible for this site again. People are leaving Stack Overflow and this is arguably the best existing alternative. Opportunity has knocked again, but in a sense it really never stopped knocking. So let's not have regrets, but instead seize the day.

Getting Noticed

There's plenty written out there about how Stack Overflow's initial SEO was so successful. I'm sure some rules have changed, but many others have not. I don't know a lot about that stuff. I'll leave the technical aspects to staff.

But what I am sure about is that a site like this needs to pop up in web searches to grow. Individual networking just isn't efficient enough. But if people start seeing that this weird "cod I'd act" site has answers for the fundamental, practical things, and puts comprehensive, organized, well explained answers at the top, and doesn't explain the same easy thing multiple times while ignoring the central cause of confusion... that's what will pull them in.

If you build it, they will come.

Just Write It Anyway

So, building.

I'm going to propose something a bit different from previous answers: I seriously think that getting the ball rolling needs to start with the experts, with self-answered questions.

You know that question, the one you're sick of hearing on all sorts of forums, the one where you're dissatisfied with the available duplicates/references you have or just can't find one version to elevate above the others? The one that plagues clueless neophytes who seemingly never have a proper idea of how to express the problem, despite how simple it is? The one where if you try to search for it, you keep finding irrelevant questions that use the same keywords? The one every "serious" developer needs to understand in order to have a smooth workflow?

Yeah. That one. Write it. And answer it.

You're the one qualified to do it. You're the one who's seen the question so often that you know every which way that people get confused, and have seen what works to alleviate that confusion. You're the one who's most invested - emotionally and from a practical perspective - in having the best possible version of it published, so everyone else can get the best possible explanation and see how smart you are not have to worry about where to refer people next time.

And right here is where you want it to exist. Right?

Don't worry about whether it's covered by the documentation. Documentation answers the wrong questions. For how-to questions, it's often only useful to people who have mostly guessed the answer already. It tells people what X does, when they want to know what does X.

Don't worry about whether it's covered by Stack Overflow, or technical forums, or any of those awful "tutorial" sites. Their questions are poorly organized; they don't factor the problem space neatly. When the question is good (and popular), it has dozens of redundant answers, making it harder to find the few good new ideas. We have the advantage of starting fresh, and the benefit of hindsight.

Nobody can keep the Internet DRY. But by leaving a clear target ahead of time, you will help keep the future, bustling Codidact DRY - and bustling, because the people you haven't met yet will appreciate that environment. And in the mean time, you'll be attracting people to the site by building something of value.

Intrinsic motivation

The best volunteer work comes from enthusiasts - the people who want to put in the work because they get fuzzy feelings from the result of the work, who don't need an automated system to increase a number associated with their name.

What I suspect helps to some extent is to make a bit of a competition out of it. I think this counts as an application of game theory. Pick something you're an expert in, and start writing. Nobody can stop you from writing, because you're producing quality, on-topic content. But if you're the only one doing it, the site gets dominated by your pet topic.

And other people don't want that. Right? So what can they do? Well, I suppose they could complain on Meta, but I'd like to think a more productive and natural response is to compete with you, and write other Q&A about their pet topics. Or if they like your topic, maybe they pick something else just for variety. Point is, fixing the problem involves doing more useful, high-quality-library-building work.

posted over 1 year ago

CC BY-SA 4.0

Karl Knechtel‭

2150 reputation 39 54 276 85

After a few years of casually using stack-exchange sites and wandering around on coda-dict, I feel there are mainly three components to the quality of the content on each of these sites:

1. the knowledge base
2. the community
3. the game

The Knowledge Base

I think most answers to this question focus on this aspect of the site. It is also an important aspect, especially if communities end up building (parts of) their knowledge base using sites like this (I think Tex - LaTeX Stack Exchange is a nice example of this).

On top of the other propositions that have been made here, I would like to propose to just ignore questions from "other sites" when looking for an answer. I believe most developers rely on a search engine and the internet to get a quick answer to their questions. Most obvious answers should be in the official documentation and can be found there. In this case, there would be no real need to ask a question. Otherwise, you would probably end up on stackoverflow where your question already has an answer. Instead of just clicking the link and reading through the available answers, I think it might not be a bad idea to ignore these links and just formulate your question on codidact. Afterwards, you can still collect your answer from other sides and move on, but at least the knowledge base on this platform will get a chance to grow.

The Community

This is one of the aspects that I personally have little experience with. I am more of a casual user that has occasionally helped with cleaning up review queues (on stackecxhange). However, this is also seems an important part for people on these sites. If I understood things correctly, the community building has been outsourced to discord. Discord might indeed be a better place to build a community, but it prove to be a barrier to enter the community. I personally do not really care for this aspect, but I do feel like I might be missing parts of the experience by not joining discord. Some sort of Fear Of Missing Out (FOMO), if you will.

From my perspective, there seem to be two issues with the current implementation of the discord solution:

1. I remember reading about discord somewhere, but at this point I can't remember where I read this and how to join. If the idea is to organise the community via discord, it might be useful to have a more accessible/prominent link to join the community. Ideally, there would be some kind of discord integration on the site so that you don't have to leave the site to check what's happening on discord.
2. I can't remember what the purpose of discord is supposed to be. Maybe it would be good to clearly specify what is happening on discord. On one side, this should reduce the FOMO if discord is just providing non-essential extras to the experience on this platform. On the other side, it allows users to make a more informed decision about whether it is worth setting up discord for these extras.

The Game

One part of the stackexchange network that seems to be less present here, is the gamification of Q&A. There is still the reputation that can be gained by asking and answering questions. However, there are no badges or other "perks" that can be earned by contributing to codadict. For people who do not care that much about the community aspect, like myself, this removes a lot of incentive to be really active.

By gamifying the experience on the site, the bootstrap might be a more enjoyable experience. After all, as long as there are no questions to answer, users could "level up" by voting on existing questions, revisiting older questions, doing moderation tasks. This might keep people active on the platform until there are more questions to answer again while keeping them active.

Design

Apart from content, there is of course also the appearance of things. I must admit that I personally do not like the design of the site:

• the content is not centred horizontally (especially annoying when scrolling down this question),
• colours are too harsh for my liking,
• no dark mode,
• too much visual noise around questions,
• reading a comment thread opens a new page,
• questions are not visible when writing an answer,
• ...

Content should of course be the focus point of a Q&A site, but I am afraid that the first impression (for many) is mostly dictated by the appearance and user experience on the site. Maybe it would be good to allocate some (more) resources into setting up a solid design.

PS: I am aware that someone put time and effort in designing the site as it is now and I appreciate the effort, but it just does not tick my boxes.

posted over 1 year ago

CC BY-SA 4.0

mr Tsjolder‭

518 reputation 6 14 61 6

S.E.O. - Stack Overflow has fantastic SEO, and this is a self-feeding cycle. Currently codidact isn't adding json-ld or oembed information so that these posts won't show up neatly in search engines or in social media posts (like discord): https://search.google.com/test/rich-results?url=https%3A%2F%2Fsoftware.codidact.com%2Fposts%2F290279

Better Nomenclature - "Codidact" is hard to say and referring to communities is harder than saying "SO" and "Security Stack Exchange". Consider a more condensed name (https://shop.gandi.net/en/domain/suggest?options=1&search=Co) like co.ninja or co.guru. This will make subs appear as more prominent - software.co.guru looks nice and is easy to say.

posted 12 months ago

CC BY-SA 4.0

rook‭

31 reputation 0 1 3 0

I realize this might not be feasible, because course all of that hinges on the possibility to get some acceptable data on a users behavior on other sites, but for me the biggest hurdle is that to contribute here I have to start over building trust and gaining abilities. That took some serious time over at the other Q&A site, the prospect of doing it all over is very daunting to me.

So my suggestion is give people with a sizable fake internet currency pile on other Q&A sites a migration bonus.

A few options how that could work:

• Plain reputation bonus (maybe a flat 100, maybe something relative to their points at the other site)
• Lower ability requirements (ie if you've proven you can do good edits (flags, ...) elsewhere you're likely to be able to do good edits here)

posted over 1 year ago

CC BY-SA 4.0

cafce25‭

131 reputation 0 3 13 0

Not having a single sign on option greatly increased the friction in adopting the site for me.

This was further compounded when the sign up process failed the first time. Which I believe was a bug that has since been addressed because I have since made my account (obviously).

Apologies if this already in the works. Another thing I don't see is a roadmap of things that are planned.

posted almost 3 years ago

CC BY-SA 4.0

3y ago

jmathew‭

275 reputation 0 6 25 9

Broaden the scope (slightly)

I noticed https://software.codidact.com/posts/292660 was closed due to being off-topic. That's fine on a site the size of Stack Overflow, where there are too many questions and filtering keeps it focused, but a site with too few questions that wants more activity should allow any question that is useful and close enough to the central topic (which this is, IMO), to prioritize growth.

If the contention is that it belongs on another Codidact site, then my recommendation is to merge similar sites until the combined site is big enough to sustain its network effect, or there are no more sites similar enough to merge. Reddit did not gain subreddits until it already had a big enough posting volume to justify splitting them - this avoids the appearance of a ghost town.

posted about 2 months ago

CC BY-NC-SA 4.0

2mo ago

user253751‭

61 reputation 1 1 10 2

A note on first impressions:

I really like that popup windows like when you click "react" are closed by clicking "react" again, not just anywhere else on the screen. This makes the site feel solid to me, and less prone to error. It also works really well with Vimium where you can't easily select a non-action part of the screen.

What I have been fighting with is how comment threads work. Maybe a popup offering a tutorial for new users would be nice. Really though, I want to be able to reply to a comment from the question page; not have to go to a separate page two clicks away.

I like that the main level header size is not obnoxious, making it usable.

Update [7.22.22]:

I like that post urls are short and shareable as is. I use Vimium so grabbing a link on codidact becomes a quick yy. As opposed to massive SO urls that require hitting share and copy to get something reasonable.

It'd be nice to be able to "follow" or "save" users. This is a fun way to follow and learn from the activity of bright users. No feed BS; just looking for a way to "bookmark" these users within the bounds of my codidact account.

I would like a quick way to return to the codidact topics page from within a topic.

I like that you have to actively hit "mark as read" or "mark all as read" in your inbox, for the same reason I like to have to click "react" again. Too often with other services have messages "disappeared" on me, causing me to have to track them down through my browser history.

Update [10.26.22]:

I think there should be voting accountability and traceability. You should be able to see who upvoted and downvoted posts. Even better would be requiring comments for downvoting or a selection from a menu of reasons. A downvote with no context does not help the reader nor the writer.

posted over 2 years ago

CC BY-SA 4.0

2y ago

mcp‭

152 reputation 8 2 21 23

This may be a minor thing to some but it's a huge annoyance / barrier to me - we need to change our scoring system to be human friendly.

score +0.57142857: How to add user and a group in Docker Container running Macosx

8 decimal places is just incredibly silly and unprofessional.

posted over 2 years ago

CC BY-SA 4.0

NobleZarkon‭

87 reputation 0 1 3 0

Emphasize and expand content that competitors fail at or deliberately exclude.

This section of What type of questions can I ask here? is already a big deal, but it could be bigger:

Best practices, as long as clear "best" criteria are provided. Examples: fastest execution, least memory use, widest tool support for a target, IDE for a certain language and operating system

Software development is an art, not a science. Attempting to reduce it to black-and-white technical facts is futile and actively harms understanding. Stack Overflow's aggressive exclusion of "opinion-based" questions severely hinders the transmission of knowledge. A medium for transmitting the deep insights of long experience is something the world desperately needs, because formal education has always failed at it, and long-term professional relationships that create a business case for employers investing in their employees are a thing of the past.

Novices often frame a question as "What is better, X or Y?" This is a teachable moment. We can factually discuss the advantages and disadvantages of different solutions, the history of their popularity, and the real-world reasons you would choose one over the other. On SO, such a question would be closed immediately.

Edit: SO is now promoting their new Discussions site, a place for opinion-based and otherwise fuzzy questions, which is what this site should have been. It already has more traffic than this site. This makes me sad, because while I want such a site to exist, I don't want those jerks controlling it.

posted about 1 year ago

CC BY-SA 4.0

9mo ago

Kevin Krumwiede‭

161 reputation 4 5 15 12

My recommendation: Follow the good parts of StackExchange, & learn from the rest.

This is long, I apologize, but I hope someone finds my treatise valuable.^[1]

Items done right (to be duplicated)

There are, a number of things they did, and to some degree, still do very well. A few of the top seem to be:

1. Questions can be asked by anyone and answered by anyone
2. Questions and answers can both be edited by those with enough reputation
3. Reputation is gained by users showing that an answer or question helped them
4. The look and feel of gaining new rep (and the gamification of all of it) is done well and is successfully motivational for many

Items that could be avoided or improved

StackExchange, while a great improvement is not the end all be all of QA... (and not just because they recently announced an agreement to giveaway all contributor content to OpenAI).

The system they created was ground breaking but has some flaws. It is NOT a perfect system.

Summary of main problems

1. SO/SE is hostile to question contributors
2. Their system lends itself to inbreeding

Elaboration

Hostility

The guarding of the specifics of how questions are asked... and rejecting/downvoting them if asked poorly, or if there is some belief that they are (or even if they are) duplicates makes for a generally hostile environment.

Example

Imagine, you are a new user, particularly one who is a non-native English speaker.

• You are not even sure how a question is meant to be asked in English.
• You do your best to search and don't find anything.
• You ask the question as best you can.
• Within minutes, you get downvoted and get a few messages telling you you did it wrong.
• In fact, you're told the question was a duplicate of a different question that doesn't seem to have answers for your question.
• You now dumped a bunch of time into this very popular site, yet you aren't much closer to getting help.

Alternative

Instead of criticizing the question as bad, link it to the one you think is an answer. Then tell the OP you think the other question has answers for them, but if they don't find them there then please clarify...

Lesson

There really is no bad question... because a question is just a grouping of words that someone might type in a search engine... the closer the 'bad' question is to something someone else might type, the closer they both are to an answer to their 'bad' question. (NOTE: bad answers on the other hand... technically bad answers)

Recommendations

• Don't allow down voting on questions... a low upvote score is plenty enough from making people think it's the best way to word the question...
• Don't close questions... a question that is worded badly will have less answers... this is a self solving 'problem' in the area of popularity of a question and the subsequent answers
• DO link questions to similar or same questions... in this way it is very easy for someone with a similar or same, but differently worded, question to get their answers... ANSWERS are the treasure in these sites, questions are just a way to get to them... so if you have 50 similar questions pointing to the same gloriously worded answer, then people are just more likely to find the ANSWER they need.

Inbreeding

If I have a question on SO that is a really well thought out wording and has a lot of good answers, I'm going to see anything similar as a duplicate and want to shut that 'duplicate' down to get more people funneled to the 'best' version of the question. This is a version of 'when you have a hammer all problems look like a nail'.

However, if I'm asking through a search engine, am I likely to ask a question in a very well worded question format? In fact, who cares how I ask? If I have the right technical parts to the question to get me to an answer, that's all I need.

There is a darker side to inbreeding, which I suspect happens on SO and SE in general, but have no proof of. But it's something I ask myself in all situations, because typically if there's nothing preventing it someone will do it.

Scenario

A user who has the privileges can shut down related questions and then can add those details to their original question... making for motivation to shut down not even duplicates but just similar questions.

But here's what that gets us:

1. One persons perspective on the technical topic... their wording and inherent opinion ... this does NOT help those who wouldn't word the question that way or who have different inherent/underlying opinions
2. Less variety of answers as well as questions for search engines to pick up...

Good Luck!

I'll help where I can... I may bring my few questions over from SO.

posted 6 months ago

CC BY-SA 4.0

5mo ago

MER‭

10 reputation 1 1 -1 5

Give software.codidact.com its own domain

Stack Overflow is by far the largest community on Stack Exchange and likely serves as the main base of users (covered in metal.carratt's answer). I see software.codidact.com as prominent enough to get its own domain name, so it shouldn't feel like just another community under Codidact and I don't like to type a long URL software.codidact.com every time to view the front page.

posted over 1 year ago

CC BY-SA 4.0

qwr‭

-13 reputation 0 2 -7 3