Mailing List Archive

While Wikipedia shouldn't be counted in such list (it looks like we
came up to the position when news will be that we have more visitors
then, for example, msn.com), it is interesting to see that
wikimedia.org (after wikipedia.org) is at the second place and
wiktionary.org at the 25th place :)

On 10/3/07, Erik Moeller <erik@wikimedia.org> wrote:
> http://www.nonprofittechblog.org/philanthropy-and-nonprofit-top-25-list-october-2007
>
> --
> Toward Peace, Love & Progress:
> Erik
>
> DISCLAIMER: This message does not represent an official position of
> the Wikimedia Foundation or its Board of Trustees.
>
> _______________________________________________
> foundation-l mailing list
> foundation-l@lists.wikimedia.org
> Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l
>

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 10/3/07, Milos Rancic <millosh@gmail.com> wrote:
> While Wikipedia shouldn't be counted in such list (it looks like we
> came up to the position when news will be that we have more visitors
> then, for example, msn.com), it is interesting to see that
> wikimedia.org (after wikipedia.org) is at the second place and
> wiktionary.org at the 25th place :)
>
I was really surprised by how high wikimedia.org ranked. Then I
looked at Alexa traffic details and realized why:

Where people go on Wikimedia.org:
upload.wikimedia.org - 73%
commons.wikimedia.org - 21%
meta.wikimedia.org - 3%
wikimedia.org - 1%
other websites - 2%

http://www.alexa.org/data/details/traffic_details?url=wikimedia.org

So 21% of wikimedia.org hits are for commons, and 73% are to
upload.wikimedia.org (for images, probably the vast majority being
inline Wikipedia images).

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 10/3/07, Anthony <wikimail@inbox.org> wrote:
> http://www.alexa.org/data/details/traffic_details?url=wikimedia.org
>
> So 21% of wikimedia.org hits are for commons, and 73% are to
> upload.wikimedia.org (for images, probably the vast majority being
> inline Wikipedia images).

Whats confusing about this data is that almost every hit to EnWp
results in an upload hit, yet it shows up much lower. I can't say
that I like their counting methods. :)

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 03/10/2007, Gregory Maxwell <gmaxwell@gmail.com> wrote:
> On 10/3/07, Anthony <wikimail@inbox.org> wrote:
> > http://www.alexa.org/data/details/traffic_details?url=wikimedia.org
> >
> > So 21% of wikimedia.org hits are for commons, and 73% are to
> > upload.wikimedia.org (for images, probably the vast majority being
> > inline Wikipedia images).
>
> Whats confusing about this data is that almost every hit to EnWp
> results in an upload hit, yet it shows up much lower. I can't say
> that I like their counting methods. :)

ISP-side image caching comes into it, perhaps? After all, the content
of pages changes a lot faster than the images they include do...

--
- Andrew Gray
andrew.gray@dunelm.org.uk

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 10/3/07, Gregory Maxwell <gmaxwell@gmail.com> wrote:
> On 10/3/07, Anthony <wikimail@inbox.org> wrote:
> > http://www.alexa.org/data/details/traffic_details?url=wikimedia.org
> >
> > So 21% of wikimedia.org hits are for commons, and 73% are to
> > upload.wikimedia.org (for images, probably the vast majority being
> > inline Wikipedia images).
>
> Whats confusing about this data is that almost every hit to EnWp
> results in an upload hit, yet it shows up much lower. I can't say
> that I like their counting methods. :)

Then it has a sense: Commons + Meta + species + local chapters etc.
are at the second place.

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 03/10/2007, Andrew Gray <shimgray@gmail.com> wrote:
> On 03/10/2007, Gregory Maxwell <gmaxwell@gmail.com> wrote:
> > On 10/3/07, Anthony <wikimail@inbox.org> wrote:
> > > http://www.alexa.org/data/details/traffic_details?url=wikimedia.org
> > >
> > > So 21% of wikimedia.org hits are for commons, and 73% are to
> > > upload.wikimedia.org (for images, probably the vast majority being
> > > inline Wikipedia images).
> >
> > Whats confusing about this data is that almost every hit to EnWp
> > results in an upload hit, yet it shows up much lower. I can't say
> > that I like their counting methods. :)
>
> ISP-side image caching comes into it, perhaps? After all, the content
> of pages changes a lot faster than the images they include do...
>

I suspect that it only thinks you are viewing a page on
upload.wikimedia.org when you chose to view an image at full size.

--
geni

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 10/3/07, Andrew Gray <shimgray@gmail.com> wrote:
> ISP-side image caching comes into it, perhaps? After all, the content
> of pages changes a lot faster than the images they include do...

Actual upload.wikimedia.org requests account for (...checking) 46% of
the HTTP requests that hit Wikimedia's servers. That doesn't suggest
that the difference could be caused by external caching.

Milos Rancic <millosh@gmail.com> wrote:
> Then it has a sense: Commons + Meta + species + local chapters etc.
> are at the second place.

Not really: If they are counting inline images you should expect
wikimedia.org to get almost the same score as Wikipedia. If they are
not counting inline images then wikimedia's traffic level shouldn't be
enough to get it on a top-sites list.

Wikimedia hosts by percentage of http requests with more than 1/100th
of a percent of the requests.

46.35% upload.wikimedia.org
29.81% en.wikipedia.org
5.85% de.wikipedia.org
4.07% es.wikipedia.org
2.51% fr.wikipedia.org
1.78% pl.wikipedia.org
1.19% it.wikipedia.org
0.99% ja.wikipedia.org
0.96% pt.wikipedia.org
0.95% nl.wikipedia.org
0.51% commons.wikimedia.org
0.45% sv.wikipedia.org
0.37% www.wikipedia.org
0.33% tr.wikipedia.org
0.27% fi.wikipedia.org
0.23% ru.wikipedia.org
0.20% en.wiktionary.org
0.17% no.wikipedia.org
0.16% he.wikipedia.org
0.14% zh.wikipedia.org
0.12% en.wikiquote.org
0.12% da.wikipedia.org
0.12% cs.wikipedia.org
0.11% hu.wikipedia.org
0.11% ar.wikipedia.org
0.10% ro.wikipedia.org
0.10% en.wikibooks.org
0.08% en.wikinews.org
0.06% wikimediafoundation.org
0.06% hr.wikipedia.org
0.06% de.wiktionary.org
0.06% ca.wikipedia.org
0.05% www.wikipedia.com
0.05% meta.wikimedia.org
0.05% bg.wikipedia.org
0.04% wikipedia.org
0.04% sk.wikipedia.org
0.04% simple.wikipedia.org
0.04% lt.wikipedia.org
0.04% gl.wikipedia.org
0.04% de.wikibooks.org
0.03% www.mediawiki.org
0.03% sr.wikipedia.org
0.03% sl.wikipedia.org
0.03% id.wikipedia.org
0.03% fr.wiktionary.org
0.03% fa.wikipedia.org
0.03% en.wikisource.org
0.03% el.wikipedia.org
0.03% bs.wikipedia.org
0.02% wikipedia.com
0.02% vi.wikipedia.org
0.02% uk.wikipedia.org
0.02% pl.wiktionary.org
0.02% pl.wikiquote.org
0.02% pl.wikibooks.org
0.02% nn.wikipedia.org
0.02% it.wikiquote.org
0.02% fr.wikisource.org
0.02% et.wikipedia.org
0.02% es.wikiquote.org
0.02% es.wikibooks.org
0.02% eo.wikipedia.org
0.02% de.wikiquote.org
0.02% de.wikinews.org
0.01% www.wikipedia.pl
0.01% www.wikimedia.org
0.01% www.de.wikipedia.org
0.01% wikisource.org
0.01% tr.wiktionary.org
0.01% tl.wikipedia.org
0.01% th.wikipedia.org
0.01% sv.wiktionary.org
0.01% static.wikipedia.org
0.01% sq.wikipedia.org
0.01% species.wikimedia.org
0.01% sh.wikipedia.org
0.01% ro.wikisource.org
0.01% pt.wiktionary.org
0.01% pt.wikiquote.org
0.01% pt.wikibooks.org
0.01% pl.wikisource.org
0.01% pl.wikinews.org
0.01% nl.wiktionary.org
0.01% nl.wikibooks.org
0.01% nap.wikipedia.org
0.01% ms.wikipedia.org
0.01% mk.wikipedia.org
0.01% meta.wikipedia.org
0.01% lv.wikipedia.org
0.01% lb.wikipedia.org
0.01% la.wikipedia.org
0.01% ko.wikipedia.org
0.01% it.wiktionary.org
0.01% it.wikisource.org
0.01% it.wikibooks.org
0.01% is.wikipedia.org
0.01% fr.wikibooks.org
0.01% eu.wikipedia.org
0.01% es.wiktionary.org
0.01% es.wikisource.org
0.01% en.wikiversity.org
0.01% de.wikisource.org
0.01% als.wikipedia.org
0.01% af.wikipedia.org

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 10/3/07, Gregory Maxwell <gmaxwell@gmail.com> wrote:
> Not really: If they are counting inline images you should expect
> wikimedia.org to get almost the same score as Wikipedia. If they are
> not counting inline images then wikimedia's traffic level shouldn't be
> enough to get it on a top-sites list.

At the list wikipedia.org is rated 40, wikimedia.org is 3.xx and the
third site is something like 3.xx - 0.03.

So, it is reasonable to suppose that wikimedia.org is really at the
second place (and without upload).

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 10/3/07, Milos Rancic <millosh@gmail.com> wrote:
> On 10/3/07, Gregory Maxwell <gmaxwell@gmail.com> wrote:
> > Not really: If they are counting inline images you should expect
> > wikimedia.org to get almost the same score as Wikipedia. If they are
> > not counting inline images then wikimedia's traffic level shouldn't be
> > enough to get it on a top-sites list.
>
> At the list wikipedia.org is rated 40, wikimedia.org is 3.xx and the
> third site is something like 3.xx - 0.03.
>
> So, it is reasonable to suppose that wikimedia.org is really at the
> second place (and without upload).

No it's not. See the list I posted. Without upload the sum of all
other wikimedia.org traffic (including commons) is less than 1/50th of
en.wikipedia.org alone and about 1/100th of the traffic from all of
the *.wikipedia.org sites.

They are measuring 'monthly distinct visitors' rather than hits. So
perhaps it's somehow possible that their numbers are right but given
the traffic levels it is unlikely.

It should be easy enough for us to actually measure that number. As
soon as I have the page counter stuff up I'll set up something to
measure distinct viewers since thats the standard sizing metric used
in the commercial web world.

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 10/3/07, Gregory Maxwell <gmaxwell@gmail.com> wrote:
> No it's not. See the list I posted. Without upload the sum of all
> other wikimedia.org traffic (including commons) is less than 1/50th of
> en.wikipedia.org alone and about 1/100th of the traffic from all of
> the *.wikipedia.org sites.

Aha... I didn't realize that there is so big difference (I didn't make
approximation). I was just thinking that 1:10-15 is significant
enough.

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l

On 03/10/2007, Gregory Maxwell <gmaxwell@gmail.com> wrote:
> No it's not. See the list I posted. Without upload the sum of all
> other wikimedia.org traffic (including commons) is less than 1/50th of
> en.wikipedia.org alone and about 1/100th of the traffic from all of
> the *.wikipedia.org sites.
>

I suspect they are doing it on the basis of the in the address bar.
Not impossible that 1:100 people decides to view an image at full
size.


--
geni

_______________________________________________
foundation-l mailing list
foundation-l@lists.wikimedia.org
Unsubscribe: http://lists.wikimedia.org/mailman/listinfo/foundation-l