It has finally happened!

Hi, Mobi,

On an old Puppy Linux, I've surfed here with Links as well. Good to see the alternative approaches living on.

Obviously, text-based porn is as heartbreaking as it implies, but at least we're up on all things H/PC.

Jake

What about Windows 8.1?! hahaha nice. I'm on Firefox here (except on Surface RT, which I use IE11 on)

The numbers are irrelevant at this stage.

hehehe, oh C:Amie. You're quite the 8.1 evangelist, aren't you?

0.2% if you must know

Ha, probably all me!

Now
50.05% for 7
30.08% for XP
7.32% for 8.0
4.05% for Vista
3.91% for 8.1

Chrome 30 was October's top browser, followed by Firefox 24 and finally IE10.

Chris, does Linux Chromium and Windows Chrome come up as the same?

Jake

Jake,

Looking at the chromium agent, it appears to send its own as well as chromes, therefore it will be aggregated in chrome. I've made a note to separate it out in 6.0.12 of Analog.

Now just figure a way to separate it out on user login. That is once a user has logged on using one system during a month, it doesn't continue to count his subsequent logins with the same system...

Maybe you could tie it into the IP address to disallow for repetitive system counts...

It would be a wholly pointless exercise for Analog. The use of NAT means that you could wind up with hundreds or thousands legitimate of user accesses vanishing.

If you want to do that, you simply need to restrict the scope of the report output i.e. generate stats hourly, daily, weekly, monthly and then work out what you want to know over a shorter period.

With the forum stats, that's easy, it's a SQL query to count unique instances of a user id inside the last 30 days. You could do the same with IP address, but that comes back to the same NAT problem.

I thought that the whole idea was just to provide information on the type and frequency of web browsers that were visiting HPCFactor. Seems to me if the data collection counted the same high-frequency visiting person over and over, then the data would be "tainted." Like CEGeek and his Firefox. He alone may be responsible for all of those visits.

I didn't realize that it had anything at all to do with user access authorization...

It depends on what matric you want. For an advertiser, 1000 page impressions means that the advert was displayed 1000 times, 100 of them could have been to the same user, but that's academic.

What you are asking after is actually session counting rather than hit counting. The HPC:Factor servers consider a session to be about 20 minutes of inactivity, rather than anything as extreme as a month. A trackable session is thus anyone who comes to the site, does anything for as long as they like up to the point that they stop doing anything for more than 20 minutes. At their next contact with HPC:Factor they are considered to be a new session.

There are only two ways that we can do that, the forum session or the ASP session. I don't log the latter, so it would have to be forums.

Non trivial percentages of the page view and hit counts are bots.

You know all those fancy words....session counting...who would have known...

You know you love it