Would a questions that says "Could someone recommend anextension that does x, y, and z", be on topic? (ie the Software Recommendations site). These questions are off topic in stack overflow, would they be on or off topic here?
Let's take a step back for a moment, and look at what we, as a site, want to accomplish. Our job here is to solve problems for people who are working in Joomla. Sometimes, a problem is solved with an extension. It doesn't matter if there are multiple extensions available - our system is designed to support the idea that there may be multiple solutions to one problem, that's why we allow more than one answer to begin with. The key is that what we have is a problem, and that problem can be solved by at least one individual answer.
There's an inherent stigma around the idea of "recommendations", and the trick here is to realize, there isn't always an issue with these things. The stigma about recommendations draws from how those kinds of questions played out when formatted a certain way, wherein the goal isn't to get one item but to build a list of items - so let's not format them that way.
I like to point out my very first question on Stack Overflow, because I find it very illustrative. This I asked before I even knew that there was some sort of "recommendations" thing, this was my first step on the site. And while it's SharePoint and not Joomla, it remains the same thing - I'm looking to get some kind of thing that will solve my problems. And that's even what I got - the top answer, and the one I ended up implementing, is a "solution pack" that did exactly what I needed.
I didn't get chastised for this because I didn't simply ask for a solution pack. I laid out exactly what my problem was, and specified what I was trying to solve and how I wanted to solve it. What follows may play like what people call recommendation questions, but the heart is that I didn't ask one. I stayed true to the form of a real question.
Again, I'll reassert what I opened up with. We're in the business of solving problems, and sometimes the solution to a problem is an extension. We focus on that direction, everyone is happy. We on the site, we get a question that is a problem with a clear solution. Sometimes it'll get a bunch of clear solutions, but they each all solve the problem. The asker, they get the extension they need. Future askers who have the same problem, what they find is a solution to that same problem. That's not just win-win, that's win-win-win.
We're in private beta now - this is the perfect time to get this kind of operation into the heart of the site. Build that direction that you want people to go in, make examples of proper questions being about the problems with their solutions, and not about a list of things that do a thing. You do the former, you get the latter by extension because that is how it manifests.
And most of all, be open to guidance about this. Especially when this site opens up, there's lots of nuances about how to post properly and all that jazz. If someone provides a concrete problem they want to solve, but their framing makes it "look like a recommendation question", then we can help. We know how to fix this - we use our powers to edit to shape it so that the focus of the question is the problem, and that it seeks a solution. We explain things to people, we work with new users, and that is how we avoid 'extension recommendations' being a thing to worry about.
Looking over the other answers here, especially web-tiki's, and I want to make a note about problem spaces. Not everything someone approaches us with as a "Find me an extension" can be transformed into an acceptable question. This is what we have to understand is that this is not about the nature as a recommendation, but about the problem space that is asked for.
If I ask for an extension about how to complete a specific task, if I lay out groundwork on what I'm trying to get done, that's something you can abstract into a singular concrete problem. That is the kind of thing that mirrors the regular questions the site asks for, and we can fix those.
If I ask for an extension about accomplishing a vast library of tasks, like managing different portions of a gigantic workflow, that is multiple problems. "E-commerce", there's a lot that needs to be done in that. That kind of situation is where the problem space is too large to manage in a single question. And that's the kind of problem space is already covered by close reasons for "being too broad" or as is used in this case, "too opinionated".
That's a lot of how these play out. Check if the transformed state makes sense on the site. If it does, we win. If it doesn't, then we can't help properly, recommendation or otherwise.
First, I think we should focus on the main aim this site should fullfill and I think in a few words it is : provide high quality information for Joomla! administrators, users, developers and designers.
After going through David's link, I ended up on this blog post : shopping questions.
It says :
...Thus, when it comes to shopping questions, don’t ask us what you should buy — ask us what you need to learn to tell what you should buy... ...As the old Chinese proverb goes, “Give a man a fish and you feed him for a day. Teach a man to fish and you feed him for a lifetime.” Every question and answer ultimately should be about teaching and learning — yes, even the shopping ones
I have the feeling that extension recommendations will often end up in the "give me a fish" category witch should be considered as low quality information.
Now lets take an example : E-commerce with a Joomla extension?
The Users asks how he should choose a E-commerce extension. So it ends up in the "teach me how to fish" category. But how can this question provide high quality information in the Q/A context this site is built on?
- The question should clearly list specific requirements/ features the user is looking for and not just "I want to sell X witch one should I choose?".
- The answer should list most (all) the available extensions that have those features and explain why and how they fit.
This would involve very long answers that may not suit Q/A.
The answer provided would only be valid for the time extensions aren't updated and as features are regularly added by develloper the answer could be wrong tomorow. Updates could be made regularly but it would be a lot of work and moderators shouldn't have to go around and update these questions. And the JED/develloper site do this very well.
This senario would provide high quality information but I don't think it can realy occur because :
- If the user can ask such specific question, he would most likely already have taken a tour on JED and seen devellopers website and therefore found the information he is looking for.
- Can someone realy answer such a question without involving his/her opinion?
If we look at the WSX site they consider these questions as off-topic because :
They add exactly no value to the site.
WSX is similar to this site and I think we can use their experience.
I agree, this is a bit of a grey area, but the answers might be useful.
An answer, for example, that discuss multiple extensions and a high-level description of how those extensions could be used to achieve the site's objective would be very valuable.
So for example, somebody might ask, "How can add Event Booking with RSVP capabilities"
Use XYZ Extension - here, I'll give a coupon code and just call me if you need any help.
You will need a 3rd party extension. I've used 3 different ones over the years
ABC extension works by.....
DEF extension works by.....
XYZ extension, the one I am using now, works by...
I think we should have a rule that you cannot promote your own extension or services for sale - but we should be able to freely ask and answer questions about available solutions.
I don't know, I think there is a degree of need for someone to say "I need to be able to present this type of information... how do I do it?" And that's going to end up with a recommendation most of the time. "Best eCommerce" is a totally different question entirely and not appropriate.
Best eCommerce for posting Cars? maybe that's a valid question... because problem-solving I would think the vertical market applications, as well as a CCK, before something like Virtuemart.
So there's a gray area.