On 12/22/07, Terin Stock <terin.stock(a)wikinewsie.org> wrote: Great! I'll ask admins of Toolserver to make unix group "wikinews" so we would be able to edit files together. Python is not a hard to learn language. Learning shell scripting is much more difficult. Note for others: If you know to edit wiki and you know to use command prompt on Windows, it definitely qualifies you for option 3 ("willing to learn"). So, don't hesitate to join. More people are in the "bot team" means smaller response time for changes.

I too would very much like to learn python... :) If I can use that knowledge to further the goals and aims of the WMF in general and Wikinews in particular, well.... Yeah! I used to be quite good with Perl... then I grew up, and like any language, when you stop using it, you lose it. -N. On Dec 22, 2007 6:49 PM, Thunderhead <wm-thunderhead(a)charter.net> wrote: -- Nathan Reed nathanreed(a)gmail.com

Great to see that there are five of us :) On 12/23/07, Thunderhead <wm-thunderhead(a)charter.net> wrote: Yes, there is a list [1]. It is with very low activity (last email was written at November 2nd), which may be a good condition. So, it would be good that all of you join this list and to continue to talk about technical issues there. Please confirm here that you joined the list. (There is also pywikipediabot list and you should join there [2], too. But this list is pywikipediabot specific and it is used mainly by their developers to discuss pwb development.) But, we are Wikimedians and we have much better place for collaborative work. A couple of weeks ago I opened a course on Wikiversity [3] with the idea of collaborative learning. I wrote there a scratch, with a subpage for people who are using Windows. Note that the course is wider then our needs and our path. While someone may be completely content with using Windows, we have to pass a complete road to active usage of GNU/Linux because our bots have to be controlled from Toolserver. (So, for people who are using Windows, Cygwin is good enough.) Basic documentation for Python is its official tutorial [4], which is very good for the start. Anyone who has knowledge of some other language may learn Python basics in one hour. (Of course, if that person doesn't read the book, but primary examples and testing them.) For using pywikipediabot you will have to know to use subversion (actually, two commands: "svn checkout" and "svn update"). Take a look at their primary page [5]. On 12/23/07, Nathan Reed <nathanreed(a)gmail.com> wrote: Actually, my plan for Wikiversity course is exactly that: Starting with Wikimedia wide project from Wikinews :) On 12/23/07, the wub <thewub.wiki(a)googlemail.com> wrote: Whenever I am online (and it is very often, but note that I am living in GMT+1 time zone), (all of) you may ask me if you have some questions in relation to GNU/Linux and bots. At least, I am sure that I am able to tell you what documentation to read :) * * * Ok. Let's summarize where we are standing: - Terin, Nathan (?), Thunderhead (?) -- you should ask for an account on Toolserver [6][7] if you don't have accounts there. (If I rightly understood, Nathan and Thunderhead know to use GNU/Linux.) - The Wub, start with option 1 or 2 from this [8] page. Whatever we learn, we should write on Wikiveristy course page. Maybe it would be better to keep the main communication on wiki talk pages (with feed reader it is easy to follow all of the subpages :) ). * * * And about my stewardship... Thanks for congratulations. And it is very interesting to see that there is no way to abuse steward rights two times. Transparency is at very high level and there are not a lot of interesting things on private lists... Actually, as I remember it from my checkuser days, emails were very boring. So, it is funny to offer bribe to one steward :) It may have a sense only if stewards become one highly bureaucratic group of Wikimedians. Then you would have to bribe a steward to do something which had to be done. * * * [1] - http://lists.wikimedia.org/mailman/listinfo/wikibots-l [2] - http://en.wikiversity.org/wiki/Programming_bots_on_Wikimedia_projects [3] - http://en.wikiversity.org/wiki/Programming_bots_on_Wikimedia_projects [4] - http://docs.python.org/tut/tut.html [5] - http://meta.wikimedia.org/wiki/Using_the_python_wikipediabot [6] - The main page for Toolserver is: http://meta.wikimedia.org/wiki/Toolserver [7] - Place for asking user accounts is: http://meta.wikimedia.org/wiki/Toolserver/New_accounts [8] - http://en.wikiversity.org/wiki/Programming_bots_on_Wikimedia_projects/Prere…

Milos Rancic wrote: A lot of how weather is implemented will depend on how we choose to present it. At the moment we've just the global map with only temperatures. It, to start with, would be a huge improvement if the map can be clicked on to bring up a map of one of the continents/regions with additional data. That could be on a page of its own with capital city conditions detailed in text underneath. Taking into account the size of Asia might mean we need futher drilldown from that. I've done work on financial software for business in the past, how are you organising the data? Have you selected a base currency and stored all rates from it to other currencies? Either USD or EUR for that, then to get from any currency to any other currency you have two conversions unless one is the base. From these results you could build a table onto a country portal listing the country's currency and rate to a variety of other currencies. I've also dealt with the software side of transitioning from a national currency to EUR, so if you've any questions on that. [[Portal:Business knowledge]] going to: [[Wikinews:Business LSE]] [[Wikinews:Business NYSE]] [[Wikinews:Business NASDAQ]] ... Do we have anyone who tracks anything on the stock market could detail what information would be useful? In the cases here I believe the data isn't copyrightable, but the presentation may be. Thus we may have to work out our own way of presenting the information. These will be pages where tables of data will be presented, presumably with some in expand/collapse boxes to keep the initial presentation reasonable. People who are interested in that will either spend a minute looking up one detail or ages reading quite a bit out of the data. An example of how you could show the NYSE 100 is in a table with alternating shaded lines to ease reading across the data. Brian McNeil

Agreed, That would be a cood idea, and would make life easy. I would also like a !list command or something, that would list the days articles {{published}, in {{ready}} etc Brian Anderton -----Original Message----- From: wikinews-l-bounces(a)lists.wikimedia.org [mailto:wikinews-l-bounces@lists.wikimedia.org] On Behalf Of Brian McNeil Sent: Sunday, 23 December 2007 10:54 p.m. To: 'Wikinews mailing list' Subject: Re: [Wikinews-l] Organizing work on bots Here's what should be a simple one... A bot which tracks recent changes; where an article is tagged {{ready}} or {{publish}} it notifies #wikinews-en, or whatever language it is tracking. Tracking the recent changes channel isn't that useful and it becomes annoying when Chatzilla beeps all the time during an editing session (stalking of own username issue). Brian McNeil _______________________________________________ Wikinews-l mailing list Wikinews-l(a)lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/wikinews-l

On 12/23/07, Brian McNeil <brian.mcneil(a)wikinewsie.org> wrote: My first goal is to make a working general algorithm. When we have data, presentation is only a matter of our decision. So, here are some notes about my approach: - Like for all other comparable things (exchange rates, and even automatic adopting of news), I have two types of sources: global and local. Local source is more relevant, but until we don't have local sources for the whole world, we have to use for "the rest of the world" some global service. - So, again, as an achievable example, I am using data from Serbian Meteorological Agency as a local source and I am trying to find some good global service (I am in the phase of finishing algorithm for parsing data from Serbian Meteorological Agency site). - A couple of days ago, I was looking what should be covered firstly. Then I went to Wikipedia and realized that we need to think about the next places as our priority: List of urban areas by population [1], World's largest urban agglomerations [2] (those two are the most important), List of metropolitan areas by population [3], List of cities by population [4] (next two are important, too) and the less important List of cities by surface area [5]. So, those cities/urban areas should be covered at the global map (the last category only if there is enough of space in the area on the map). There would be a lot of space in Siberia, Canada, Arctic, Antarctica, even in Australia etc. So, we should find representative places in those areas (while it is quite easy for Australia, we will have to find the most important places in North Canada, Siberia, Greenland and the most important scientific bases on Antarctica. After that, we would be able to think about continental maps or by using some more natural type of partitioning Earth: Europe, Mediterranean, Middle East, South-East Asia, South Asia, Eastern Coast of USA and Canada etc. In the mean time, we need to have good resources. Primary sources are meteorological agencies of particular country/state/region. As many of them we have, our data will be more accurate. Note that those agencies have a lot of data which are interesting only at the country level (which may be our next goal for presentation)... I've done it a couple of days ago [6] and I may send to you the code (while it will be accessible on Toolserver). In short, there are "global" and "local" currencies. USD and EUR are "global" and other are "local". Local currency is giving only their rate to one of the global currencies and all other rates are done via global currencies. In this moment I have data from European Central Bank and National Bank of Serbia. As ECB doesn't have exchange rates for Serbian Dinar (as far as I understand economics, RSD is not traded currency or something like that), information for RSD is based on its rate compared to EUR, after which all other rates are calculated according to EUR rates. (Actually, National Bank of Serbia is doing the same: it defines relation between RSD and EUR, after which it calculates other rates on ECB rates.) It is possible to make similar algorithms for all other currencies. But, it is good to know that you are introduced in that. I am sure that I'll have questions for your. The most important is how to keep factually correct rates when part would be based on USD and part on EUR. So, it would be good to have someone who knows all of those things. I tried to find something, but I found data only on CNN's site, which may be accurate, but it shouldn't be our primary source of informations. * * * At the end, please think about presentation! I am so much in parsing and numbers so I am have a little time to organize presentation. Making a concept at one static wiki page as an example of your concept would be a very good input. [1] - http://en.wikipedia.org/wiki/List_of_urban_areas_by_population [2] - http://en.wikipedia.org/wiki/World%27s_largest_urban_agglomerations [3] - http://en.wikipedia.org/wiki/List_of_metropolitan_areas_by_population [4] - http://en.wikipedia.org/wiki/List_of_cities_by_population [5] - http://en.wikipedia.org/wiki/List_of_cities_by_surface_area [6] - http://en.wikinews.org/wiki/Template:Exchange_rates

On 12/23/07, Milos Rancic <millosh(a)wikimedia.org> wrote: Correct address is: http://lists.wikimedia.org/mailman/listinfo/pywikipedia-l :)

I know some basic Python and Perl, but my main languages are VB.NET and PHP. Obviously I know (X)HTML, JavaScript, CSS. I have experience in writing IRC bots in PHP and in modifying eggdrop bots with TCL. This can probably be ported to Python. ~ Paul Williams ~ en:n:User:Skenmy

OK. The course has minimums for start. Here is the structure: * Start page [1] * Advices for people who are using Windows [2] * Introduction to GNU/Linux [3] * Python tutorial is at their site [4] * First seven lectures in Pywikipediabot [5] (you may learn Python only with learning Pywikipediabot, if you don't understand something, feel free to ask on relevant talk page, via email, IRC or so... However, talk page is the best because other people would be able to find answers to your questions easily). Also, maybe we should write at the talk page of the main project page who are participants? And to start with sharing ideas and code there? [1] - http://en.wikiversity.org/wiki/Programming_bots_on_Wikimedia_projects [2] - http://en.wikiversity.org/wiki/Programming_bots_on_Wikimedia_projects/Prere… [3] - http://en.wikiversity.org/wiki/Programming_bots_on_Wikimedia_projects/Prere… [4] - http://docs.python.org/tut/tut.html [5] - http://en.wikiversity.org/wiki/Pywikipediabot

I've moved the old wikinewsbot framework forward since the time that page was posted, fixing a few bugs and creating a better API. The new code is at http://code.google.com/p/mediawikiengine/ and is used by the current version of the Wikimedia Commonplace app. The framework (and the app) work quite well on Mono under Linux and other environments where Mono runs. -ilya On Dec 25, 2007 8:09 PM, bawolff <bawolff+wn(a)gmail.com> wrote: