Should we be allowed to resurrect years-old threads?

Perhaps it would be best to lock posts after a few months?, a year?

Resurrecting such an old post is probably not good way of continuing the conversation. It might be best to start a new thread and reference the old one. What do you think?

4 Likes

Iā€™m not sure fully locking is the right answer. Maybe an are you sure? Prompt

1 Like

Weā€™ve considered doing this in the past, but as far as Iā€™m aware the only auto-lock option is currently on ā€œsolvedā€ posts. I donā€™t think thereā€™s a setting to do it automatically in general.

Personally, Iā€™m torn on this one ā€” and in my view the tie-break goes to the lower-effort and less-intrusive option. We donā€™t experience necroposting all that frequently, and it can indeed sometimes be very helpful and appropriate updates/corrections/addenda.

Iā€™ve been a member on other discourse boards that were far more aggressive in locking posts and Iā€™ve found them to feel much more stilted and unfriendly. Your feelings may vary, of course.

13 Likes

Would it be possible to show a message instead, if someone tries to reply to an old post? That way, we could still nudge new users not to bump really old posts.

12 Likes

It would also be nice to print a banner at the top, to save readers from puzzling over the first postā€™s question before seeing the date. (Could even be on all old posts, not only resurrected ones.)

For the most part, I think it is just new users that do this (perhaps even more so those unfamiliar with forums in general). I donā€™t think locking is really a great way to go about this, but I agree that a message would be very useful as a nudge.

ā€œAre you aware that this post is over 2 years old, which is equal to 100 Julian years?ā€

4 Likes

That warning already appears.

Iā€™ve tried to get the Discourse software to support putting a banner across old threads, but I got a negative reaction from the devs there about that.

8 Likes

Ha! Evidence that nudges arenā€™t perfect. Not sure much more is needed, makes it even more likely that my first point is correct (that the worst offenders are probably not familiar with basic internet forum etiquette). Nothing terribly wrong with that, but maybe it would be a good policy to just steer them to make a new topic rather than continue any discussion in the existing old thread (which I think is sometimes informally done anyway).

1 Like

On a related question, do you think that Discourse provides any feature to automatically tag new discussions with the current Julia release? That could help with ranking search results, and perhaps also help with avoiding resurrection of old discussions.

2 Likes

I think in some forums the policy is to never have more than one thread on a given topic.

I have defiantely seen people who act like creating a new thread (or new issue on github for that matter) has much greater cost than adding a comment to an existing one ā€“ even if it pulls a dead thread in a new direction

4 Likes

How about an option for users to automatically ā€œunfollowā€ threads when they have been inactive for a few months? Users with a strong dislike of necroposting would then no longer be bothered, while others can continue the discussion should they feel like it.

4 Likes

Is it possible to link this to trust level, so that you have to be ā€œBasicā€ or maybe ā€œMemberā€ to open old threads?

2 Likes

As a beginner, I see the Julia discourse posts as more than just ā€œconversationsā€, but primarily as repositories of knowledge and solutions. I see many valid reasons for their resurrection including: (i) the code is not working on a new version of Julia or on a different operating system; (ii) the solution was well understood by the experts but not by the beginners who need additional explanations; (iii) There is always a different angle of attack on a problem or related questions that might not have been discussed. Thank you.

21 Likes

I hate when some forums (such as Reddit) automatically close threads with unanswered or even wrong answers. You find them and want to reply but itā€™s impossible.

As others have pointed out, many times the thread is still valid, and creating a new one with the same content is very inefficient.

Maybe we just need to tag the thread with the affected Julia versions and packages. That way we wonā€™t be reading a Julia v0.6 threads if we need to solve a problem related with v1.5.

7 Likes

I agree with @mbauman, @rafael.guerra and @Juan that there are often legitimate reasons for adding posts to old threads. Absolute solutions, like automatic locking or unsubscribing participants, fail to serve these valid use cases, for little benefit (eliminating a very small amount of thread ressurrection that currently happens).

I think the ā€œRevive this topic?ā€ notice is the right way to address this problem, and for all we know (and countering @tbeasonā€™s suggestion), the fact that we do have a small amount of thread revival may well be evidence that the nudge works well ā€” though I concede, it isnā€™t 100.00% perfect. :slight_smile:

If anything, Iā€™d suggest tweaking that notice to add that, in addition to notifying a lot of people, reviving an old thread that has dozens of posts is more problematic (as it may require a lot of context-gathering) than an old one that has only a few. But thatā€™s just my 2Ā¢. :wink:

1 Like

People resurrect old topics only sporadically on this forum, so perhaps the warning is sufficient. No system is 100% perfect, and a little noise is OK to deal with occasionally as long as it is not overwhelming.

So I would just leave the setup as it is now.

9 Likes

It could be done in such a way that only those who have posted at least 5 times could reply to old threads.
They will likely know the rules and only resurrect if they have good reasons.

And I still think the best option is to tag each thread with the related version (or range of versions) of Julia/packages, or tag it as generic.

3 Likes

I have resurrected old posts twice now. They were directly related to the problem I had and yet the issue had not been fully resolved.

I have been on this discourse forum for over a year now.

I post on these ā€œoldā€ threads because I feel there is a better chance of someone helping address my problem. After all all people on the thread were genuinely interested in the problem and likely have more experience than I, to help resolve the issue.

Time should not be the factor, but whether the issue is solved as @rafael.guerra points out . Why start a new thread when an old one that is exactly on the same issue remains unresolved? At least this is not how I organize my notes.

Whenever I get reminders from posts that I am no longer interested in, I simply unsubscribe from that line of conversation.

A reminder like ā€œThis thread has been inactive for X years. Reviving this thread may not ā€¦ā€ and for ā€¦ fill in with why one would be against reviving old threadsā€“of which I donā€™t see why unless subject matter is very different.

3 Likes

good point.
I think its a bit of a compounding thing.
Its when someone post on a old thread something that is not closely related to the topic of that thread that is is problem.

I think there is kind of trade-off surface.
The older the thread, the more ontopic a new post should be to resurect it.

but a lot of threads donā€™t have clear topics, and so nothing is surficently on topic to make it good ro resuract them.
and a reason they donā€™t have clear topics is because when they are still fresh it is ok to post stuff not too closely related.


Though the other problem is that things might have ceased to be relevent.
While thigns might look similar other things have changed in the between time rending a lot of the older commentry misleading

1 Like

I kind of liked my own suggestion of tying this to trust levels, to avoid typical ā€˜drive-byā€™ resurrection of ancient threads that is sometimes done by people who apparently sign up just for that. Not sure if itā€™s possible.

3 Likes