# Is the beta community representative of the intended audience?

The people involved in this beta were likely recruited somewhat through social networking. I am curious if this has resulted in the beta community being a particularly isolated subcommunity within scientific computing.

For example, I notice that there are four articles tagged with PETSc but no articles tagged with Trilinos despite these being comparable packages. This is unfortunate. scicomp.stackexchange should be helpful to incoming Trilinos users as well.

Q1: Is it the case that those currently involved in the beta are from an isolated subcommunity?

Q2: Is this a problem?

Q3: If so, are there ways we can reach out to the broader scientific community?

• We're still in private beta. It happens that among the people who committed were the people with interest in that package. Let's worry about reaching out when we're in public beta. Which is in ~ 2 days... – J. M. Dec 5 '11 at 1:25
• BTW: this counts as social networking? – J. M. Dec 5 '11 at 1:27
• Comment 1: Who should we be reaching out to and what are some good ways to do this? Comment 2: I was directed to that page by people I know. – MRocklin Dec 5 '11 at 2:09
• Oh, okay. I was thinking that the people who committed to this beta fall roughly into two groups: those guys pointed to here via social networking, like you, and those guys who were already active on other SE sites (and also Area 51). In any event, the site will be promoted by SE when we're open for the public, so I won't be too worried about it now. For now, we should just concentrate on seeding the site with quality questions/answers so that it looks attractive to people who'll see it when this site opens up. – J. M. Dec 5 '11 at 2:58
• An early bias toward just one package like PETSc when others don't get a mention is worrying, and obvious. I noticed it too, and I'm not the observant type. Perhaps it would be polite for the developers to put a disclaimer in any answer that could be interpreted as advertising. This makes things fully transparent. – qubyte Dec 5 '11 at 17:02

I think the answer for any and all private beta sites is "No, this isn't representative". Or more importantly for your issues, it's not big enough to ensure a broad base of opinions even if it was representative. After all, the private beta community is small enough that even if it was a perfectly random sample of "computational scientists", you could have one PETSc user tagging things, and noone familiar with Trilinos.

Whether this remains true or not once we hit public beta, and hopefully have some forward momentum, remains to be seen.

• +1 for the last sentence. – J. M. Dec 6 '11 at 1:07

I will only give information about Q1, as I'm brand new to SE and don't presume to know about Q2.

The current top 4 people (in terms of rep) all know each other very well, and 2 (not me) are PETSc developers. However, we work at 3 different institutions, in different (but related) disciplines.

It is unfortunate that the community is so small that a handful of people are responsible for a very large proportion of the existing questions and answers. Obviously this will have to change over time if the site is to be successful.

### Is it the case that those currently involved in the beta are from an isolated subcommunity?

I expect there are a number of loose clusters formed via social networking, but many like myself will have found scicomp through Area51.

### Is this a problem?

No, whatever cliques there may be now will soon be subsumed as the site grows. You have to remember that we currently only have 188 users and only 16 of these users are considered avid.

### If so, are there ways we can reach out to the broader scientific community?

The best thing we can concentrate on now is to ask good questions and provide good answers.

If we don't have good content, it doesn't matter how much we reach out, people will never return. It we do have good content, people will trust us with their questions and the site will grow.