What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 133 Stack Exchange communities.

I posted a question. This question is kind of important to me, and there doesn't seem to exist a lot of resources on the web for this topic.

While at first glance it may seem to be a question over "religious zeal" of programmers about their favorite respective frameworks (that way calling for endless debating), it actually asks about technical aplicability / best suitable tool to achieve something.

On basis of technical details of python frameworks & tools.

The information is needed and important.

Can I, and how, get the question reopened?

share|improve this question
My favorite tool to hammer in a nail may be a rock, yours may be a hammer, and someone else's may be a frozen banana. Is one the best? You might be able to narrow is down by asking if there is an industry standard, but even that can be iffy. –  jprofitt Jun 20 '13 at 0:36
I am asking what is technically the shortest path to achieve something. Different tools are designed with different emphasis and for solving different groups of problems. –  toninoj Jun 20 '13 at 0:42

5 Answers 5

Think about this in the context of Gorilla vs. Shark. What you're asking can be distilled into a FFA cage match between Django, Pyramid, and Flask. Also, if someone decides that it's Shark Week, you're going to also get recommendations for web2py and CherryPy as well. Or the network decided that it wants to retool, so you're getting a suggestion to work with Ruby on Rails. Heaven help you then.

In essence, here's what I see from your question:

Dynamic subdomains with Python

How do subdomains relate to Python specifically? That would be handled more by the backing web server than anything else.

What would be the shortest path, technically, to create a website, using python, with a lot of dynamically created subdomains, for different users etc.

Two problems right here:

  • "Shortest path" is highly subjective. The shortest path to developer A may seem like an unnecessary jaunt through the forests and trees for developer B, and a simple hack will do, which will send developer C's blood pressure through the roof.

  • Dynamically creating subdomains makes it sound more like a problem dealing with Apache or whichever backing web server you're using - and there is a treasure trove of answers and articles on the Web for this particular problem domain. There may even be some out there on Serverfault, too.

And also, assigning real user domains to different users / other "objects" within a web app.

I'm not sure what you're asking here. Why would the web application care about domains so long as it can serve content?

Are other python framewrorks like Pyramid, Flask, etc, designed to be a better fit for problem like this?

So at this point I do want to give you a ton of credit - you have a general idea of what the problem you're trying to solve is - you want dynamically generated subdomains. But you're still looking for love in all the wrong places (or so it seems).

At this point in time, the question you have there is simply not answerable without getting into a long-winded debate about which framework is better suited to solve an Apache/backing web server problem.

Don't feel too bad about it though - there's a chance you can save it. Remember when I identified the specific problem? You should be asking the question related to dynamically creating subdomains. You could ask the side question, "would my specific choice of Django have anything to do with/help/hinder the ability to dynamically create subdomains?"

This way, your question is now pointed, with a specific focus, giving a potential answerer the domain they need to sufficiently answer your question, and for you to avoid it being closed again.

share|improve this answer
I am asking about a way to programatically create subdomains. So, its not practically solved with servers. I am asking for a way to serve dynamic content, from a same application, for different subdomains (also domains). It is not a merely "personal pereference" question. It is technical. –  toninoj Jun 20 '13 at 0:57
@toninoj - You're asking what essentially amounts to a shopping question. If your question is of the form "What's the best way to do X?", then it's not a good question for Stack Overflow. Period. Try Quora or Yahoo! Answers. –  Jack Maney Jun 20 '13 at 0:58
Alright. First off, don't get discouraged by others' comments. I'm here to help you understand, in some capacity, why this question should've been closed. Second, from my own professional experience, the only involvement any webapp I've used/written has had with a subdomain is making sure that the domain it's pointing to is the domain it should be writing to. I have strong opinions that it's more a webserver problem than a Python/framework problem, and there are articles/answers on the web/Serverfault.SE to support that line of thought. –  Makoto Jun 20 '13 at 1:00
Bear in mind, even technical questions can be highly tangential and subjective, so it's important to limit the scope of what you're asking to something that is specific. Involving multiple frameworks undermines the very definition of "specific". –  Makoto Jun 20 '13 at 1:02
@Makoto it isnt a server issue since I need to create subdomains programatically and programatically spit-out different content for them. So it HAS to do with servers, but it also involves a framework. Thats why I need to hear voices "from the trenches". –  toninoj Jun 20 '13 at 1:03
And specific tools are just better fit for certain applications. It doenst inevitably involve "my tool is the best" debates. –  toninoj Jun 20 '13 at 1:04
Feels like it is, though - despite my opinion (whoops!) that a web server configuration would be better (and there being answers to support this ), you wish to insist that it has to be done programatically. I'm merely telling you that there exists an answer, and with the right bit of phrasing, you can get that answer either on SO or from Serverfault. I'm not saying you're saying some tool is the best. I'm saying that you're saying too many. –  Makoto Jun 20 '13 at 1:08
Mind you, even though I didn't want to dive into this on Meta, we've kind of demonstrated that this sort of question can lead to arguments, extensive polling, debates, etc etc etc - all things that Stack Exchange wishes to avoid. I'm not upset/mad/frustrated or anything like that, but I do wish for you to understand this close reason - it's leading to a debate. I don't wanna debate; I want to help. My advice: narrow your question scope, and be open to different avenues to accomplish your goal. –  Makoto Jun 20 '13 at 1:09
Okay. Not bad answers you gave.. I say programatically, btw, cause I will take usernames or other dynamycally created objects from database and serve subdomains from that. Ill check the links you gave.. Thanks. (I kind of think it is mostly about server-framework simbiosis. So Im not asking for "opinions", more for experiences.) –  toninoj Jun 20 '13 at 1:13
@toninoj - You may find this Stack Exchange blog article helpful: Good Subjective, Bad Subjective –  JDB Jun 20 '13 at 2:20

Your question raises the 'not constructive' flag because it has some phrases we commonly associate with those types of questions.

Underneath all that, there is a great question. If you edit it, it's likely to be reopened.

Here's how you can do that:

I need to be able to construct subdomains programmatically per user. I've tried using Django to do this (insert what you've tried here), but it wasn't straightforward and didn't work for me for (insert reasons why it didn't work, or is not easy to maintain/etc).

How can I use Django (or any other python web framework) to programmatically give a user their own subdomain?

This takes out the framework fight; it replaces a "Best way" to, "Help me get this working", and because you have your own skin in the game (with explaining what code you've written, and why it doesn't work), it gives the community the idea that you've done some work on your end, and don't just want us to do your work for you.

share|improve this answer

I dont think a question was “non constructive”. Should I improve it in any way?

Yes! Absolutely! The point of closing a post is to give the author, and the community, time to revisit it and edit the post with the goal in mind of getting it reopened. Sometimes, posts may be closed that you disagree with. As a user on several Stack Exchange sites, I witness this all the time.

If we disagree with the closure, we have two choices:

  1. Try to convince others that the post should be reopened.
  2. Find out what other users find wrong with the post, and fix those problems, even if you disagree they are actually real problems with the post.

I find that option 2 tends to work out really well. If there are potential reopen voters who are sitting on the fence, we can convince them to join our world by solving enough of the problems with the post to sway their opinion in our direction. We may even convince some close voters to vote to reopen.

Of course, we can try to convince others that the post was wrongly closed, and sometimes this may get a post reopened. Unfortunately, we lose a huge opportunity to make vast improvements that would only make the post better, regardless of whether or not the post was really closed prematurely. Thus, if a post is closed, and you see why people closed it, and you know how to fix those problems, even if you don't personally view them as a problem, then by all means, please fix them!

This sometimes leads to some really great Q&A! :)

With that said, as many others have pointed out, comparing gorillas and sharks is tough to provide an objective, factual answer. Thus, to improve the post, you may actually need to throw out the entire thing and start from scratch.

There's a reason you're asking this question though, and it is sort of broad, but it's also answerable for someone who has done what you're trying to do. Since I don't think you're really asking for a step by step tutorial on how to implement from A to B a domain management system, I did what I can to edit this for you and get the ball rolling. I'm hoping the way I reworded it works for you. If not, please continue editing.

I will say that I personally don't think this is Gorilla vs Shark. The Gorilla vs Shark question simply asks "Who would win in a fight, a Gorilla or a Shark". This particular question is entirely theoretical, as the likelihood of these two entities meeting is quite slim.

However, trying to determine which tool will better solve a problem is not the same thing. Some tools are indeed built for specific purposes, and there may very well be an answer to that problem that is based on facts, references, and specific expertise, one where an open source tool developer perhaps wrote something to specifically deal with this problem.

One suggestion, compose your post as clearly and as organized as possible. If you pay attention to details like spelling, grammar, and clearly stating your problem, posts are less likely to attract negative attention. I'm not sure this can be reopened as there arguably could be other problems with it, but I do hope this helps!

share|improve this answer

The question was not constructive because:

As it currently stands, this question is not a good fit for our Q&A format.

This is because your answer does draw on specific expertise, or facts. Instead, it solicits debate, argument, and extended discussion; specifically, "What is the best way to do this?" solicits debate and argument:

We expect answers to be supported by facts, references, or specific expertise, but this question will likely solicit debate, arguments, polling, or extended discussion.

Your question: "Why is it not constructive?" can be answered by reading the close reason. I recommend reading this article which gives an overview of which questions are acceptable. Specifically, see:

  • a specific programming problem


  • practical, answerable problems that are unique to the programming profession

You haven't really posted a specific programming problem. This is more of a general, "How do I do this?" problem, which cannot be reasonably answered by anyone. You have provided project specifications, and are asking for guidance from the community as to how you should implement those requirements.

This line indicates that your question is highly subjective. It also could cause extended debate between developers over which tools are best, and which methods are best.

What is the best way / tools to code this, and to implement this (servers)?

You then ask about multiple tools which are used in this type of development; this becomes very general, and turns into a comparison between tools.

Your answer in its current, edited form happens to be more of a shopping question; please see "Q&A is Hard, Let's Go Shopping!" from the StackExchange blog for more information as to why these types of questions aren't generally appreciated.

share|improve this answer
-1; This answer is not useful (click again to undo). –  Jeremy Banks Jun 20 '13 at 0:26
The problem I posted is specific. What is the best tool (among 3-5 tools available), that offers technically the shortest path to achieve a very specific problem. I asked if anyone has any experience in solving that specific problem with those specific tools. What is wrong with that? –  toninoj Jun 20 '13 at 0:33
@toninoj It's not a specific problem. It depends on your implementation, your desired architecture, as well as about a hundred other things. There have been (albeit short) books written on selecting the best architecture and frameworks for web applications. Additionally, your question will solicit debate between users over which solution is best. –  Emrakul Jun 20 '13 at 0:35
I edited the question. I tried to make it more specific. Is it better suited now, or will I need to seek answer elswewhere? –  toninoj Jun 20 '13 at 0:40
@toninoj Please see my answer edits; currently, it fits more the paradigm of a shopping question. I suggest reading this: blog.stackoverflow.com/2010/11/qa-is-hard-lets-go-shopping –  Emrakul Jun 20 '13 at 0:43
I just need a freaking answer or a suggestion. Great. It isnt "pure" enough. Cool. Ill go elsewhere. Thanks guys. –  toninoj Jun 20 '13 at 0:44
@toninoj Pure is not the problem here; it doesn't have a specific problem, it's too broad. Did you read any of the articles linked? –  Daedalus Jun 20 '13 at 0:49
Yes, I did. My question asks about which framework has best support, regarding to routing and dynamic creation of subdomains. I also wanted to get some experiences of users who tried doing that with django, pyramid, flask etc. What is unspecific, unconstructive or subjective aboput that? Some tool may offer ten miles workaround for doing that, while some other tool may be designed with such goal in mind. I want to find out about it & get specific experiences. What is unconstructive or vague about it??? –  toninoj Jun 20 '13 at 0:52
This feels like a mathematician's answer. Yes, technically it's correct, but this says nothing of how to improve the question, or for the OP to avoid having their questions closed in the future. –  Makoto Jun 20 '13 at 0:52
@toninoj: Every single person ever could answer that with something different and it would be completely valid, so it’s subjective and therefore not a great fit for Stack Overflow. –  U2744 SNOWFLAKE Jun 20 '13 at 0:53
@toninoj - If you need suggestions or recommendations, go elsewhere. That's not what Stack Overflow is for. –  Jack Maney Jun 20 '13 at 0:54
@minitech not true. There are problems that different tools are designed to solve better. They are desinged with certain problems in mind. For example, RoR is more suitable for creating restful interfaces than django. And for creating one-page applications. It isnt a matter of personal preference, it is technical fact. Just like my question. –  toninoj Jun 20 '13 at 1:00
@toninoj: RoR is designed for creating RESTful interfaces, but I will gladly explain to you how to do that in Node and why it’s better. It’s also my opinion. See? –  U2744 SNOWFLAKE Jun 20 '13 at 1:02
jQuery is the best tool for everything. –  Emrakul Jun 20 '13 at 1:34

Many users of StackOverflow seem to automatically deem question unconstructive if it has anything resembling "X vs Y" dilemma.

It is often justified, but not always.

Sometimes that approach is unconstructive, and denies us the answers to some valid questions.

As one commenter on "Gorilla vs Shark" blog post correctly noticed,

Yeah, commonly questions “Gorilla vs Shark” suck, but taking the rule “everything with ‘\S+ vs \S+’ in title should be closed” is very wrong, but very usual mistake. There are questions of “What is better tool to solve this task: Gorilla or Shark”, which can be objectively answered by giving cons and pros about both tools, so question could choose. It’s very normal question as I think. But people are to afraid of them and close them. It’s not right!

And as one more commenter wrote,

People will have different opinions as to which one is better for doing one task over the other, but listening to them helps a person to decide which one they would prefer. Even after 31 years I continue to learn from reading how others use these languages. I can filter out the “my language is better then yours” people. It’s easy to to that. To not have the question because these kinds of people exist doesn’t make sense. Answering them rather then trying to word smith or eliminate the question is more productive. To disallow these because you don’t need it is wrong. You say you can’t read the site anymore is ridiculous, you don’t need to read the questions and answers that don’t apply to you. That doesn’t mean others don’t want to read them.

share|improve this answer
How does this relate to the OP's specific question? –  Robert Harvey Jun 20 '13 at 1:31
@Robert The answerer is the OP. –  Emrakul Jun 20 '13 at 1:35
Ack. If you already knew the answer, then why did you ask the question? –  Robert Harvey Jun 20 '13 at 1:35
Well, I discovered the answer from the moment I asked till the moment I answered it. All these wonderful, helpful and deeply constructive answers here reiterated it for me :) –  toninoj Jun 20 '13 at 1:36
Have a look at my answer. :) –  Robert Harvey Jun 20 '13 at 1:39
Yeah, it is thorougly insightful. I would never have thought of that myself. –  toninoj Jun 20 '13 at 1:40
Assuming that you're being sarcastic, why wouldn't the point of view that I outlined be legitimate? Read this, if you're mind hasn't already closed: meta.stackexchange.com/a/164981 –  Robert Harvey Jun 20 '13 at 1:42
I read it. The answer to that is - Stack Exchange solves the issue with voting & accepting answers. That is quite enough to not need to close questions that can bring some useful experiences & insights to users. And its quite enough to filter the noise. And Im still searching for the answer to my question, regretfully I will hardly find it on SO, as it is. –  toninoj Jun 20 '13 at 1:47
Try the Python Chat Room. Maybe they can offer some insight outside of the Q&A environment. –  Robert Harvey Jun 20 '13 at 1:50
My mind isnt closed. Its just that the question is still valid and worth answering. Thats all. It isnt a very smart thing to close every "X vs Y" question out there. –  toninoj Jun 20 '13 at 1:50
-1 A pros and cons list for each option is not even close to objective. –  U2744 SNOWFLAKE Jun 20 '13 at 2:32

You must log in to answer this question.

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