Skip to main content
Triage
Source Link

A suggestion to keep technical Q&A's relevant as standards/specs Re: tech questions where correct answers change over time - is there a way to show only the votes after a certain date?

This isNote: In an old topic, e.g. A better answer was posted long after accepting another. Should I accept the new one?. Although there are useful features such as ordering by 'activity' and 'oldest' (... btw, shouldn't that be 'newest'?), I feel that I am heuristically estimating what parts of answers and votes same before some relevant cutoff date - and that that estimation is error prone and time consuming. It's not helped by the fact that votes are probably clustered timewise around the timeact of the original post.

Many answers/discussions ontriage, this topic seemquestion was retitled and edited to focus on changing/not-changing accepted answers/votes, but in fact that rarely happens, or only sometimes happens.

My suggestion is a graph of votes over time,become a slider on the time axisrequest for technical support instead of that graph, and a slider response which filters all the votes earlier than the specified time, and then lists the answers in the order of remaining (newer) votes. That might be a server CPU killer if done mostly on the server side, but perhaps it could be pushed mostly to the client sidefeature request - just sending an array of dated votes, and rearranging answer order and vote count on the fly with ajax (not my specialty but I am aware it canyet still be done).

EDIT: As @JourneymanGeek♦ mentioned in their comment, older answers are sometimes updated, and thus comea question to match the top in activity. Sometimeshigh quality answer that is useful, but sometimes the update is not (especially) useful, but due to earlier votes it looks "20 x better" thanwas given. In case of a more recentquality answer which is actually more relevant (this is my subjective experience). If only we could select to view the more recent votes, that new weighting would become apparent. That is the motivation for the proposed solution.

It would be interesting and informative to get more feedback with the downvotes. E.g.,SE urges not deleting the question -

  • disagree with my above described subjective experience, so no need.
  • it's unfair to those posted long ago
  • perhaps rarely useful but not enough to warrant change

If your question has good answers, though, it's not fair to have those answers removed along with your question: other users put effort into helping you and even if you no longer want the answers, somebody else might. This is why the system prevents you from deleting answered questions most of the time.

A suggestion to keep technical Q&A's relevant as standards/specs change over time

This is an old topic, e.g. A better answer was posted long after accepting another. Should I accept the new one?. Although there are useful features such as ordering by 'activity' and 'oldest' (... btw, shouldn't that be 'newest'?), I feel that I am heuristically estimating what parts of answers and votes same before some relevant cutoff date - and that that estimation is error prone and time consuming. It's not helped by the fact that votes are probably clustered timewise around the time of the original post.

Many answers/discussions on this topic seem to focus on changing/not-changing accepted answers/votes, but in fact that rarely happens, or only sometimes happens.

My suggestion is a graph of votes over time, a slider on the time axis of that graph, and a slider response which filters all the votes earlier than the specified time, and then lists the answers in the order of remaining (newer) votes. That might be a server CPU killer if done mostly on the server side, but perhaps it could be pushed mostly to the client side - just sending an array of dated votes, and rearranging answer order and vote count on the fly with ajax (not my specialty but I am aware it can be done).

EDIT: As @JourneymanGeek♦ mentioned in their comment, older answers are sometimes updated, and thus come to the top in activity. Sometimes that is useful, but sometimes the update is not (especially) useful, but due to earlier votes it looks "20 x better" than a more recent answer which is actually more relevant (this is my subjective experience). If only we could select to view the more recent votes, that new weighting would become apparent. That is the motivation for the proposed solution.

It would be interesting and informative to get more feedback with the downvotes. E.g.,

  • disagree with my above described subjective experience, so no need.
  • it's unfair to those posted long ago
  • perhaps rarely useful but not enough to warrant change

Re: tech questions where correct answers change over time - is there a way to show only the votes after a certain date?

Note: In an act of triage, this question was retitled and edited to become a request for technical support instead of a feature request - yet still be a question to match the high quality answer that was given. In case of a quality answer SE urges not deleting the question -

If your question has good answers, though, it's not fair to have those answers removed along with your question: other users put effort into helping you and even if you no longer want the answers, somebody else might. This is why the system prevents you from deleting answered questions most of the time.

EDIT to clarify problem and ask for downvote feedback
Source Link

This is an old topic, e.g. A better answer was posted long after accepting another. Should I accept the new one?. Although there are useful features such as ordering by 'activity' and 'oldest' (... btw, shouldn't that be 'newest'?), I feel that I am heuristically estimating what parts of answers and votes same before some relevant cutoff date - and that that estimation is error prone and time consuming. It's not helped by the fact that votes are probably clustered timewise around the time of the original post.

Many answers/discussions on this topic seem to focus on changing/not-changing accepted answers/votes, but in fact that rarely happens, or only sometimes happens.

My suggestion is a graph of votes over time, a slider on the time axis of that graph, and a slider response which filters all the votes earlier than the specified time, and then lists the answers in the order of remaining (newer) votes. That might be a server CPU killer if done mostly on the server side, but perhaps it could be pushed mostly to the client side - just sending an array of dated votes, and rearranging answer order and vote count on the fly with ajax (not my specialty but I am aware it can be done).

EDIT: As @JourneymanGeek♦ mentioned in their comment, older answers are sometimes updated, and thus come to the top in activity. Sometimes that is useful, but sometimes the update is not (especially) useful, but due to earlier votes it looks "20 x better" than a more recent answer which is actually more relevant (this is my subjective experience). If only we could select to view the more recent votes, that new weighting would become apparent. That is the motivation for the proposed solution.

It would be interesting and informative to get more feedback with the downvotes. E.g.,

  • disagree with my above described subjective experience, so no need.
  • it's unfair to those posted long ago
  • perhaps rarely useful but not enough to warrant change

This is an old topic, e.g. A better answer was posted long after accepting another. Should I accept the new one?. Although there are useful features such as ordering by 'activity' and 'oldest' (... btw, shouldn't that be 'newest'?), I feel that I am heuristically estimating what parts of answers and votes same before some relevant cutoff date - and that that estimation is error prone and time consuming. It's not helped by the fact that votes are probably clustered timewise around the time of the original post.

Many answers/discussions on this topic seem to focus on changing/not-changing accepted answers/votes, but in fact that rarely happens, or only sometimes happens.

My suggestion is a graph of votes over time, a slider on the time axis of that graph, and a slider response which filters all the votes earlier than the specified time, and then lists the answers in the order of remaining (newer) votes. That might be a server CPU killer if done mostly on the server side, but perhaps it could be pushed mostly to the client side - just sending an array of dated votes, and rearranging answer order and vote count on the fly with ajax (not my specialty but I am aware it can be done).

This is an old topic, e.g. A better answer was posted long after accepting another. Should I accept the new one?. Although there are useful features such as ordering by 'activity' and 'oldest' (... btw, shouldn't that be 'newest'?), I feel that I am heuristically estimating what parts of answers and votes same before some relevant cutoff date - and that that estimation is error prone and time consuming. It's not helped by the fact that votes are probably clustered timewise around the time of the original post.

Many answers/discussions on this topic seem to focus on changing/not-changing accepted answers/votes, but in fact that rarely happens, or only sometimes happens.

My suggestion is a graph of votes over time, a slider on the time axis of that graph, and a slider response which filters all the votes earlier than the specified time, and then lists the answers in the order of remaining (newer) votes. That might be a server CPU killer if done mostly on the server side, but perhaps it could be pushed mostly to the client side - just sending an array of dated votes, and rearranging answer order and vote count on the fly with ajax (not my specialty but I am aware it can be done).

EDIT: As @JourneymanGeek♦ mentioned in their comment, older answers are sometimes updated, and thus come to the top in activity. Sometimes that is useful, but sometimes the update is not (especially) useful, but due to earlier votes it looks "20 x better" than a more recent answer which is actually more relevant (this is my subjective experience). If only we could select to view the more recent votes, that new weighting would become apparent. That is the motivation for the proposed solution.

It would be interesting and informative to get more feedback with the downvotes. E.g.,

  • disagree with my above described subjective experience, so no need.
  • it's unfair to those posted long ago
  • perhaps rarely useful but not enough to warrant change
Source Link

A suggestion to keep technical Q&A's relevant as standards/specs change over time

This is an old topic, e.g. A better answer was posted long after accepting another. Should I accept the new one?. Although there are useful features such as ordering by 'activity' and 'oldest' (... btw, shouldn't that be 'newest'?), I feel that I am heuristically estimating what parts of answers and votes same before some relevant cutoff date - and that that estimation is error prone and time consuming. It's not helped by the fact that votes are probably clustered timewise around the time of the original post.

Many answers/discussions on this topic seem to focus on changing/not-changing accepted answers/votes, but in fact that rarely happens, or only sometimes happens.

My suggestion is a graph of votes over time, a slider on the time axis of that graph, and a slider response which filters all the votes earlier than the specified time, and then lists the answers in the order of remaining (newer) votes. That might be a server CPU killer if done mostly on the server side, but perhaps it could be pushed mostly to the client side - just sending an array of dated votes, and rearranging answer order and vote count on the fly with ajax (not my specialty but I am aware it can be done).