incubator-ooo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rob Weir <robw...@apache.org>
Subject Re: [DISCUSS] Is it worth looking at Confluence Wiki Again?
Date Thu, 29 Sep 2011 17:37:37 GMT
On Thu, Sep 29, 2011 at 5:33 AM, TJ Frazier <tjfrazier@cfl.rr.com> wrote:
> On 9/27/2011 20:37, TJ Frazier wrote:
>>
>> On 9/27/2011 17:03, Rob Weir wrote:
>>>
>>> <snip>
>>> I'm assuming this is the link:
>>>
>>> http://ooo-wiki.apache.org/wiki/Wiki/statistics
>>>
>> That should be a good assumption, but no. It's busted. I'll fix it.
>> Meanwhile, you can see it on the main page of the live wiki:
>>
>> http://wiki.services.openoffice.org/wiki/Main_Page
>>
>> The problem is complex. Give me a day or so.
>>
>
> OK, try the "statistics" link. Now showing top 100. Page is still messy
> because of difficult working conditions (ATS problem).
>

Thanks. Just looking through the top pages, I'm not seeing very
complex pages.  I know there are highly customized parts, that might
not convert easily to moin moin or CWiki.  But I don't see it in the
top pages.

Or am I missing something?


>>> Do we have any sense of what % of page visits are comprised by the top
>>> pages? The above numbers look impressive, but I have no idea if the
>>> top 10 pages account for 20% of the hits, or 1%.
>
> AFAIK, the wiki is keeping a hit count per page. Someone with SQL karma and
> fu could total the field. It's a big number: the 100th page has 10E5+ hits.
>

It is hard to interpret the absolute number without knowing the time
scale we're measuring.  10,000 hits in 5 years is not so impressive.
In one day, it is.

In any case, we have some very popular pages.  But it drops rather
quickly.  After 25 pages we're down a factor of 10.  This is as you
would expect.

The interesting thing (and something that is hard to look at in detail
without web analytics) is how people are getting to these pages, and
why.

For example, the Dictionaries page is a good compilation of open
source dictionaries for ispell/hunspell.  As such, it is of interest
to several OSS programs.  So that page is linked to from:

PostgresSQL:  http://www.postgresql.org/docs/8.4/static/textsearch-dictionaries.html

NetBSD:  ftp://ftp.netbsd.org/pub/pkgsrc/current/pkgsrc/textproc/hunspell-lt_LT/README.html

TortoiseSVN:  https://apps.uillinois.edu/TortoiseSVN-1.6.6-en.pdf

Even commercial products like Altova:  http://www.altova.com/de/dictionaries/

So we have a page here that is important beyond users of OpenOffice.
We should try to preserve both the content and the URL



>>>
>>> I can crunch the numbers if there is some way to derive them. I could
>>> even crunch the http logs directly, if they are saved for a couple of
>>> days.
>>>
>>> In any case, the top pages you've reported are not typical "community"
>>> pages.
>>
>> True.
>>>
>>> They would work well as markdown pages.
>>
>> Only if we ignore the valuable feedback from user changes on the
>> manuals. I find this a significant quality enhancer.
>>
>>> Except maybe the
>>> FAQ's we want to be more dynamic and wiki-enabled. Might even be
>>> something the Forum volunteers would be interested in maintaining
>>> directly, since they would have the best sense of what questions are
>>> "frequent".
>
> I suggest that a lot more info belongs on a wiki, rather than elsewhere.
> Please see my upcoming reply to a commit by Dave Fisher, on a web page.
>
>>>
>>> I think this is encouraging.
>>>
>>> -Rob
>
> --
> /tj/
>
>

Mime
View raw message