The watching 'vBulletin's market share nosedive' thread.

When will vB drop below 50% market share?

  • Total voters
    173
rudedog

rudedog

Jim
Joined
Apr 8, 2013
Messages
577
Happy New Years everyone.

Just checked the stats (Digitalpoint) and it seems vB is racing to 3rd place with 27.8% of the market share. with Invision closing in with 23%. XF is at 39.7%

Cookie Search 2019-01-01 09-35-40.png
 
Boothby

Boothby

Enthusiast
Joined
Sep 25, 2012
Messages
132
Interessingly, after almost 2 years of stagnation with little ups and downs XF makes a big jump of almost 3 percent within the last 3 months.
 
Alpha1

Alpha1

Administrator
Joined
May 28, 2007
Messages
4,268
Im not sure the cookie tracker is fully functional anymore. It showing clearly untrue stats for server stack and php. (NGINX jumping overnight from 38% to 85% and NGINX version at the same time jumping to 99% use of ancient version 1.1. Also at the same time PHP, IIS have a drastic shift ) Its not measuring IPS 4.1-4.4
There seems to be something wrong.
 
haqzore

haqzore

Devotee
Joined
Dec 6, 2012
Messages
2,654
Alfa1 said:
Im not sure the cookie tracker is fully functional anymore. It showing clearly untrue stats for server stack and php. (NGINX jumping overnight from 38% to 85% and NGINX version at the same time jumping to 99% use of ancient version 1.1. Also at the same time PHP, IIS have a drastic shift ) Its not measuring IPS 4.1-4.4
There seems to be something wrong.
Click to expand...
Digitalpoint
 
rudedog

rudedog

Jim
Joined
Apr 8, 2013
Messages
577
cheat_master30 said:
He means it's gone down, and if it continues, it'll eventually end up lower than IPB's market share.
Click to expand...

Correct Cheat master. There is a good chance IPB will overtake vB in 2019.

Boothby said:
Interessingly, after almost 2 years of stagnation with little ups and downs XF makes a big jump of almost 3 percent within the last 3 months.
Click to expand...

I know Digitalpoint mentioned a while ago that his spiders where catching up or there were some issues when the last jump happened. It was not capturing data for a few months and then caught up all at once when it was fixed.
 
R0binHood

R0binHood

Habitué
Joined
Nov 23, 2011
Messages
1,606
The stats on Wappalyzer are hugely different to DPs tracker.

It puts XF and IPB at about the same market share, with VB still double both combined, leaving a huge amount of room for growth in all new platforms.

Interestingly it looks like Discourse is pretty prominent on this graph in relation to XF and IPB at about half of the number of forums that either of the other two platforms have. While Burning Board looks like a blip on the radar.

upload_2019-1-2_11-24-42.png
 
R0binHood

R0binHood

Habitué
Joined
Nov 23, 2011
Messages
1,606
It seems Wappalyzer gets its data mainly from their browser extension though, not sure if they have any crawlers. I'm not sure how that might affect the results.
 
Banxix

Banxix

Enthusiast
Joined
Jan 13, 2018
Messages
138
R0binHood said:
It seems Wappalyzer gets its data mainly from their browser extension though, not sure if they have any crawlers. I'm not sure how that might affect the results.
Click to expand...

They do not have crawlers but that is why their results are more trustworthy from my POV.
Because they are mostly live sites which people visit for real, not dead graves if you just crawl with no check.
 
R0binHood

R0binHood

Habitué
Joined
Nov 23, 2011
Messages
1,606
Hmm...yeah, but what if the site is old and active or just doesn't have the kind of visitors demographic that would have extension installed?

Their data could be missing a huge number of smaller sites, with fewer visitors, none of which have the browser extension installed.

A site that is publicly visible and crawlable is still a live site with readable for content and discoverable in search engines even if traffic or current posting activity is low.
 
Banxix

Banxix

Enthusiast
Joined
Jan 13, 2018
Messages
138
R0binHood said:
Hmm...yeah, but what if the site is old and active or just doesn't have the kind of visitors demographic that would have extension installed?

Their data could be missing a huge number of smaller sites, with fewer visitors, none of which have the browser extension installed.

A site that is publicly visible and crawlable is still a live site with readable for content and discoverable in search engines even if traffic or current posting activity is low.
Click to expand...

No method is perfect. With crawler, you then have too many expired data, unless you recrawl frequently so it requires huge resources and a good crawling technique.

On large scale, crawler is better for sure but I do not go into micro detail so Wappalyzer result is enough for me.

Also if you are talking about smaller and old sites with very few visitors then phpBB and vBulletin obviously lead the chart, there is no chance for IPB or Xenforo overtake them in this aspect. Base on that, I really doubt Digitalpoint crawler.
 
Digitalpoint

Digitalpoint

Brain
Joined
Aug 29, 2012
Messages
309
Alfa1 said:
Im not sure the cookie tracker is fully functional anymore. It showing clearly untrue stats for server stack and php. (NGINX jumping overnight from 38% to 85% and NGINX version at the same time jumping to 99% use of ancient version 1.1. Also at the same time PHP, IIS have a drastic shift ) Its not measuring IPS 4.1-4.4
There seems to be something wrong.
Click to expand...
Internally the spider itself works as expected... although from time to time an issue pops up in how that data is reported.

For example, for awhile the reporting engine was truncating Nginx versions after 3 characters. This wasn't an issue until Nginx 1.10.x came out. At which point it started counting everything that *started* with 1.1* as Nginx 1.1... so 1.10, 1.11, 1.12, etc. I don't watch it close enough to notice stuff like that until someone says something. It was corrected a couple months ago, which is why you see the Nginx chart reverting to what it should have been (the data was there, it was just a reporting issue).

upload_2019-1-15_10-12-14.png

The "current" numbers should be good... was just some of the historical months where it was only looking at first 3 characters.

upload_2019-1-15_10-13-28.png

The reporting/charts isn't really what it's used for primarily so I don't pay much attention to it. :)

All systems are going to have different methods of looking at stuff... in this case, it's a spider that looks at cookies set and JavaScript objects on the sites.

Some of the other stuff you mentioned, like it not looking at Burning Board versions higher than 4.1, it's because I didn't know there was a newer version and also don't know of any Burning Board side to look at to figure out a way to extrapolate that data. I'm not a member of any sites that use IBB myself, so just not something I'm checking for every day (or ever). Most stuff doesn't need different logic built for different versions, so all that stuff doesn't need maintenance (things like jQuery version, PHP version, etc.)
 
You must log in or register to reply here.
Top