number of Guests

[winemaker81]

I don't normally pay attention, but I scrolled to the bottom of the forums list and spotted the number of logged in members. That wasn't surprising -- it was the number of guests:

Total: 492 (members: 40, guests: 452)

10 times as many guests as registered members ...

 

[crushday]

I guess they like ads...

 

[Johnd]

I've seen similar numbers before as well. My guess is that there are plenty of folks who use the forum to research the topics of interest and harvest the information without ever joining as a member. There are probably also lots of robo-computers digging around, spamming, information mining, and the like, not real people using the forum. Just a guess on my part, some of our more computer driven members probably know way more about the automated computer activities.......

 

[winemaker81]

Various pundits say web traffic is about half bots, e.g., some type of non-human visitor. Even so, that's 225 visitors vs. 40 members.

This makes sense. When I research a topic, I may hit half a dozen forums, just browsing. Especially technical forums, I may visit one infrequently and have no need to post, so I don't register, especially since so many forums are a toxic environment.

The tremendous imbalance of visitors to members caught me off guard.

 

[balatonwine]

Various pundits say web traffic is about half bots, e.g., some type of non-human visitor.

That is really only a guess. Not even an average. Because it actually varies quite a lot. It depends on too many factors, one of which is the popularity of the site. An unknown site, for example, may have almost only bots and not real visitors, for example. So "about half" for that site is an incorrect estimate. I have managed my business site for 20 years, and "about half" is not accurate for my site either. And it can even vary a lot -- by year, month, quarter, etc.

The real answer to this question is for the site admins to look at the logs and report the actual traffic sources (people, bots, unknown). All else might (or might not) be invalid speculation

 

[Johnd]

If you go back to the main screen, down in the lower left corner, and click on "Members On Line", it pops up a screen that you can actually see everybody, not just members................. When I looked at only the first page just now, 5 of 20 were robots, the other 15 were guests or members. You can tab between Everyone, Members, Guests, and Robots............................

 

[Johnd]

Looks like most of our robots are Baidu, Proximic, Bing, GoogleAdSense, Yandex, Brandwatch, and Google, with Baidu and Proximic making up most of the the robot activity.

 

[balatonwine]

If you go back to the main screen, down in the lower left corner, and click on "Members On Line", it pops up a screen that you can actually see everybody, not just members................. When I looked at only the first page just now, 5 of 20 were robots, the other 15 were guests or members. You can tab between Everyone, Members, Guests, and Robots............................

Even if web applications, such as forums, provide some data, it only provides one form of "who visits". And that is only one simple statistic. It might take a bit more analysis from the server logs to get the real numbers. Because it can be more complicated... As to who exactly is a "guest". As not all robots state they are robots (even if they are suppose to). Not to mention all the other quite tests of any web service, to probe for vulnerabilities, from many zombie servers. All looking like unique visitors, but all actually controlled by one entity. Again, it is not simple.

 

[sour_grapes]

If you go back to the main screen, down in the lower left corner, and click on "Members On Line", it pops up a screen that you can actually see everybody, not just members................. When I looked at only the first page just now, 5 of 20 were robots, the other 15 were guests or members. You can tab between Everyone, Members, Guests, and Robots............................

Wow, I dunno why, but I find that feature cool! It is funny to see what the robots are reading at the moment.

 

[balatonwine]

Wow, I dunno why, but I find that feature cool!

Yes. Cool. Fun. Entertaining....

But I stress the last: Entertaining. Because they really do not give much more than the basic and entertaining information to a real IT professional.

 

[Boatboy24]

Wow, I dunno why, but I find that feature cool! It is funny to see what the robots are reading at the moment.

 

[sour_grapes]

Yes. Cool. Fun. Entertaining....

But I stress the last: Entertaining. Because they really do not give much more than the basic and entertaining information to a real IT professional.

And I stress that I don't give a rat's ass what information this conveys to an IT professional.

 

[winemaker81]

I took a look at the visitor's list, in the first half dozen pages, bots appear to be about 10% of the whole.
https://www.winemakingtalk.com/online/
However, these numbers may not mean what we think they mean. [Yeah, I had to throw in this reference! ]

The forum software tracks registered users well -- I verified that it keeps only my last entry in the list, removing previous entries. But the software knows who I am, so it's easy.

For guests? Unless the software is tracking by IP address, it may be counting page hits, so one guest who looks at 10 pages in 10 minutes will be listed 10 times. Counting site access is really easy to misunderstand, since there are numerous ways to count.

That's my guess regarding the wildly huge number of guests listed. Also keep in mind that the counts go back for 59 minutes, so the counts are not fully current -- someone who viewed a page 58 minutes ago will be counted.

 

[jswordy]

WAIT! There's a place where you can tell how many people are logged in here? Does it tell if they are drunk or not? Asking for a friend...

 

[heatherd]

I suspect that's the reason you see the large delta between replies and views on threads.

 

[winemaker81]

I suspect that's the reason you see the large delta between replies and views on threads.

Agreed. A count of replies is easy. A count of views is probably the count of distinct page hits in a given time frame, so if the interval is 24 hours and I view a thread a dozen times each day, in 5 days that's distinct 5 hits.

I consider views to be more of a barometer than a count of views, e.g., a thread has few, many, or TONS of people reading.

 

[Handy Andy]

Agreed. A count of replies is easy. A count of views is probably the count of distinct page hits in a given time frame, so if the interval is 24 hours and I view a thread a dozen times each day, in 5 days that's distinct 5 hits.

I consider views to be more of a barometer than a count of views, e.g., a thread has few, many, or TONS of people reading.

It depends how many times you read the same thread, I hadnt seen the members on line feature at the bottom of the forums page, so went to look and came back to comment, thats 2 views in 2 minutes. from the same person, with one reply.

11 members were on line > copy paste

• Handy Andy
• bstnh1
• Phil*iceman
• RButch
• Ldypink
• Dustyfj
• BarrelMonkey
• Basilhaydens
• everbuddy

Useful replies or number of posts today might be a better measure of forum activity. A member might be permanently logged in, or may logout and show up as a none member on the web counter.

 

[balatonwine]

Unless the software is tracking by IP address

Most web servers track and record IP addresses by default. But one can turn this off.

A forum is simply an application sitting on top of the web server. So your IP is usually recorded by the server, and simply can be accessed by the application using simple code.

In some countries, the web server logs, including all IP addresses, have to kept for up to two years.

However (and I will over simplify here), the IP address accessing a server need not be the IP address of the person actually accessing the server if that person is using a VPN (legal in many countries) or a zombie server (illegal in most countries).