The Icon Bar: How popular are RISC OS sites online?

How popular are RISC OS sites online?

Just as in life, there are lots of different ways of measuring and estimating popularity. Online one of the ways you can do this is to use a tool called Alexa. This gives sites a ranking based on how popular Alexa thinks the site is (so number 1 is google.com).

It is not an exact science (and it can be misrepresentative on some sites where Alexa has less data), but it is a useful 'guess'. So I typed in some RISC OS sites (and non-RISC OS sites which you may have heard of as a comparison) to get some numbers. Here is what Alexa reported for global rankings.....

apple.com 65
bbcbasic.co.uk 2,564,449
cjemicros.co.uk 3,463,770
drobe.co.uk 19,898,135
iconbar.com 3,913,170
linuxmint.com 4,450
netsurf-browser.org 1,165,775
osnews.com 114,759
orpheusinternet.co.uk 17,233,044
raspberrypi.org 3,186
riscos.com 2,866,998
riscos.org 9,126,309
riscository.com 11,268,284
riscosopen.org 366,518
stardot.org.uk 827,545 (41,450 in just uK)
ubuntu.com 1,493
xara.com 88,840

It is not a total surprise that ROOL is easily the top RISC OS site I could find. We have some work to do with Iconbar (as do the RISC OS vendors if they want to grow their sales online).

What do you make of the numbers?

Weblink to lookup a website on Alexa.

How popular are RISC OS sites online?

VincceH (09:58 30/6/2017)
  flibble (18:54 30/6/2017)
  VincceH (18:49 5/7/2017)
    helpful (23:55 10/7/2017)
      VincceH (09:27 11/7/2017)
        VincceH (10:21 11/7/2017)
          VincceH (13:28 11/7/2017)
            flibble (21:17 12/7/2017)
              arawnsley (10:09 13/7/2017)
                nunfetishist (10:27 13/7/2017)
                  CJE (11:08 13/7/2017)
                    markee174 (11:59 13/7/2017)

VinceH

Message #124105, posted by VincceH at 09:58, 30/6/2017

Lowering the tone since the dawn of time

Posts: 1600

We have some work to do with Iconbar

Unless you're actually paying Alexa, there's absolutely nothing you can do because the numbers are pretty much made up.

(They call it estimated, and have some blurb about it being based on real data from other sites - but ultimately if they have no data on your site, then the figures may as well be plucked from thin air.)

[ Log in to reply ]

Peter Howkins

Message #124106, posted by flibble at 18:54, 30/6/2017, in reply to message #124105

Posts: 892

Here's a few more

armini.co.uk: We don't have enough data to rank this website.
ident-online.co.uk: We don't have enough data to rank this website.
riscos.info: 4,845,247
elesar.co.uk: We don't have enough data to rank this website.

[ Log in to reply ]

VinceH

Message #124107, posted by VincceH at 18:49, 5/7/2017, in reply to message #124105

Lowering the tone since the dawn of time

Posts: 1600

I was asked (by Bryan IIRC) at Recursion this weekend what visitor numbers were like for RISCOSitory. (Perhaps his question was prompted by this article?)

I forgot until now - but I've just looked. My hosts provide some nice analytics data, but their interface for selecting a date range (other than the simple defaults) is a bit broken.

Anyway, here's a little info for June (and May) this year for RISCOSitory.com:

Visitors: 11,668 ( 12,036 )
That gives a daily mean of: 388 ( 388 )
Page impressions: 57,345 ( 55,272 ) - see notes below.
Page impressions with NetSurf: 333 ( 323 )
Page impressions with "RISK OS": 328 ( 313 )

Those numbers are miniscule in real terms - but I don't know whether they're good, bad, or somewhere in between in terms of RISC OS sites.

Notes:

1: The site's feed was accessed 30,020 times in June, and the login page 3,940 - so these should be knocked off that total.

2: The number of impressions by search engine robots is 53,650 (56,963). The fact that May's robot impression total is higher than the page impression total suggests the page impressions given above has these already taken out.

[Edited by VincceH at 19:51, 5/7/2017]

[ Log in to reply ]

Bryan Hogan

Message #124108, posted by helpful at 23:55, 10/7/2017, in reply to message #124107

Member
Posts: 255

Yes it was me that asked the question

Does the number of visitors/page impressions each month correlate with the number of new articles that have been posted? May and June were fairly quiet on the news front IIRC.

Do you have the figures for last October, because in the run up to the London show there were dozens of new items.

[ Log in to reply ]

VinceH

Message #124109, posted by VincceH at 09:27, 11/7/2017, in reply to message #124108

Lowering the tone since the dawn of time

Posts: 1600

No - it was trying to look back at older figures that led me to the conclusion the date selection was broken!

I wanted to look at older figures because it seems fairly obvious to me that the figures would correlate to the amount posted, and most of this calendar year the amount posted has been down.

(This is because I've only been posting announcements that have been sent to me, and mostly ignoring things posted elsewhere by WROCC syndrome sufferers*.)

I'll take another look in a while and see if I can get the date thing working.

* Keeping forum posts flagged in my feed**, and so on, so that I can do a round-up or two at some point.

** Assuming I'm even up to date with the feeds in question. With ROOL's forum, for example, I have 2672 unread posts. (Ignoring my occasional glance at the recent posts page on the website).

[ Log in to reply ]

VinceH

Message #124110, posted by VincceH at 10:21, 11/7/2017, in reply to message #124109

Lowering the tone since the dawn of time

Posts: 1600

Okay, today it seems to be working a little better - but it does limit things if I go back that far; setting 1/10/2016 to 31/10/2016 actually gives me 26/9/2016 to 6/11/2016, with the detail broken down by week rather than by day. i.e. it's shifting the date range to a number of weeks that encompasses those dates. (26/9/16 was a Monday - so I guess its historical data is weekly from Mondays).

The overall number of visitors for that period is 11,294 - that's 42 days, so around 269 per day; obviously less than the numbers for June this year.

However, looking at that broken down by week, I see:

26/9-2/10 = 1459 = 208/day (0)
3/10-9/10 = 1538 = 220/day (0)
10/10-16/10 = 1554 = 222/day (5)
17/10-23/10 = 2039 = 291/day (7)
24/10-30/10 = 2712 = 387/day (23)
31/10-6/11 = 1992 = 285/day (6)

The number in brackets is the number of posts on the site in the same period - so based on that, yes, more posts does seem to have a bearing.

(Although the fact the show was looming might itself have had an effect - more people may have looked at the site for news, even if it turned out there wasn't any!)

[Edited by VincceH at 11:31, 11/7/2017]

[ Log in to reply ]

VinceH

Message #124111, posted by VincceH at 13:28, 11/7/2017, in reply to message #124110

Lowering the tone since the dawn of time

Posts: 1600

A little bit later, and another chance to take a look. The available stats only go back as far as 1st August 2015. From there until 31st January 2016, only the month totals are available.

From 1st February 2016 until 31st January 2017, only weekly figures are available - and from 1st February 2017, I can specify any date range.

The obvious conclusion is that as they 'expire' data, they preserve some, but less granularly - but the oddity there is if you only have per week numbers, how do you reduce granularity to per month numbers when the weeks and months don't actually coincide. I guess there must be some approximating/fudging.

However, since 1st February this year:

February: 13 posts, 10275 visitors (367/day)
43440 page impressions (15787 feed, 4147 log-in)

March: 14 posts, 13086 visitors (422/day)
54425 page impressions (18104 feed, 4957 log-in)

April: 20 posts, 13895 visitors (463/day)
56966 page impressions (22169 feed, 4963 log-in)

May and June (10 posts each) are given above.

Unsurprisingly, I didn't visit the log-in page that many times in those months!

Also: I mentioned at Recursion that I was getting an emailed report for webchange.co.uk - I've found where that's set, and I should now get it monthly for riscository.com instead.

[ Log in to reply ]

Peter Howkins

Message #124114, posted by flibble at 21:17, 12/7/2017, in reply to message #124111

Posts: 892

orpheusinternet.co.uk 17,233,044

Company dissolved.

https://beta.companieshouse.gov.uk/company/05359727
https://beta.companieshouse.gov.uk/company/05359727/filing-history

I hope all the customers were moved to an alternate supplier.

[ Log in to reply ]

Andrew Rawnsley

Message #124115, posted by arawnsley at 10:09, 13/7/2017, in reply to message #124114

R-Comp chap
Posts: 600

I'll check with Richard, but I'm guessing this is simply a case that there's no point him filing for both Genesys and Orpheus as separate companies, since at this point "Orpheus Internet" is merely a trading name.

You'll note that all the domains are live, and so on. A phone call would also have cleared this up.

[ Log in to reply ]

Rob Kendrick

Message #124116, posted by nunfetishist at 10:27, 13/7/2017, in reply to message #124115

Today's phish is trout a la creme.

Posts: 525

Compulsory strike-off? That doesn't sound good.

[ Log in to reply ]

Chris Evans

Message #124117, posted by CJE at 11:08, 13/7/2017, in reply to message #124116

CJE Micros chap
Posts: 228

I gather Orpheus the limited company hasn't been actively trading for years.
I recall in the past being told that to shutdown a company the official way is quite expensive so many companies get themselves into the position of having no assets or liabilities and then stop filing returns. The company will then be struck off with no extra cost.

[ Log in to reply ]

Mark Stephens

Message #124118, posted by markee174 at 11:59, 13/7/2017, in reply to message #124117

Does all the
work around here
Posts: 155

Orpheus Internet are very much alive and kicking and we will be running an interview on iconbar as part of our interview series in the near future.

[Edited by markee174 at 13:00, 13/7/2017]

[Edited by markee174 at 13:01, 13/7/2017]

[ Log in to reply ]