Mailing List Archive

I would like to suggest a few direction of thoughts...

Last year, a couple of concerns erupted before/during/after the elections.



First, some wondered what the role of the board was.
I would be pleased that some feedback is given regarding that topic
during the election. So that the next board may try to do its best for
taking community opinion into consideration.

I have also wondered if it would not be interesting that some of you
prepare a sort of short list of questions, which each candidate would
have to answer or comment.




Second, the participation rate of languages have been very diversed.
English participants represented a huge number of voters.
German were second and french third. Other languages had basically not
participated but for a very few people.
Link : http://meta.wikimedia.org/wiki/Image%3AElection_participation2.png

Our project is international. It is not very suitable that such a
discrepancy exists.
I would like that all non english editors on foundation-l take special
attention in involving their projects.




Third, last year, some rather heated discussions occured when results
were not fully displayed. I would be pleased that this is set before the
election, so that editors are not surprised when results are not
published. Hence the questions : which results should be published ?
Interest and disadvantages of not publishing certain results ?
Publication of results per projects ? Only limited to bigger projects ?



Fourth, do you have overall some feedback to give on last year
organisation, so that this year organisers can take them into account ?



Thanks.


Anthere

Anthere:

> Second, the participation rate of languages have been very diversed.
> English participants represented a huge number of voters.
> German were second and french third. Other languages had basically not
> participated but for a very few people.
> Link : http://meta.wikimedia.org/wiki/Image%3AElection_participation2.png

Yes, that's a very good point. The graph you link to does not actually
show the participation *rate*, but the number of participants per
language. It might be interesting to compare this against, say, the
number of very active contributors per language in June 2004:

http://en.wikipedia.org/wikistats/EN/TablesWikipediansEditsGt100.htm

Doing this, we get the following rates (roughly, as I'm reading the
number of votes per language from the graph):

41% EN
41% DE
91% FR
20% JA
32% NL
75% ZH
17% PL
31% IT
etc.

Taking just these languages, English and German had average rates, but
the French participation rate was extraordinary by any measure. Japanese
and Polish in particular were depressingly low. Hopefully, Datrio and
Britty will be able to help with that.

In general, I can only really think of one solution -- getting the
relevant text translated into as many languages as possible. For the
sake of fairness, we should announce the election in the same location
in all languages (e.g. Recent Changes). We won't be able to stop local
"get out the vote" efforts, so we should encourage them instead and hope
that as many projects as possible make an effort to go beyond the minimum.

> Our project is international. It is not very suitable that such a
> discrepancy exists.

Agreed, though it's always important to look at the rates rather than
the absolute numbers.

> Third, last year, some rather heated discussions occured when results
> were not fully displayed. I would be pleased that this is set before the
> election, so that editors are not surprised when results are not
> published. Hence the questions : which results should be published ?

I'd say at least
- number of votes per candidate
- number of voters per wiki project / language.

> Fourth, do you have overall some feedback to give on last year
> organisation, so that this year organisers can take them into account ?

I'd like the allowed length of the candidate statements to be clearly
defined, and the length limit to be enforced. (I suggest 1000 characters
of rendered text total.) Every candidate will of course be allowed to
link to a detailed statement without a length limit. I'd also like the
1000 character statements to be fairly free-form, i.e. every candidate
should be able to decide for themselves how to use that space.

All best,

Erik

Is this going to be a vote for another one year term?? As I feel a one
year term is to short and very bad for consistency.

I personally feel we should give Angela and Anthere another year btw.
But I am afraid we passed that station already.

Walter/Waerth

On 4/30/05, Anthere <anthere9@yahoo.com> wrote:

> Second, the participation rate of languages have been very diversed.
> English participants represented a huge number of voters.
> German were second and french third. Other languages had basically not
> participated but for a very few people.
> Link : http://meta.wikimedia.org/wiki/Image%3AElection_participation2.png

I don't agree with the conclusion you draw from this graph. The
English and German Wikipedias are much larger than those in other
languages, so it's only to be expected there will be more votes coming
from those. Taking the number of active editors in April 2004 as my
measure, I find that French participation is remarkably large, but
English and German is not much more than would be expected from a fair
division over the languages. In numbers (top 16 Wikipedia languages;
numbers are number of editors, number of voters, and the second as a
percentage of the first):

French 321 89 28%
Finnish 33 5 15%
Norse 27 3 11%
Italian 69 7 10%
German 1613 145 9%
English 2746 238 9%
Dutch 191 16 8%
Chinese 143 11 7%
Esperanto 44 3 7%
Polish 124 8 6%
Swedish 98 6 6%
Danish 67 4 6%
Japanese 360 18 5%
Spanish 123 6 5%
Hebrew 69 3 4%
Portuguese 67 0 0%

Andre Engels

Erik Moeller (erik_moeller@gmx.de) [050501 04:12]:

> I'd like the allowed length of the candidate statements to be clearly
> defined, and the length limit to be enforced. (I suggest 1000 characters
> of rendered text total.) Every candidate will of course be allowed to
> link to a detailed statement without a length limit. I'd also like the
> 1000 character statements to be fairly free-form, i.e. every candidate
> should be able to decide for themselves how to use that space.


So I can write a statement in 1000 Chinese characters, and get that
translated back to English... you sure about this one?


- d.

On 4/30/05, Andre Engels <andreengels@gmail.com> wrote:
> On 4/30/05, Anthere <anthere9@yahoo.com> wrote:
>
> > Second, the participation rate of languages have been very diversed.
> > English participants represented a huge number of voters.
> > German were second and french third. Other languages had basically not
> > participated but for a very few people.
> > Link : http://meta.wikimedia.org/wiki/Image%3AElection_participation2.png
>
> I don't agree with the conclusion you draw from this graph. The
> English and German Wikipedias are much larger than those in other
> languages, so it's only to be expected there will be more votes coming
> from those. Taking the number of active editors in April 2004 as my
> measure, I find that French participation is remarkably large, but
> English and German is not much more than would be expected from a fair
> division over the languages. In numbers (top 16 Wikipedia languages;
> numbers are number of editors, number of voters, and the second as a
> percentage of the first):
>
> French 321 89 28%
> Finnish 33 5 15%
> Norse 27 3 11%
> Italian 69 7 10%
> German 1613 145 9%
> English 2746 238 9%
> Dutch 191 16 8%
> Chinese 143 11 7%
> Esperanto 44 3 7%
> Polish 124 8 6%
> Swedish 98 6 6%
> Danish 67 4 6%
> Japanese 360 18 5%
> Spanish 123 6 5%
> Hebrew 69 3 4%
> Portuguese 67 0 0%

Very interesting. Thanks for that analysis! Fr: users are also
unusually well integrated with IRC. Clearly the conclusion is that
eating well makes you productive.

SJ

Thank you for your interesting analysis, Andre

On 5/1/05, Andre Engels <andreengels@gmail.com> wrote:
> (top 16 Wikipedia languages;
> numbers are number of editors, number of voters, and the second as a
> percentage of the first):

I would like to add another (more recent) index, caliculated by
BjarteSorensen from Erik Zachte's stats: square (active editors [5+
edits in a month] * article counts). I call it "WMF activity index",

> French 321 89 28% 9425.6
> Finnish 33 5 15% 1754.4
> Norse 27 3 11% 1590.6
> Italian 69 7 10% 3174.9
> German 1613 145 9% 25048.5
> English 2746 238 9% 61466.5
> Dutch 191 16 8% 5118.8
> Chinese 143 11 7% 2579.9
> Esperanto 44 3 7% 994.5
> Polish 124 8 6% 4989.5
> Swedish 98 6 6% 3824.4
> Danish 67 4 6% 1359.4
> Japanese 360 18 5% 8693.8
> Spanish 123 6 5% 4133.8
> Hebrew 69 3 4% 1691.7
> Portuguese 67 0 0% 2791.1

The order of high score projects are a bit differe from rank by
article numbers. Remarkable differences are on this index
-Fr is over Ja (9425.6pt vs 8693.8)
-Es is over Sv (4133.8 vs 3824.4)
-Ru with 17K articles and He with 18K over four projects with over 20K
articles; more over Ru doesn't appear in the Andre's original list.

Generally we can say active projects are more involved into grobal
matters. There are some exception, of course. I expect some of them
will volunteers to organize translation to involve more users than in
the last year.

Besides those, there are some projects or language communities which
seem to be possible to be more involved. Specially I think so for less
than 5% vote rate projects; that is, Japanese, Spanish, Hebrew and
Portuguese. I would like to add Russian because they are relatively
rare in our grobal community. Is any Japanese, Spanish, Hebrew or
Prtuguese editors who would like to oversee the translaion?

Now we call for translators and their coordinators on meta.
Translation cooridnators will care for
-Notice translation
-Candidate profile translation (I agree Erik; profile up to 1000
words seems fine)
-FAQ from/to their favorite language

If you are interested in supporting vote organisation and promotion,
please list yourself on meta,
http://meta.wikimedia.org/wiki/Election_notice_translations_2005
and please don't be suprise if the organiser team will contact you
personally. when nobody apply this task. ;-)

Cheers,

--
Aphaea@*.wikipedia.org
email: Aphaia @ gmail (dot) com

David Gerard a écrit:
> Erik Moeller (erik_moeller@gmx.de) [050501 04:12]:
>
>
>>I'd like the allowed length of the candidate statements to be clearly
>>defined, and the length limit to be enforced. (I suggest 1000 characters
>>of rendered text total.) Every candidate will of course be allowed to
>>link to a detailed statement without a length limit. I'd also like the
>>1000 character statements to be fairly free-form, i.e. every candidate
>>should be able to decide for themselves how to use that space.
>
>
>
> So I can write a statement in 1000 Chinese characters, and get that
> translated back to English... you sure about this one?
>
>
> - d.

good point ;-)

ant

Sj a écrit:
> On 4/30/05, Andre Engels <andreengels@gmail.com> wrote:
>
>>On 4/30/05, Anthere <anthere9@yahoo.com> wrote:
>>
>>
>>>Second, the participation rate of languages have been very diversed.
>>>English participants represented a huge number of voters.
>>>German were second and french third. Other languages had basically not
>>>participated but for a very few people.
>>>Link : http://meta.wikimedia.org/wiki/Image%3AElection_participation2.png
>>
>>I don't agree with the conclusion you draw from this graph. The
>>English and German Wikipedias are much larger than those in other
>>languages, so it's only to be expected there will be more votes coming
>>from those. Taking the number of active editors in April 2004 as my
>>measure, I find that French participation is remarkably large, but
>>English and German is not much more than would be expected from a fair
>>division over the languages. In numbers (top 16 Wikipedia languages;
>>numbers are number of editors, number of voters, and the second as a
>>percentage of the first):
>>
>>French 321 89 28%
>>Finnish 33 5 15%
>>Norse 27 3 11%
>>Italian 69 7 10%
>>German 1613 145 9%
>>English 2746 238 9%
>>Dutch 191 16 8%
>>Chinese 143 11 7%
>>Esperanto 44 3 7%
>>Polish 124 8 6%
>>Swedish 98 6 6%
>>Danish 67 4 6%
>>Japanese 360 18 5%
>>Spanish 123 6 5%
>>Hebrew 69 3 4%
>>Portuguese 67 0 0%
>
>
> Very interesting. Thanks for that analysis! Fr: users are also
> unusually well integrated with IRC. Clearly the conclusion is that
> eating well makes you productive.
>
> SJ

I suggest that Erik and André's numbers (voters/very active users and
voters/user numbers) be added on meta for numbers analysis.

Though this can not be proved by numbers, I think these participation
rates recover two realities.
First involvement in meta topics. And if we consider number of editors
involved on meta, numbers of people in #wikimedia, numbers of people on
board related list, I think the high and lower percentage of
participation are clearly reflected. Typically spanish editors are not
very widespread on meta related issues while french are.

The second relaty is the existence of a candidate or not. When there was
no candidate in one language, this language did not participated much.
It is my understanding there is a candidate on nl wiki this year, so we
can probably expect higher participation rate. I however think
participation rates of es or ja will not be high :-(

ant

David Gerard:

> So I can write a statement in 1000 Chinese characters, and get that
> translated back to English... you sure about this one?

I'm not sure it's a good way to start a campaign by deliberately looking
for loopholes in the rules ;-). I don't think this will be a problem in
practice, but if you want, you can define that the most commonly used
language is used for the count, or you can retain some flexibility in
the enforcement. I'll leave it to the organizers to decide this. What is
important is that there is at least a basic requirement so that people
don't go overboard.

Erik