[Wikimedia-l] Wikimedia (Foundation) endowment

Theo10011 de10011 at gmail.com
Fri Mar 15 06:36:53 UTC 2013


Hi Michael

On Fri, Mar 15, 2013, Michael Snow <wikipedia at frontier.com> wrote:

> I'm not sure why you would use traffic ranking for financial analysis,
> even the envelope-and-napkin kind of analysis we're engaging in here. I'm
> pretty confident that just because Google has been sitting at #1 for some
> time, it doesn't mean that their core operational costs have remained flat
> over that period.
>

I'm actually not using the traffic for financial analysis. I'm only using
the trend in traffic to compare the hosting costs - I think it would be
fair to assume that both are intrinsically linked. :)

The analysis of 6M/ year wasn't based on traffic at all, it was from the
annual budget and expenditure I saw in the reports, though that was an
envelope-and-napkin kind of analysis, it wasn't entirely based on
conjectures either. I also think its unfair to compare Wikipedia with
google, but if you were to take a top 10 traffic website and separate
their infrastructure and cap-ex, and look at annual operational costs
especially with things like bandwidth cost, it would have to be comparable.
(Maybe not for google but let's say for twitter or linked.in - comparable
bandwidth usage *is* the reason they are in the same league.)


> Aside from that, it's only recently that Wikimedia sites have approached
> having the kind of redundancy and failover capabilities we've talked about
> needing for a long time. That's at least one example of something that can
> add pretty significant costs without having a material impact on traffic
> (except in emergencies, of course).


I wouldn't know about the redundancies or those capabilities, it seems
fairly the same. My location and perspective might have more to do with
that but I just don't see the change as that dramatic. I wouldn't say there
isn't any change, from a performance stand-point, it just seems
incrementally better - outages still happen[1], there
are occasionally things that break, and minor lag issues persist on the
other side of the world. I'm grateful for the improvements but I wouldn't
really know what changed under the surface.

On Fri, Mar 15, 2013, James Alexander <jamesofur at gmail.com> wrote:
>
> Aye, I know for example that our page views have more then doubled in the
> last 5 years (since 2008) and I believe grew even more dramatically in the
> years before that.


They increased a lot, but I don't think they more than doubled, or even
doubled[2]. The rise is pretty steady from Feb 09.

Regards
Theo

[1]http://blog.wikimedia.org/c/technology/operations/outage/
[2]
http://stats.wikimedia.org/reportcard/RC_2012_02_detailed.html#fragment-31


More information about the Wikimedia-l mailing list