Network Engineering is not a welcoming place

Question

This may seem like a click bait title but it's actually not. We all know that Ron is the most actively guy here on this community, but that ended up locking the place as one moderator with an strict set of rules that many times does not make any sense.

If you go to the main page of Network Engineering the majority of the questions are always closed as off-topic. Sometimes they are closed with, in my opinion, some reasons that are sometimes pathetic. Yes the word is harsh, but take this one as example: How to filter unused routes in MikroTik routers?

It was closed because there's no paid support from the upstream vendor. That's sad. We all know that Mikrotik is heavily used on enterprise networks, and it's very big on developing nations. I myself don't like Mikrotik devices, I don't think they are reliable/good enough, but closing the question with this vague argument that it was off-topic is just bad.

I'm around this site since 2016 and roughly 50% of the time I've posted something here that I've regretted later on due to some similar issues, I will not dig on everything that happened, but one issue in particular got me really pissed off with the directions of this community that Ron enforces, and was this thread: Force Catalyst 2960X Boot Loader upgrade to the same version

Ron tried everything to disqualify me and my question. Downvotes were cast mainly because we've ended up with falsified devices and I was trying to at least make them usable since Universities on developing countries, which unfortunately I live, don't have resources to just throw away equipment and buy new ones, we keep it or we don't have anything to replace. That's our reality.

That wasn't the first time this happened. I've recently opened a thread here that was immediately closed as off-topic and now it's deleted because I can't find it anymore, but the reasons is that was a server question and not a "networking" question.

Well this may sound as as surprise to some folks but servers consumes networks provided by network equipment and if a given server is behaving errantly it may be related to the networking equipment. And guess what? After two weeks of digging on the issue, by myself and with Dell ProSupport, it was nailed down to an issue with Dell Networking OS 10 and not the server itself. Curiously the Dell technician didn't said that the issue was server related and closed the ticket. So after solving the issue, I naturally would promptly answer my own question and leave it there for new comers, but as you know, it was closed because it was off-topic.

So yes, I'll not extend myself, but Network Engineering may be doomed as community if we continue as is. I got the Mikrotik question literally in 60 seconds, so the pattern is there for anyone to see.

That pattern goes back easily to 4 years ago, it's easy to find people complaining about the same issues, as example: Why was my question closed then deleted?

So there we are, years later and Network Engineering failed to grow as a welcoming community like other Stack Exchange sites.

As today I don't have any reason to help people out or even waste my time writing proper questions just to be disqualified here as a professional or ended up with the question closed as off-topic. So yes, I'll probably continue active in ServerFault where those things don't happen.

Thanks for listening.

PS: The criticism may be hard to read, yes I know, but when you're in a position of power you should know how to deal with it. It's not personal, but it's targeted to the atitudes of the moderation. I don't know the lives of the people behind moderation, I can only infer from attitude.

EDIT after first Ron response:

Regarding the question to ignore the rules: I'm not saying to ignore the rules, but if the rules need to be enforced on almost every post something is just wrong. The rules are wrong, and they need to be changed. Clearly a decent chunk of considered off-topic questions are not off-topic per se, they are for an "enterprise business network" but it does not comply with the closed mind aspect of the rules. Take the Mikrotik topic as an example. The efforts to change the rules never really happened.

Still on the rules, specifically Ron, I was not saying that you've created the rule. I just say that you always quickly enforce them, so fast that even questions that may fit in the rules are immediately closed as off-topic without further way to the community question the OP to see if it was viable or not. That's pattern is only here on NE.

The strict set of rules is the real culprit here, and as I said on the PS, that's wasn't targeted to you as a person, but as the one who enforces the rules. That's not personal.

tl;dr: the rules does not leverage the community, in fact they limit the community.

Regarding I was madding up: Again don't take it personally because is what it seems.

I've waited a long time to write this topic so I could be the most unbiased possible. But the facts are there, you don't need to close the question to disqualify.

Read again the comments, and that happened on other questions too, as example here: The equivalent of "ip dhcp snooping information option allow-untrusted" on NX-OS

Both questions are open but the way those questions are handled on the comments is just to remove the focus on the question and try to find anything wrong on the reasoning of the question. This link above shows exactly this point, the focus on the question is changed to: you don't know how to design proper network cores. And that became personal if you look it closely enough.

So yes the issues are on the attitude. That happened on both questions in different ways, on the Nexus question a random guy with low reputation provided a proper answer, the moderation instead tried everything to disqualify the question, when finally gave up. The requirements are wildly different depending where you work for, and technology is changing, we need to address it and not disqualify whoever comes questioning issues on non standard Cisco topologies.

When this happens more than once, you as a users tend to avoid the community, and here we go on the first page of Meta with the exactly same issue: https://networkengineering.meta.stackexchange.com/a/1945/10528

We lost a properly capable user that will never come back and will never recommend NE.

When I joined ServerFault on 2013 (I think) I was 9 years younger, on my 20's. I've learned a lot from the site, contributed everything that I can, and today I have a reasonable good score there with more than 5k points. In 2016 when I discovered Network Engineering I was really happy to found it in first place because the gap regarding networking question would finally be solved, but that didn't happened. And that's my point on writing this.

Again, to clarify, it's not personal. I just cited you Ron because you're the most active user here, and this is not a complain, actually being active is a good thing, it was a compliment to you.

Finally about the question that you've recovered, yes that was the initial text of it. It was clear to me that the issue wasn't server side, so what I've done:

1. I've searched for similar questions on NE, and there's none.
2. So I carefully crafted a question to list the rules of proper jumbo frames on networks, which in my opinion is on-topic.
3. I didn't mention any software or server OS on the question to be on-topic.
4. Strict group of features that only exists on enterprise grade equipment, so it seems on-topic.

So yes that would be the first step of debugging. The server question is irrelevant, I would open another question to add more information to the site, to keep it consistent, just like on other StackExchange sites. That we have a simple questions, delimited without too much specifics so it can reach a broader audience. The next questions would be MTU configuration specific to the devices, but that never came to reality, due to the closing of the first one.

Again here we have a strict group of rules that does not help the community, rules that didn't even make sense. And those are extremely enforced. Common sense is advised, but the rules are always processed literally. I know that for us, engineers, are really difficult to take account of the human factor, but as a moderator it should be done.

Finally on the argument that I delivered more details to the Dell Technician, yes, in fact that's true. But I started with the same question to him, and the following events led with to the solution of the case, and in that case to be honest, the technician only assisted. The bug on the software was found by me when I compared the device to the infamous Nexus 3048 that I had on other network. The technician wasn't aware of the 14 bytes ethernet frame bug on MTU larger than 1500 bytes.

And to answer everything, when I said: downvotes are cast I wasn't referring to you, downvotes is plural and not singular, so that's for more than one person. When I've opened that question I remember that it reached quickly 4 votes, and after the whole argument regarding Cisco SmartNET it was downvoted to 1 vote. So that's what happens, people tends to read an argument with a moderator and just downvotes it, because we are automatically biased towards it. I'll not be saying things like: what if. But the reality is that a proper, and interesting question, was dismissed with the most ludicrous arguments. Come on.

Final considerations after the first edit: If the set of rules where working, all the networking related questions on ServerFault should be here, right?

But that wasn't what we seem: https://serverfault.com/questions/tagged/spanning-tree

It's hard to argument that a specific thing to networking, like STP is viable on ServerFault and not viable here because you're running STP on a FreeBSD box, with support from a company that sells support, linked to a device that although it is enterprise, the company does not provide paid support for it, and you're using STP in your home because you've a homelab an it's not On-Premises on your company.

Can you see how ridiculous the last paragraph was?

So that's why ServerFault still receives a lot of networking related questions. We as a community failed. Five years has passed and NE didn't grow as it should have grown.

In the other hand those clearly off-topic questions, like this one: https://serverfault.com/questions/1049944/cisco-spanning-tree-configuration aren't closed. You may argue that's moderators on ServerFault are lazy. But the reality is that when we search for anything related to networking NE never came up first, an this may be nailed down to:

1. Excessive of closed questions.
2. Deleted questions.
3. Lack of interest from users.
4. The attitude of the the moderation.
5. Not being a welcoming place (which is the topic of this question).

It may nails down to the following question: to whom this community is targeted towards to? As now it seem to be for the "Ciscotist" that works inside telcos and live in a bubble where only rack mounted network equipment exists.

Again thanks for listening and at least considering what I have written.

EDIT after next questioning by Ron

Ron, this will be my last edit so we don't get lost on the original motivation of this topic: NE not being a welcoming place and it failing to grow as, I think, we both envisioned 5 years ago.

I think the most important points were explicit described and shown with examples. It will be the reader duty to get its own conclusions.

Just want to comment on the two questions that I've cited, because I left the impression that, as you said: You are reading something not written.

1. Force Catalyst 2960X Boot Loader upgrade to the same version
2. The equivalent of "ip dhcp snooping information option allow-untrusted" on NX-OS

On the first cases there's a philosophical fallacy mainly know as The Straw Man fallacy. Which is exaclty what happened on that specific question: A straw man (sometimes written as strawman) is a form of argument and an informal fallacy of having the impression of refuting an argument, whereas the real subject of the argument was not addressed or refuted, but instead replaced with a false one. As described on: https://en.wikipedia.org/wiki/Straw_man

The questioning were irrelevant to the issue.

On the second one another type of fallacy was used, and was the Loaded Question Fallacy, where one of the interlocutors try to disqualify with a question that is clearly to input that the other side is incapable of something, when you say: Are you enabling DHCP snooping on the core?

That's exactly what is the loaded question fallacy: https://en.wikipedia.org/wiki/Loaded_question

In that question is totally irrelevant if I'm enabling it on the core or not, I was just questioning why DHCP snooping does not work on Nexus 3048 platform. If I've had just written that I had issues with DHCP snooping on the Nexus 3048 you'll never know if I was running it on the core, which I know that 3048 should not be used on the core.

The story would be totally different if an answer to the subject was made and an observation like: don't do it on the core please. It would be totally different instead of just changing the focus off the question and answering with another question.

Regarding I not knowing how to use SE sites, well, I can't say that I know how to use the site properly, but I think I've enough reputation on other SE sites to state that point. So again it's the reader exercise to elaborate its own conclusions. I may be totally wrong and worthless to the community, yes, but I'm here stating my opinion to at least make this community better in terms that I believe.

I'll not extend myself anymore, that's my last edit. To finalize I think that's something really generous to spend time managing a community without any financial support, done as is. As I said before, it was a compliment to be that active on a given website just to make the community strive. Thanks for that.

If I said anything that may sounded offensive wasn't the intention. I ask sorry for that if it's the case. I'm not a native english speaker and language barriers with different cultures may lead to false impression of the true image of a given person.

Thanks for the last time for taking the time to answer.

Answer 1

If you go to the main page of Network Engineering the majority of the questions are always closed as off-topic.

That is because they are off-topic. Should we just ignore the site rules?

It was closed because there's no paid support from the upstream vendor. That's sad. We all know that Mikrotik is heavily used on enterprise networks, and it's very big on developing nations. I myself don't like Mikrotik devices, I don't think they are reliable/good enough, but closing the question with this vague argument that it was off-topic is just bad.

That is not a rule I created, but it is certainly one of the caveats on the What topics can I ask about here? page. See the second caveat below:

and meets the following requirements

• under your direct control (if the network is not under your control you will not likely be able to provide the information required to answer your question);
• hardware that has a paid support option from the manufacturer (enterprise/provider class products, some small business class devices);

I also asked here, in Meta, about such things, and part of the answer I received explains:

The line we have drawn is if the vendor of the product classifies it as at least small business and provides a paid support option for their product(s).

-and-

The key difference is that vendors who provide paid support for their own product have a vested interest in the product.

It may be an imperfect rule, and it has received complaints over the years, but, so far, nobody has proposed a better one that the community will accept.

---

Ron tried everything to declassify me and my question. Downvotes were cast mainly because we've ended up with falsified devices and I was trying to at least make them usable since Universities on developing countries, which unfortunately I live, don't have resources to just throw away equipment and buy new ones, we keep it or we don't have anything to replace. That's our reality.

You are making that up. I never closed that question, nor did I ever vote down on that question or any other of your questions. I have cast 21,130 up votes and only 106 down votes over the nearly seven years I have been on the site. In fact, I looked, and you have not received any down votes from anyone on any of your questions, and I have cast up votes on most of your questions.

---

That wasn't the first time this happened. I've recently opened a thread here that was immediately closed as off-topic and now it's deleted because I can't find it anymore, but the reasons is that was a server question and not a "networking" question.

Well this may sound as as surprise to some folks but servers consumes networks provided by network equipment and if a given server is behaving errantly it may be related to the networking equipment. And guess what? After two weeks of digging on the issue, by myself and with Dell ProSupport, it was nailed down to an issue with Dell Networking OS 10 and not the server itself. Curiously the Dell technician didn't said that the issue was server related and closed the ticket. So after solving the issue, I naturally would promptly answer my own question and leave it there for new comers, but as you know, it was closed because it was off-topic.

That particular question only included server information and asked about the server network card, and hosts/servers/VMs are off-topic for NE; they are handled on Server Fault. There was no mention of network equipment or the network in your question. You could have included such information, or edited the question after it was closed to include the information and start an automatic reopen vote. There is really nothing on-topic in the question, and nobody cast a reopen vote:

What I should be looking for when trying to enable Jumbo Frames in networks that relies on PXE for server booting?

As far as I know I must have:

Network switch capable of MTU >= 9000 The ability to enable Jumbo Frames not only on the network interface on the switch but also on the Port Channels and VLANs A compatible SO to deal with the higher MTU after the boot process with network bonding and VLANs if needed. What I don't know exactly is how the network card on a given server would know that the connection is not using the default MTU = 1500 value during the PXE phase. Is it supported? Varies by vendor? It should work out of the box? There's something that should be added to the DHCP server the handles the addresses during boot?

Thanks

We could only speculate or guess at your network and that it was your problem, but that is off-topic for SE sites. I am sure you gave the Dell technician far more information about your network than was included in your question. Remember that SE sites are explicitly not discussion forums (we have Network Engineering Chat for that, and you have enough reputation to participate and can ask off-topic or speculative questions there. The main SE Q&A sites are for specific questions that have specific answers, not speculation, opinions, or discussion that is allowed on the chat.

---

the pattern is there for anyone to see.

Yes, it is the pattern of closing off-topic questions according to what the community has decided for NE. There is always the opportunity for the OP to edit the question to try to rehabilitate it, but most of the closed questions are simply asked on the wrong SE site and are off-topic for the NE community.

---

PS: The criticism may be hard to read, yes I know, but when you're in a position of power you should know how to deal with it. It's not personal, but it's targeted to the atitudes of the moderation. I don't know the lives of the people behind moderation, I can only infer from attitude.

My only attitude is to help NE by enforcing the rules that the community has decided for itself. I did not write the rules. If you ask off-topic questions on any other SE site, the you will find those questions closed, too, often without any explanation.

You must understand that NE was split off from Server Fault and Super User by the professional network engineers so that they did not need to wade through all the host/server/VM and home networking/consumer-grade device question to see the questions that interest them. Allowing such questions on NE completely defeats the reason for NE, and it simply duplicates other SE sites.

---

Response to your edit:

if the rules need to be enforced on almost every post something is just wrong.

What is wrong is that people are not taking the Tour or reading the What topics can I ask about here? page, and the problem is not limited to NE. Many questions asked on Stack Overflow with the Networking tag get closed because people are asking off-topic questions.

---

I was not saying that you've created the rule. I just say that you always quickly enforce them

That is the specific guidance from SE. Close questions quickly to let the OP rehabilitate them and prevent opportunists from posting answers on off-topic or incomplete questions.

---

the rules does not leverage the community, in fact they limit the community

The rules were set by the community.

---

you don't need to close the question to declassify.

That is how SE works and the way it is done.

---

Both questions are open but the way those questions are handled on the comments is just to remove the focus on the question and try to find anything wrong on the reasoning of the question.

Not at all. In fact, the comments are meant to be helpful and focus the question or provide additional information to the OP that does not necessarily constitute an answer. You are reading something not written.

---

So yes the issues are on the attitude.

I would argue that the only attitude is yours. You seem to say that stating facts is a bad attitude. I am not sure you understand how SE sites work. SE sites are not discussion forums or help sites. The premise is that if you have a question that fits the community rules, others may have the same question, so SE creates a searchable archive of questions and answers, but they must be explicit questions that have explicit answers.

---

Finally on the argument that I delivered more details to the Dell Technician, yes, in fact that's true. But I started with the same question to him, and the following events led with to the solution of the case, and in that case to be honest, the technician only assisted.

Basically, you started a discussion that led to an answer, but, again, SE sites are explicitly not discussion forums. We have chat for that, and you have enough reputation to chat.

Answer 2

For what it's worth I think Ron does a great job keeping the signal-to-noise ratio managed. The reason you see so many closed questions is that there is a lot of noise - a seemingly never-ending stream of people who don't understand subnetting, students copying and pasting homework questions, etc.

My observation is that there tends to be observations like this posted periodically, with a common consensus that the rules are too strictly enforced. What's really missing from these discussions are concrete suggestions about how the rules should change to make things better.

Should we allow questions for non vendor-supported hardware or software? Is there some level of host-specific question that would make sense to allow? Maybe. I'd be open to looking at proposals and I'm nearly positive that others reading this answer would be as well.

If you don't like how things look then make some constructive suggestions. You may find that lots of folks agree with your ideas. You also may find objections that you hadn't considered.

TL;DR That the rules are being enforced isn't our problem. If there's an issue with the rules themselves then improvements should be proposed and discussed.

Answer 3

I would like it if we could move questions to more-relevant Stack Exchanges as an alternative to closing them. I think that might help with user frustration.

AFAIK Super User is a good section for all the home networking questions we see, several per week.

What's the reason we're not able to move questions as part of the moderation process? Can this be improved?