You're doing it all wrong!

Question

Before this proposal went into beta, Tim Post laid out some ground rules on Area 51. If you haven't read them yet, I strongly encourage you to do so. The problem, a lot of you guys aren't following them, and it's really bringing down the overall quality of a just-blooming site. It's beginning to look an awful lot like Yahoo! Answers.

My main concern here is with the answers being posted on this site, and how much like pure spam they look. Let's keep in mind here that the name of the site is Software Recommendations, not Software Lists. That means you should be recommending software, not just going to Google, searching for something that matches, and posting a link to it with some vague description or "this does what you want." That's just plain not useful, and some of you might notice that I've cast quite a handful of downvotes already on the main site.

Let's quote a very important line from Tim Post's Area 51 guidelines:

Good answers on this site will be in the form of honest testimonials that share first hand experience with something and why it meets the needs of the asker.

You should answer questions when you have first hand knowledge of things that satisfies at least most of the constraints given in the question. You should be ready to talk about your experience with it, what you liked about it, what features it does (or perhaps doesn't) have, within the context of the question. You should also talk about any potential quirks that you can think of, and even things that you found to be a down side of the product. In short, share your experience with something, not simply your knowledge that it exists.

Now, I don't even need to link answers here because you can browse through just about any question that has an answer so far and see examples of users not doing this.

Some questions to consider answering when you're recommending your software:

• How does it meet the needs of the asker? Don't just say "it has these features" or "it meets all of your needs" - who cares? Describe the features in more detail.

• Why do you, personally, like the product? What other niche features does it have that the asker didn't even think about that make this product so great? Don't limit yourself to the constraints of the question. Remember that the question should be useful to future visitors, who may be looking at other things in addition to those criteria. Sell it to everyone.

• Why is this product the one the asker should choose? What makes it stand out above all the other possible products? You should be taking the stance that this is the only option and you are trying to convince someone to use your product, even though it's not yours (or maybe it is, in which case, provide disclosure).

• Back up your details! If you throw out any statistics about the product whatsoever, for heavens sake have a source!

Of course, there are some other basics that I've seen:

• Include the name of the product. Like seriously, why would you ever post an answer that links to a product and does not ever mention the name of it in the post? I shouldn't even have to mention this...

• Include a link to the product. Just naming a product is like a reverse-link-only answer. We shouldn't need to go to Google and search for the product you mentioned. Again, seriously?

Some of you know I came into this proposal skeptical and assuming it would fail. I'd love to be proven wrong, but what I've seen so far is not very promising.

Answer 1

I agree with the general sentiment, but not with all your recommendations.

How does it meet the needs of the asker? Don't just say "it has these features" or "it meets all of your needs" - who cares? Describe the features in more detail.

Yes, definitely.

Why do you, personally, like the product? What other niche features does it have that the asker didn't even think about that make this product so great? Don't limit yourself to the constraints of the question. Remember that the question should be useful to future visitors, who may be looking at other things in addition to those criteria. Sell it to everyone.

Why is this product the one the asker should choose? What makes it stand out above all the other possible products? You should be taking the stance that this is the only option and you are trying to convince someone to use your product, even though it's not yours (or maybe it is, in which case, provide disclosure).

This is dangerous advice. An answer is not a soapbox for a product. On the contrary, answerers should attempt to consider their product objectively and mention limitations as well as benefits.

Back up your details! If you throw out any statistics about the product whatsoever, for heavens sake have a source!

Yes, definitely.

---

Here's my advice on writing a good answer — a sort of answer template. Not every good answer has to follow that model, but if you're unsure how to write an answer, start with this.

1. If there is any major caveat with the answer that may make it unsuitable in some realistic circumstances, lead with that. For example, if you're recommending a very expensive product, or if you're recommending online software for a scenario where users may not have an Internet connection, mention that right off the bat. (Of course, if the question requests a free tool or an offline tool, your answer would be wrong.) Most answers shouldn't need this bit.

2. Mention the name of the product, linking to the web page that is the best introduction to it. That's usually the project's home page or the vendor's product page, but sometimes the Wikipedia article, a page on the project wiki or some other page may be more appropriate. Link to other resources that are generally interesting, for example third-party Windows binary for an open source project, give the name of Ubuntu/Fedora/… packages, etc. Good tutorials for the audience defined by the question can be nice as well.

3. Go over each of the requirements in the question. One by one, explain how they are met. Some requirements are trivial — if the requirement is “runs on Android” and the product link you give is on the Android market, that's enough. Other requirements may need more details.

   • It's often good to link to the part of the documentation that describes the feature that implements a requirement.
   • If the software and the question use different terminology, match up the vocabulary.
   • If the software needs to be configured in a certain way, explain how to do it (or link to a stable resource that explains it if it's too long, e.g. on the project's wiki page).
   • If a requirement requires third-party add-ons, link to them and explain how to install them.
   • If there are multiple ways to achieve a requirement (multiple features that can be used for the same goal), discuss them and explain how one would pick between them.
   • For nice-to-have requirements, explain to what extent they are met.

4. Mention anything else that's important to know, and that is likely to be relevant to the audience implied by the question. This is not the place to tout how the product is the cat's pyjamas, but to inform the reader about things that are good to know. Be honest: mention downsides as well as upsides. For example: “FooBarSoft has a unique feature which can help you in that …”. Or “feature X can be hard to configure, but once you manage that it can greatly help you by …”.

5. If you think that the requirements are a poor match for the overall goal, state that, and explain how to better achieve that goal. (Most questions shouldn't need this.)

---

Also, I don't think the content from the first couple of days is that bad. I'd say it's on par with most other Stack Exchange sites, not the cesspool that we were fearing. This site needs to do better — it needs to do better than average — but there is definitely hope.

We have a decent volume of questions now, almost 200. Most sites go public with less than that. We should spend the rest of the private beta focusing on quality.

Answer 2

Successfully executing this site's vision will be a matter of raising the bar of quality on this site for both questions and answers. By that I mean, the quality of answers that will be acceptable on this site, in order for it to graduate from beta, will have to exceed the typical level of quality that we have come to expect on other sites.

Posts like this are helpful in that they point out what we're doing wrong, but what we also need to consider is how we can help raise the bar even for users who haven't read this post.

It doesn't help to be cynical and think that raising the bar is impossible. I for one think it is possible. Consider that all of the successful SE sites have raised the bar several kilometers above the typical level of forums, Yahoo Answers, Quora, etc. If it can be raised once, it can be raised again.

The challenge is thus:

• Get information out to the people who are most active on the site and make sure that each one of them understands the level of quality we expect on this site, in particular making sure that each answer produced by our core group of users contains the attributes outlined by Tim in his post;

• Document for posterity in the Help Center the essence of what consists of quality on this site;

• Expeditiously edit or delete questions and answers that do not meet our level of quality, so that users do not imitate or emulate bad content in developing their own content (bad content spreads like a plague, but if the overwhelming majority of the content on the site is good, and people are looking for something to emulate, they'll be hard-pressed to find a bad example to go on.)

When I say "quality", I mean really that our questions and answers need to fit a rather specific format. I consider this to be a higher demand on post quality, because it's less of a free-form discussion as you'd expect in a chatroom or a forum, and more about "know the rules before you post". I use the term "quality" loosely here; things like grammar, post formatting, etc. are less important than the content (aka "meat 'n potatoes") of the post.

Consider just how much learning and preparation a user has to go through in order to participate in each of these:

• A chatroom or topic forum: Pretty much none, except for basic competency in the language being used in the chat. The user will need to understand that they're apt to be banned if they post spam links, links to illegal downloads, or insult other users or use egregious foul language, but other than that, they don't have to go through any sort of orientation before they can participate.

• Most technology-oriented SE sites (SO, SU, SF, etc.): A basic amount of "training" or "orientation", above and beyond what you'd need for a chatroom, is necessary to participate. The natural tendency of most people is to be very free-form, ad-hoc, and casual, and to do the simplest, laziest thing possible that accomplishes the task. For instance, if they're trying to answer a question, the most casual thing a user could do is to post a link to an article or a software download that solves the problem. It may well do it, too, but we know that this is not acceptable on any SE site.

• Software Recs: We need to raise the bar of quality here once more. Not only is it not acceptable to just link-drop and run; the problem is fundamentally worse because the questions that are asked here can often be answered (in the sense of "solves what the OP is trying to do") by just link-dropping. So the naive user will assume that, if what they're posting helps the OP, this is a net-win and they'll get upvotes and everyone is happy. Right? Wrong! It's almost like those trite quotes from various authors about how life is not about the destination but about the journey. Answers on SR are not about providing the link, but they're more about documenting why (ideally, based on personal experience with the software) you think the think you're recommending fits the bill. In a word, for those who read the SO blog, good-subjective answers.

I liken it to the amount of training you need to walk, vs drive a car, vs fly an airplane. Forums are comparable to walking, if SE is comparable to driving a car and SR is comparable to flying an airplane. If this site is going to fly, the pilots need to check all the fluids, the mechanics need to make sure the plane's in good shape, the weather has to be just right, all the switches and dials have to be in the right place, the FAA has to authorize the plane as airworthy, you have to get your pilot license, et cetera.

Let's not be cynical and think that people won't put in the effort. Rather, let's focus on how to get the message out to this site's community, especially new users, that they need to get their pilot's license before hopping in the cockpit and grabbing the yoke.

Answer 3

I see the issue two-fold. First, I totally agree simple link-dropping isn't the way it should go. And yes, as allquixotic put it: ideally it should be backed by personal experience. So far, nobody will argue I guess.

But as with other SE sites, I think we are after collecting knowledge here. So if I've just heard of a good piece of software matching a question, can back my information from an existing source, but did not (yet) try it out myself – I think it would still be of service answering that question as to my best knowledge, linking to more information (if possible), and (of course) including information on what makes me think it's a worthy answer.

For sure, we should "reject" answers containing nothing but "Take a look at X < link>." But we should not frown upon answers just because their author hasn't used the software his/her own – correct me if I'm wrong.

Answer 4

We can vote our way out of crappy content!

In fact, given what happens on all the other SE sites, that's what will happen more-or-less anyways.

Been on SO lately? Almost every new question is basically a thinly-veiled request to send-me-teh-codez-please. And people answer them! Because they get rewarded with points for doing so, and because people like being helpful.

And I, like many of the users concerned enough about Doing The Right Thing to even look at the meta version of the various sites, think "uggh, another crappy question – they're doing it wrong".

But you know what? I still get lots of results from SO when I search for stuff. Hopefully some day I'll get results for other SE sites consistently, but here's why we shouldn't care too much about these Discipline and Purity and Integrity issues:

My search results, from Google or Bing or whatever, and from the SE sites themselves, aren't going to be overly polluted by bad questions (and their inevitable answers).

I mean, feel free to spend your time tidying up a tiny little portion of The Vast Waste of Bad Questions, but me, I've got better things to do. [Like ranting on meta-SE sites!]

Answer 5

The answer by Gilles perfectly explains how an answer on Software Recommendations should look like. Besides this theoretic description, the following list of answers, provide a good impression of what a high quality should look like (keep in mind the guidelines provides in the mentioned answer.

• What pixel graphics program supports creating seamless tiles?

Answer 6

A couple points.

My most controversial answers have been based on first hand experience, but the field is so wide that it is often impossible to ground every aspect of an answer in first-hand knowledge.

As I thought was clear from my answer regarding editors for shell scripts, I had some experience with Eclipse even though I tend to avoid IDE's. The big strength of Eclipse IMO is its plugin ecosystem and that means that if I need (as I rarely do) IDE features that's where I go (maybe a couple days a year). I dont have experience with the specific plugins I recommended checking out but at some point nobody is going to have experience with everything in a recommendation. Heck when I do work as a consultant, I don't have first-hand experience with all the technologies I recommend.

In other words, there is a huge difference between grounding in first-hand experience and being expected to have experienced, first hand, every aspect of the solution. If you insist on the latter, you miss insight that people who have a lot of experience with related technologies who can walk through issues and recommend technologies based on that, and that's very valuable. Indeed, that's even valuable when the specific recommendations of pieces of software go out of date so, for enduring value, that's the sort of answers that we need.

Personally I think this forum is doomed, not because it isn't a good idea but because:

1. The field is frankly too broad for a private beta to produce enough answers. Too many questions are going to go unanswered.

2. There's too fine of a balancing act between shopping-list questions and good questions, and too fine of a line between a really good answer and one that runs amok with the guidelines.

I would like to see this forum be a success, but I think we need some real thought into what makes a QA here to be valuable five years later.

Answer 7

Four points:

1. When a question gets no answers, most askers would prefer link-dropping to something they were not aware of to having no answer at all.
2. Requiring personal experience decreases the probability of a question being answered, until the site has a *^&%load of answering users.
3. There are few answering users. :-(
4. One should distinguish between questions looking for any piece of software for performing a certain task and questions asking for such software which is somehow "good": High-performance, intuitive to use, takes up very little disk space etc - out of a field containing other alternatives as well. I tend to agree with OP about the second kind of questions (well, modulo what I said above), but I disagree about the first kind. In that first kind, whoever asked is obviously not able to locate such software themselves, so - maybe they're just not searching for the right thing, and your link-dropping facilitates what they would like to do.

Answer 8

Yes, we are doing it completely wrong. When this beta will become public and meta will start closing question because they are too broad, too opinionable or too.. users will think: "this meta are suffering of some mental disorder if they close question for being opinionable in a section called "software reccomendation". So it will end very soon.

And bye bye. We should ask for the pro and contro when we do something!!!