Lately there's been said a lot about factionalism, self-determination and existential fears in community, especially with respect to Developer Testing website. It seems, lots of people are overwhelmed by sheer number of questions on SO and feel like their particular field (software testing in this case) becomes belittled or unimportant there. (Regardless of whether there's a technical way to watch for interesting tags only)

Now, let's turn to SQL. There we have a concept of views: we can select whole table or just some set of rows and call it a 'view'. And this view we can treat just like a normal table: execute select, join, update or insert. We don't even have to know it's not a table.

Applying same concept here, what if all SO questions tagged 'testing' and 'unit-testing' were automatically displayed on a 'faux website' testing.stackoverflow.com? And questions created through testing.stackoverflow.com interface were actually saved in SO database with additional 'testing' tag.

I understand that this is big, labour-intensive proposal which can't be implemented on a whim. And still, it feels like the current way of chopping off communities, 'all or nothing', doesn't always work. Maybe we could benefit from an interim solution, offering both small cozy home for testing experts and one big community for all programmers.


2 Answers 2


I believe the best of both worlds is to shore up and add features to the Stack Exchange engine that allow (and ENCOURAGE) groups of users to form user groups (sub-communities) so they can commune in their own dedicated areas.

Similar to what you were suggesting, the tag wikis and tag grouping is just the very beginning of tapping that functionality. Chat is another resource. It's not just about seeing your own virtual subset of the questions. It's about users feeling like they are part their own community; having that sense of place.

The Testing People aren't going to have that same feeling of "group of my peers" on Stack Overflow like they would on their own site. But you don't want to create an ineffective Q&A system either just to fulfill some sense of segregated ownership.

The ideal situation is to enhance these systems to give each group their own personal slice of a much larger community. That's the middle ground I would strive for.

  • Excelllent answer. There's another issue around segregation - take the tex.sx porposal - [latex] tag was active on SO and attracted some experts, but the "long tail" of Latex qns tended to be a better fit for SU and had repeatedly ran into problems on SO & there was also a problem with tags - some tex.sx questions, such as those about ConTeXt and Metapost, did not fit the tag. The larger group may not always be the superset it at first appears. Oct 3, 2010 at 8:43
  • Thanks. It's good to know admins are watching this issue too: for a moment it seemed the whole 'community splitting process' went a little bit too far. Oct 4, 2010 at 2:43

Interesting tags can be setup, but they still won't ensure that niche questions are viewed and answered. I have often found that some obscure and pointed questions fail to get answers or even views. Even adding a bounty doesn't help. SO currently has 160-170 featured question and obscure ones are quickly relegated to the last page. Unless SO actively promotes less interesting questions, I don't see a solution to this. And SO sister sites though they attract specific crowds, still do not have enough traffic to actually make a difference.

(The stackexchange pull down menu on the top left is a good thing.)

Update: @ Nikita Rybak

SO => massively more eyeballs although non-specific NewTesting.SE.com => specific eyeballs

  1. Difficult testing question on SO

  2. Difficult Testing question on NewTesting.SE.com

  3. Easy testing question on SO

  4. Easy Testing question on NewTesting.SE.com

I guess 2 will beat 1, but 3 should beat 4(atleast in the shortterm, in terms of answers). Although for that(2 vs 1) matter, a new website is recommended.

In the long term a new website should work. Defining long will be difficult though.

  • 2
    Well, 'hard questions' is a different issue. People seek to optimize upvotes/effort ratio and 'obscure answers' aren't likely to get any upvotes. I just don't see how's it related to the testing community feeling deprived. Even on a separate testing website, obscure, wordy and difficult to understand question won't get a lot of attention. Oct 2, 2010 at 17:47

You must log in to answer this question.

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