9

This is a huge request, but I would love to see http://data.stackexchange.com/ built on top of the Stack Exchange Q&A engine.

The data explorer is great for exploring on your own, but really bad at searching for other people's answers or even your own past queries because everything becomes duplicated for every time the search is run, final product or otherwise.

One of the issues that can be addressed is the ability to delete your own query on data explorer.

If there's a need to keep records of searches for auditing purposes, surely this can be done on the backend.

People without a great understanding of SQL could explain what they're trying to do as a question. And other people could provide the execution by answering the question. We'd have automatic version control for people to update their own answers and a mechanism for other people to provide different answers or different questions structures. Voting on answers could help get good answers to the top.

If the query itself was built with a tool like Stack Snippets, each post could have a syntax highlighted version of the code and the ability to run inline. If you wanted to edit the query, you could open up the post in a full screen editor similar to current query page.

People who just wanted to write an ad hoc query could post a question and answer at the same time. By adding room for free text in the post description, past queries could be much more easily searched and explained.

An integrated service could provide better linking between accounts. In the future, perhaps that could be opened up to queries against things that require authentication. 10K users could query deleted posts. Individual users could query their upvote and downvote history on a particular site using that sites.

It's a big request, but could provide a much more consistent experience and branding across different Stack Exchange tools.

Here's a design concept of the idea

Proof of concept

|improve this question
  • 1
    SEDE is open-source so you could help out yourself... – rene Dec 22 '14 at 18:53
  • 1
    @rene, but the stack exchange engine isn't. Why reinvent the wheel? – KyleMit Dec 22 '14 at 18:55
  • 1
    I think you should rephrase you feature request to avoid confusion. What I think you really want are data.se snippets to run on any (meta) site vs. a single site for that purpose. – m0sa♦ Dec 22 '14 at 19:03
  • @サモライ, that's only a very small part of my request. Snippets would provide an adequate vehicle to preserving runnable code, but the request, as written, is really to structure the data explorer as a Q&A style site based off the stack exchange engine so it can leverage the power of searchable, revisionable questions and answers. – KyleMit Dec 22 '14 at 19:08
  • "searchable, revisionable questions and answers" - that's exactly why we added snippets @KyleMit – m0sa♦ Dec 22 '14 at 19:14
  • 2
    Related: Running queries to StackExchange Data Explorer from anywhere, Add tags to categorize the queries on SEDE, Can we have a feature to allow commenting on Data Explorer queries? – Tim Stone Dec 22 '14 at 19:18
  • 6
    Personally, I'd rather fix the issues with Data Explorer's query management and then also maybe integrate running queries from posts. – Tim Stone Dec 22 '14 at 19:19
4

I definitely understand the need for better organization with SEDE, but I don't think that Q&A is going to work (at least the way you've presented it). Mainly because I picture it often going like this:

The question and the answer I wrote are downvoted, because the question is only useful to me and the answer's not finished. Pleads of "stop downvoting" are ignored.

I'm not sure what everyone else is doing, but a lot of the queries I make are not the type I need other people seeing or downvoting. Here are the potential problems I see with how my queries are typically written:

  • The query I wrote (or edited) only returns results that will matter to me.
  • I only designed the query to get "good enough" results and have known bugs in the code.
  • The code isn't formatted.
  • Parts (or, more likely, most) of the code has been taken from elsewhere without attribution.
  • There is no title or description.
  • I'm writing something that has definitely been done before (i.e. a duplicate), but it's faster to do it from scratch than to search, find, and understand someone else's stuff.

Although it's not a problem for me, it's also possible that SEDE is being used by people who don't speak English well enough to participate in an English Q&A site. Keep in mind that SEDE covers four sites where the UI language isn't English and countless other bilingual sites, where it's possible to post entirely in a language other than English.

Were we to change the system to Q&A, freedoms like these (including the ability to be lazy without penalty) would be lost for those who want to save their queries.

What I can support is the ability to embed SEDE queries in meta posts. Meta is already the place to go if you have a query that you want to formally publish. Meta also allows you to suggest other methods of doing something: regular search, userscripts, etc. In addition, each site has a meta, so queries that are only interesting to one site can be posted to that meta.

I can also support the ability to delete your own queries and better searching.

|improve this answer
  • I think that raises a fair point about how much querying would need to be done essentially in a "draft" or "private" mode which isn't currently a capacity that the Q&A engine has, so there would be another heavy lift to make sure it was hitting those user stories – KyleMit Jan 15 at 21:01

You must log in to answer this question.

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