-1
$\begingroup$

Recently Matter Modeling SE became a site that attracts lots of off-topic questions (see a recent one here that is more suitable for SO and has nothing to do with matter modeling: Is there a way to do array broadcasting faster, or without using loops?, but you can find lots of other examples by searching in it).

For this very particular example, one thing which is very interesting is that asking more off-topic questions even encouraged publicly in answers just for sake of having more questions (see it here):

If you need to run this on a GPU you can ask a separate question about that and I'll answer that

In my opinion, moderators of Matter Modeling SE don't care about off-topic questions and they think that as long as they have a high traffic in terms of questions and answers, it's fine in their eyes.

I'm using SE for almost 6 years now and I thought that the basic concept of SE is to discourage off-topic questions in each SE to make it more cleaner and clearer for the users to search and find information about specific topics included in each dedicated SE website.

Specifically, moderators of Matter Modeling have a quite hostile behavior towards marking questions as off-topic and nominate them for closure, which is quite surprising to me. Even, I was punished and banned from my review privileges in the Matter Modeling SE twice last year for nominating off-topic questions for closure, which again is pretty surprising and is not very well aligned with the SE code of conduct.

My question: Why moderators of Matter Modeling SE encourage off-topic questions and scarify clarity for higher website traffic? It's quite weird that SE doesn't do anything about it and let this behavior continue, which I think is not very well aligned with what we see in other SE websites.

Other Examples:

$\endgroup$
7
  • 4
    $\begingroup$ I'll try to add an answer sometime today, but I would like to note that we have had an number of Meta discussions about the scope of the site, what sort of questions we want to allow, and how to best handle borderline questions (e.g. whether to redirect them or reshape the question to fit the site). See for example: $\endgroup$
    – Tyberius Mod
    Jan 5 at 18:45
  • 4
    $\begingroup$ mattermodeling.meta.stackexchange.com/questions/291/…, mattermodeling.meta.stackexchange.com/questions/222/…, mattermodeling.meta.stackexchange.com/questions/38/…, mattermodeling.meta.stackexchange.com/questions/230/…. If you disagree with the conclusion reached in any of these posts, you should voice your concern there. $\endgroup$
    – Tyberius Mod
    Jan 5 at 18:50
  • 1
    $\begingroup$ @Tyberius Note that just because a question have a keyword of "materials", "matter", "modeling", or "computational materials science", it doesn't mean it's on-topic, even tangentially. With this mindset, I can ask a question like: I was reading computational materials science journal and I got a headache, what should I do then? (Yes, some of these questions are exactly like my hypothetical example...) $\endgroup$ Jan 5 at 20:17
  • $\begingroup$ Many of the examples you gave, were questions written by Diamond Moderators, who were elected democratically to enforce decisions about what's on-topic/off-topic. For many (or all?) of these examples, excellent and very helpful content showed up in the answers. What is the problem? $\endgroup$ Jan 5 at 23:40
  • 1
    $\begingroup$ @NikeDattani It doesn't matter if a question or answer is written by a moderator, or a new user, or even an unregistered user. So, I'm not sure what's your point here by saying that those questions are written by Diamond Moderators. Just because moderators are elected, it doesn't mean that every single question or answer they post here should be worshiped blindly without further review. Inspecting relevance of a question to be off-topic or on-topic has nothing to do with the rank of its owner. $\endgroup$ Jan 5 at 23:47
  • 1
    $\begingroup$ @NikeDattani Last but not the least point: Where I said that the quality of this questions or answers are not good? The main topic of this discussion is about what is on-topic and what is off-topic. If these questions were asked in the relevant place (e.g. SO), some of them were quite on-topic and might help other users to find them more easily in their correct place. $\endgroup$ Jan 5 at 23:48
  • $\begingroup$ @MithridatestheGreat In the first pair of comments by Tyberius, links to 4 different Meta posts which discuss on-topic vs off-topic were provided. 3/4 of them were "Hot Meta Posts" and were therefore featured on the site, and the other was posted during Private Beta. We have 3700+ users. What will it take for you to agree that the questions listed above, are on-topic? $\endgroup$ Jan 5 at 23:58

3 Answers 3

8
$\begingroup$

I'm guessing my response is going to be fairly long, so I'll try to keep it at least somewhat organized by putting it in sections.

Meta and Feedback

I wanted to lead off by just mentioning that one of the things I have found tricky with moderating thus far is the relative lack of feedback and discussion on Meta. There have been a couple of Meta discussion, a few of which I mentioned in the comments, focused on defining the scope of the site and establishing policy. After the early days of the site, these have received few votes and even fewer answer/comment responses. So I did want to thank you bring this up for discussion, but I want to emphasize that it would help us to have more consistent/frequent feedback/discussions on the direction of the site and how users want us to moderate. Without feedback, we can only really respond to issues on the fly and take the lack of complaints as approval.

Assessment

Here I want to discuss my personal view on the "on-topicness" of your examples. I have tended to avoid this previously because I had hoped to not monopolize discussion, with users passively accepting my view rather than forming their own consensus.

That's also been true of close voting. I more or less only use my vote to close abandoned or spam questions. On the keep/reopen side, I have done this in a few cases, but these were, as was described in the previous section, instances of handling a special case on the fly without an established policy.

With that out of the way, here is how I would classify the example questions. I won't go through all of them individually, but I'll pick out a few representatives.

Probably off-topic: To me, a few of these have too tangential a connection to matter/materials/modeling where I don't think they really make sense here. While they might be useful to someone in the field, they are more general and we want to avoid "boat programming" questions. This would include What software will allow me to combine two images?, How do I search for a particular string in Linux?, Is there a way to do array broadcasting faster, or without using loops?, High-Performance Computing: What does "Mio CPUh" mean?, where you could remove references to matter modeling (and some actually already do) and the question wouldn't fundamentally change.

However I think very few questions fit in this category. These few questions seemed to be highly upvoted and never received enough close votes and I didn't feel it was my place to overturn what the community found useful.

Judgement call: This category is of questions that there was at least some ambiguity about it being on-topic, but prior site discussions, their reception, and their close proximity (if not direct connection) to matter modeling have convinced me they should be on-topic here. As mentioned, this is my opinion and I'm open to the community over time deciding to include/exclude these sorts of question.

I would include among these Supercomputers around the world, What is a chemputer?, Did the 2019 discovery of O(N log(N)) multiplication have a practical outcome?, Is it better for me to study chemistry or physics?. I could see a scope for the site where some of these aren't allowed, but I think they have proven useful and brought people here that may not have otherwise joined.

Multiple sites: These are questions that may be on topic on another site(s) and may even fit better there, but I would also say pretty confidently that they should be on topic here (again, my opinion).

What are good protocols for creating a database based on python?, How should I organize and keep track of a huge number of calculations systematically?, Suggestions on laptops for matter modeling. These sorts of questions might also work on Software Engineering, Chemistry, Hardware Recommendations, Physics, Computational Science, etc, but I think the connection to matter modeling is clear and in some cases they are more easily answer by someone with our background.

On topic To me, a lot of these examples seem exactly within the scope of the site: helping users to understand the principles and software behind matter modeling.

How to simulate a furnace computationally, How to compute the overlap matrix in Python, How to properly define %maxcore in ORCA, How to manage disk space for Gromacs XTC (trajectory) file output.

Going forward

To me at least, I don't see a huge problem with the way the site is operating. I think we have had questions that I wouldn't necessarily have considered in scope, but I think these are far from the norm. I think even some of these borderline questions have been beneficial to the community and brought in new users.

However, I feel my job as mod is to help carry out the community consensus. So I'd love to see more discussion about how users feel about the direction of the site (preferably across several new Meta posts).

One area that I think is relevant to this post is: how inclusive do we want to be? Should this be a site where we accept questions from all-levels of matter Modelers, even complete newcomers, with all the potential issues of quality and on-topicness? Or do we only want expert in the field, possibly increasing quality and narrowing the scope at the expense of a much smaller community? In a loose analogy, do we want to be more like Math SE or MathOverflow?

We can also continue considering questions of scope. The typical StackExchange approach is to consider only questions directly on the topic. But sites like MathOverflow work quite well and have a number of questions that are "useful to professional mathematicians" without being strictly about professional level math.

$\endgroup$
4
  • $\begingroup$ Thanks for your answer. At least your answer is somewhat fair and open a way for future discussions of what is on-topic and what is off-topic. But, the fact that your answer, despite its fairness, is not the most upvoted answer here shows that there is not much interest from the community to talk about this topic. Thank you again! $\endgroup$ Jan 6 at 4:33
  • $\begingroup$ @MithridatestheGreat Meta can be a bit slower in traffic and I answered pretty late in the day (not to mention well after the other answers). It may even take a few days before a significant number of people have a chance to vote on the various responses. $\endgroup$
    – Tyberius Mod
    Jan 6 at 4:36
  • 2
    $\begingroup$ "But, the fact that your answer, despite its fairness, is not the most upvoted answer here shows that there is not much interest from the community to talk about this topic." It's unfair to say that about the community. This answer came 7 hours after the other two answers! $\endgroup$ Jan 6 at 4:41
  • 1
    $\begingroup$ In my last comment, I probably should have also mentioned that the post had only been up for 2 hours. Now you can see it's gaining traction. The net score was +2 (versus +5 on the answer by Camps) before being pinned to the top recently. Now it's +3 while Camps is still at +5. Pinning has a tendency to change the game too. $\endgroup$ Jan 6 at 21:47
6
$\begingroup$

I will give my own personal opinion (without previous conversation with the moderator team) and it is based only on my experience as a researcher, teacher and advisor.

When you start working in Matter Modeling, you find a zoo of methods, programs (well- and poorly-documented, non-free, free, opensource, etc.), written in several languages (C++, Fortran, Python, Java, Julia, etc.), available for different operating systems (Mac, Windows, Linux), etc. It is a headache for a beginner!

If you are (very) lucky (or have a lot of financial resources, so very lucky), you will need only a few sets of tools (for example the Material Studio marvel that cost around 250k with all the tools). If you can only use free/open-source software and a few commericial ones, you will need to use several tools: for preparing the inputs, for running the calculations, for visualizing the results, for automating the workflow, etc. Finally, you will need to present/publish all the results.

As most of the community members already know about much of that, Matter Modeling is, from my point of view, a singular site very different from other much more specific SE sites: here you can design a pipeline that mixes programming, queue submission, script running, automation of calculations, analyze the results, create the figures, prepare a manuscript, etc.

If you look carefully for all those questions that, at first sign appear off-topic, you will find that in some point they are related to the matter modeling workflow/pipeline.

Based on my knowledge and experience, I usually read the question and if I find a link with matter modeling, I welcome it.

$\endgroup$
10
  • 2
    $\begingroup$ I would ask then, with this specific question, how would another user ever find this at a time when it is relevant. The example given is so vague that its essentially a question of "How do I do linear algebra in numpy". $\endgroup$ Jan 5 at 20:06
  • 4
    $\begingroup$ Maybe you should start asking the OP where will he/she will use that code and for what. Again, I personally needed to manipulate a huge amount of data (from entropy calculation to millions of molecules) where optimizing a code in Python is a must. $\endgroup$
    – Camps Mod
    Jan 5 at 20:16
  • 2
    $\begingroup$ I agree we need the usage of the code in the question, however, it is not up to moderators / answerers to build context for future use. This question was definitely off topic, it was closed, then reopened for some reason. As it stands right now, we made a HNQ which is improperly in our SE and not the main stack overflow where it belongs. $\endgroup$ Jan 5 at 20:18
  • $\begingroup$ I think I need to agree @TristanMaxson here. Making up context by throwing answers and make questions tangentially on-topic by heavy editing is not a correct mindset. Also, need for a wide range of skills from programming, knowledge about Linux, HPC, Computational Materials Science, etc. to have a successful matter modeling do not mean that all these different topics should be asked and answered in this very website. $\endgroup$ Jan 5 at 20:23
  • $\begingroup$ @TristanMaxson The question was closed and re-opened with only 4 seconds in between. If being closed is your only evidence for it being "off-topic" then I don't think that evidence holds water, unfortunately. $\endgroup$ Jan 5 at 21:01
  • 1
    $\begingroup$ Well, if that is really an example of how closure and reopening nominations are considered here, I think it's just a self-explanatory example of the moderation situation in this SE. I believe it needs more than "4 seconds" time to decide if a question should be closed or reopened. $\endgroup$ Jan 5 at 21:06
  • $\begingroup$ @NikeDattani I simply meant that its confusing why it would be closed then reopened? To clear close votes maybe? $\endgroup$ Jan 5 at 22:16
  • $\begingroup$ @TristanMaxson I've seen that done to clear close votes when close votes go against the site policy (which is that if there's an upvoted answer, leave it open). You and I can probably/hopefully? have a constructive conversation about whether or not it's harmful or helpful to leave such questions open, but I'd appreciate if the tone were less aggressive. Telling us what "should be closed" or that something "should be on another site" comes across to me like you don't want to have a conversation and you just want to cause trouble. Could discussion lead you to change your opinions at all? $\endgroup$ Jan 5 at 22:22
  • 1
    $\begingroup$ @NikeDattani This is a meta SE talking about if questions are off topic, if I cannot say that a question "Should be closed" without being accused of causing trouble then this states a lot about the current moderation approach. I am always open to discussion, but I am surprised that somehow the entire moderation team thinks this question is on topic when it doesn't pertain to matter or modeling in any sense. $\endgroup$ Jan 5 at 22:27
  • 1
    $\begingroup$ @TristanMaxson You also said "You need to properly profile this code and provide a test dataset etc. We cannot improve this code in its current form (and this isn't a coding SE)" to a relatively inexperienced user. Telling someone that they need to to something and that we cannot help is a bit aggressive. I've only seen 2-3 users on this site use that type of language. "I am surprised that somehow the entire moderation team thinks this question is on topic", please see the links Tyberius gave in his 2nd comment on OP (e.g. Anyon's question and dwhswenson's reply). $\endgroup$ Jan 5 at 22:35
-2
$\begingroup$

I just came here to make this exact post. While it's okay for us to answer questions which are semi-related to matter modeling (maybe publication advice in the field etc), this particular question is confusing.

We have given a couple answers to the question, which we do not even fully understand the context of. The question also became a hot network question, which is concerning as people seeing this will wonder why we are approaching this question when it is also cross posted on the actual SE it should be on (see this post).

I know there is this feeling behind some of the moderation team that closing questions is bad and downvoting is also bad, but these features exist on SE for a good reason and they are not being used when they clearly should be. This mentality does not align well with the general community of stack exchange. I think we are past the point of needing to get a core userbase and we need to ensure experts who participate in this SE are answering and processing high quality questions and answers. This is something I have gotten feedback on from a few people on why they do not actively participate.

Side note: I have also noticed the original poster added the atomic-simulation-environment tag to this question. This is definitely not the case for the question being asked, even if it is likely that ASE is being used. Improperly tagging can also be dangerous, there were tag edits made that did not remove this for some reason.

$\endgroup$
3
  • 3
    $\begingroup$ Yes, I stopped contributing to this website almost one and half year ago exactly because of this confusing mindset... $\endgroup$ Jan 5 at 20:18
  • $\begingroup$ I've responded in chat. $\endgroup$ Jan 5 at 21:30
  • $\begingroup$ @MithridatestheGreat Since the comments were getting long and it was trying to create a new chatroom, I cleared the discussion between you and Nike after he copied it to the Meta chat room. If there is comment you thought was particularly relevant to remain on this answer, I can put it back. $\endgroup$
    – Tyberius Mod
    Jan 5 at 21:44

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .