On Mon, Mar 2, 2009 at 8:42 AM, Syagrius syagrius@gmx.fr wrote:
The recent mass bot addition on Portuguese Wikisource (whose "number of articles" overtook French and Chinese) should make us think about changing the main page system. There was a discussion there : http://wikisource.org/wiki/Talk:Main_Page#Changing_the_main_page . Must we adopt the "page views" system, as Wikipedia did, http://stats.wikimedia.org/wikisource/EN/TablesPageViewsMonthly.htm ?
Page views is based on the expectation that those views are *readers*, and is appropriate for Wikipedia which has become part of typical Internet usage the average person.
A large percentage of our page views will be *editors* and *bots*.
If we do go to page views, we might also want to weight the page views based on average page size.
-- John Vandenberg
John Vandenberg a écrit :
On Mon, Mar 2, 2009 at 8:42 AM, Syagrius syagrius@gmx.fr wrote:
The recent mass bot addition on Portuguese Wikisource (whose "number of articles" overtook French and Chinese) should make us think about changing the main page system. There was a discussion there : http://wikisource.org/wiki/Talk:Main_Page#Changing_the_main_page . Must we adopt the "page views" system, as Wikipedia did, http://stats.wikimedia.org/wikisource/EN/TablesPageViewsMonthly.htm ?
Page views is based on the expectation that those views are *readers*, and is appropriate for Wikipedia which has become part of typical Internet usage the average person.
A large percentage of our page views will be *editors* and *bots*.
there is no perfect system.
for small subdomains with few page views, it is true that pages views are likely to be caused by editors and bots (I suppose you mean search engine bots rather than our wiki bots).
If you consider this table : http://stats.wikimedia.org/wikisource/EN/TablesPageViewsMonthly.htm you can see that the top 10 domains all have had months with more than 1million page views.
For these subdomains, the bias induced by bots should be small.
This is not true for the traffic caused by editors. However, I would say that this part of the traffic is not completely meaningless.
Thomas
On Sun, Mar 1, 2009 at 4:42 PM, Syagrius syagrius@gmx.fr wrote:
Must we adopt the "page views" system, as Wikipedia did, http://stats.wikimedia.org/wikisource/EN/TablesPageViewsMonthly.htm ?
No, you don't *have* to do anything. The decision rests with the local Wikisource community, the wikipedia.org portal was changed because of a poll on Meta-Wiki: http://meta.wikimedia.org/wiki/Top_Ten_Wikipedias.
Syagrius wrote:
The recent mass bot addition on Portuguese Wikisource (whose "number of articles" overtook French and Chinese) should make us think about changing the main page system. There was a discussion there : http://wikisource.org/wiki/Talk:Main_Page#Changing_the_main_page . Must we adopt the "page views" system, as Wikipedia did, http://stats.wikimedia.org/wikisource/EN/TablesPageViewsMonthly.htm ?
In this instance I must concur with the prevailing view that there is no need for change. I believe that most regular contributors have little need to look at the Main Page. Their access to the system begins from a page that is key to whatever they are doing. Thus, the ones who are more likely to need the Main Page are the pure readers, and casual visitors. Before we consider the proposed change we need to consider why we have a Main Page in the first place. The number of pages in the project is far more informative to the casual user than the number of page views. Knowing that there are a lot of articles that might satisfy his needs is useful; what can he do with the knowledge of how many people have looked at these pages?
The most useful function of the Main Page is to help the user find what he is looking for. That should be the overwhelmingly primary principle that guides the design of that page.
Occasional anomalies such as what you mentioned about pt:wikisource should not be driving the agenda of what happens on the Main Page. Even if some kind of massive gaming or mischief was intended such anomalies tend to be overwhelmed by more steady and consistent behaviour.
Ec
On Mon, Mar 2, 2009 at 2:12 PM, Ray Saintonge saintonge@telus.net wrote:
Occasional anomalies such as what you mentioned about pt:wikisource should not be driving the agenda of what happens on the Main Page. Even if some kind of massive gaming or mischief was intended such anomalies tend to be overwhelmed by more steady and consistent behaviour.
Ec
Thankyou for your nice words about adding valid and referenced public domain content in a language spoken by 191 milion of peoples.
I strongly agree with this message: http://wikisource.org/w/index.php?title=Talk:Main_Page&diff=247906&o...
If anyone is interested, I can start in the next weekend a draft for a vote on main page changes.
Hi. Though i did not discus to this matter yet, here some thoughts. I do not matter much how the main page is graphically done. But it is fully insufficient how the activity is beeing measured. The number of pages is a very bad method. In the past, many projetcs used this to get more on the top by bot edits (e.g. the sk.wiki did some more thousands pages of Czech villages or meteorits, everything one line), and also some Wikisource domains did: I remeber one that published the Bible on some more hundred od thousand pages (just one page a verse), now the cs.source published some 770 pages of The Rubaiyat of Omar Khayyam (four verse a page in all). Such decisions can have some real reasons but I guess there is also the fact to make many pages and to get a good place in the top ten or top twenty or so.
Therefore there should be a change that shows more precisely the activity of a subdomain.
Regards,
Jan
-jkb-
----- Original Message ----- From: "ThomasV" thomasV1@gmx.de To: "discussion list for Wikisource,the free library" wikisource-l@lists.wikimedia.org Sent: Tuesday, March 03, 2009 2:09 PM Subject: Re: [Wikisource-l] Changing the Main Page
Luiz Augusto a écrit :
If anyone is interested, I can start in the next weekend a draft for a vote on main page changes.
that would be great. I will be glad to help you
_______________________________________________ Wikisource-l mailing list Wikisource-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikisource-l
jkbwiki wrote:
insufficient how the activity is beeing measured. The number of pages is a very bad method. In the past, many projetcs used this to get more on the top by bot edits (e.g. the sk.wiki did some more thousands pages of Czech villages or meteorits, everything one line), and also some Wikisource domains did: I remeber one that published the Bible on some more hundred od thousand pages (just one page a verse), now the cs.source published some 770 pages of The Rubaiyat of Omar Khayyam (four verse a page in all).
The number of wiki pages is indeed a bad measure for an e-text project, exactly because each page could contain a whole book, a chapter or a single verse. But for a book scanning project, the number of book pages (pages in the Page: namespace, facsimile images) is a very good measure, since it gives equal weight to one thick book of 400 pages as to two thinner books of 200 pages each.
You can also roughly translate 20,000 book pages (or 100 books of 200 pages each) to one metre of shelving, which is a measurement that any librarian or archivist can immediately understand.
The New Student's Reference Work in 5 volumes has 2516 pages, http://en.wikisource.org/wiki/The_New_Student%27s_Reference_Work
That was 5-6 times more work than Meyers Blitz-Lexikon with 443 pages, http://de.wikisource.org/wiki/Meyers_Blitz-Lexikon
Lars Aronsson wrote:
jkbwiki wrote:
insufficient how the activity is beeing measured. The number of pages is a very bad method. In the past, many projetcs used this to get more on the top by bot edits (e.g. the sk.wiki did some more thousands pages of Czech villages or meteorits, everything one line), and also some Wikisource domains did: I remeber one that published the Bible on some more hundred od thousand pages (just one page a verse), now the cs.source published some 770 pages of The Rubaiyat of Omar Khayyam (four verse a page in all).
The number of wiki pages is indeed a bad measure for an e-text project, exactly because each page could contain a whole book, a chapter or a single verse. But for a book scanning project, the number of book pages (pages in the Page: namespace, facsimile images) is a very good measure, since it gives equal weight to one thick book of 400 pages as to two thinner books of 200 pages each.
Yes, I agree, but I would have reservation to use the nomber of uncorrected pages and only that, as it doesn't show the real work done, only how much bots work ;o). So I see several possibilities: 1. number of proofread pages; 2. number of validated pages; 3. a mix from the 3 numbers, f.e. (number of validated pages X 4) + (number of proofread pages X 2) + total number of pages
Regards,
Yann
You can also roughly translate 20,000 book pages (or 100 books of 200 pages each) to one metre of shelving, which is a measurement that any librarian or archivist can immediately understand.
The New Student's Reference Work in 5 volumes has 2516 pages, http://en.wikisource.org/wiki/The_New_Student%27s_Reference_Work
That was 5-6 times more work than Meyers Blitz-Lexikon with 443 pages, http://de.wikisource.org/wiki/Meyers_Blitz-Lexikon
On Tue, Mar 3, 2009 at 10:09 AM, ThomasV thomasV1@gmx.de wrote:
Luiz Augusto a écrit :
If anyone is interested, I can start in the next weekend a draft for a vote on main page changes.
that would be great. I will be glad to help you
I've created the page http://wikisource.org/wiki/Wikisource:Changing_the_main_page , that intends to be a starting point for those with no time or patience to read the entire discussion on wiki and here at the mailing list. This may be helpful specially for those that will be known this issue only when spamming their local Village pump, for example. My intention is to make a detailed page like the http://meta.wikimedia.org/wiki/Top_Ten_Wikipedias
You and everyone on this list is invited to improve that page with more details about the "worries" mentioned in the current page version, with pros and cons to the specified options or adding new options to change the main page.
On Wed, Mar 11, 2009 at 10:30 AM, Luiz Augusto lugusto@gmail.com wrote:
On Tue, Mar 3, 2009 at 10:09 AM, ThomasV thomasV1@gmx.de wrote:
Luiz Augusto a écrit :
If anyone is interested, I can start in the next weekend a draft for a vote on main page changes.
that would be great. I will be glad to help you
I've created the page http://wikisource.org/wiki/Wikisource:Changing_the_main_page , that intends to be a starting point for those with no time or patience to read the entire discussion on wiki and here at the mailing list. This may be helpful specially for those that will be known this issue only when spamming their local Village pump, for example. My intention is to make a detailed page like the http://meta.wikimedia.org/wiki/Top_Ten_Wikipedias
Thank you! That is exactly what we need to come up with a good solution.
-- John Vandenberg
Luiz Augusto a écrit :
I've created the page http://wikisource.org/wiki/Wikisource:Changing_the_main_page , that intends to be a starting point for those with no time or patience to read the entire discussion on wiki and here at the mailing list. This may be helpful specially for those that will be known this issue only when spamming their local Village pump, for example. My intention is to make a detailed page like the http://meta.wikimedia.org/wiki/Top_Ten_Wikipedias
In addition to the debate about how we should rank subdomains, I think we should discuss whether it is really useful to display a main page that gives so much importance to ranking.
I believe the main page should not be a 'hall of fame'. It should not be an incentive to inflate number of pages.
To illustrate this, I have made a new proposal for the layout, where only the top 10 languages are displayed. I think this simplified layout is much easier to read.
This relates to my previous comment : http://wikisource.org/w/index.php?title=Talk:Main_Page&diff=247906&o... http://wikisource.org/w/index.php?title=Talk:Main_Page&diff=247906&oldid=247810
This proposal is only about layout : it is independent from the choice of criterion that we use to select the top 10 languages.
Thomas
ThomasV a écrit :
To illustrate this, I have made a new proposal for the layout, where only the top 10 languages are displayed. I think this simplified layout is much easier to read.
sorry, I forgot the link :
ThomasV wrote:
ThomasV a écrit :
To illustrate this, I have made a new proposal for the layout, where only the top 10 languages are displayed. I think this simplified layout is much easier to read.
sorry, I forgot the link :
Wow! I actually agree with Thomas on something. I do find this layout preferable, and I also agree that the layout question is distinct from how we determine the criteria for choosing the top ten.
Ec
wikisource-l@lists.wikimedia.org