Great Plans Abound: or how we're going to launch a Q&A site quicker

The eagle-eyed among us have already noticed that a few hours ago, I added a qpixel repository to our organisation. For the unacquainted: qpixel is the software that’s running writing.codidact.com - the site that’s hosting the Writing community to avoid it dying out completely on Stack Exchange.

Prompted by some discussion around our tech stack in the past week or so, we’ve been kicking ideas around between the leads. It seems that a number of the concerns folks have expressed come back to the central idea that we’re not working quick enough. The impetus that Stack Exchange has given us by being terrible won’t last forever; we need to have something to show communities when they come knocking.

So, putting those two things together: here’s the plan.

Point the First: we take what we have in qpixel, and update it to work for us. There’s a small number of significant changes we need to make for it to work for us, and a larger number of small polishes. We’ll use this to host communities for now, while

Point B: qpixel isn’t Codidact; it’s still very much SE-style Q&A, even with updates. So we’re going to keep working on Codidact, because that’s our new format, and because that has the potential to be much better than SE-style Q&A. It’ll take longer to launch, but that’s okay - we’ll be able to host communities in the meantime, and let them switch over once we’re ready.

Point III: The point (ha!) of this plan is twofold. One: we get the ability to get communities on board soon (a few weeks to a month, with any luck). Two: the folks we currently have who can’t contribute to .NET Core, for some reason, may be able to contribute to a Rails app instead: that way, we engage more of our people, and keep ourselves growing.

There’s bits of work to be done before we’re quite ready for this, but stay tuned :smiley:

17 Likes

Key questions, which may have been discussed and answered previously, but which will significantly affect how qpixel works compared to SE & Codidact:

  • Does it currently support multiple Communities in one instance?
  • If it does not, is there any practical/reasonable/not-a-ton-of-work way to have multiple instances shared user record/login credentials?

This is a key issue that has been discussed (more than once…) regarding Codidact. I am not concerned at the moment about the “database per community vs. shared database within an instance” question (though I have made my opinion pretty clear on that). I am just concerned with “will it work” for a reasonable user-across-multiple-communities experience.

1 Like

Not presently. This is one of those “major changes” I mentioned: while we could get away with hosting a very small number of communities on separate instances, we need the network functionality for it to be a workable stop-gap solution.

5 Likes

100% agreed. As it is your code and you know it best, I’ll leave the details to you (I see three different ways of handling it). But it is a key feature. And since I don’t know Ruby/Rails, I can’t help much :frowning:

2 Likes

You’re doing just fine. I’m more of a bystander here since I don’t have the software knowledge for the kind of web development this project requires, but I’ve seen over and over again how pushing for something quick is bad for the result, and can even make things slower in the long run.

You have the advantage that there isn’t some hard deadline. It’s far more important to get it right than quick. Take whatever time you need to get it right. Don’t succumb to instant gratification and end up with a compromised design that will haunt you years into the future. There will be enough of that anyway just because things change as you get into the details. Don’t make it worse by starting out quick and dirty.

Also, don’t ignore or postpone discussion of features just because you don’t plan on them being in the first release. I do know enough about software development to know that thinking about the big picture up front is vitally important, even if the first release only implements a subset. You have to get the architecture right, and have a vision for how all imagined features will be implemented, else you’ll paint yourself into a corner by implementing the first few. Even the first few lines of code need to written within the context of the large vision.

This looking over your shoulder is not productive. Forget Stack Exchange. Their screw-ups may be what got you all here in the first place, but that’s not relevant now. Never mind what they might or might not fix or when they might do it. Focus on building the bestest most awesumest Q&A site ever. In a few years, that is all that will matter. Long term, that’s what will attract the users. You don’t have a bunch of investors breathing down your neck expecting a return on their investment in 24 months. Don’t get caught in that trap.

Thinking you have a short window to capture SE users is counter-productive, and probably wrong anyway. The best platform will win out in the end. Yours will be powered by open software and run by the community, two things SE can’t ever replicate. A few years from now, nobody will remember which month the first Codadict instance was spun up. Once you have something good, the users will come.

Building up a site is something I can help with. But, the software has to support all the necessary features because you only get one (maybe 1½ if you’re lucky) opportunity to bring key contributors from elsewhere. A rush to a partial implementation doesn’t serve that purpose. People you invite will tune out, and it will be a lot harder to get them to take a look after you’ve got what they really want. You can’t keep saying “We fixed it. No, this time for real.” over and over again.

17 Likes

For those wanting to contribute to qpixel (that’s Ruby on Rails), I’ve set up some tasks. Take a look at the issues for a list of things that need doing; the most critical issues are listed in this project board. Each issue also has a priority tag.

6 Likes

I agree. The company will almost certainly muddle through. It alienated people, but the site is pretty much on autopilot for now. At some point leadership will wake up to the problems of unanswered questions and aging information. When they do, they will find solutions to those problems unless they have run out of money or something. The mistake they made was to drain the moat that is the community. Filling that moat will be difficult 'cause it’s more than just building new features or reallocating resources. So you have an opening and will have it for a long time.

Joel wrote about Fire and Motion, which is the concept of moving forward a little bit every day as a strategy. What you gotta do is ignore the covering fire. Don’t worry about what Stack Exchange is doing. (This is the same advice I’d give if I were still working there. I think a little competition would do them good. They’ve had a long run on top which always makes a company complacent.)

Now if there are communities about to die on Stack Exchange, it would be good to provide a home for them. Your project values start with communities, so that’s putting community first. If you can accomplish it without too much effort, that’s a good plan in my opinion. You won’t be able to match Stack Exchange feature for feature. But you might succeed by finding a better formula for motivating contributions.

13 Likes

Worthwhile advice there - we’ve talked about what we’re trying to build, but I don’t think we’ve yet looked at contributor motivation. Perhaps as we get further along, we should start up a conversation about that: it’s partially down to the communities we host, but as with all good things our software shouldn’t be getting in the way of it.

This, on the other hand, we have talked about - this very thread, no less. The software that’s currently running Writing is being expanded to be able to host more communities at once, and we’ll soon be able to host more sites as and when they leave the SE network (or, for that matter, any others who want to join us).

6 Likes

Administrative notice: I have deleted four posts that were off topic. Please don’t discuss “reputation” (in any form) in this thread.

2 Likes

Then what did you mean by “motivating contributions”?

“Contributor motivation” or any other type of reputation or whatever is off-topic for this thread. It’s about launching communities with a stop-gap solution (qpixel) to save them from being harmed. Reputation as in SE is a part of that software, but is open for discussion for our proper Codidact software. Please open a new topic, though, to start a discussion.

2 Likes

I would like to suggest to launch not only the Writing community, but the Programming (StackOverflow equivalent), Server, SuperUser, Ubuntu and Design as well.

I really don’t want to post, ask and answer anything on SE sites anymore. I also want to transfer all my existing questions and answers to the proper Codidact communities, if this is allowed.

(even if the only option will be to rewrite everything from my questions and answers)

3 Likes

One catch, aside from that QPixel does not have the level of features that we plan to have in Codidact (but it has a very good start with all the key elements) is moderation. Any site we support (whether in QPixel or Codidact) will need to have reasonable moderation in place. My understanding with Writing is that enough key people were interested in moving to QPixel/Codidact to take care of any moderation needs. I don’t know if that would be the case at this time for the more general programming/IT sites.

3 Likes

We’ll launch the communities that have users who want to build communities with us. When QPixel is ready, we should offer this stopgap to anybody who seems interested, but we need a user base to build here. Some communities will presumably jump onto the stopgap, others will wait for Codidact so they only go through change once, some will go to TopAnswers instead, and some will decide that inertia is a powerful force and SE isn’t that bad so they’ll stay put.

Nothing looks worse to visitors than a community with no activity, so we absolutely need people who want to participate on each community here. I sure hope that my (and your) favorite communities are among the ones that join us.

6 Likes

For Writing, the current and recent mods were all interested in the QPixel site, so we just appointed the people who had already been mods on SE. If a group of users comes here to set up a community and they don’t bring SE moderators with them, we’ll need to discuss with them how to choose some moderators.

6 Likes

When I’m sitting at my computer and it’s doing some long-running process, I’ll quite often open a browser and type the URL of some familiar website to pass the time. The other QA site has long been among them. The same fiery spirit that prompted devs to create this site prompted me to seek it out. What I find in myself is that as time goes on, that fire cools slightly. There’s a memory of the burn, and an expectation that at any moment new fuel or wind will be added to that fire.

When I find myself visiting the other site, I often wonder “How’s Codidact coming along?”
I don’t have the experience to help it come along any faster. I hope that Codidact, whether a brand new product or through qpixel has a viable alternative soon enough to capture those whose spirits still burn!

4 Likes