Libcom forum stats for March

Submitted by libcom on April 12, 2007

Comments

nastyned

17 years 7 months ago

In reply to by libcom.org

Submitted by nastyned on April 12, 2007

I can't make out the words. What do the lines means and what are the axes labelled?

Joseph Kay

17 years 7 months ago

In reply to by libcom.org

Submitted by Joseph Kay on April 12, 2007

d'oh, i'll do a bigger image, give me 5 minutes.

pingtiao

17 years 7 months ago

In reply to by libcom.org

Submitted by pingtiao on April 12, 2007

On the bottom it reads: posts, new users and then linear regressions for the two
The axes are: posts on the left and new users on the right.

So the red line and broken line are to be read on the left hand scale, and the black and broken black on the right.

Roughly, this month the graph says we had around 10,000 posts and another ~260 users register.

john

17 years 7 months ago

In reply to by libcom.org

Submitted by john on April 12, 2007

isn't there a problem with measuring new users, in that a slump in new users will make it look like the site's failing, when actually it's doing ok but not attracting new people.

I think there might be a tendency for the rate of new users to fall! -

Jacques Roux

17 years 7 months ago

In reply to by libcom.org

Submitted by Jacques Roux on April 12, 2007

John - yes i would agree thats a limitation, but its in the interpretation - something to bare in mind when looking at the graphs.

AndrewF

17 years 7 months ago

In reply to by libcom.org

Submitted by AndrewF on April 12, 2007

Is there any measurement of active users (ie users who have posted or even logged on in the previous month). That would probably be the most useful measure of actual growth

Jacques Roux

17 years 7 months ago

In reply to by libcom.org

Submitted by Jacques Roux on April 12, 2007

Not atm. When it comes to users they aren't strictly forum stats - people register to submit news/library and never look at the forums for example. We are working on a way of killing off inactive users though to try and make it more accurate.

Mike Harman

17 years 7 months ago

In reply to by libcom.org

Submitted by Mike Harman on April 12, 2007

Drupal does keep records of how recently people have logged in, so could count manually. Could probably also do an sql query that'd work it out from the timestamp. But nothing one clickable.

edit:

OK I went to have a look....

In the past hour and 27 minutes we had 50 different logged in users access the site.

About 190 in the last 24 hours

550 unique logged in users in the past 28 days. I can't tell how many times each of them logged in, just that they logged in at least once in the past four weeks.

Some people register but never actually log in because it sends an e-mail first, so those 230 new registrations won't be anything like 230 of the accesses. Quick count and around 30-40% of new users in the past month haven't logged in yet. A bit random, especially since our auto-e-mail's playing up, but assuming that was accurate though, it'd be about 430+ users who registered more than a month ago who came back

Higher than I expected, that's about 1/6th of the 3,500+ total registered users since enrager (3 years, 29 weeks according to drupal) visiting within a month.

Admins if you're interested go to admin / users - then sort by last access, it's 50 users/page

Steven.

17 years 5 months ago

In reply to by libcom.org

Submitted by Steven. on June 8, 2007

can these images just be put on libcom, so that people whose works block imageshack can see?

actually can we enable image attachments on blog posts?

Mike Harman

17 years 5 months ago

In reply to by libcom.org

Submitted by Mike Harman on June 8, 2007

Just need to add imagefield to blog posts, easy to do.

jack white

17 years 2 months ago

In reply to by libcom.org

Submitted by jack white on September 12, 2007

I don't know so much about this but this blog post I came across highlights some tools that might be useful?
http://feeds.feedburner.com/~r/SmashingMagazine/~3/155547855/

Mike Harman

17 years 2 months ago

In reply to by libcom.org

Submitted by Mike Harman on September 13, 2007

That looks really good, in the same way that heroin looks really good.

jack white

17 years 2 months ago

In reply to by libcom.org

Submitted by jack white on September 13, 2007

eh?

like i said I don't know much about this stuff so why's it like heroin? Cos its addictive or cos it'll make you do the junkie shuffle instead of walking?

Mike Harman

17 years 2 months ago

In reply to by libcom.org

Submitted by Mike Harman on September 13, 2007

the former. Perhaps cocaine would've been a better analogy.