> That would be good. But what would be that parameter? Would it be unique to > installation, and is that anonymous?
The parameter would just be a boolean, e.g.: https://netbeans.apache.org/nb/updates/11.1/updates.xml.gz?firstAutoupdateThisMonth=true These URLs would show up in the update center's web server logs, together with the server-side request timestamp. To find the number of active monthly users, just grep for "firstAutoupdateThisMonth=true", and count the number of such requests per month. No information about the client is submitted, and you wouldn't be able to identify any single client. But the totals for each month would still be correct (assuming clients set the firstAutoupdateThisMonth parameter correctly). -- Eirik -----Original Message----- From: Zoran Sevarac <[email protected]> Sent: Saturday, February 8, 2020 5:02 PM To: [email protected] Subject: Re: Statistics re Apache NetBeans That would be good. But what would be that parameter? Would it be unique to installation, and is that anonymous? On Sat, Feb 8, 2020 at 7:10 PM Eirik Bakke <[email protected]> wrote: > If we could find a reliable way to measure active users, a good way to > respond to all the various unscientific surveys would be to simply > say, on our website, "43,210 active users!" and explain in fine print > how the number is calculated. That alleviates the main concern--that > users may be turning away from NetBeans because they think nobody else is > using it. > > One anonymous way to measure monthly active users would be to count > the number of requests to the update center. By default the IDE checks > for updates every week. To avoid double-counting clients, a HTTP > parameter can be added to the update center request that indicates > whether the request is the first one made in the current calendar month. > > Once monthly active users are measured, a running average over 6 or 12 > months can be calculated for website display purposes, to dampen > fluctuations due to academic semesters and such. > > -- Eirik > > -----Original Message----- > From: Geertjan Wielenga <[email protected]> > Sent: Friday, February 7, 2020 10:36 AM > To: [email protected] > Subject: Re: Statistics re Apache NetBeans > > Well, this is problematic, too. Firstly, all months are not equal, > e.g., every March and October usage of NetBeans is higher than other > months (probably connected to academic years), and secondly, downloads > do not equal active users (three pings to plugin infrastructure equals > an active user, i.e., starts of NetBeans. > > Let’s leave this whole survey discussion now. It’s clear surveys suck > and every year we’ve been getting further in making this point. > > Gj > > On Fri, 7 Feb 2020 at 16:32, Zoran Sevarac <[email protected]> wrote: > > > > > > > Also those statistics don't show anything, unique visitors to a > > > site is not users. > > > > > > > Actually, statistics provided in Excel file by Emilian for Dec and > > Jan include also the number of downloads per month which is very > > likely highly correlated with the number of active users > > > > -- > > Zoran Sevarac, PhD, Associate Professor University of Belgrade, > > Faculty of Organisational Sciences, Department for Software > > Engineering Java Champion > > <https://community.oracle.com/community/java/java-champions> > > | Oracle Groundbreaker Ambassador | Deep Netts > > | <http://www.deepnetts.com> > > Co-founder & CEO > > > -- Zoran Sevarac, PhD, Associate Professor University of Belgrade, Faculty of Organisational Sciences, Department for Software Engineering Java Champion <https://community.oracle.com/community/java/java-champions> | Oracle Groundbreaker Ambassador | Deep Netts | <http://www.deepnetts.com> Co-founder & CEO --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected] For further information about the NetBeans mailing lists, visit: https://cwiki.apache.org/confluence/display/NETBEANS/Mailing+lists
