In the early days of Photo-SE, there were lots of questions like “What books should a new photographer read?” or “What’s your favorite website for learning about photography?”
These questions are kind of a Q&A engine trap — it seems like the vote-on-answers format would be well suited for this, but it turns out to really not. Answers can’t be easily marked as duplicates of each other, there’s no way to enforce the format, and, crucially, ongoing maintenance is hard. So, most of those early questions turned out to be a mess and are now closed, locked, or deleted.
But, I think these things really could add value! In the background, everyone on the site was thinking “Michael Freeman’s books are really great!” and “Well, of course, The Online Photographer, and If you want to learn about lighting, the Strobist 101 blog series is a must-read” — but we had no good place to put that.
I’m envisioning a type of post where:
Answers are sorted by votes but there is no solution
Answers are of a more limited length
You don’t get “points” for answers (but maybe you do for curation actions)
There are tools for merging duplicate answers
Some higher level of trust is required to answer
These posts would be specially identified, and could be sorted and filtered separately. Keeping them more prominent would help one of the other problems (bit-rot, basically).
I think this would address a lot of the problem with these kind of posts historically
and be pretty useful for a lot of situations.
Except for the (significant) difference that the proposed “Canonical Post” would not have separate Q vs. A but rather all be in one community-edited post, I think that would actually work well. It would not provide “answers sorted by votes”. But it would have:
No accepted answer (aka solution)
Answers of a more limited length - arguably, but flexible, since answers would just be sections of the post
No points
No need to merge - just edit.
Higher level of trust to start a canonical post (and possibly to edit)
“What books should a new photographer read?” definitely sounds to me like an appropriate Canonical Post.
Yeah, Canonical posts might do, but community-wide sorting and validation would be nice. Here, the various options aren’t necessarily facts or agreed-upon by everyone. For example in Photography, many people think Brian Peterson’s Understanding Exposure is a great introductory book — I happen to think it’s terrible and misleading. With a wiki-like canonical post, how do we resolve that? Hopefully the answer is not “edit wars”. With an answer-voting system, we can let the wider group sort it out.
I would approach that problem as if I were writing a page on Wikipedia: neutral point of view (NPOV), describing each option with its positives and negatives.
Hmmm — actually, maybe what I’d really like here is a special index page for collections of discussion topics! The topic could be something like “Books on Photography”, and everything with a matching tag (or other classifier) could be shown under that in a vote-ranked order, and then people would be free to click through to discuss.
I’d like to point to https://slant.co as an example of the format that works better for the “what is the best…” type poll. That even happens to be the prompt in their search. For example, best cameras for a beginner or java IDEs.
The issue isn’t so much “they aren’t useful questions” but rather “in the Q&A format, they can produce rather poor answers.” Consider Windows equivalent of the Unix tail command on SO and look at how many times cygwin is listed, or GnuWin32, or the powershelgl one liner… or out of date systems that no longer work.
Its not so much “slant is great, it is so much better” but rather “if these questions are to be allowed, it prevents something from being so much worse and helps format the material so that its once and only once and allows for updating/outdating.”
Hmmm. I like that conceptually, but I’m not sure it’s working in the example you give. Part of it is the UI — I find the relation between the list at the top and the entries confusing. (And “screen size” and “processor” are the weirdest details to put front and center for a camera!)
Most importantly, it’s a really poorly curated list. The top recommendation is an entry-level model from 2014 which isn’t even available new anymore (and the list overall is a kind of random assortment of entry-level models from the last five years).
Unfortunately, this is a big problem with these types of posts. The “which is best” or “what should I buy” questions inevitably result in obsolete content over time, and are often very opinion-based in the first place. You can try to make answers subjective, but that’s really hard, doesn’t happen often, and isn’t even possible in most real world cases. Cameras are a great example. Many features are personal preference and can’t be evaluated as better or worse. Should the f-stop wheel rotate clockwise or counter-clockwise to stop down? Some may have strong preferences either way, but neither is inherently better than the other.
So while I see the attraction to what you are trying to do, I’m not sure it’s ever really going to work well in any on-line format. SE’s solution was to just ban these types of post. That may in the end turn out to be the only workable answer.
I think in particular, “what camera is best” is terrible — the list changes too fast and it’s really too broad to be meaningful. However, “great books for beginner photographers” might be different.
Sounds like such posts (even if nicely organized) will age poorly, resulting in broken links and obsolete recommendations. It also seems like a popularity contest, rather than expert opinion. These questions will probably need ongoing maintenance.
At Chinese.SE there’s a horrible mess of old answers giving highly upvoted, now-broken links.
I think “List of good things for X” are a good thing if approached in a readable way. And this should definitely be a separate section of the site from the main Q&A to not clutter it if it turns out it’s not very useful to see on the main page, or difficult to maintain this portion of the site, and it should be hidden from Q&A viewers.
Simply listing off items is not great and should be discouraged, but sometimes it might be better than nothing, so I don’t think such answers should be deleted.
Ideally there should be lists of Pros and Cons abstractly, and a short comparison to other options on the list.
One example I know are sites like “Alternative To” https://alternativeto.net/software/adobe-photoshop/. There you can add a product/site/thing and add Pros/Cons to it and compare it with other items on the same list.
Wikipedia has tables with columns of various properties and features, for example https://en.wikipedia.org/wiki/Comparison_of_raster_graphics_editors which is a data heavy page, and I don’t really know how to improve this. But again, it’s better than no such data or data without structure.
And there also needs a way to make it easy to keep the lists up to date or make it clear that the current items might be outdated. Maybe promoting more recent additions to the top for new visitors to see and vote on.
As much as I am in favor voting on Q & A, etc., I think the “popularity contest” model of “lists” is a bad fit for Q&A. My understanding of one of the original goals of SE and now of Codidact is to be a repository of knowledge. Popularity contests are not a “repository of knowledge”.
Lists can be a repository of knowledge if they are either done really well or are not as time-sensitive. Done well, these can answer a lot of otherwise “another one of those…” questions. For example:
List of various travel-related things - e.g., visa & passport requirements for country A to country B (I occasionally look at Travel SE and see a lot of those questions). Time-sensitive (the rules change) but not product-related.
List of photography books - again, key is to not make it a popularity contest but rather a curated list of “here are classics that have stood the test of time and/or are reasonably objectively useful for learning the topic material”
List of basic electrical (or other) DIY instructions - how to replace a receptacle, how to wire a 3-way switch, how to wire a GFCI receptacle, etc. A couple of us tried to do this on DIY SE but because there is no way to make a Canonical post that “sticks around at the top” it doesn’t get seen, yet the questions it would answer keep coming.
etc.
There are many more I can think of. But the key is that while they are “lists”, they are not popularity contests for products. Even free software has that problem - a list of “best” Linux distros tends to devolve into a popularity contest and becomes nearly useless. OTOH, an actual curated comparison article explaining the differences between distros, could be very useful.
Authorship is important, but not the key, IMHO. I think a Canonical post should show the original author and perhaps a list of people who contributed more than ‘x’ % (based on # and size of edits). But showing authorship is not a reason to turn it back into Q&A (each A having an author) because the format of “one coherent curated post” is much better, IMHO, for Canonical posts then a bunch of separate answers.
Yes, this. I think the model for a good “list post” is a Wikipedia page that lists and discusses options/products/books/whatever. We can show authors’ contributions somehow, but the value is that the information is organized and presented consistently. Sometimes there are even comparison tables. The focus should be on what the consumer of this information needs to make decisions, not on what works best to manage users’ individual contributions.
With an answer-voting system, we can let the wider group sort it out.