Jump to content

User talk:Henrik

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Welcome! This is my user talk page. If you want to communicate with me for any reason, especially about pages I have made or edited, or any administrative actions I've performed, feel free to do so here.
I will generally reply here, unless you request otherwise. Please start a new section for new conversations.
I've been very busy in real life recently. I may suddenly go for long stretches without editing with no warning, however if something urgent comes up or the stats service is malfunctioning, I'm reachable through e-mail.

Bulk queries against stats.grok.se

I occasionally have people ask me how to make a list of the most frequently viewed articles in a topic area. Assuming that they can make a list of all the articles they are interested in, the next thing they have to do is to get hitcounts for all of them. Would it be possible to set up some sort of bulk query of the stats.grok.se data (say 100-500 articles at a time) to make this efficient? That would save these people from needing to download the raw data themselves. — Carl (CBM · talk) 17:01, 14 February 2009 (UTC)[reply]

The json interface should be reasonably efficient for this. The main overhead isn't the web request in any case, it's the database lookup: providing a web interface to do bulk queries wouldn't really help that much other than making it more convenient. Doing runs of 100-500 articles against the json interface occasionally isn't a problem. henrik•talk 08:41, 16 March 2009 (UTC)[reply]
I did not know there is a JSON interface, that is exactly what I am looking for. Is it documented somewhere? — Carl (CBM · talk) 21:02, 17 March 2009 (UTC)[reply]
Not really, but it's pretty simple: prepend json to the url, like so: http://stats.grok.se/json/en/200902/Main_Page henrik•talk 21:47, 17 March 2009 (UTC)[reply]
What if I want to get the data for 1000 pages at a time? That's really sort of the feature that I need. But I appreciate you may not have the bandwidth for that sort of thing. — Carl (CBM · talk) 12:18, 18 March 2009 (UTC)[reply]
Loop over the set of pages you want to query and make a thousand requests to the json interface. You don't really need to worry about bandwidth, the json requests are tiny. The main problem is probably performance: a lookup takes between 0.1 to a second or more, depending a bit (old months are archived to a more compact and faster format once they are complete). But as long as you are patient and only run a single query thread you probably won't melt my server :-)
There's no real good way around the performance unless someone wants to give me money to build a real datacenter to run the stats service: doing random lookups in large amounts of data will be limited by hard disk spindle speed. henrik•talk 13:36, 18 March 2009 (UTC)[reply]
How big is the database? Is it possible to load it into RAM? Binary search in RAM shouldn't be too slow? Cyberfishee (talk) 22:19, 30 March 2009 (UTC)[reply]
As of right now, it's 294 GB. henrik•talk 22:23, 30 March 2009 (UTC)[reply]
Ah... SSDs could possibly work (2-3?), since throughput is not as important as seek time. But that gets expensive. Cyberfishee (talk) 19:42, 31 March 2009 (UTC)[reply]

please change my user name : mvsrao as abdullah

please change my user name : mvsrao as abdullah

About the "wikipedia article traffic statistics" Average by day

If possible should have an average by day. Thanks. Sergio Kaminski (talk) —Preceding undated comment added 16:05, 23 April 2009 (UTC).[reply]


About the "wikipedia article traffic statistics"

Dear Henrik,

I found your website--http://stats.grok.se/zh/200902/%E9%9B%B7%E4%BA%BA--is very helpful for me to get the visit counts. But I found the statistics seemed to be inaccurate. So I am wondering whether the unit of the counts statistic is 100 or 1000? Do you have a better way for me to get a more accurate statistic?

Thanks in advance for your kind help


Grain @ China

Advice request

Sure. Feel free to have him contact me on my talk page. I've got a little experience with guys who come in and want to do good/help, but don't know the wiki ways... — BQZip01 — talk 17:39, 15 March 2009 (UTC)[reply]

I'll let you know. Thanks! — BQZip01 — talk 05:37, 13 April 2009 (UTC)[reply]

stats.grok.se

THANK you for creating this great tool. I wonder if it would be possible to add option to show most popular articles within a scope of a given Wikiproject and/or category? There is a tool for scanning categories to generate article lists here: CatScan, I think it wouldn't be that difficult to tie those tools so that CatScan would generate lists of articles, and stat tool would check their popularity and rank them. User:Mathbot also can generate list of articles in categories and subcategories. Alternatively, we could scan by using the Wikipedia 1.0 system (which looks at talk page tags and generates such stats) but I am not sure how exactly it is done. The reason I am asking for this feature is that WikiProject Poland would like to improve the most popular articles related to Poland and WWII before the upcoming anniversary of WWII (1st Sept), but while we can speculate what articles are the most important, we don't really know which ones are the most popular. Such a tool would be a boon to many a wikiproject, as well as to scientists studying Wikipedia (and I am also speaking as one of them). Once again, thank you for your great work! --Piotr Konieczny aka Prokonsul Piotrus| talk 20:04, 29 March 2009 (UTC)[reply]

Hi Piotrus. I agree: this would be great; unfortunately the database isn't structured in a good way, it would likely take a long time to do all the processing for a large category. I'll give it some more thought, it would be a great feature. henrik•talk 17:28, 1 April 2009 (UTC)[reply]

Re:User_talk:WilyD#FYI

Thanks, such as it is. WilyD 12:58, 30 March 2009 (UTC)[reply]

Is there possibility to register or to know how many times was specific page displayed ?

Hello. I would like to ask you, if there is possibility to register or to know how many times was specific page displayed ? Thank you. Softvision (talk) 18:14, 30 March 2009 (UTC)[reply]

Yes, go to http://stats.grok.se and enter the page title. henrik•talk 17:28, 1 April 2009 (UTC)[reply]
Well, you sure picked the right guy to ask. --Steve, Sm8900 (talk) 19:30, 1 April 2009 (UTC)[reply]
:-) henrik•talk 19:34, 1 April 2009 (UTC)[reply]

question...

Your tool that shows how often pages are read generates handsome tables. I'd like to generate some handsome tables too. Can I ask if you made the source code for your tool available?

If you were writing your tool, from scratch, would you use the same graphics package? If not, which would you recommend?

Thanks! Geo Swan (talk) 17:16, 31 March 2009 (UTC)[reply]

The tables are simply html and css, swiped and adapted from somewhere else which I can't remember exactly where from now. Just look at the html source and you'll see the CSS in all it's glory. henrik•talk 17:30, 1 April 2009 (UTC)[reply]

The Wikipedia SignpostWikipedia Signpost: 30 March 2009

Delievered by SoxBot II (talk) at 20:05, 31 March 2009 (UTC)[reply]

Stats

I got this line: Megan_Baker has been viewed 739 times in 200902. Does that mena: Megan_Baker has been viewed 739 times in 2002-2009? Debresser (talk) 13:04, 1 April 2009 (UTC)[reply]

Nope, '200902' means 'February 2009'. Emw2012 (talk) 15:40, 1 April 2009 (UTC)[reply]
Yes, it's even a standard. henrik•talk 17:26, 1 April 2009 (UTC)[reply]

That "yes" was to 2002-2009 or to 02/2009? Debresser (talk) 21:49, 1 April 2009 (UTC)[reply]

I understand now. Thank you. Debresser (talk) 23:33, 1 April 2009 (UTC)[reply]

KARACHI STOCK EXCHANGE

DEAR SIR,

I WANT TO ONLINE VIEW THE KSE 100 INDEX AND DAWNLOAD THE DATA AT THE END OF DAY. HOW CAN I DO .

BEST REGARDS IRFAN RIASAT ALI —Preceding unsigned comment added by IRFAN RIASAT ALI (talk • contribs) 17:50, 1 April 2009 (UTC)[reply]

Page view analysis

Hi Henrik,

I think you've done a great service to Wikipedia by adding the page stats within the history tab of each page. I use it regularly to check the hits on various pages. I'm wondering, would it be difficult to add basic statistical analysis measures, such as Mean, Median, Mode, and Range to the page? I know that I, along with many others, would appreciate the small addition. Either way, great job.

Junko —Preceding unsigned comment added by Junko (talk • contribs) 23:06, 1 April 2009 (UTC)[reply]

April figures

Hi Henrik, sorry to bother you. Do you know when the April stats will start to become available? Cheers.-- Myosotis Scorpioides 11:58, 2 April 2009 (UTC)[reply]

Answered this user's question at WT:DYK. rʨanaɢ talk/contribs 13:32, 2 April 2009 (UTC)[reply]

Article traffic statistics monthly and yearly

Hi Henrik: Is there a way to see the statics by month and year? Thanks --Arcibel (talk) 16:35, 3 April 2009 (UTC)[reply]

By month see http://stats.grok.se/en/2008/Main_Page [in year /2008/] (July is missing: outage). By year: not yet, as there is only one full year (2008), so nothing proper to compare with ;-) --- Best regards, Melancholie (talk) 09:18, 4 April 2009 (UTC)[reply]
Thanks! --Arcibel (talk) 20:14, 5 April 2009 (UTC)[reply]

Yearly stats

Have you thought of adding an option to get the stats for the entire year with monthly bar graphs since you now have a full year of info for 2008.--TonyTheTiger (t/c/bio/WP:CHICAGO/WP:LOTM) 06:31, 4 April 2009 (UTC)[reply]

See http://stats.grok.se/en/2008/Main_Page [/2008/] (July is missing: outage). --- Best regards, Melancholie (talk) 09:18, 4 April 2009 (UTC)[reply]

Adminreview lead

Hi, thanks for your edits; there are a few good things there, but it's now longer than I think is desirable. Did you think admins' role is not expressed in terms that are not positive enough? I'm open to suggestions, but I think it needs to be trimmed a little. What do you think generally about the process (which still has a few issues to settle—one of them, the role of third-parties). In particular, do you approve of the tight word limits for submissions? I want to avoid the dog's mess that the current ArbCom hearings process is. Tony (talk) 09:08, 6 April 2009 (UTC)[reply]

The Wikipedia SignpostWikipedia Signpost: 6 April 2009

Delievered by SoxBot II (talk) at 19:07, 6 April 2009 (UTC)[reply]

Stats

Will the stats be lost for the days that they are unavailable? Until Wednesday, April 8th? http://stats.grok.se/ Daniel Christensen (talk) 18:08, 7 April 2009 (UTC)[reply]

No, the server should sync up those days that are unprocessed once it is up and running again. henrik•talk 20:11, 7 April 2009 (UTC)[reply]
Oh man, they're still not working. I just created a bunch or articles in this past week, too, that I'd like to check on... Daniel Christensen (talk) 13:23, 9 April 2009 (UTC)[reply]
It took a bit longer than expected, but the server is actually up now. Now we're just waiting for the DNS changes to propagate through the net so it can be accessed. henrik•talk 13:29, 9 April 2009 (UTC)[reply]

Stats have already made a lasting impact

09-April-2009: (subtopic) While the stat-server is being renamed, this is a good time to recognize the impact that it has already made:

  • I have seen new articles, whose titles had zero prior interest, become viewed hundreds per month.
  • The stats have proven an article-title doesn't need to be requested or spammed in navboxes, but just by existing, an article title can get hundreds of hits per month.
  • Articles about U.S. subjects translated to German, can get 25%-50% of the English-WP hit level, even though they don't concern Germanic culture (and vice versa: German articles can get 50% of German-level hits when in English).
  • Some article-titles, requested hundreds of times per day (for 2 years), only got created due to the known stat-counts.
  • Even if an article is requested hundreds of times per day, "no one" else will create it, even after a full year of requests.
(Few people realize they want the same article as others, so no one creates it.)
  • An article with a grandstanding vanity-box "This article has hideous problems, please help" might go more than 28,000 pageviews with no improvements.
  • A hot-topic article-title, requested by thousands of hits in one week, might become totally forgotten the next week, with hits less than 8 per month, afterward.
  • Rare articles, with some marginal interest, typically log more than 50 hits per month.
  • Articles about common words (such as: river, spoon, table or sky) typically receive 500-1500 hits per day.
  • Popular articles, viewed 1000 times per day, might not get edited often: editing or vandalizing is not directly tied to reader popularity, but rather the cultural impact of the article.
  • Due to the stat-counts, article-titles now exist that handle thousands of hits per day (which were previously missing).

Those are some of the many issues that the stat-counts have already revealed since December 2007. -Wikid77 (talk) 15:26, 9 April 2009 (UTC)[reply]

Very interesting. Will there ever be a listing of "most requested" pages that don't exist? A listing of most viewed pages with major "needing work" tags might also be nice. Anyway, sorry to drive by comment, but I think the stat results are neat. ChildofMidnight (talk) 16:04, 9 April 2009 (UTC)[reply]
@ChildofMidnight: For the first issue see Wikipedia:Most missed articles (to be updated very soon), the second is a nice idea, will see to include it there eventually. --- Best regards, Melancholie (talk) 21:05, 10 April 2009 (UTC)[reply]

Daedalus969

It seems as though the type of behaviour displayed to User:TPTanque, for which Daedalus969 was criticised recently at ANI, was repeated very recently here. User:Daedalus969 is edit warring on Skywalker's talk page. He is aggressively threatening Skywalker with a block and he is inappropriately accusing him of personal attack. What should be done? Nothing has been learnt. Paul Beardsell (talk) 18:31, 7 April 2009 (UTC)[reply]

This is a complete fabrication. I was not edit warring. Every time I left a note on SkyWalker's page, it was new, not once did I revert, or save a version close to a revert. SkyWalker was saving his reverts with edit summeries of user is total nuts. That is a personal attack.— Dædαlus Contribs 20:34, 7 April 2009 (UTC)[reply]

Offensive

How am I the one going on the offensive? I wasn't the one accusing and insulting people like User:SkyWalker did when he said I was crazy.— Dædαlus Contribs 20:27, 7 April 2009 (UTC)[reply]

Re: Nuts

First you need to understand why iam angry. That user keeps on adding csd template on newly created. Time must be given to establish notability. I tend to get angry on such people who does not understand other people. Seriously i have added many sources on the article below the external links to establish notability and yet that user added a A7 csd template. Adding that template is unwarranted and uncalled for. Iam not sure how many article he keeps on adding csd template and how people he provokes. How i can create a friendly environment when they are people who provokes other people unnecessary?. I do hope you will understand. --SkyWalker (talk) 02:22, 8 April 2009 (UTC)[reply]

I do understand; I'd probably be none too happy myself. But ChildofMidnight posted some good advice on your talk on a good way to respond, even when provoked. Also: These days articles really need to be able to stand on their own legs from the beginning, developing stubs in userspace and then moving them is a good plan - that's what I usually do. henrik•talk 05:25, 8 April 2009 (UTC)[reply]
I didn't add it multiple times. I reverted your removal of it because article creators are not allowed to remove CSD templates from articles they created themselves.— Dædαlus Contribs 02:28, 8 April 2009 (UTC)[reply]
Your rigid interpretation of policy is none too helpful. SkyWalker is a longstanding editor, why didn't you just post a question and explain your rationale on the article talk or his talk page instead? Wikipedia won't fall over if deleting an article takes a few minutes longer. henrik•talk 05:25, 8 April 2009 (UTC)[reply]
Because he insulted me in an edit summery.— Dædαlus Contribs 05:46, 8 April 2009 (UTC)[reply]
Responding to provocation by escalating the situation is strongly discouraged. In any case, looking further in your edit history, I must say I am troubled - I'll follow up with a longer comment on your talk page. henrik•talk 05:51, 8 April 2009 (UTC)[reply]
I will await to hear what you are referring to. As to further escalating, I do not see how what I did further escalated anything. He insulted me twice, in response, I left two warnings, one for the removal CSD template of a page he created himself, and another for personally attacking me. I didn't think of him as an experienced editor, as he attacked me twice because I nominated his article for deletion.— Dædαlus Contribs 05:56, 8 April 2009 (UTC)[reply]
Further, if a user insults me, I have been told to warn them, and not respond back with incivility. You seem to suggest that I do something different, what exactly would that be?— Dædαlus Contribs 05:58, 8 April 2009 (UTC)[reply]
As a note, I shall not make any further edits, as of this post, until you have replied to me. I do realize that I have made several bad mistakes recently. I realize that about 40 percent of my edits are in the user talk space.. that is mostly from reverting vandalism, or arguing with others, but it is also because I don't trust in my ability to write articles. They're just like the school essays.— Dædαlus Contribs 09:21, 8 April 2009 (UTC)[reply]

Re: Advice

Alright, I shall take note, and listen to what you have told me. Thank you.— Dædαlus Contribs 21:11, 8 April 2009 (UTC)[reply]

Raw numbers for statistics rather than a graph

Hi, how can I extract the raw numbers for a page for importing into a spreadsheet? I can see them on the graphs provided but I'm just wanting the raw numbers. Alex Sims (talk) 00:56, 10 April 2009 (UTC)[reply]

You can use the JSON output: http://stats.grok.se/json/en/200904/Main_Page --- Best regards, Melancholie (talk) 21:06, 10 April 2009 (UTC)[reply]
Thanks that's exactly what I'm after. How did you find this out? Is there some documentation somewhere? Alex Sims (talk) 01:06, 11 April 2009 (UTC)[reply]
No, not really - I've only mentioned it here some times and Melancholie is very helpful. I should update the site to mention the json output. :) henrik•talk 06:31, 11 April 2009 (UTC)[reply]
Regarding the json, why do some months start with 0? Like 200802 Britney Spears. Also, do you know if there is any pre-December 2007 page view data out there? Thanks! --Odie5533 (talk) 23:37, 11 April 2009 (UTC)[reply]
There are some days in which the stats collection has been broken (for various reasons) - those show up as zero views for the day. No, as far as I know there wasn't any systematic collection before Dec 2007. henrik•talk 05:19, 12 April 2009 (UTC)[reply]
It seems to happen a lot on the first day of the month, and there is also an extra day for that month. For the link I gave, there are 32 entries. For many of the months there are extra days and random zeros at the beginning or end. Perhaps a broken for loop somewhere? --Odie5533 (talk) 13:54, 12 April 2009 (UTC)[reply]
Ah, I see what you mean. Yes, it's a bit broken - the numbers in the json output are zero based, but I counted days starting at one. That resulted in the first number always being zero and there being 32 instead of 31 entries. That's now fixed. But to make it a little bit easier to handle automatically all months are still have same length of 31 days; any extra entries will be zero. henrik•talk 15:13, 12 April 2009 (UTC)[reply]
I think you've introduced a new bug in your fix. The graph has an extra entry at the beginning while the json is missing the first of the month's entry. This only happens on some months. For instance, Nov 08 is fine. Also, if you ever get bored, a Y-M-D\tCount listing for a full article would be much appreciated. Thanks again for the site. --Odie5533 (talk) 16:12, 12 April 2009 (UTC)[reply]

(unindent) Argh! I knew having slightly different database formats for different months come back to bite me. Grumbl. Thanks for pointing it out though! I'll work on a fix. henrik•talk 16:18, 12 April 2009 (UTC)[reply]

How do you get the stats from Wikipedia? Is the site run on the toolserver or something? Do you mind if I download lots of page view data archives from you? Also, what's your database look like, perhaps I or another database designer could help. --Odie5533 (talk) 17:19, 12 April 2009 (UTC)[reply]

Good work

I just have to say that stats.grok.se is still very much beta, but keep up the good work. South Bay (talk) 05:28, 10 April 2009 (UTC)[reply]

Thanks. :) henrik•talk 15:31, 12 April 2009 (UTC)[reply]

the cult of herodotus

henrik,

I am confounded as to why my article: The Cult of Herodotus, has been deleted.

At the bottom of the article, I clearly stated its relevance to Wikipedia and demonstrated why it meets the all the requirements of inclusion. If there is some point that you still find to be unclear, please inform me immediately and I will surely be able to rectify it immediately, otherwise there is no reason for it not to appear on Wikipedia.

Please write back to let me know when you have put it back up.

thank you —Preceding unsigned comment added by Jellyphant (talk • contribs) 11:02, 12 April 2009 (UTC)[reply]

henrik,

thank you for your response. it is very difficult for me to provide a credible source for this information as this is a "secret" cult. I am not, however, just making it up on a whim with my mates at a pub or whatever. This is indeed a very real happening that takes place in the woods near my home every year on the last full moon of autumn.

as it is both factual and spiritual, i really believe it has every right to be on wikipedia, not just as some sort of a joke. it is an annual event! also, there is a link to a video included that proves this does happen, with the masks and everything.

the cult of herodotus is fact, a mysterious fact, i grant you, but it is a fact.

if you still really, really have a problem with it and won't allow it back onto wikipedia, please humour me with a more detailed explanation as to why or as to a means that i might verify this by definition through a 'source'. i mean, how do people writing about apparitions quote sources? how is that credible? i'm not talking about a real supernatural experience here, just about an actual gathering of people whose intentions are to summon a spirit, not dissimilar to seances or invoking the holy spirit at an evangelical church.

thank you for not ignoring me.

jellyphant —Preceding unsigned comment added by Jellyphant (talk • contribs) 14:48, 12 April 2009 (UTC)[reply]

Ideas for stats tool

I was pointed to your tool after a query here. It is quite an achievement! It answers one question very well. I'd be interested in knowing a couple of other things: how many people (or unique computers) access the main page on an average day? Sometimes I go back several times in one session or day to the main page, so that will boost its apparent popularity. Secondly, is there a tool to show how many click-throughs an article gets? Let's say it was the featured article of the day, or one of the others mentioned on the main page. What effect will that have on the traffick? Thanks for any help you can offer! BrainyBabe (talk) 14:54, 12 April 2009 (UTC)[reply]

Ah, a question in two parts. :)
(i) No, there's no way to filter out non-unique requests from the data we have now. It's difficult both from performance and privacy reasons, plus it is not obvious it would increase the accuracy: A lot of people sit behind proxy servers which cause them to share a single IP, balancing out those who for whatever reason accesses the page more than once.
(ii) If you're just interested in the traffic stats for the featured article of the day, you'll have to do some manual work and find the article from the archives and click the stats link in the history tab. If you're interested in click-throughs ("what percentage of main page viewers read the featured article?") the easiest way to check would be to divide the main page numbers with the ones for the featured article for a given time period.
In general, the tool only presents the raw data: More sophisticated interpretation of that data is left to the people viewing it. :) henrik•talk 15:31, 12 April 2009 (UTC)[reply]
Re (i), fair point about privacy. It is a guess that the two factors balance each other out, but a reasonable guess.
Re (ii), I've had some interesting answers from the Helpdesk query (linked above), which provided links to previous experiments. BrainyBabe (talk) 23:09, 13 April 2009 (UTC)[reply]

Wikipedia Stats

Hello Henrik,

You have done awesome work with the stats visualization of page views of the various wiki articles. I did have one question though. Is it possible to further tag this traffic data by Geography?

Thanks, Vaibhav —Preceding unsigned comment added by Whoisvaibhav (talk • contribs) 08:08, 13 April 2009 (UTC)[reply]

I'm not quite sure how that would work? See the most viewed stuff within a radius? Interesting idea, that. henrik•talk 17:48, 13 April 2009 (UTC) [reply]

Page rank question for your stats.grok.se

Hi Henrik,

Many thanks for the excellent service you have provided at stats.grok.se.

I have a question about ranking, however. Two, in fact.

First, I assume the rank is independent of the actual data slice selected? This is because the rank stays the same for any data slice selected. So I assume the rank is averaged out over all available data?

Second, I wanted to compare a set of pages (covering religious topics). However, when I came to "Secular Humanism", the data slices were available (e.g. for 200903), but no ranking was shown.

While I could construct my own ranks -- simply by accessing your site for all 2008 data slices and then collating the results for the pages I am interested in -- I thought I would ask you first if there is any "cut-off" in your rankings -- say at 10,000 or something?

Hope you have a moment to reply soon!

Cheers,


Ed —Preceding unsigned comment added by Nrubdarb (talk • contribs) 11:14, 13 April 2009 (UTC)[reply]

Yes, only the first 10k articles are ranked. henrik•talk 17:46, 13 April 2009 (UTC)[reply]

The Wikipedia SignpostWikipedia Signpost: 13 April 2009

Delievered by SoxBot II (talk) at 16:21, 13 April 2009 (UTC)[reply]

Talk:Dano-Swedish War (1658–1660)

I posted a comment there. I see you nominated it for DYK, don't forget to add it to WP:MILCON.--Piotr Konieczny aka Prokonsul Piotrus| talk 00:06, 14 April 2009 (UTC)[reply]

Macedonia

FYI, the Macedonia issue on which you commented yesterday is now at arbitration - see WP:RFAR#Macedonia naming dispute. -- ChrisO (talk) 07:10, 17 April 2009 (UTC)[reply]

ITN for The Pirate Bay trial

Current events globe On 17 April, 2009, In the news was updated with a news item that involved the article The Pirate Bay trial, which you substantially updated. If you know of another interesting news item involving a recently created or updated article, then please suggest it on the In the news candidates page.

--BorgQueen (talk) 12:00, 17 April 2009 (UTC)[reply]

Article Traffic statistics 2

If the same user visits an article more then one time a day: is it counted as 1 or every time.--de:user:Hfst —Preceding undated comment added 20:20, 17 April 2009 (UTC).[reply]

Every time; it's a hit counter. You may test this with a random non-existent page. --- Kind regards, Melancholie (talk) 10:05, 19 April 2009 (UTC)[reply]

Statistics page

Nice feature. How about adding cumulative entries rather than just monthly? For total visits by year (with an accompanying graph to visualize traffic rates and patterns) and just plain total to date, all visitors all years. These would be most useful and user-friendly additions to your beta release. Thank you. Keep up the good work. Wikiuser100 (talk) 14:57, 19 April 2009 (UTC)[reply]

There's some preliminary support for viewing entire years, just remove the month from the URL and you'll see the entire years worth of stats. I'm missing some data in the right format, so some months are missing. The other is a good idea, I'll consider it when I have time to do implement it. It's a bit tricky, since it'll need to be fast enough. henrik•talk 15:42, 19 April 2009 (UTC)[reply]

Häddsupp

I made an attempt at a copyedit of the war article. I also included a few hidden comments concerning some unclear. If I did something stupid, just revert it. If you think any of my queries are unclear, just let me know and I'll clarify on the talkpage.

Peter Isotalo 18:14, 19 April 2009 (UTC)[reply]

Wow! Great! *bows to your superior copyediting* I probably went a bit overboard on the superlatives, good work on trimming them back. I'll work on trying to expand the unclear bits, the comments are very helpful. BTW, just one minor thing: Fulfillment is also an acceptable spelling, it's one of those UK vs US things: the two-'l' version is mainly UK spelling, while the three-'l' spelling is US English. Funnily enough, travelled vs. traveled is the opposite. I'll probably try to go with a consistent US English spelling in this case. henrik•talk 19:25, 19 April 2009 (UTC)[reply]

Stats defaults to previous month

When I check stats /latest/ defaults to " in 200903", which is last month. Does this usually display the last 30 days? Also, is it possible to use json to get a page's entire history, and its monthly stats, rather than just last month's stats? –MT 18:48, 19 April 2009 (UTC)[reply]

No, it currently displays the previous month: The /latest/ URL was meant to eventually display the last 30 days: I'm working on a new version of the site which will do just that. henrik•talk 19:27, 19 April 2009 (UTC)[reply]

The Wikipedia SignpostWikipedia Signpost: 20 April 2009

Delivered by SoxBot II (talk) at 18:34, 20 April 2009 (UTC)[reply]

unblock of User:Medisave

I don't understand why you did this. The user was posting a profile of Medisave UK on behalf of the company; their comments admit as much. They had standard poor arguments for why they should be allowed to list themselves. It is not the username only that is a problem; the problem is WP:COI. I'll leave it up to you to decide if this was a mistake or not; if not, I hope you'll keep a close eye on the situation. I don't see it ending well. Mangojuicetalk 20:24, 20 April 2009 (UTC)[reply]

I'm indeed keeping a close eye on the situation and had coordinated the unblock with blocking admin User:Juliancolton off-wiki before taking any actions. It is an attempt to test out a different approach to initial problematic (COI/spam/promotional) users. henrik•talk 20:30, 20 April 2009 (UTC)[reply]
Ah, ok, that makes sense. I hope you keep your test on a small scale, though. :) Mangojuicetalk 20:35, 20 April 2009 (UTC)[reply]
Nah, small scale is for small people. My next move will be an unblock of all Willy on Wheels socks *grin*.
Actually, me and Chillum are trying out the wording of a new template, feel free to chime in if you have any suggestions on improving it. henrik•talk 20:41, 20 April 2009 (UTC)[reply]
I think that's a fine template to use when a user really has done nothing wrong, as in a case where a user has a company username but hasn't edited yet. But in a lot of cases, these users are pushing for directory entries on their company... if we are to treat that gently, that's one thing, but I think it's not good to tell them they've done nothing wrong in such a case. Mangojuicetalk 23:41, 20 April 2009 (UTC)[reply]
Yes, that's a good point. It would probably be better to explain that we're not really being a business directory with a similarly friendly message rather than focus on the username. henrik•talk 05:09, 21 April 2009 (UTC)[reply]

Stats page

I've noticed some irregularities on the stats page...

My first is with [1] which shows 1 visit for a page that was never created(unless it was deleted and hidden).

Could you explain how this visit came to be=P.Smallman12q (talk) 21:54, 21 April 2009 (UTC)[reply]

Sure, all accesses are recored: not only the ones for existing pages. If you click this link, Iraq wae a hit will show up again in tomorrows stats. henrik•talk 08:21, 22 April 2009 (UTC)[reply]
Oh. I didn't realize that. Is this noted somewhere?Smallman12q (talk) 22:30, 22 April 2009 (UTC)[reply]

Statistics - Bots?

I tried to search if someone had already asked you this question, but I couldn't find anything like it, so sorry if everyone asks you the same question like this. Does your statistics site pay attention, ie makes difference, between human views and Wikipedia bots, or maybe search engine spiders? I took a look at some of pages, and I wander is it all "real" views, or I have to take into account that web spiders and wikipedia bots crawl over all Wikipedia pages all the time and make us think a page is viewed more than it really is. Thank you in advance for your response. --Дарко Максимовић (talk) 00:54, 22 April 2009 (UTC)[reply]

All accesses are recorded, there are no provisions that I know of for filtering out search engine crawlers (most wikipedia bots use the API which I don't believe is counted). I would expect that noise to be fairly insignificant on popular articles though: It seems barely noticeable on even very minor articles (Xiafs and Versorium are two very obscure articles, getting maybe 10 views a day). Even if all those views are web crawlers and bots trawling through wikipedia, it wouldn't alter the statistics noticably on a more popular page. henrik•talk 09:18, 23 April 2009 (UTC)[reply]

Thank you for your explanation. --Дарко Максимовић (talk) 15:06, 25 April 2009 (UTC)[reply]

Random...

I was talking with my friend, and he says that ninjas are more powerful than a Spartan Warrior. Who do you think would win in a fight? Sentry2 (talk) 13:52, 22 April 2009 (UTC)Sentry2 april 22, 09 9:51[reply]

User:Mrmerlot

If you want to mentor and help him, I'll unblock (after he's said who his sockpuppets are, anyway) but truth be told, you might want to think about how much of your volunteer time you'd want to give towards dealing with an editor with that kind of contrib history. Gwen Gale (talk) 14:59, 22 April 2009 (UTC)[reply]

It's part of a longer term project I have to engage with problem users and see what kind of approaches that work. Is trying to reform them successful? How should we best explain our policies and principles to new users? How can trolls and misguided users be told apart? It might be moot in this particular case though, last comment he gave was somewhat resigned. (I imagine the sock might simply be his IP address, it has some contributions over a decent amount of time) henrik•talk 15:04, 22 April 2009 (UTC)[reply]
Yes, this is more or less what I thought you might be thinking about. Let me know! Cheers, Gwen Gale (talk) 15:09, 22 April 2009 (UTC)[reply]
With a background in psychology and a deep interest in sociology, I have recently taken to the fascinating study of the development of online communities. Obviously, the Wikipedia community one such community. As much as I might be a "project," consider yourselves "projects" as well. I have extensive contacts within the "government 2.0" sphere and intend to widely circulate my observations to compare my experiences with others.
FYI, with the transparency afforded by Talk pages, it would behoove gardeners to refrain from posting comments intended to offend other users when they themselves are criticizing potentially offensive behavior of others.
User:Henrik, look for an updated post on my blog shortly... :) Waykup (talk) 21:21, 22 April 2009 (UTC)[reply]
In some aspects, we're all projects of others in the study human interactions, so please feel free to study me and others - I'd be very much interested in the results. I didn't mean any offense in somewhat circumstantially calling you a 'project', but my initial goal in engaging this was to gain a better understanding on how to improve the experience for you and people like you in general. As it turns out, you have plenty of good ideas and are interested in the same subject, so that was lucky. :) henrik•talk 09:54, 23 April 2009 (UTC)[reply]

Appreciation for your steadfastness

I want to commend you on your help in sorting out the admin ugliness that recently occurred. I'm surprised at the continued lack of civility, maturity, and humility. I suppose I should simply come to expect that responses like yours are now the minority on Wikipedia and that most gardeners would simply like to ban any more users from entering the space (we call this the "please let me in! OK, now that I'm in, lock the door!" effect). While I will continue to use and track the evolution of Wikipedia, I will keep my contributions to a minimum as I simply do not have the time to invest in learning all of the requirements necessary for participating without getting a talk page full of hate mail. Thanks again! Waykup (talk) 19:23, 22 April 2009 (UTC)[reply]

Thank you. I hope you'll amend your blog to include the fact that not all of us gardeners are bad :) Seriously, bad experiences can happen and do happen far too frequently - but not all is bad at Wikipedia. Most of my interactions with other users here are pleasant. A few bad experiences isn't a reason to stay way for ever, so I hope you'll reconsider and give us another chance in the future. henrik•talk 19:34, 22 April 2009 (UTC)[reply]
While it isn't a new trend, a number of users and admins alike are leaving their light of AGF at the grave and are instead patrolling Wikipedia with steel blades obsessively nominating articles for deletion and using sensationalism to counter logical arguments. Still, there are admins and users such as henrik who do remain polite and genuine despite the growing incivility, asinine behavior, and ignorance of others. Hopefully you will as Henrik said reconsider and give Wikipedia another chance.Smallman12q (talk) 22:41, 22 April 2009 (UTC)[reply]
Ah-ha! The "resistance" emerges! :) Yes, I'm beginning to realize there are still those on the good side of the Force. Perhaps my faith in Wikipedia can one day be restored... Waykup (talk) 23:55, 22 April 2009 (UTC)[reply]

DYK for Dano-Swedish War (1658–1660)

Updated DYK query On April 23, 2009, Did you know? was updated with a fact from the article Dano-Swedish War (1658–1660), which you created or substantially expanded. If you know of another interesting fact from a recently created article, then please suggest it on the Did you know? talk page.

Ruhrfisch ><>°° 13:16, 23 April 2009 (UTC)[reply]

Wikipedia article traffic statistics

I think that this is a fantastic service - I hope that is stays and will not disappear! Thank you!

--Yoavd (talk) 12:13, 25 April 2009 (UTC)[reply]

Hello, Henrik. You have new messages at MichaelQSchmidt's talk page.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

RfA4

Ok, I think I'm ready for an RfA now. Per this comment, you indicated you'd like to be a nominator/conominator. If you'd be so kind as to fill some words in on my RfA4 draft page, I'll post it on WP:RfA in a few days. If you need more time, just let me know. — BQZip01 — talk 20:14, 26 April 2009 (UTC)[reply]

Ah, nice! I might need a few days more, this week already has very little room for Wikipedia and I think you deserve a through nomination statement. After/during the weekend? henrik•talk 18:46, 27 April 2009 (UTC)[reply]
Sure. I'll hold off the RfA and will continue my work. Thanks for the support/co-nom. — BQZip01 — talk 22:16, 27 April 2009 (UTC)[reply]

Hi Henrik, the traffic-counter is superb. Couldn't one create another tool that shows links from outside: Webpages that link to a certain article - not from within wikipedia, nor from one any of the clones but from regular sites. It would be immensely interesting to see whether journalists or professionals discuss a certain Wikipedia article. I know it would need spiders to crawl the web for that purpose, but maybe Google could offer the information, and maybe the main problem is to delete all Wikipedia-produced references (i.e. those of other Wikipedias and those of clones). Maybe you are the wrong person to ask and you know who should know.

PS my own web provider offers a special analysis of traffic and where it comes from - that would be the ideal thing, to know how much traffic my article gets from where. --Olaf Simons (talk) 14:37, 28 April 2009 (UTC)[reply]

The Wikipedia SignpostWikipedia Signpost: 27 April 2009

Delivered by SoxBot II (talk) at 04:18, 29 April 2009 (UTC)[reply]

Adding masri ( egyptian ) to Wikipedia article traffic statistics

Hi Henrik , please add masri ( code ARZ ) to your wonderful article traffic statistics. Thank you. Samsam22 (talk) 05:04, 30 April 2009 (UTC)[reply]

Retrieved from "https://en.wikipedia.org/w/index.php?title=User_talk:Henrik&oldid=286993877"