Should we start a Concrete5 Q&A site on Stack Exchange?
Permalink 4 users found helpful-----
Concrete 5's forums are decent, but they are no match to Stack Overflow's Q&A engine. Understandably - the forum is not part of C5's core business. Stack Overflow has dozens of developers working on the engine full time.
Stack Exchange, SO's sister network featuring Q&A sites on a variety of topics, has the Area 51 process for creating new Q&A sites specific to a technology or topic:http://area51.stackexchange.com/... Because SE doesn't want to create ghost towns, they require a pretty big community to gather *before* a new proposal is allowed to even go into private beta, during which time the site is seeded with questions and answers, and its viability determined. The community that wants to form the site also largely decides what is on topic and what is an example of a good question (check out the proposal linked above).
Currently, C5 programming questions live on Stack Overflow proper. It would be cool to have a C5 Stack Exchange community, though. It would be more inclusive to all kinds of questions around C5, including use, and the system would have a lot of advantages over the current forum:
- A mature, world-class Q&A platform with excellent, commercial-quality search, ease of use, community editing, tagging...
- Easy formatting options, easy insertion of images, etc.
- The opportunity to create "canonical" entries for questions that come up frequently - that can be edited over time
- A voting system that is very helpful in gauging the quality of a contribution
- An addictive reputation system that greatly boosts contributions, and allows contributors to make a name for themselves
- A great chat product that (AFAIK) is bolted onto each Q&A site
- CC-Wiki licensed contributions (so anyone can take and republish them at any time)
Many CMS communities already have their own very successful Q&A sites there: Drupal (http://drupal.stackexchange.com/), Joomla (http://joomla.stackexchange.com/), Expression Engine (http://expressionengine.stackexchange.com/), and more.
Would it be worth starting a Concrete5 Q&A site on Stack Exchange, and outsource most Q&A there, keeping the forums for real *discussion* only?
The guys and I discussed this last month when I was in town. How can we help promote this? Tweets, a blog post?
I guess it would be worth giving it a new try - but only if C5 is willing to support it with an orchestrated publicity effort (Newsletter/Facebook/Twitter etc.)
There is now a dedicated "StackOverflow" page and navigation link on concrete5.org. The link is found under Community.
https://www.concrete5.org/developers/stack-overflow...
If enough people get in the habit of using Stack Overflow, it could make a future proposal easier to pass.
I can start a new proposal, but we should organize it so a couple people can seed questions right away, and your newsletter can mention it.
Would it make sense to start, say, a week or two before your next newsletter is due?
There is now a Stack Overflow link in the concrete5.org navigation under Community. The goal of this was to get people used to using Stack Overflow for certain types of questions.
https://www.concrete5.org/developers/stack-overflow...
I think it might be better to wait a few months before starting a new proposal.
My impression is that the current concrete5.org forums and howtos are not as well used or as in depth as they used to be and it would be detrimental to the community if they received even less high-level input as a consequence of a new communication (unless everything switched).
It all comes down to striking a balance between how widely to wave the flag and diluting the resources needed to support that flag waving so far as to become ineffectual.
1. typical reply times, looks like the Wordpress one (http://wordpress.stackexchange.com/) is very active and replies are typically within minutes or within the hour. The Drupal one is also fairly quick (http://drupal.stackexchange.com/) but appears to be a little slower than the Wordpress one.
2. Other communities like Wordpress appears to be supporting their StackExchange portal officially/semi-offically:
http://make.wordpress.org/support/handbook/stackexchange/...
However, looks like there's a proposal to close down their forums in favor of stack exchange
https://groups.drupal.org/node/313083...
Magento seems to have a very active developer community too.
* The system is proactive in determining duplicate questions
* The sheer popularity of the system ensures a strong community can grow, and feedback is much more immediate. It's not uncommon for a question to attract really high-quality answers within minutes of being asked. In my experience, this is much, MUCH faster than traditional forums.
So here's my thoughts on the issue of many "communication mediums", firstly stackoverflow is not a medium for communication, it is a place to ask questions and get answers. We will not be communicating anything in there at all, people will just get their questions answered.
Take a look at this stack overflow question:
http://stackoverflow.com/questions/34305548/activate-a-topic-for-a-...
This same question was posted on our forums:
https://www.concrete5.org/community/forums/5-7-discussion/add-a-topi...
Notice that there is a VERY clear difference between the content of the questions, the stack overflow question has a lot more information and even has a bounty attached to it.
I am confident that had Daenu not posted to stackoverflow, he'd still be waiting for an answer.
Stackoverflow gave him:
* Tools to write a solid question
* Very specific requirements to follow that FORCED him to write a good question
* Proof reading by people outside of concrete5's community
* A way to assign a bounty and get a faster answer
* A community of mixed developers who are looking for developer problems to solve
* Exposure to people who do not go into our forums
* A place for that question and answer to live forever in google search results
* A path for anyone to update the answer when it's no longer valid
Now on the other hand, lets look at a different question: http://stackoverflow.com/questions/34763931/concrete-5-7-displaying...
This question does not have any code, there's no real answer to give. This question violates the rules of stackoverflow,
so it should go in our forums where we have site owners and developers willing to help other site owners and developers.
TLDR:
StackOverflow is for when:
1. You have tried to fix the problem yourself
2. You have clear code that shows what you are trying to do and what you expect should happen
3. The answer you get will be a short code snippet or a few paragraphs with deeper knowledge about core functionality
concrete5 forums are for when:
A. You have no code to show
B. You have an open ended question that requires back and forth to answer
C. You have a question about concrete5.org or concrete5 the community
D. If you're given code you won't necessarily know what to do with it
E. You want help implementing the answer
Great writeups @jessicadunbar and @Korvin. They echo exactly my thoughts why this might be a net benefit to the C5 community.
As to what needs to be done, spreading the word through Twitter / a Blog post would be really helpful. We still need 49 followers for the proposal, and 40 more highly voted example questions. Anyone who signs up in this phase will be able to participate in the private Beta, and shape direction the site will take.
Examples of questions that would NOT work (say, because they'd be more on topic in the forum under the separation Korvin outlines) would be great as well!
> There are already several communication mediums that have been enthusiastically started for concrete5 and pretty much atrophied. In the course of atrophying they have spread resources and placed information in places outside the usual search scope.
That's a serious concern for sure. I thought about this before making the suggestion - but like others in this thread have said, the SO infrastructure so many advantages that could make this actually work *better* than the forums that it seems like it's worth a shot.
Also, in a worst case scenario, the Stack Exchange site won't be going live at all if the Beta doesn't show healthy community interest and activity. It won't be left out there to languish. (Any content created until that point we could even reintegrate into the forums thanks to the CC license.)
There are potential benefits though:
- It helps promote concrete5 to a new audience.
- I imagine many people view a product with a dedicated StackExchange community as being more "legitimate".
- For many, there will be a much greater incentive to answer questions on StackExchange than the concrete5 forum. Due to most StackExchange users using their accounts for getting jobs and networking. Which makes them very focused on earning StackExchange points and badges.
Currently there is an issue with the forums generally serving to answer low level and non-technical questions. The developers who have the knowledge and skill to answer more technical questions often avoid the forums. This leaves technical questions unanswered and pushes the question asker to other venues like IRC.
Considering Korvin's outline of how StackExchange question asking works, this could provide a separate outlet serving a different purpose and a different set of users.
"StackOverflow is for when:
1. You have tried to fix the problem yourself
2. You have clear code that shows what you are trying to do and what you expect should happen
3. The answer you get will be a short code snippet or a few paragraphs with deeper knowledge about core functionality
concrete5 forums are for when:
A. You have no code to show
B. You have an open ended question that requires back and forth to answer
C. You have a question about concrete5.org or concrete5 the community
D. If you're given code you won't necessarily know what to do with it
E. You want help implementing the answer"
A great majority of the questions posted on the forums would not qualify for StackExchange. They are often broad and lack specific details.
This still leaves one area unaddressed, technical questions that are suited for the concrete5 forum.
Why move all developers forum to Stack and keep only editor and other funky forum on C5 ?
A skilled user is not necessarily a developer - getting an answer like that could draw him away from the concrete5 forums, what would also mean losing a potential supporter.
The current problem is that many skilled developers who know a lot about concrete5 avoid the forums. I think part of this has to do with many forum questions not including specific details, are not written clearly, require a lot of work to coach the question asker to provide useful information, and could be self-answered by doing a search of the forum, documentation, and tutorials. This makes answering questions much more time consuming and requires a lot of patience. Patience some don't have or have the time to invest.
The core team currently does not have enough time to work on the core and look through the forum questions to see what needs answering. This means many of the technical questions on the forum go unanswered.
There are patient and skilled developers answering questions on the forum, but there are only so many of them and they only have so much time. They can't get to all questions, which means some fall through the cracks.
If StackExchange could provide a channel for clear, well formed questions that can be more easily understood and answered. I think it could be a compliment to the forums and attract skilled developers to ask and answer questions.
As an experiment, there is no harm in trying.
"but who will then filter out 'unsuitable' questions and will these then get rejected, and by whom?"
If we use exchangecore's idea to create an area of concrete5.org that has StackExchange questions, we can include question guidelines for forums and for StackExchange. Something like Korvin's descriptions could be at the top of each page.
"StackOverflow is for when:
1. You have tried to fix the problem yourself
2. You have clear code that shows what you are trying to do and what you expect should happen
3. The answer you get will be a short code snippet or a few paragraphs with deeper knowledge about core functionality
concrete5 forums are for when:
A. You have no code to show
B. You have an open ended question that requires back and forth to answer
C. You have a question about concrete5.org or concrete5 the community
D. If you're given code you won't necessarily know what to do with it
E. You want help implementing the answer"
If users choose to ignore the guidelines and post the wrong type of question on StackExchange, these "unsuitable" questions will be rejected by StackExchange moderators and users with high reputation scores.
One example of this is the ProgressTalk community, which has an entire forum section that's strictly pulling in external resource feeds and republishing to their own website to help gain community exposure:http://progresstalk.com/categories/external-resource-feeds.90/...
I think this is an excellent idea. The StackExchange section could have a pinned message describing the distinction between the regular forum and StackExchange and how the questions are different.
what about the concrete5 forums then? Should there be another pin saying "don't ask too advanced questions here, go to StackOverflow instead'...?
Just wondering how this should be handled in practice.
I'm not sure if there are enough questions asked on our CMS at SO.
What I'm missing a bit in this thread is the fact that SO is having tags not only as topics but also for subjects that aren't big enough to have an own site.
If you look at the number of followers/questions our tags are having it shows that we aren't by far big enough to get our own site.
"concrete5 5.7" has 11 followers & 40 questions
"concrete" has 6 followers & 50 questions
"concrete5" has 113 followers & 395 questions
For comparison the "drupal" tag has 4.3K followers & 16.4K questions.
"joomla" has 2.3K followers & 13K questions.
So if we take the amount of questions that would fit for SO (Korvin: "A great majority of the questions posted on the forums would not qualify for StackExchange. They are often broad and lack specific details.") and the amount of questions being asked in here that wouldn't fit on SO at all, In my opinion we aren't a community big enough to have our own site on SO.
Also an editor really can't be bothered to ask code questions on SO.
The question Korvin mentioned is mine, BUT I surely wouldn't have had to go on SO if our docs were complete. I remember when working with c5 5.6 I had always an answer in here or when reading our proper docs (googling around usually led me to concrete5.org).
BTW As soon as I have some spare time (it will be next week) I'll write some docs too.
What I'm trying to say is that we should first complete our own docs, especially the developer part, and if then there is still the need of an own SO site, we could think about.
Don't forget that very often people don't invest time in Google researches nore they are reading the docs nore they are byting thru the source code. It is easier to cry "Help Me!" in the forums than googling, byting & trying around to get a solution from the CMS itself.
Also I'm agreeing with JohnTheFish that more channels aren't necessarily better to find a solution.
I personally would be happy if we'd improve our own tools to be able to write better questions (MarkDown, being able to edit other questions if enough Karma, etc)
On SO, you are limited by the extremely strict on-topicness guidelines of the whole site. Those guidelines are enforced by moderators and users that don't necessarily know anything about C5.
On a SE site of our own, we could work out for ourselves what is on topic and what isn't. Chances are we are going to be slightly more lenient than SO, allow good questions that focus on non-programming aspects of the system that would likely be closed as off-topic on SO (usage, customization, configuration...) and generally be less harsh than SO has to be with its 10,000+ new questions every day.
And I totally agree that on a SO site we could make the rules and it would be better than any tag will ever be.
In terms of documentation, I agree we should continue to work on documentation but I disagree that we need to halt any other community activities until that is resolved. Part of what made v6 documentation more "complete" is that we had a lot of questions and answers developer or not in the forums. Encouraging more of that can only help.
On the question of multiple avenues of communication, IMHO that is a red herring. Stack overflow is not a way to communicate, it's a platform for asking and answering questions.
I think part of the misunderstanding here is that forums are for discussion while stack overflow is for answers. We have tools to make a discussion forum work as an answer platform, but we have no way to restrict extra discussion or prevent the thread from running off onto a different topic.
When I search google for an answer and find a result on concrete5.org, I end up having to wade through tons of discussion to find a suitable answer. Stackoverflow doesn't have this issue because stackoverflow doesn't allow discussion. For that reason any time I'm googling for an answer to ANYTHING and I see a stackexchange site, I go there instead of any other resource.
If we don't get enough followers and questions to move into the next phase, then I think we should look into creating a StackOverflow page under the Community navigation link. On this page we could link to existing StackOverflow concrete5 related tags and include information that describes what StackOverflow is, is not, and what it is suited for. This could address the issue of diluting the concrete5 communication channels.
http://area51.stackexchange.com/proposals/94596/concrete5?referrer=...
We need:
30 more followers
39 more questions with a score of 10 or more
Following can be done by anyone, users, designers, and developers.
27 more folowers,
36 more questions with 10 upvotes
We need:
24 more followers
36 more questions with a score of 10 or more
http://area51.stackexchange.com/proposals/94596/concrete5?referrer=...
We'd really love to have our own SO site, so please go to
http://area51.stackexchange.com/proposals/94596/concrete5...
and do the following:
1. Click on the orange button "Follow"
2. Post up to 5 questions you'd ask on the concrete5 SO-site
3. Upvote 5 questions that haven't got 10 upvotes
Please use your 5 votes on questions with less then 10 votes!!
Thank you very much!
We still need 20 more followers & 35 more questions with a score of 10 or more
Please sign up and vote for questions that have LESS than 10 votes.
http://area51.stackexchange.com/proposals/94596/concrete5?tab=votes...
I like this idea.
One note though, it is "concrete5" and not "Concrete5". Are you able to update the name?