14
$\begingroup$

A few days ago, I asked this question. On the one hand, it received a couple of upvotes and got a good answer, which also received a few upvotes. On the other hand, several people voted to close it as off-topic, and it is now on hold.

There are a few major potentially problematic things with this post. First, it's about software related to engineering, not exactly engineering. Second, it leans heavily toward being a recommendation question, as addressed by this earlier meta question. Finally, it is rather vague.

This brings up the question: what sorts of software questions should be allowed on this site? I am of the opinion that specific questions concerning specific software should be permitted, queries such as: "My Creo Parametric model is giving a short-edge error. How should I go about tracking it down?" or "How do I create a variable-radius round in Autodesk Inventor?" However, I fully understand that the question I mentioned earlier does not fall into this category, and may be undesirable.

What is the community's stance on this? What software questions do we want, what do we not want, and where do we draw the line?

$\endgroup$
  • $\begingroup$ I'm really glad this is on meta; it's a different issue than the other post. $\endgroup$ – HDE 226868 Jan 30 '15 at 21:24
  • 2
    $\begingroup$ By the way, a related proposal on Area 51 has been reborn: area51.stackexchange.com/proposals/80927/…. $\endgroup$ – HDE 226868 Jan 30 '15 at 23:05
  • $\begingroup$ Good question. I don't think I fully addressed the broader questions you raised, but I hope I nudged the community towards making some decisions on the points you raised. $\endgroup$ – user16 Jan 31 '15 at 1:21
  • $\begingroup$ @HDE226868: In my opinion CAD questions related to engineering (with a certain minimum level of specificity) belong on this site. The area between "CAD questions" and "engineering questions" is probably too broad and gray to make it anything but a mess trying to split them between sites. $\endgroup$ – Matthew Jan 31 '15 at 1:40
  • $\begingroup$ I didn't downvote or vote to close your original question, but I could see why someone would. And if it wasn't on hold, I might consider flagging it for being too broad. A quick google search for "open source cad" reveals no less than 7 different programs on the first page alone. If you had come in asking about the maturity and usability of a certain program you're considering, or specifically how one open source program compares to another (free or paid) I think you would have gotten more traction. You can always edit your question, and if you make significant changes it will be reopened. $\endgroup$ – CBRF23 Aug 12 '15 at 11:22
  • $\begingroup$ Fwiw, I've restarted the CAD proposal. $\endgroup$ – Jason C Dec 17 '15 at 1:51
15
$\begingroup$

This is pretty much going to be a concurring opinion with Glen, but it sounds like he doesn't have a lot of CAD and modeling experience, so this comes from the perspective of a mechanical engineer whose primary function at work is creating and editing models and drawings.

My gut tells me that the example questions posed would be on-topic. But it also tells me that there should be a base-level of complexity to CAD questions. I don't love that concept, because I wouldn't apply that criteria to all types of questions on the site. What an expert in the field might consider the most basic information to his field, someone else might not even know about the existence of. Someone without any significant engineering or physics knowledge may not know that F=ma, and we shouldn't send them away for asking a question which can be answered with a simple explanation of Newton's Second Law of Motion.

But with CAD, there are a bunch of readily available (and on the internet, often free) resources to learn the basics. It's not that I'd consider the questions beneath us, but someone learning to model a simple tube would be better off using the pre-existing resources, not only because they're already there, but because they'll take you through the whole process, explaining everything along the way.

I think the criteria I have in my head is "If it's something I'd be Googling at work because an issue cropped up during modeling, it's ok to ask here." I think this is certainly influenced by reading on some meta site that the goal for an SE should be to be the top result when a question that we've answered is searched on Google, which I like the idea of. This is a hard criteria to enforce though, and everyone's will be different. I feel like Justice Stewart; I can't tell you what a bad CAD question is, but "I know it when I see it."

I definitely worry about a flood of overly-simplistic questions, but I have no clue if those fears are founded at this point. Thus, I think it's best to opt on the side of allowing more questions early on, and if they become a problem, we can address it then.

However, the question Matthew asked and first links to in his question here, I agree with it being off-topic. It's not entirely opinion based, because he simply asks if such software exists. But that information can be found elsewhere (Wikipedia is good about collating lists of similar software packages, Google otherwise) and isn't truly about engineering. Questions about the actual process of modeling are fundamental to engineering though. Part of being an engineer is knowing how to realize and communicate your vision and your design.

$\endgroup$
  • 1
    $\begingroup$ So a tutorial would be "Too Broad" and therefore off-topic here. But a targeted, specific question ought to work. Right? Personally, I'm okay with that level of subjectivity, especially with community driven enforcement of it. From experience on programmers, we'll get a lot of meta complaints about "why was X allowed but my question Y wasn't?" But those complaints can be addressed too. $\endgroup$ – user16 Jan 31 '15 at 14:20
  • $\begingroup$ I would say that basically the programming question must have a strong background/relation to engineering. So "How to use this CAD software" should not be part but "How can I solve this engineering problem with software" would be okay. $\endgroup$ – Trilarion Feb 3 '15 at 17:34
  • $\begingroup$ I think your concern about "a flood of overly-simplistic questions" is well-founded. However, would it be acceptable to allow well-worded and specific, yet simple, questions, and then mark as many others as possible as duplicates? $\endgroup$ – Matthew Feb 8 '15 at 21:44
  • $\begingroup$ @Matthew In theory, that sounds like it would work, but there still needs to be a line between "I need help" and "Please teach me how to model" and your suggestion sounds to me like its the start of our own personal tutorial, which is not the point of this site. $\endgroup$ – Trevor Archibald Feb 9 '15 at 14:06
6
$\begingroup$

My answer comes from the perspective of being an active member of Programmers.SE; a sometimes contributor of StackOverflow; and having some knowledge of Computer Science.SE as well as Theoretical Computer Science.SE1

For questions like:

I am of the opinion that specific questions concerning specific software should be permitted, queries such as: "My Creo Parametric model is giving a short-edge error. How should I go about tracking it down?" or "How do I create a variable-radius round in Autodesk Inventor?"

Of the SE sites I listed, both of those would be most on-topic at StackOverflow as they are implementation issues. Progs, CS, and TCS would all likely close as off-topic.

That said, I think a question on StackOverflow about parametric models would receive crickets in response. Seriously, just go search on "Creo" there.

There are currently 456 questions on SO tagged with Autocad, so the variable radius round question may pick up a decent answer from there. But 456 questions is pretty darn small for a tag on StackOverflow.

So it's pretty clear to me that those questions are in a gap that none of the existing software SE sites are addressing3. We've already established that we're not going to worry about overlap with other sites at this point, but I think it's worth pointing out when there could be an overlap but there isn't.


Blah, blah, blahTM - let's cut to the chase.

I think the parametric modeling question would be very much on-topic here. It seems like a question that's begging for an audience of Engineers to address. So, I would encourage questions like this one.

I'm a bit more lukewarm about the Autocad question, but I think that could be due to my "Programmers vs. StackOverflow" bias2 towards conceptual questions. I personally prefer to see more conceptually oriented questions and answers as I think they build higher quality content that's likely to have longer lasting value. This one seems more like a technician type question. OTOH, I could have also just revealed horrific ignorance there with using Autocad and other CAD systems.

TL;DR - Let's give 'em a run and see how they do. We can always narrow scope later.

1 What's surprising is the modest amount of overlap between the various sites... You would otherwise think there's lots of overlap there.
2 Note, I do like StackOverflow, but I much prefer Programmers. A simple glance at my profile will confirm that. I personally value understanding the why behind a particular approach more than I value an individual case of how. That said, I do lots of research on SO to solve problems I run into during my work.
3@Gilles - please don't drop a network wide suspension on me for even remotely implying that CompSci is "just about software." :-)

$\endgroup$
  • 1
    $\begingroup$ I can see how the CAD questions would be on-topic at SO, but I would NEVER think to ask them there. I'm not on SO or any of the CS SE's, and I don't associate CAD programs with programming. So I think it's definitely something that should be around here. $\endgroup$ – Trevor Archibald Jan 31 '15 at 3:30
4
$\begingroup$

Questions about CAD should not be completely off-topic here. Perhaps, these should have a higher bar.

@Glen drew analogies with SO and Programmers.SE. I'll draw an analogy with EE.SE. Questions about specialized software for electronics design1 are on-topic there2. On EE.SE the following characterized EDA questions:

  • The number of EDA questions is around 3 a week on average, which is moderate.
  • EDA packages evolve at a moderate rate compared to, say, JavaScript frameworks. The questions and answers have a better reuse value, compared to other software usage questions.

Having said that, specialized stack would do a better job of covering CAD. For example, it can cover issues like CAD graphics, CAD hardware. So, I hope that they successfully launch.

1 Electronics Design Automation (EDA) software. Equivalent of CAD in EE.
2 post on meta.EE.SE to that effect
3 DIsclaimer: Nick is a user of SolidWorks CAD and several EDA.

$\endgroup$
  • $\begingroup$ +1 for gratuitous use of footnotes. $\endgroup$ – user16 Jan 31 '15 at 14:16
3
$\begingroup$

This question is really about engineering tools, of which CAD software is just one example.

What tools exist, what situations each is good for, and how to use them properly are all part of engineering, and often the answers may not be so obvious. However, that does not make all questions about tools for engineering use on topic. To be on topic, questions about tools must be one of:

  1. About a general type of tool, not a particular model or brand, relative to a specific problem. For example, Is a oscilloscope, voltmeter, or electrometer more appropriate to measure voltage in xxx specific application? is OK, but not What is better for measuring voltage, a Floop 900-A Voltmeter or Hafwitt-Peckard 9137B oscilloscope?.

  2. A specific problem you face with a specific tool. We aren't the help desk for that tool, but if it's common enough others may answer here anyway. For example, Every time I specify a #4-40 hole in MechBarf 1.3.7, the screen goes blank and my drive is reformatted. How do I avoid this?.

  3. A question about a type of tool, like I have to replace my soldering iron tip every week, but my tech does the same amount of soldering and his tips last a few months. What techniques and tricks are there to get longer tip life? He says I shouldn't leave it on 24/7 at 800°F. Is that really true? If so, why?

Generic what tool is best, unquantifiable comparisons between models, or questions that are just opinion polls or bait for a religious wars are squarely off topic. We don't want Which are better scopes, Hafwit-Peckard or Techno-Trix? or What is better, algebraic notation or RPN? . However, What is the shortest key sequence to solve xxxx in RPN, and how does that compare to the shortest using algebraic notation? would be on topic.

$\endgroup$
1
$\begingroup$

Engineering software is part of engineering. Therefore we should encourage engineering software related questions.

Disclaimer: I work mostly in Embedded SW, EE, Project Management, Product Development

Reason: Assume someone modeling a transistor. CAD software like PSpice has many parametric variable that need to be defined properly to obtain a good engineering model. We should encourage these questions.

On the other hand it is highly unlikely that you will find a answer to how to configure PSpice on Ubuntu machine. In such a hypothetical situation the poster should be encourage to post the question on askubuntu where an effective answer could be sorted very quickly.

The CAD software in questions was asking recommendations which I think belongs in a different forum like linkedin.

$\endgroup$
1
$\begingroup$

My personal opinion on the matter:

Engineering software that isn't CAD should be on-topic for the most part. I don't know of a good repository of information about programs dealing with structural analysis, thermodynamics, simulations, etc. This seems like as good a place as any to start that repository. Also the issue of restricting an infant site comes into play here.

CAD questions could be treated on a per-question basis. A great deal of us have to use some form of CAD software on some kind of regular basis. Personally, I use it daily, but I wouldn't be surprised if the next guy used it once a month. "How do you make a circle in AutoCAD?" questions should be off topic, but "How is [item x] typically displayed in drawings?" lean more towards on-topic to me, as they are about communication.

If the CAD proposal takes off, then we can draw a better line between Engineering and CAD. I just don't like putting too many restrictions this early.

$\endgroup$
  • $\begingroup$ I don't think basic questions should be off-topic. Rather, we should allow one particular question to be asked, and then flag any and all similar ones as duplicates. Look at this question and this question. Both are pretty basic programming questions, but they are both very popular and approved-of, as evidenced by the upvotes. Stack Overflow is great because you can go there with any question - not only those that are sufficiently complex. $\endgroup$ – Matthew Feb 2 '15 at 20:04

You must log in to answer this question.

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