[drupal-devel] Community density
Hi! Last night I asked Steven to send me data about how many users have 1 node, 2 nodes and so on. Of the 22695 users, 18219 have zero nodes. This means that most of the users are just commenting or not contributing at all. If we want to size the "hard core" of the community, there are 66 users who have written 25% of all nodes, they have at least 39 nodes and 78 nodes in average. A bigger circle is the 302 user who have written 50% of the nodes. They have written at least 12 nodes, the average is 32 nodes. Given that the definition of "belongs to the community" is a very subjective one, I say we have a 300 user community which is real amazing. 22K, of course sounds better, but those are not active. Regards NK
-1 on purging lurkers. Karoly Negyesi wrote:
Last night I asked Steven to send me data about how many users have 1 node, 2 nodes and so on.
Of the 22695 users, 18219 have zero nodes. This means that most of the users are just commenting or not contributing at all.
These numbers are not surprising, in fact, they verify that the Drupal community is a scale-free network as are most on-line communities. I did a similar analysis last year of an on-line community for social entrepreneurs and got similar numbers. I was excited as this showed that this community was behaving as would be expected for a scale-free network. When I presented the analysis as part of a review of Albert-László Barabási's book, "Linked: The New Science of Networks" (http://www.nd.edu/~networks/linked/), I was summarily read the riot act and told that the data was private and the article was not to be published on the community's web site nor anywhere else. In other words this on-line community took the "appalling results" of lots of non-participants as an indictment of the failure of the community. On the contrary, the community was behaving exactly as a scale-free network behaves. "Non-participant" members are members of the community as much as anyone. Lurking (and learning) is a form of silent/invisible participation, but it is participation nonetheless. To delete someone for not posting nodes in an inadvertent slap in the virtual face. Joining an on-line community is a way of defining who we are. I would sure hate to tell somebody that their decision to be part of the community was rejected by the community because his or her style of participation was somehow "not worthy" to borrow Wayne and Garth's turn of phrase. For more on this phenomena, see Clay Shirky's excellent blog article "Power Laws, Weblogs, and Inequality" (http://shirky.com/writings/powerlaw_weblog.html). -- Jim Salmons and Timlynn Babitsky Founders and Research Directors Sohodojo - http://sohodojo.com
On Fri, 6 May 2005, Sohodojo Jim wrote:
-1 on purging lurkers.
Nobody proposed this. I'd like to see a more general evaluation of the drupal.org users nevertheless. While it will be difficult to trace cvs contributions (often somebody commits patches for his project that somebody else wrote, I have added a lot of translation files without speaking the language etc), we should track the comments in a similar number that we tracked nodes. I'd be especially interested to see if there are users that have a lot of comments, but nearls no nodes. Ideally, this would be correlated with the time the user is a member of drupal.org. Lots of fancy graphics can be done. :) Cheers, Gerhard
While it will be difficult to trace cvs contributions (often somebody commits patches for his project that somebody else wrote, I have added a lot of translation files without speaking the language etc
We currently have a way to link drupal.org node ids in cvs commit comments (e.g., 'Applied patch from issue #12345', which will appear as a link on drupal.org). Were we to do something similar with drupal user names, we could do tracking. E.g., 'Applied patch from %dries'. Then we could explicitly track this sort of contribution. Of course, we could simply search for user names without this type of selector, but there would be more false matches.
I definitely agree here, I consider myself to be a lurker, not only here, but several other places as well. However, after I feel that I have been brought up to speed on things, my goal is to jump in feet-first and become an "active, participating" (rather than "passive, lurking") member of the Drupal community. Having been just introduced last week to the Drupal system (and immediately adopting it over the several previous CMSs that I've worked with), I'm still learning how the software works and how the community works. I have gained invaluable knowledge from being a "passive, lurking" member of other communities, where I later became one of the most active members. --Matthew Hildebrand p.s. My background is in editing/tech writing & software QA, so I'm trying also to decide where I can be most useful in the community (and still feel like I'm enjoying myself as well). -----Original Message----- From: drupal-devel-bounces@drupal.org [mailto:drupal-devel-bounces@drupal.org] On Behalf Of Sohodojo Jim Sent: Friday, May 06, 2005 8:46 AM To: drupal-devel@drupal.org Subject: Re: [drupal-devel] Community density... And the Lurkers shall Inherit the Web -1 on purging lurkers. Karoly Negyesi wrote:
Last night I asked Steven to send me data about how many users have 1 node, 2 nodes and so on.
Of the 22695 users, 18219 have zero nodes. This means that most of the users are just commenting or not contributing at all.
These numbers are not surprising, in fact, they verify that the Drupal community is a scale-free network as are most on-line communities. I did a similar analysis last year of an on-line community for social entrepreneurs and got similar numbers. I was excited as this showed that this community was behaving as would be expected for a scale-free network. When I presented the analysis as part of a review of Albert-László Barabási's book, "Linked: The New Science of Networks" (http://www.nd.edu/~networks/linked/), I was summarily read the riot act and told that the data was private and the article was not to be published on the community's web site nor anywhere else. In other words this on-line community took the "appalling results" of lots of non-participants as an indictment of the failure of the community. On the contrary, the community was behaving exactly as a scale-free network behaves. "Non-participant" members are members of the community as much as anyone. Lurking (and learning) is a form of silent/invisible participation, but it is participation nonetheless. To delete someone for not posting nodes in an inadvertent slap in the virtual face. Joining an on-line community is a way of defining who we are. I would sure hate to tell somebody that their decision to be part of the community was rejected by the community because his or her style of participation was somehow "not worthy" to borrow Wayne and Garth's turn of phrase. For more on this phenomena, see Clay Shirky's excellent blog article "Power Laws, Weblogs, and Inequality" (http://shirky.com/writings/powerlaw_weblog.html). -- Jim Salmons and Timlynn Babitsky Founders and Research Directors Sohodojo - http://sohodojo.com
Hi - very interesting analysis! Another form of contribution is via the mailing lists - it might be interesting to look at posting distribution there, too. Another point - I realize that Karoly wasn't trying deprecate the role of non-posters in the community, but it's probably worth emphasizing that the 'tail' of this distribution is an important part of this 'user-community', namely the users. Perhaps we should think about ways to encourage the silent majority to contribute useful input? I'm also thinking that we humans spend much of our lives 'consuming' before we start 'producing', and this is a typical behaviour when we are introduced to new contexts. Lurking before posting is generally recommended etiquette. So, it might be interesting to look at the distribution of 'lurking while waiting to post' times - the time between signing up and making the first posting. This might help us to get a sense for the future growth of the community. Nice work! Djun On 6 May 2005, at 3:57 AM, Karoly Negyesi wrote:
Hi!
Last night I asked Steven to send me data about how many users have 1 node, 2 nodes and so on.
Of the 22695 users, 18219 have zero nodes. This means that most of the users are just commenting or not contributing at all.
If we want to size the "hard core" of the community, there are 66 users who have written 25% of all nodes, they have at least 39 nodes and 78 nodes in average.
A bigger circle is the 302 user who have written 50% of the nodes. They have written at least 12 nodes, the average is 32 nodes. Given that the definition of "belongs to the community" is a very subjective one, I say we have a 300 user community which is real amazing. 22K, of course sounds better, but those are not active.
Regards
NK
does the "node" analysis include comments? Dan
Hi - very interesting analysis!
Another form of contribution is via the mailing lists - it might be interesting to look at posting distribution there, too.
Another point - I realize that Karoly wasn't trying deprecate the role of non-posters in the community, but it's probably worth emphasizing that the 'tail' of this distribution is an important part of this 'user-community', namely the users.
Perhaps we should think about ways to encourage the silent majority to contribute useful input?
I'm also thinking that we humans spend much of our lives 'consuming' before we start 'producing', and this is a typical behaviour when we are introduced to new contexts. Lurking before posting is generally recommended etiquette.
So, it might be interesting to look at the distribution of 'lurking while waiting to post' times - the time between signing up and making the first posting. This might help us to get a sense for the future growth of the community.
Nice work!
Djun
On 6 May 2005, at 3:57 AM, Karoly Negyesi wrote:
Hi!
Last night I asked Steven to send me data about how many users have 1 node, 2 nodes and so on.
Of the 22695 users, 18219 have zero nodes. This means that most of the users are just commenting or not contributing at all.
If we want to size the "hard core" of the community, there are 66 users who have written 25% of all nodes, they have at least 39 nodes and 78 nodes in average.
A bigger circle is the 302 user who have written 50% of the nodes. They have written at least 12 nodes, the average is 32 nodes. Given that the definition of "belongs to the community" is a very subjective one, I say we have a 300 user community which is real amazing. 22K, of course sounds better, but those are not active.
Regards
NK
On 6 May 2005, at 12:41 PM, Dan Robinson wrote:
does the "node" analysis include comments?
Dan
Judging from Gerhard's message in this thread, no... it would be interesting, though! Djun
I'd like to see a more general evaluation of the drupal.org users nevertheless. While it will be difficult to trace cvs contributions (often somebody commits patches for his project that somebody else wrote, I have added a lot of translation files without speaking the language etc), we should track the comments in a similar number that we tracked nodes. I'd be especially interested to see if there are users that have a lot of comments, but nearls no nodes. Ideally, this would be correlated with the time the user is a member of drupal.org.
Lots of fancy graphics can be done. :)
Cheers, Gerhard
participants (7)
-
Dan Robinson -
Gerhard Killesreiter -
Karoly Negyesi -
Matthew Hildebrand -
Nedjo Rogers -
puregin -
Sohodojo Jim