17

This was brought up on Area 51, but I don't think it was sufficiently addressed.

Taking into consideration the following:

  • The primary goal of a Stack Exchange site is not to simply provide an answer only to the user who has asked the question—the "mission statement" of all Stack Exchange sites is to build a repository of high-quality Q&A for future visitors. From the Tour:

    Hardware Recommendations Stack Exchange is a question and answer site for people seeking specific hardware recommendations. It's built and run by you as part of the Stack Exchange network of Q&A sites. With your help, we're working together to build a library of detailed answers to every question about hardware recommendations.

  • The majority of traffic to Stack Exchange sites comes from search engines. Therefore, if the questions on a site generally don't continue to be useful for... ever (or at the very least a few years), then your site simply isn't going to work on Stack Exchange.

  • Hardware outdates incredibly quickly. The A51 post I linked mentions that Stack Overflow and other SE sites (most prominently, Software Recommendations, of course) have the same problem, but they simply don't have it on the same scale. Hardware outdates exponentially more quickly. A simple example, from mkomo.com:

    http://www.mkomo.com/assets/hd-cost-graph.png

    And one from preshing.com:

    http://preshing.com/images/integer-perf.png

    Note the log scales on these graphs. This is a problem unique to Hardware Recommendations, even though other sites may have it on a much smaller scale—and if we don't find a really good solution, I have very significant doubt as to the success of this site.

An outdated question is a useless question. This probably goes without saying, but especially in the case of HR, a question such as "what's the best frobble that I can get for under $200 with feature X and specs Y and Z?" that gets answered and then ignored for a few years will become completely worthless. This is the polar opposite of the Stack Exchange mission.

So, what's the solution to this problem? How will the community handle this?

|
  • 1
    I think this is going to be the biggest long term issue that we face. – Red Shift Sep 10 '15 at 4:50
  • @RedShift, Not just on hardware.SE, even on other sites. – Pacerier Nov 8 '15 at 2:59
10

Contrary to the question, actually.

As long as one of the requirements of the question isn't "must be the most modern tech possible", then answers don't actually outdate - think about it.

What's the best frobble I can get for under $200 with requirements X and Y?

The product that fits that specification will in most cases always fit that specification. If the OP hasn't specified that they want the most modern frobble possible, then the frobble suggested in the answer will always be a correct answer. It's not like programming languages, where features are modified and deprecated, and a code snippet that worked once no longer does.

The thing about recommendations is that the Stack Exchange Mission was never really built for them. They're an outlier in the system - and as such, they have different criteria for being useful. They're also incredibly selfish - a recommendation with the kind of specification we require to be answerable is also only ever going to be useful to one or maybe a few people.

What we do need to focus on is keeping advice in questions such as "what should I consider when buying my frobble?" correct. These questions are widely applicable, and the advice may well change over the years. In these cases, it's just a case that we have to edit them as we find incorrect information.

|
  • I can imagine lots of questions that meet this criteria. For example, in the competitive gaming community, lots of old technology is used to lessen latency and because of pure nostalgia (CRT's, etc). – Cilan Sep 9 '15 at 21:52
  • I can see your point about preferring older technology for one reason or another, but that is a special case and might even be a different question. Even though answers still may be technically correct in the more general case, Hardware.SE will lose any edge it could have had over traditional forums where it's difficult to filter out old answers. If I return to this site in 5 years looking for the best frobble, I will be very frustrated if I have to dig through a bunch of answers about frobbles that don't even use holographic storage or quantum entanglers. – rob Sep 10 '15 at 5:07
  • 1
    After rereading your answer, I think it's worth pointing out that the product that fits the specification, "What's the best frobble I can get for under $200 with requirements X and Y?" often does not continue to fit that spec because "best...under $200" will change as technology improves. – rob Sep 10 '15 at 15:22
  • @rob Debatable, because the answer may well still be the best one matching the specification the OP set out. It highly depends on case, I think. – ArtOfCode Sep 10 '15 at 15:27
  • I agree it depends on the case, but again that's a special case in which someone requests a specific dated or limited feature, not the more general case (see my answer for examples). For example, the best 3G Android smartphone under $200 today may continue to be in the top three a year or two from now, but the best Android smartphone in general (or even if you specify 4G LTE) under $200 today will undoubtedly be surpassed by several other sub-$200 Android phones in the next 1-2 years. Most people will list requirements as X or newer/better, not X or older/worse. – rob Sep 10 '15 at 16:15
  • @ArtOfCode, What rob said. While old products would still fit the requirement, they are no longer considered best or even good as time goes on and new products are manufactured. Fast-backward a few years and Windows 98 can be the answer to "What's the best computer for text-editing you can get for under $x with requirements Y and Z?" – Pacerier Nov 8 '15 at 3:01
6

It's not so much the questions that become obsolete, but rather, the answers. Some questions are timeless but the answers to many of these are perishable (like fresh food) and have a limited lifespan during which they are useful. For example:

These questions could be asked over and over again every 6-12 months with different answers. One year from now, any answers that were written today will no longer be the best answers. Simply hoping that newer, then-current answers will be upvoted past the old answers will be an exercise in futility.

I have a couple possible solutions which we could use independently or together:

Add review-again date tags to questions with perishable answers

In the Area 51 proposal I suggested tagging questions with review-again dates, and using these dates to help with performing maintenance on these questions. My original idea was to eventually delete old copies of questions (and their answers), or perhaps just the answers. Now that I'm revisiting this topic, I think there may be a better solution than deleting the old ones and creating new copies, but the review-again-<year>-<month> tag could still be helpful in identifying which questions have answers that need to be refreshed.

Convert questions with perishable answers to community wikis

Another idea I suggested was to convert every one of these timeless questions (whose answers will rapidly become obsolete) to a community wiki. Originally I had a different motivation, but my current thought is that newer recommendations can be added at the top of the accepted community answer, pushing the older answers to the bottom. Including a last-reviewed or next-review tag would help with identifying questions with stale answers, as I mentioned previously.

|
2

The cost of desired "frobbles" (under $200 in your example) is a big variable that will likely change over time. So the pool of included devices will change too.

However specs. of a specific "frobble" probably won't change that much.

Cost might be referred to in an answer (for consideration), but should not be a criteria in the question.

|
2

Arbitrarily pick a time period you can re-ask questions by (without being closed as duplicates).

Also have a custom close reason which is something like, "This question was deemed to be relevant only for an [arbitrary period] of time. Answers to this question may no longer be useful/relevant."

|

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .