I tend to agree. Unless, perhaps, it’s a sticky thread, in which case it may be preferable to keep it open.

I tend to search answers to my questions, and many times it leads to posts in forums like this. So, I think that it is good to keep the posts, so others can find answers to their questions without even asking it…

Yes, I was talking about keeping the posts (absolutely!). Just closing the threads so people cannot answer to a (very) old thread, but they’re forced to start a new one.

It sounds reasonable to me, I’m not sure if 3 months is ok or if the time should be a bit longer. Forum admins should be able to disable the automatic close for epic threads, though.

I disagree. I think that it is reasonable to show a warning that you need to click away before you get the Post a reply box, but I think topics should not be automatically closed. Sometimes it is just reasonable to follow up on a topic after a longer time. How about the case where somebody would like to give feedback on the SimpleCrypt class I published on the wiki? I put a link to a forum topic for that in the wiki page. Why should such a topic close all of a sudden because it has been three months since the last comment there?

Personally, I don’t mind people posting in old topics, as long as it is on the same issue, not to ask a new but related question. I find necroposting less of a problem than people just dumping questions without searching. Necroposting at least shows that people were using the search facilities to try to find an answer first. That is commendable.

I totally agree wit Andre’s point of view.

IMO there is nothing wrong with answering an old post with a solution or interesting new development info.
Sometimes topics don’t get solved and can be later on.

Necroposting is only annoying if it doen’t add anything to the original post or additional questions are asked. In the latter case it is recommended to start a new one as Peppe suggested

I think this is difficult to handle automatically. Isn’t that what moderators / admins are for ? ;)

Eddy wrote:

I think this is difficult to handle automatically. Isn’t that what moderators / admins are for ? ;)

We can spilt threads, but if someone does necroposting, it happens. But I also agree with Andre. I would not close those threads. What could be thought of that solved threads are closed or thus that came out of some misunderstanding. But that would need to be done by hand.

Perhaps we could mark those thraeds with a closable tag and check after some time. If the tag is still there, it can be closed.

Ok, so basically a warning next to the reply textarea would be ok? Something like “if you’re adding relevant informations to the thread please go on, otherwise please <link>start a new thread</link>?”

Sounds like a good idea.
Or if you click post —> open a new window which states:

you are posting to an old thread. do you really want to continue instead of creating a new one?

Yes, that popup warning idea sounds good since most people don’t read all static text besides a topic.

I hope it is easy to implement. Shouldn’t we make a jira suggestion and vote for it?

Done: QTWEBSITE-303 [bugreports.qt.nokia.com]

Thanks guys, makes sense to me. I’ll add it to the road-map.

Great, thanks.

A forum I am also on from time to time uses this approach:
Topics have what they call a “quick reply” area that is basically equal to the reply area that you have at DevNet. However, it also has a more advanced editor for posts, that is on another page. For topics that have been silent for more than two weeks, the quick reply is disabled, and warning is displayed instead that it might be a better idea to post in a new topic.

Another additioal feature that it has, is that there is a warning at the top of the quick reply area if you were the last user to post in the topic. In that case, it requests you edit your last comment instead (though it does not stop you from making a new comment, replying to yourself).

I rather like that setup, though you don’t actually need the separate page for posting. You can just hide the editor area and replace it with a message, and only show it again after an explicit action.

The situation described in the last post of Andre sounds very reasonable. We must make sure that a user who has entered some text somewhere is not urged to re-type it because of some warning coming into the way.