flushing a long useless thread
 

Sometimes, people just take the wrong approach to a problem, perhaps due to not understanding it precisely, and wrong suggestions get made. Then when it leads to a long thread with no real end, and particularly when the people replying really didn't have the answer (once they realized what the real issue is), it usually means no other answers will be coming. I know that even for things I know the answers for, if I see a thread is long, I don't bother reading it (takes to long, usually it is answered, and if it isn't, that's hard to see). I focus on the short threads, especially with no replies (if I believe I might know the answer).

But what is one to do when a thread they started gets long with no answer, to encourage answers to be tried? Should a new thread be started? Or is there a way to flush a thread and reset it (start over)?

Great question.

The most important things to focus on IMHO are what we're trying to achieve, who we do it for and how to succeed. Good troubleshooting in my opinion starts by asking questions (OK, and being methodical about things yourself). Only once you have gathered information and analyzed it you make good suggestions. Good troubleshooting may also include providing structure. This may benefit new Linux users, who may not know their way around the system, and those who read the thread later on. At times it may be important to take stock of what's been posted and summarize things to show the OP where he's at. Then asking for missing links or pointing out suggestions not tried yet may provide the impetus the thread needs. (At times I may email an OP asking for him/her to revisit the thread to reply or provide information. You don't have to but please feel free to do that if you care.) If a new thread should be created, which I think should be done sparingly, then the OP minimally should summarize what advice was offered, what he/she tried, the outcome of that (in a way that makes sense so error output and not "doesn't work") and provide a link to the previous thread. IMHO in the end what makes the difference is your own stance towards troubleshooting, if you're willing to stay with the thread and how much time and effort you're willing to invest.

Hmm. It's rather usual (for me) to ask the wrong question in the thread's title. When I then learn that the problem might be something else, I start a new thread with the new question (and so on...). When the problem is solved, I post the link(s) to the previous thread(s) (or I cross reference them) to show the reasoning on the way to the solution. That way the effort of the posters in the other thread is acknowledged, too.

Cross-referencing threads is an excellent way to provide background information. However the act of doing that itself does not (very often) justify creating new threads IMHO. People can handle only so much fragmentation (OPs forgetting to revisit threads too) and some tend to create new threads just to attract attention...

So if a new thread is justified, cross referencing a related thread you previously started would be a good thing (to fill in the background, perhaps). But the fact that you can cross reference doesn't make for that justification. Maybe one way to consider it is, if the original question was all wrong, and is reflected in the title that wrong way, then maybe a new thread is justified just so it is usable for others to search for. Then they can find the answer via the right question.

But what about the case of a thread where no answer has yet been determined, but the thread, by virtue of its great length (such as because of many suggestions that are not leading to solutions) is no going to attract as much attention from those who might know the answer? The case I'm interested is more where someone tries to answer, and the answer may well even be right for some question, but isn't a solution for the original problem. One possibility is the question was misunderstood. Or maybe the question was incorrectly or badly phrased.

A thread can be marked SOLVED. Too often threads are not so marked even if solved, for whatever reason. Maybe a special flag/mark for "this -old- thread still needs to be solved" might be one way to help.

I didn't mean trivial misunderstandings. I asked lately about a problem with the hardware clock. In the end it had to do something with creating a new intrd. Now that is not immediately or self evidently connected IMHO ;) so I created a new thread with what I thought to be a question more to the point (which led to the resolution in the end).

Hi,

Interesting topic.

It does depend on the experience of the OP, most times the user doesn't provide detail or compose the query in understandable terms. Sure, some of this is attributed to newbieness, fear from poor communications skills or just down right laziness. To break into another thread would just cause more confusion for this type of thread.

Sometimes we must guide the OP to extract or request usable information to lead to the solution by diagnosis of presented information. Thus extending the thread which can be a frustration on all parties. Then someone jumps in with a far out suggestion and the thread starts spinning. As a participant, we should try to bring it back on track. If the OP creates another related thread then that too would not be justified since things can again be fragmented thus more confusion.

The forums here at LQ are members from around the world. This world community has different societies that sometimes don't communicate easily. We use English as the required posting language. Members that use English as a second language don't always have the communication skills with English to get the required details to resolve a problem. I know at times my misinterpretation of a post has caused trouble for parties involved because of simple grammar or abbreviate text errors from the original post. I now try to include links for definitions of words that seem to cause people problems. Hoping the definitions will help the parties to understand my meaning thus able to discern the description by detail or diagnosis of the problem by understanding the content presented. If the parties or participants communicate via PM then that will sometimes solve the isse but the thread will not benefit. This has caused the opening of a new thread and is not necessary.

I do like to open doors for a OP so that person will be able to solve or work to the solution of the problem. At times it will be referenced links for the poster to review or read for understanding so that they can get to a solution. Sure, some effort on their part but these forums are not 'Help Desks'. Users are responsible to get things done. Sometimes they too will create another thread because of the errant misgivings when the poster should remain within the original and possibly back track a bit.

LQ members participate as volunteers, yet sometimes the OP expect direct resolution or implies that detailed information or solution is a must. I've seen to many threads of this type and sometimes I'll just unsubscribe when the thread is of this type. My mail box is filled with notifications that are just down right wrong or silly.

Sometimes the post(s) do not break things down to the simple terms that are understandable. When they do extend from this point, they think or imply that information is not related. I will see loads of new threads created from this situation. When they should remain within the current thread because of related detail. Poor communication has caused more alien threads from the original because of injection by errant posters within a thread.

I wish that a flusher/strike through was allowed at times but then why would we have moderators.

As I said, interesting thread and thoughts.
:hattip:

That's one way to say this self-inflicted problem stemmed from you loading a custom initrd instead of a stock one :-]

In hindsight there's room for improvement: your first thread http://www.linuxquestions.org/questi...r-boot-844173/ started and ended on 14-11-10 but no link was added to your newly created http://www.linuxquestions.org/questi...reboot-844338/. Your second thread got created on 15-11-10 and ended on 16-11-10, but only just after you created the new http://www.linuxquestions.org/questi...utdown-844592/ a few minutes before and again without adding a FUP link. I disagree about creating a new thread while the old one is still active as this may lead to duplicate efforts on your fellow members behalf. Not adding follow-up links makes it harder or impossible for people to follow you and when and where they do they have less information than they are entitled to. I hope you can see those points. Now when I say there's room for improvement I do realize we're not all information management specialists so it's not like I'm chiding you for doing things the way you did it. It's fab you added related threads and it's grand things got fixed successfully. (BTW the last thread IMHO is still incomplete as it lacks a listing of modules that were missing from the initrd.)

A new title can be edited in (report the Original Post if you think you can't) and the OP could be appended to (not changed, especially when there are replies). OTOH if that leads to a complete mess then yes, a new thread would be in order.


There's at least two aspects here. The first one should invoke the "complete mess" clause w/o probs. The second one, not attracting attention from knowledgeable people, is something I have not answer for because interest in certain sub fora or topics and reading behaviour and perseverance vary per person.


The problem with the "solved" marker is that having it does not guarantee that the problem actually is solved (or solved in a way that is useful) and not having it doesn't mean the thread isn't. It's not like we have nine out of ten years worth of unsolved threads...

Thinking about the special flag/mark I can't see any single solution:
- You can always report a post and ask a moderator to bump it to the top of the forum. Sure that isn't much but it's a start.
- If you encounter a topic you knows somebody else is knowledgeable about you could always PM or email the member and ask them to have a look.
- You could web log about a certain thread to give it exposure.
- You could even organize yourself on LQ and group together as "Most Honourable Defenders of the SATA Bus" and between you tackle all SATA-related problems.
It certainly would be interesting to brainstorm about this some more as LQs power relies on both individuals and groups of individuals.

Tsk tsk :D. The problem stemmed from a crashed system update which killed the kernel and initrd. Those kindly supplied but since with the originals replaced items saved my (still running) system.

Huh? There is my post dated 18-11-10 with those links :scratch: ...


It doesn't. Please scroll down to my last post dated 18-11-10.


Here you are right, I didn't give back-references to the other threads (which where centered on the wrong topics in my regard). I Amended this right now.

Well, members can't close threads and I wasn't sure at all whether the last thread would solve so tenacious a problem, thus I linked the other threads two days later with the solution at hand. But I do see your points.

Please don't forget, that we aren't all linux specialists; I was sorely pressed to get that system up and running since it held my wife's data.

Lastly, I can't add the information about the modules, that is knowledge of udaman which I don't share.

Hi,

It's even happened here!

See how the drift occurred? Then the resultant.

:hattip:

Even experienced people can have difficulty with something new. Sometimes just a new way of using something can be problematic. For example my current issue (do not answer this in this thread) involves DHCP which I have used before only to serve up dynamic IP addresses. But now I need to make DHCP serve up PXE references to hosts asking specifically for that while co-existing with a separate existing DHCP server that serves up just dynamic IP addresses. The documentation is found to be incomplete, fails to explain semantics, fragmented (lots of pieces in various places that explain only parts ... some of which are in conflict). I simply don't know what it all means, yet, and am still trying to find out. And I have to be careful when doing tests to be sure not to disrupt the existing network.

One recurring issue I find with lots of programs and projects is that documentation is just not clear enough to avoid having to carry out test scenarios to see what it REALLY does. In quite many cases I have found that the program and the documentation are not even consistent (often, but not always, because the documentation is for an older version).

One of the things I have found is that some people will try to take that information extracted and formulate the solution. That's very often what the OP wants (as in "just make it work"). OTOH, at least at the beginning, what I want is to know enough about it to figure out how to make it work for my situation. Many years experience tell me I may have to tweak that later, or use it for a different situation. As the saying goes, teach me how to fish (and cook it), instead of giving me a prepared meal.

But sometimes, there is no way to get the answer without someone else figuring it out. I wish that were not so. But documentation is so often a weak point (in reference to questions about initial setup, as opposed to the class of questions about figuring out what's broken somewhere).

I've also seen people that totally butcher their one and only language.

One possibility, which I have seem as very common practice on a couple legal forums, is for the initial responses to an OP be to rephrase their question into a well written form, and ask the OP if this is a correct description of their problem. Maybe that could work with those whose command of English, for whatever reason, is weak.

And documentation is usually what I want. Sometimes I already have the documentation and it is either incomplete or not well understood, and merely referencing it again doesn't help. These kinds of situations should probably focus on identifying the misunderstanding and correct that (then let the OP go back to solving the problem with now good information).

I don't even subscribe to anything. I just leave tabs open to threads I've asked or answered something in. I can always click on "My Posts" or "My Threads" on the right side to get back if I lose the browser window when Firefox dies, has to be killed, or I need a reboot (rare, but I did that yesterday to verify if a weird mounting issue was persisting).

The opposite problem is one I sometimes see (not just here ... and really not very much here). I or someone asks a question on a topic not well understood. The question was poorly phrased. The answers come in, but they were not targeting the real problem. At some point in the exchange, the issue gets clarified. Those giving the answers initially have either quit the thread, or just don't know the answer to the now correctly understandable problem. But the big difficulty is, that another member who might well know the answer won't jump in because the thread is already long. Maybe the see the original bad question and see it answered "correctly" and just move on.

That's the case I think a new thread might be called for. Yet (in other forums, notably in the past that being Usenet), when I've done this, I get negative responses about having created another thread and that I should stick to the original thread (one where fewer people will visit to see if they can answer). Then it's now TWO threads that people with answers try to avoid.

I do like to give answers about things I know about. But I also do avoid longer threads because I don't always have a lot of time to sort out the issue. I won't know if the OP got a meaningful answer or not.

In some other cases I have tried to just augment the answers given. If someone gets a short correct answer that may be non-intuitive, I've made posts just trying to explain it, or why. If that instills some understanding of what is going on to the OP, I think I've done some good.