Hi Brad Patrick, Mailinglist
and local wikimedia/foundation responsibles.
Jim wants to create a search engine for the wiki community as strong as google. To fight central authorized search engines like google and to support the community help of searches in the internet
As it is good to have a wiki-relation, I sent the attached idea as well to you.
I read, that wiki.com is financially independent from the wikifoundation. And that the new search-engine wikisari or wikiAsariA, dunno, is with the purpose of earning money with ads in cooperation of amazon.
http://www.google.de/search?hl=de&q=wikisari&meta=
(yesterday google had only 5 results, today 238)
http://www.google.de/search?hl=de&q=wikisaria&meta=
(today 307 results)
Ok, maybe a parallel model or integrated model with the wiki-enterprise is possible in collaboration woith wikimedia foundation.
The idea is, to use the p2p indexing search engine yacy (see link below). (speek blacky like jacky Onasis)
Wiki and Amazon create some big servers and get a central webadress. Users can search the database (as a p2p node) from that central website and get advertisements. This central website is in cooperation with amazon and as well the amazon website gets a search box.
Then.. as yacy search engine is a p2p engine, we need a lot of nodes, running yacy.
To be strong as google we need at elast 20.000 peers running yacy. If there is less bandwith dedicated, we need maybe 50.000 users every minute online.
The idea is now (already posted with a draft in the yacy forum) to combine yacy with an instant messenger, users always run, Yacy could be stick to the messenger as a deamon and doing the indexing job of websites.
The bandwith for that could be adjusted with a slider in the Instant Messenger friendslist.
As the friendslist CSpace could be used. http://cspace.in
CSpace is an open Source and SERVERLESS Buddylist (so no icq aol msn yahoo or google talk, which are all serverbased). Cspace is fully decentral using a dht kademlia lookup for the buddies online in a p2p style.
If you stick both together, yacy and cspace, then users run it all the time and yacy could index and the buddylist coud offer as well a search box to run searches decentral on every computer direct accessing the p2p yacy network.
Many users will use the buddylist cspace then from wikipedia and support this way the backgroud daemon yacy indexing websites.
As well users then can acess the search engine using the wikiasaria homepage of amazon and wiki.com in a browser.
Furthermore we can make of the new application (yacy + cspace) an toolabr for firefox, thunderbird and songbirdnest.com
then we have the buddylist integrated and teh yacy background is doing website indexing in all the buddies an performing a big p2p net.
Wikiasarias Idea is to add manual page ranking to the website index. This is already coded in yacy and a browser toolbar exists as well.
We just need to distribute yay as a daemon of a budylist / Cspace and bring it into the community or into the branding of Wikimedia.
Yacy is written in java, CSpace in Python, but a common gui or a XUL-Toolbar for firefox should be no problem.
As both code branches exist, we have less money to take to get both together and bring the p2p net for buddies and website indexing out.
If anyone needs futher ideas, write a feedback.
Thanks for considering a plan, how to decentralize the search engines and websiteindex, in the wikimedia foundation baord,
As I see, more then 300.000 Dollars are already donated, maybe a coding team can bring both applications into one gui and wiki can distribute it, if wiki.com or wikiasaria then supports this with a central wesite search (as a big serverfarm-node of the p2p network yacy) this would be a perfect cooperation.
Thanks a lot
tom
PS1: Maybe you know the german-french search engine QUAERO project, which goal was to fight the majority of google.
http://en.wikipedia.org/wiki/Quaero
This project is broken and not done. So the need to have a community build search engine still exists.
PS2: Besides now: Please forward this mail to the techical administrator lists of wikipedia, then each wiki server can set up a yacy node already today and display as well yacy searche ngine results, as well the node can index the wikimedia site of the individual country and get the external links as well, as they are already the manual edited best links for each wiki-article. Would be perfect, to see in each wikipedia the yacy saerch window as well to have an external search option as well. maybe the donated money could be used to have some big yacy servers as well.
-------- Original-Nachricht --------
Datum: Mon, 25 Dec 2006 19:39:14 +0100
Von: "Thomas Müller" <thomasasta(a)gmx.net>
An: jwales(a)wikia.com
Betreff: wikipedia: wikiasaria the p2p way
Hello Jimmy,
I want to suggest to do the wikiasaria the p2p way.
There is alreay an open source software to have a p2p search engine to fight google.
http://www.yacy.net/yacy/Demo.html
You with the wikipedia background and amazon can support the open source java development.
Then you can built a big server to acess this p2p net over a central homepage and place here the advertisements you need to earn money and found it (and all servers supporting this).
YACY (speek Jacky like Blacky) is open source and could be customized to have an url rating from the users while they run an indexin node or aces your central webserver to search the index or make a toolbar (which already exists) to rate that page in the browser.
Already many nodes of yacy start with the wikipedia urls and links indexing.
To get the p2p indexing nodes popular, I suggested to have the indexing yacy node as an background deamon in an instant messenger, This could be Cspace, as it is serverless and open too. but it is python written, not java.
Http://cspace.in
If you need more proposal, how to code, I can send a draft of how they can interact.
kind regards and if you like, sent feedback.
Tom
--
Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer
--
Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen!
Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer
When people upload files to commons it would be nice if they would
consider that not everyone out there has the latest PC hardware.
My pc is a pentium IV, 2 ghz, with 1gb of RAM and a 64 mb graphics card.
Not the worst system out there I would think. I found a new map added to
the India article on the nl.wikipedia. Which looked quite nice so I
wanted to see the original version of it. I clicked on it and crash
there went firefox. Now I have managed to open the file but it
completely eats all of my computer resources! It eats 100% of my CPU
capacity and 500 mb of RAM for just this file (add to this the 250 mb of
RAM that gets eaten by the usual programs on my pc and I get very close
to my max). So when people upload files please see to it that people
with lower configurations can use it! Most people simply do not work
with the latest hardware available!
It is this file btw:
http://upload.wikimedia.org/wikipedia/commons/7/78/India_roadway_map.svg
Waerth
Regarding, http://en.wikipedia.org/wiki/User:Gmaxwell/nonalphausers
On 12/21/06, Yann Forget <yann(a)forget-me.net> wrote:
> If you want this to be meaningful, you need to remove accounts with only
> Arabic numbers. Obviously these can be recognized easily by English-only
> speakers.
Some are, some are not. 863929163 is not too easily remembered,
although I can type it. While 555 is a fine name for an English
speaker to communicate with... Feel free to skip over the ones that
seem too English compatible. With 1000 in the list there are plenty to
look at.
> > Obviously it does not count users who were blocked before they could
> > "contribute".
>
> This would be useful too, as a separate list, maybe.
Based on your suggestions I have created a new list consisting of all
usernames which have been blocked since Dec 2005 that contain any
non-ascii characters.
http://en.wikipedia.org/wiki/User:Gmaxwell/nonalphausers/blocked
They have been grouped based on the existence of edits by the users.
It would be useful if someone would go through and fix the places
where a loose RTL character is making everything backwards (don't
remove the RTL mark, just insert a matching LTR mark). I have no idea
how people edit mixed directional text with any ease, and I can't seem
to find help on any of our metapages.
Of 3,049,903 usernames in the enwiki database 454,122 contain at least
one non-ascii character. Of those, 2,016 have been blocked at least
once in the last year.
Of the 454,122 accounts with non-ascii characters 367,501 were created
in the last year.
I am somewhat surprised that so much fervor is being made of a matter
which appears, by the numbers, to of nearly insignificant frequency.
Erik Moeller wrote:
> On 12/14/06, Birgitte SB <birgitte_sb(a)yahoo.com> wrote:
>
>> I don't believe you understand how useful it would be
>> to just have some one say "That particular case is
>> unknown. The most similar case to this is Foobar."
>> Even if there were a table of questions that people
>> have asked in the past with yes/no/unknown and no
>> futher advice would be extremely helpful. I think it
>> a false expectation of yours that we are expecting
>> clearcut answers. Really we have been mucking through
>> copyright questions as best we can for some time; we
>> are all well aware there are often not answers only
>> arguments. Just being able to eliminate some
>> arguments as invalid would be very helpful.
>
> Aside from the potential issues with the WMF "officially" giving such
> advice to the communities, Brad (our GC and ED) simply doesn't have
> the time to do this. Let's brainstorm about how we can get juriwiki-l
> going, i.e. a functioning, community-driven group of advisors with
> demonstrable legal expertise.
>
> At the moment juriwiki-l is configured so that postings from the
> outside are moderated and replied to by a group of insiders. Is there
> any real issue, from a legal point of view, with making it a public
> mailing list? This is perhaps something Brad can answer.
In the context of rethinking the various internal, private mailing
lists, I'm not sure whether it will be considered appropriate to
continue juriwiki-l in its present form. As Anthere noted, the list is
not all that active at present. However, it has had sporadic use and a
variety of sensitive subjects were discussed there with an expectation
of confidentiality. So I would strongly oppose simply converting it to a
public list and making the accompanying archives public.
If a public list dedicated to legal issues is thought desirable, it
should be launched separately. We had wikilegal-l before, but that too
dwindled into oblivion.
--Michael Snow
Dictionary translations are infamously bad, as anyone who has worked in
translation or even travelled overseas can tell you. In Israel, my friends and I
used to love checking out the menus of cheap restaurants for the bad
translations. Two favorites come to mind:
1. Difficult egg -- hard boiled egg
2. Grenade juice -- pomegranate juice
>From film subtitles (which I got a lot of money to correct) -- You sun of
the beach -- You son of a bitch ( I remember this especially because the
original translator argued furiously that I did not know popular idioms in English
because I had lived overseas for too long, while he had an MA in English
Lit.).
I could go on and on. To translate adequately, you need to speak both
languages. To translate well, you have to be immersed in both cultures.
Danny
(10+ years translating professionally)
In a message dated 12/20/2006 5:32:27 PM Eastern Standard Time,
gerard.meijssen(a)gmail.com writes:
habj schreef:
> 2006/12/18, Brad Patrick <bradp.wmf(a)gmail.com>:
>
>> Most of it
>> should be possible to translate with merely a dictionary.
>>
>
> I must warn against this attitude. Hardly anything can be translated
> with merely a dictionary, lacking context. If one tries, one ends up
> with a certain percentage of errors which sometimes just look a little
> funny and sometimes look very bad.
>
> /habj
Hoi,
Translation with the knowledge of only a dictionary gets you a
translation that is probably not as good as a machine translation. For
material that needs to be just so and this is true for the whole of
marketing / advertisement it PAYS to have proper translations. It even
PAYS to have it done by a professional translator because it is pennies
to the pound. When you have the connections with translators you can
even get these translations for a very reduced rate. It needs however a
proper understanding and approach. I am sure you do not want to have a
translation into English for a message like this on the English language
projects !!
I have been told that a quality translation improves the reach of your
communication by at least 15%.
Thanks,
GerardM
What about setting things up so that to create a SUL username with say,
Tegulu characters, you *must* register at TE, no other language's
registration form will accept it. That will significantly cut down on vanity
and vandal usage of the such characters.
Nick
Date: Wed, 20 Dec 2006 22:50:33 -0500
From: Stephanie Erin Daugherty <stephanie(a)sosdg.org>
This is especially true when abusive users have deliberately taken advantage
of this fact in order to make the lives of administrators and of other
Wikipedians as difficult as possible, and it's also true for those handful
of users that will mix character sets to "look cool" at the inconvenience of
others.
While this may well be important, it doesn't seem to be quite the same thing as what we've been discussing, i.e. whether usernames should be transliterated into the script that is local to the Wikipedia in question, not whether the result of the transliteration will seem correct in a given language. It may be that лл never appears in Serbian and Montenegrin, but nobody on en would object to a username containing "Qv", even though that sequence never appears in English. A topic which may have to be discussed at some point but which seems rather tangential right now is whether the Cyrillic script (or the Latin, for that matter) is one alphabet or several.
Cheers,
Nat Krause
On 12/23/06, Milos Rancic wrote:
Hm. In Serbian and Macedonian we are using strictly transcriptions,
not transliterations. So, it should be something like
English->IPA->Serbian/Macedonian. Howerver, the process is not so
simple and in some cases we are using combinied transcription and
transliteration, as well as somtimes historic names. The best examples
are name Angela and surname Wales -- for both of them we are using
historical names (not transcription Ејнџела but historical name
Анђела; not transcription Вејлз, but historical name for the country
Walest -- Велс). However, engine is
possible.
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
In Polish edition of Newsweek from 31.12.06 there is a supplement called
"Future of information". Although it is not pointed clearly this
supplement is obviously sponsored by Larousse Poland and in fact it is a
kind of "smart" advertisement of Polish edition of "Popular Larousse
Encyclopedia".
Among other articles describing advantages of Larousse over Polish
Scientific Publishers (PWN, http://en.pwn.pl/) encyclopedias and other
potential "paper" competitors of Larousse in Poland, there is also large
article comparing Larousse's encyclopedia with Polish Wikipedia.
The comparison is based on three articles "Anatomia" (anatomy), "Islam"
and "Ogniwo słoneczne" (solar cell). One of these articles is marked
"stub" in Polish Wikipedia, the second one is disambig, the third one is
normal, long article. What is interesting, there is no claims that there
are factual mistakes in Wikipedia, however the comparison points out
that the consistency of text, amount and quality of pictures and ease of
reading is much better in Larousse than in Wikipedia. As I personally
can agree with better consistency of text and ease of reading, the
claims that amount and quality of pictures is better is based only on
the pictures one can see in articles, completely ignoring what can be
found in linked galleries in Commons and after clicking on bigger
versions of pictures. Of course the selection of articles for comparison
was also not random but made by choice of guys from Larousse. Other sets
of articles could probably give completely different results.
Anyway, I think the fact that this sponsored supplement put much more
place to compare their encyclopedia with Wikipedia (three pages long
article) and not to the other formally more serious competitors like PWN
encyclopedia and Polish edition of Brittanica is interesting by itself.
It shows that Larousse is more afraid of us than of PWN or Brittanica...
or we were simply easer target of not very fair negative comparison.
--
Tomasz "Polimerek" Ganicz
http://pl.wikimedia.org/wiki/User:Polimerekhttp://pl.wikipedia.org/wiki/Wikipedysta:Polimerekhttp://www.poli.toya.net.pl/http://www.ptchem.lodz.pl/pl/TomaszGanicz.html
Hi all,
is it *please* possible to stay civil to eachother? Is it *please* possible
to keep the discussion on the main points? When you are unable to agree with
eachother you can start calling eachother names, and saying the other one
does so, you can also either try to find a solution suitable for both,
either a compromize either just accept that you disagree. These long threads
are disruptive for this list, especially as more and more emails are about
how people are, and not about the subject. I think very little people will
stay reading this, but will also not read anymore the subjects that actually
*do* have a relevant discussion as in talking about the subject. Please dont
be distructive to eachother, and respect that people can sometimes have
another opinion, and if they are not able to be persuiated, dont go on for
ever on this list, but find other methods.
Effeietsanders
2006/12/22, Bryan Tong Minh <bryan.tongminh(a)gmail.com>:
>
> I fully understand it isn't. The point I wanted to make, but
> formulated rather impolite was that hatred will never be solved by
> hatred. It has been shown in the past, and will be shown in the
> future. I will abstain from any further comments regarding this
> thread, as it is its nature to grow into a flamewar.
>
> Bryan
>
> On 12/22/06, Gerard Meijssen <gerard.meijssen(a)gmail.com> wrote:
> > Hoi Bryan,
> > It is not polite nor helpful to say that someone is "immature and
> > moreover a hypocrite". The only result is that you inflame the situation
> > even more. It is bad as it is.
> > Thanks,
> > GerardM
> >
> >
> > Bryan Tong Minh schreef:
> > > * Don't disrupt Wikipedia to prove a point
> > >
> > > Let me myself add something: You're hate for anglosaxon related
> > > subjects is disconcerning. Your prejudgment prevents any constructive
> > > discussion. You have just proven yourself to be immature and moreover
> > > a hypocrite. Please notice that your comments are in no way helpful,
> > > and prevents any further discussion.
> > >
> > > Bryan
> > >
> > > On 12/22/06, Walter van Kalken <walter(a)vankalken.net> wrote:
> > >
> > >> After Gregory's action I want to ask all wiki's who use non-latin
> > >> languages as their standard to start blocking all latin character
> user
> > >> accounts and demand they use accountnames in your script, especially
> > >> those of en.wikipedians actively supporting this en: policy.
> > >>
> > >> Maybe this way they will understand on en: that there action is
> > >> extremely biased and very very bad as they seem to insist to keep
> > >> blocking good users like admins from other wiki's simply because
> people
> > >> want to use their REAL NAME.
> > >>
> > >> Start with Gregory's account btw.
> > >>
> > >> Waerth
> > >>
> > >>
> > >>> Gregory Maxwell schreef:
> > >>>
> > >>>
> > >>>
> > >>>> On 12/22/06, Birgitte SB <birgitte_sb(a)yahoo.com> wrote:
> > >>>>
> > >>>>
> > >>>>
> > >>>>
> > >>>>> The en.WP Username policy has been changed.
> > >>>>>
> http://en.wikipedia.org/w/index.php?title=Wikipedia:Username&curid=168684&d…
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>> Unfortunatly, it was changed without any valid basis.
> > >>>>
> > >>>> It is not acceptable to remove longstanding policy simply because
> some
> > >>>> people on a mailing list decided to speculate randomly and accuse
> > >>>> enwiki of "ethnocentricism" without substantiation.
> > >>>>
> > >>>> As a result I've reverted the change and contributed some
> suggestions
> > >>>> to the talk page.
> > >>>>
> > >>>> I agree that the policy should be *improved* to address the real
> > >>>> concerns brought up, and to be more friendly to users who work
> across
> > >>>> many languages. Simply removing the section does not, however,
> > >>>> address the very real issues which caused the creation of the
> policy
> > >>>> in the first place.
> > >>>>
> > >>>>
> > >>>>
> > >>> Hoi,
> > >>> Some people congratulated me with the outcome of this affair. My
> > >>> reaction was that it did not feel like a victory because I feel
> bruised
> > >>> by the experience. Having said this, your suggestion that there has
> been
> > >>> no substantiation about the grievance that was put forward is
> seriously
> > >>> wrong. The notion that it is up to you to revert this change
> continues
> > >>> to make this affair even more painful. Simply reverting this change
> and
> > >>> thereby denying that many of the reasons why this policy was put in
> > >>> place is seriously wrong.
> > >>>
> > >>> I have to say I am increasingly upset by this affair.
> > >>>
> > >>> Thanks,
> > >>> GerardM
> > >
> >
> > _______________________________________________
> > foundation-l mailing list
> > foundation-l(a)wikimedia.org
> > http://mail.wikipedia.org/mailman/listinfo/foundation-l
> >
> _______________________________________________
> foundation-l mailing list
> foundation-l(a)wikimedia.org
> http://mail.wikipedia.org/mailman/listinfo/foundation-l
>
On http://fundraising.wikimedia.org there are now next to English some
translations made. How can we make a translation in another language?
Hans (JePe)