Re: [Wikimedia-l] Invitation to WMF November 201 4 Metrics…

2014-12-05 Thread Per A.J. Andersson
Thanks, Asaf and the others, for the talk about the global south 
in deeper terms. That's both an important and interesting perspective.


I've tried to get a grip on the issue of the two major languages 
in Ukraine, and how they affect the use and development of 
Wikipedia. But I haven't yet read any detailed info of those 
device defaults. Do you know more about what kinds of devices, 
what kinds of applications, the reasoning involved and any talk 
of trends? I'd be very happy to know more on the subject. :-)


Best of wishes,
/Per A.J. Andersson, Göteborg, Sweden
(Wikipedia user: Paracel63)


Date: Thu, 4 Dec 2014 16:18:47 -0800
From: Asaf Bartov abar...@wikimedia.org
To: Wikimedia Mailing List wikimedia-l@lists.wikimedia.org
Subject: Re: [Wikimedia-l] Invitation to WMF November 2014 Metrics 
Activities Meeting: Thursday, December 4, 19:00 UTC
Message-ID:
caamrcwccdtug7_dwa90k+nfn-s2sghhxuwpakpo7kh0qfgg...@mail.gmail.com
Content-Type: text/plain; charset=UTF-8



/snippet/


In Anglophone Africa, for example, most people are used to looking for
information online in English and not in indigenous languages.  But in
Brazil, people consume information in Portuguese, but many (16%) also refer
to the English Wikipedia (and intriguingly, 1 in 3 *edits* from Brazil is
to ENWP!), presumably for its broader coverage or higher average quality.
In Ukraine, 70% read the Russian Wikipedia and only 17% read the Ukrainian
Wikipedia; interviews tell me this is largely due to device defaults,
beyond the obvious different in size and average quality.

This page reveals some of those breakdowns:
http://stats.wikimedia.org/wikimedia/squids/
SquidReportPageViewsPerCountryBreakdown.htm



___
Wikimedia-l mailing list, guidelines at: 
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, 
mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe

Re: [Wikimedia-l] Erik Zachte on bot-generated articles

2013-06-23 Thread Per A.J. Andersson

2013-06-23, 10:34, skrev Jane Darnell:


I totally agree
(disclaimer: I don't have a bot, but I have been actively flooding
the engish wikipedia with artist stubs for years, and have used
municipality stubs created by bots to anchor their hometowns, working
places, and death towns)


I'm hoping so too. Apart from Lsjbot, Nasko.bot has actively 
been producing quite substantial svwp articles on lakes for some 
time. They are very useful in my local/regional article work – 
re for example http://sv.wikipedia.org/wiki/Sandsjö,_Halland


And I believe others at svwp can equally benefit from these 
stubs in their local/regional editing.


My two cents,
/Per
Wikipedia user Paracel63


___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Erik Zachte on bot-generated articles

2013-06-23 Thread Per A.J. Andersson

2013-06-23, 12:33, skrev Ilario Valdelli:


I am not in the opposition of the use of the bots, if these bots will have
their place and the articles generated by bots are considered like
*populated templates* and not like articles.

I think that a solution like this (the pages generated by bots are not
included in the sum of articles), may be a really good compromise.


Interesting idea. I'd be willing to accept an article counting 
first when non-bots have been involved. However, it's up to the 
local community (and those able to change the arcticle counting mechanism).


Best of wishes,
/Per
Wikipedia user Paracel63


___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Erik Zachte on bot-generated articles

2013-06-23 Thread Per A.J. Andersson
Sorry. Just a clarification. I hope you mean bot-generated 
articles are to be excluded in the tally only as long as they 
remain non-humanly edited. I see no point in indefinitely 
excluding some articles from a global tally, as all articles 
regardless of creation are bound to change content and style 
over time.


Best of wishes,
/Per

2013-06-23, 13:06, skrev Per A.J. Andersson:


2013-06-23, 12:33, skrev Ilario Valdelli:


I am not in the opposition of the use of the bots, if these bots will have
their place and the articles generated by bots are considered like
*populated templates* and not like articles.

I think that a solution like this (the pages generated by bots are not
included in the sum of articles), may be a really good compromise.


Interesting idea. I'd be willing to accept an article counting 
first when non-bots have been involved. However, it's up to the 
local community (and those able to change the arcticle counting mechanism).


Best of wishes,
/Per
Wikipedia user Paracel63


___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l




___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Wikipedia in jail!

2013-06-19 Thread Per A.J. Andersson

2013-06-19, 10:13, skrev rupert THURNER:


Hi Robert, no censoring, filtering, or monitoring. Its the full French
wikipedia. Afaiu emmanuel it was not discussed, but he explicitly pointed
out that all articles are there  including the ones about arms.


Interesting. Do you know how much data that involves? The last 
count here:

http://stats.wikimedia.org/EN/TablesWikipediaFR.htm
is from 2009 (3.5 GB), so I guess even now it will be easily manageable.

Best of wishes,
/Per


___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Swedish Wikipe dia reach (…)

2013-06-18 Thread Per A.J. Andersson

2013-06-18, 00:57, skrev Ziko van Dijk:


Hello,
I am also unhappy with the mail from Hubertl, and also some remarks that
good be understood as a criticism of the German Wikipedia editing
community. Actually, both opinions coexist also in de.WP, although the anti
bot faction is obviously stronger.
My concern is that bot articles usually stay the same and don't grow much.
They give a bad impression about a Wikipedia language version, and there is
no one to update them. Maybe it would be better to support WikiData and
later find a solution with WikiData to provide data to small or large
Wikipedia language versions.
Kind regards
Ziko


Many thanks for your clarifying reply. I just thought it a bit 
strange that these kinds of arguments (including bad language in 
a mail communication) were/are coming from a certain direction. 
No offense intended from my part.


Your idea of the extended use of Wikidata is also discussed in 
the svwp community, and there are some obvious possibilites. 
Right now we are getting accostumed to the new way of interwiki 
linking, where often (through the proportionate weight of enwp) 
the specific English definition of words and concepts are taken 
for granted.


Best of wikiwishes,
/Per


___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Swedish Wikipedia reach 1 million (with support of bots)

2013-06-18 Thread Per A.J. Andersson
I am a bit suprised about the lack of trying to understand and 
develop mindsets here. I joined this mailing list the other day 
in the belief that it was directed at discussing current topics 
in the Wikimedia community in a constructive manner. So far I 
have not found overwhelming proof of this.


Hopefully I can be proved wrong.

Best of wishes,
Per
Wikipedia user Paracel63

2013-06-18, 09:32, skrev Federico Leva (Nemo):


Anders Wennersten, 18/06/2013 09:00:

I am happy to see that you, as well as Hubertl, is happy to go on as you
always have done. I do not have the ambition to change your view on
life, [...]


To clarify, my view on life didn't include any item about bot 
article creation last time I checked.
There's also nothing I always have done on this topic, as far 
as I remember: I'm just observing global patterns in our wikis. 
I really don't care about single instances of the patterns, 
especially when I can't do anything about them.


Nemo

___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l




___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Swedish Wikipedia reach (...)

2013-06-18 Thread Per A.J. Andersson
Very good point. (Personally I think building a localised 
Wikispecies – with the potential to grow – inside a 
Wikipedia version is something good.)


Many thanks,
/Per
Wikipedia user Paracel63

2013-06-18, 14:50, skrev Gerard Meijssen:


Hear,  hear
Thanks,
GerardM
Op 18 jun. 2013 12:08 schreef Emilio J. Rodríguez-Posada emi...@gmail.com
het volgende:


Seriously, are we discussing again about bot stubs yes, bot stubs no?

Those users who want to submit complete 10-page articles, can move to the
defunct Nupedia or the 'vibrant' Citizendium.

This eternal discussion is so boring.



___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Swedish Wikipedia reach 1 million (with support of bots)

2013-06-18 Thread Per A.J. Andersson

2013-06-18, 06:34, skrev Hubertl:

Loosing our original idea for which this project is donated by 
thousands of donaters! From which you are paid for. As an 
unpaid, long term Wikipedian in Residence I do know what I´m 
talking about.


Hello!

I'm interested in knowing more of your experience (as a WiR) in 
relation to our current topics. Do you have a link to relevant 
data och texts/reports?


Many thanks in advance,
/Per
Wikipedia user Paracel63


___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Swedish Wikipedia reach 1 million (with support of bots)

2013-06-18 Thread Per A.J. Andersson
Many thanks for the resumé. I thought it would be like this, as 
I've followed the development of the different wikipedias 
(content and key factors) since 2008. BTW, is there a 
(searchable) list archive?


Best of wishes,
/Per

2013-06-18, 17:29, skrev Nathan:


Hi Per,

The discussions on this list can be heated, and not everyone is always
on their best behavior. All the caveats of debate on the Internet
apply, and any personality problems you might encounter on Wikimedia
projects occur on the lists as well. This is particularly so when, as
in this case, the topic has been debated for many years without real
resolution.

~Nathan

___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l



___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Swedish Wikipe dia reaches (…)

2013-06-18 Thread Per A.J. Andersson
Great idea! I'm going to try a way of doing this, through svwp and/or WMSE.

Best of wishes,
/Per
Wikipedia user Paracel63

2013-06-18, 17:53, skrev David Cuenca:

Instead of just discussing this topic as always has been done, it would be
great to take advantage of the opportunity to monitor changes in the sv-wp
community as a result of this massive automated article creation.

For me it would be interesting to know:
1) current perception in the community about automated articles correlated
with the number of contributions by the person answering.
2) impact of the newly created articles: views, human edits done to those
articles (specially nr of bytes added), correlation with nr of previous
edits
3) changes in the community after 1 year, survey of new members, of those,
how many registered after seeing a bot article vs how many after seeing a
human-created article?
4) some metrics about the article quality (maybe with AFT)

Otherwise the same arguments are going to be repeated over and over again,
year after year, without ever reaching a conclusion, or without ever
learning which impact those articles might have.

Cheers,
Micru
___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Fwd: Fwd: Re: Swedish Wikipedia reach 1 million (with supportof bots)

2013-06-17 Thread Per A.J. Andersson

Hi!

Sad for seeing you so disappointed. Good to see you have insight 
in butterflies. So I hope you will make your knowledge visible 
on dewp, where neither genera so far have found its way into 
articles. enwp (together with viwp!) has at least the genus 
Jameela as an article.


That aside, is it really a disaster that svwp relies on 
scientific texts that was unquestioned until three years ago? I 
think maybe there are more pressing concerns in the development 
of our wikipedias.


Best of wishes,
/Per
Paracel63 at svwp

2013-06-17, 22:29, skrev Kurt Kulac:


i don´t want to repeat all those arguments already quoted (lennard
already mentioned achim raschkas criticism, which i fully agree with),
since what happend, already happend, and will unstoppably happen in the
future.  but i would say first of all the strategic goal is not to
contribute hundreds of mistakes a day. just to take the 1 millionth
article Erysichton elaborata: after an advice on the talk page a HUMAN
added, that the species is probably synonymized with Erysichton palmyra.
so far, so good. but even this correction is not enough. in 2010 the
genus erysichton was redefined and a new genus, jameela was described.
both taxa now desrcibed as in sv.wikipedia are invalid.

so just concerning this tiny tiny group of articles, there is already a
bunch of mistakes, the bot copied out of outdated databases. wouldn´t be
a big deal, if somebody mentioned that the articles sticked to the old
view. but that´s something a bot can´t handle. so how reliable is the
rest of the articles?

a bot can be a convenient helper for authors, who know, how to handle
it, as it seems has happened with the creation of the articles about
swedish lakes (i´m no expert with that though). but it is a desastrous
tool for our whole movement, if you create hundreds of thousands (!)
articles, without the slightest idea, how to handle the contradictions,
that will appear doubtlessly?

it´s a sad thing, that you mention quality and this action in one
centence...

truly utterly disappointed encyclopedic greetings,
kurt









___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l


Re: [Wikimedia-l] Swedish Wikipedia reach 1 million (with support of bots)

2013-06-16 Thread Per A.J. Andersson
Milestones are nice. They're one big reason for celebrating, 
once in a while. There are lots of others, but not all are this 
definable and publicly palatable. Quality projects of various 
kinds are ongoing all the time, and we need different kinds of 
carrots to keep up the community interest at svwp. We need to 
communicate our efforts in different dimensions, and this was 
also the idea in our press release statement some hours ago. 
Quantity and quality are no born enemies to each other, IMHO.


Personally I enjoy making small articles (in various subjects) 
larger and better. I think this is a common motivator to many of 
us, where the article creation can be a hassle with all the 
multidimensional stuff involved. Lsjbot makes my enlarging a bit 
easier. And the fact that Lsjbot now creates a fully sourced and 
categorised Swedish-language wikispecies inside of Wikipedia 
is a bonus.


Best of wishes,
/Per A.J. Andersson
(user: Paracel63)

2013-06-16, 16:03, skrev Federico Leva (Nemo):


Article count milestones are nice!
However, I remind those who were distracted in 2008 that 
article count is a largely irrelevant piece of our statistics 
since 2008; both stats.wikimedia.org and www.wikipedia.org 
(plus the other project portals) don't use it at all to rank 
projects, etc.

https://meta.wikimedia.org/wiki/Top_Ten_Wikipedias


Anders Wennersten, 16/06/2013 15:32:

Is it a coincidence that we now see the growth of these competent
botgenerated article in nl:wp, sv.wp and it:wp?


{{citation needed}}
To me it seems the other wikis are rather stable. There is no 
wiki in top ten with 50 % bot created articles and after 
sv.wiki (13th) you have to reach positions 20th (vi) and 24th 
(vi) to find similar amounts.


Nemo

___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l




___
Wikimedia-l mailing list
Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l