A few days ago I posted the question, Revisiting producing structured PDFs from LaTeX. This asked how I could generate a tagged PDF directly from latex. Fairly quickly, the question was flagged as already having an answer here, How can tagged PDFs be created that support Universal Accessibility and reflowing?

In that case, I didn't think that the answer to the original question was actually an answer. The answer didn't contain something that I could use to achieve the goals I set out in my question, and didn't cover the same ground. Unfortunately I think in this case that those reviewers who flagged it as duplicate didn't understand the nuances of the question. I could have done a better job explaining it, maybe?

Flagging new questions as duplicates is particularly problematic when there are new packages that might have come out since the original question was asked, if requirements change with time, or there are fine nuances to a question. (As the aggrieved newbie) I don't think the current process responds to that challenge.

On to the meta aspect of this. I would appreciate some thoughts on the points below. Partly this is me learning how tex.se works, but partly I think that there are some issues with the 'all answers in the same place' model that we're trying to use.

1. How do we make it clearer to people posting here that their question is being flagged as duplicate and give them the chance to fix / respond to that before it gets confined to oblivion? Can we get those notifications in our inbox?
2. If the new question is sufficiently well phrased and differentiated from the original question, surely the new question should remain?
3. How do we encourage knowledgeable people to look again at old questions? Is a bounty of 50 reputation really enough?
4. What happens when the OP has vanished, and the 'original' question is no longer being maintained? Does it really make sense to keep the old question as the main source of answers?
5. Where packages, requirements, or standards, are changing, what about having a back link from the original question to the new question that says

This question was answered in year. A related question was asked here in more recent year which may be helpful.

A few days ago I posted the question, Revisiting producing structured PDFs from LaTeX. This asked how I could generate a tagged PDF directly from latex. Fairly quickly, the question was flagged as already having an answer here, How can tagged PDFs be created that support Universal Accessibility and reflowing?

In that case, I didn't think that the answer to the original question was actually an answer. The answer didn't contain something that I could use to achieve the goals I set out in my question, and didn't cover the same ground. Unfortunately I think in this case that those reviewers who flagged it as duplicate didn't understand the nuances of the question. I could have done a better job explaining it, maybe?

Flagging new questions as duplicates is particularly problematic when there are new packages that might have come out since the original question was asked, if requirements change with time, or there are fine nuances to a question. (As the aggrieved newbie) I don't think the current process responds to that challenge.

On to the meta aspect of this. I would appreciate some thoughts on the points below. Partly this is me learning how tex.se works, but partly I think that there are some issues with the 'all answers in the same place' model that we're trying to use.

1. How do we make it clearer to people posting here that their question is being flagged as duplicate and give them the chance to fix / respond to that before it gets confined to oblivion? Can we get those notifications in our inbox?
2. If the new question is sufficiently well phrased and differentiated from the original question, surely the new question should remain?
3. How do we encourage knowledgeable people to look again at old questions? Is a bounty of 50 reputation really enough?
4. What happens when the OP has vanished, and the 'original' question is no longer being maintained? Does it really make sense to keep the old question as the main source of answers?
5. Where packages, requirements, or standards, are changing, what about having a back link from the original question to the new question that says

This question was answered in year. A related question was asked here in more recent year which may be helpful.

Yesterday I posted the question, Revisiting producing structured PDFs from LaTeX. This asked how I could generate a tagged PDF directly from latex. Fairly quickly, the question was flagged as already having an answer here, How can tagged PDFs be created that support Universal Accessibility and reflowing?

I contend that in this case, the answer to the duplicate question is not actually an answer. The answer does not contain something that I can use to achieve the goals I set out in my question, and I tried so hard to write my question clearly and distinguish it from the original.

So, to the meta aspect of this. Flagging new questions as duplicates is particularly problematic when there are new packages that might have come out since the original question was asked, if requirements change with time, or there are fine nuances to a question. (As the aggrieved newbie) I don't think the current process responds to that challenge.

Can we figure out a way to get the closing process to be a little less final? At the moment the OP has no easy way to get the question re-opened or argue against the closure, as far as I can tell. Did I miss something? Or was I too wrapped up in another question (Replace lstlisting with verbatim depending on a boolean) that I missed a chance to comment before it was closed? Certainly nothing came up in my inbox.

A few days ago I posted the question, Revisiting producing structured PDFs from LaTeX. This asked how I could generate a tagged PDF directly from latex. Fairly quickly, the question was flagged as already having an answer here, How can tagged PDFs be created that support Universal Accessibility and reflowing?

In that case, I didn't think that the answer to the original question was actually an answer. The answer didn't contain something that I could use to achieve the goals I set out in my question, and didn't cover the same ground. Unfortunately I think in this case that those reviewers who flagged it as duplicate didn't understand the nuances of the question. I could have done a better job explaining it, maybe?

Flagging new questions as duplicates is particularly problematic when there are new packages that might have come out since the original question was asked, if requirements change with time, or there are fine nuances to a question. (As the aggrieved newbie) I don't think the current process responds to that challenge.

On to the meta aspect of this. I would appreciate some thoughts on the points below. Partly this is me learning how tex.se works, but partly I think that there are some issues with the 'all answers in the same place' model that we're trying to use.

1. How do we make it clearer to people posting here that their question is being flagged as duplicate and give them the chance to fix / respond to that before it gets confined to oblivion? Can we get those notifications in our inbox?
2. If the new question is sufficiently well phrased and differentiated from the original question, surely the new question should remain?
3. How do we encourage knowledgeable people to look again at old questions? Is a bounty of 50 reputation really enough?
4. What happens when the OP has vanished, and the 'original' question is no longer being maintained? Does it really make sense to keep the old question as the main source of answers?
5. Where packages, requirements, or standards, are changing, what about having a back link from the original question to the new question that says

This question was answered in year. A related question was asked here in more recent year which may be helpful.