User talk:LArron/Archive1

What is going on? World Blogs Clogs Elections The Bar (talk) (talk) (talk) (talk) (hic) Welcome to RationalWiki, LArron! Check out our guide for newcomers and our community standards! Tell us how you found RationalWiki here! If you are interested in contributing: Read what our articles are intended to be Check out the to-do list See our help page and Wikipedia's formatting cheatsheet If writing a new article, consider writing up a draft And most importantly: Pull up a goat and have some fun!

ħuman 03:30, 23 August 2008 (EDT)

Ignore that soulless automaton welcome above, LArron. I welcome you with a warm heart and wide open arms. Welcome to the lockpick, LArron! Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 05:49, 29 August 2008 (EDT)

Userpage

Your userpage (as it is right now) is quite impressive. You should turn it into an essay—it is certainly worthy of one. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 00:53, 31 August 2008 (EDT)

Yes, I agree. Moar!!! One minor quibble - jpg is a poor way to store images like those, gif is better. Smaller, loads faster, no weird artifacts at sharp edges. Keep on keeping on! ħuman 01:10, 31 August 2008 (EDT)

Good tip, poor habit to create jpgs - the new pics will be pngs (yes, there will be moar!!) --LArron 03:26, 31 August 2008 (EDT)

Excellent, thanks, I did my time being schooled over image formats back during web 1.0... and believe me, I got schooled! ħuman 03:38, 31 August 2008 (EDT)

"CP started with a peak, but lots lost interest shortly after"

Wasn't that the Great Purge of unwanted editors?

And why do I have fewer edits than TK? It's just not on... Totnesmartin 15:24, 31 August 2008 (EDT)

I don't know much about the history of CP, my experiences with it are now ca. three weeks old.

When I took the data, I got 1237 comments for TK - only 1210 for you. I hope that at least the order of magnitude for these numbers is correct --LArron 15:47, 31 August 2008 (EDT)

One issue is that the database is irrevocably "corrupted". If a page is deleted, so is the history of anyone's edits to it, to the best of my knowledge. So, say Aschlafly whitewashes his user talk page (deletes, then recreates copying in most of what was there), all those edits vanish. I'm not certain, but I'm pretty sure, that my edit count there is far lower than it used to be (I thought I was pushing 3k or so by the time of the purge, I'm around 1200 now) - simply the act of deleting my user & talk pages, which was done, would have vaped probably hundreds of my edits. ħuman 16:00, 31 August 2008 (EDT)

I don't know the wiki-software well, but I think that on the Special:Statistic page all edits ever made are counted (~480,000 at CP). As I can only find ~408,000 edits made by registered users (and all users over there have to register, haven't they), it seems to be a save bet that the missing 72,000 edits are mainly purged ones. --LArron 16:10, 31 August 2008 (EDT)

Oh, also, when they upgraded the MW software earlier this year, it "reset" the stats from the raw data. So deleted edits from before then are truly "lost" in every way. ħuman 16:24, 31 August 2008 (EDT)

Did anyone take a screenshot before it happened? Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 01:44, 1 September 2008 (EDT)

No one thought to, and it would have had to have been a data dump. CP was screwed up for weeks, hard to load, impossible to edit. ħuman 02:10, 1 September 2008 (EDT)

I ♥ data

We love pie charts. Sterile^snore! 16:47, 3 September 2008 (EDT)

I just popped in to check out your userpage with all its incredible diagrams. Fascinating work LArron, congrats for your rich data-mining and graphing chops. DogP 12:11, 29 March 2009 (EDT)

My sincerest apologies

My condolences, you poor soul, but it seems I have made you a sysop. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 04:14, 25 September 2008 (EDT)

You'll need some special equipment. Let's see what we got here for you... yes: here's your rag, spray cleaner, and the key to the dustbin. Enjoy! Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 04:14, 25 September 2008 (EDT)

Oh, and before I forget, here's the manual. Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 04:14, 25 September 2008 (EDT)

Perfect: I just saw Dr. Horrible's Sing-Along Blog --LArron 06:29, 25 September 2008 (EDT)

Sorry

I already did that. Plus, my version has chronology :P Pink_{(Astronomy Domine)} 14:02, 25 September 2008 (EDT)

Great! Where can I find your superior version? --LArron 14:08, 25 September 2008 (EDT)

I have it as a word document at the moment. Creating it drained me of all my lifeforce, so I'm intending to send it to my commissioner and let him sort out the wikiformatting stuff.

Did you see Raul's hidden comment? Beautiful, wasn't it? ^_^ Pink_{(Astronomy Domine)} 14:15, 25 September 2008 (EDT)

Wow, you did it by hand? Impressive! I followed the discussion via the diffs, so I laughed when I read Raul's sycophantic displayed comment (and the more honest feeling in it)... --LArron 14:22, 25 September 2008 (EDT)

Er... Yes, I suppose my method was quite... painstaking. But I can't see how else I might have done it. What did you do? Pink_{(Astronomy Domine)} 14:32, 25 September 2008 (EDT)

Um, what are you guys talking about? Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 14:36, 25 September 2008 (EDT)

Dear RA, about this - and more amusingly, about this --LArron 14:51, 25 September 2008 (EDT)

It wasn't as simple as I had thought at first: I wrote a perl-program, which expanded into something quite ugly... It wasn't absolutely trivial to decode the logical/chronological dependencies of the different diffs, but at least my understanding of wikis improved ;-) --LArron 14:51, 25 September 2008 (EDT)

Ooh, nice. Good work, LArron! Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 15:02, 25 September 2008 (EDT)

Update?

Would you mind terribly if I asked you to update the stats for Active users? Radioactive afikomen ^{Please ignore all my awful pre-2014 comments.} 03:08, 2 October 2008 (EDT)

Active users

Hadn't noticed you were working on it, sorry. 18:16, 24 October 2008 (EDT)

There appears to be something wrong with the verdict column. There are a lot of people on Ubermensch but very fewa actually meet the ccriteria.  Lily ^{Ta, wack!} 17:44, 1 November 2008 (EDT)

I always need a couple of tries :-) --LArron 17:59, 1 November 2008 (EDT)

Obvious mistake

That's bizarre. I can't think of how that got there, since I wasn't in an edit conflict, and I was editing a totally different section. Thanks for the fix. --Kels 17:27, 15 December 2008 (EST)

And I marveled quite a while about some deeper, hidden meaning, unless I found the same sentence a few lines further down :-) --LArron 17:33, 15 December 2008 (EST)

Point me in the right direction....?

Any chance you could point me in the direction of some sort of resource explaining how to extract stats from a wiki? I've been impressed with some of the stuff you've created, and I wouldn't mind having something of a fiddle using some base data that I get hold of myself - only my Google Fu must be lacking becuase I can't seem to get my sticky mitts on any kind of documentation ... :( Worm ^{(t | c)}

There is no google in it - all perl and R --LArron 17:03, 5 January 2009 (EST)

And lo, I did investigate, and find that installing MediaWiki::API on Gutsy Gibbon is a pain in the rear. Still, it gives me a project to fiddle with :) Worm ^{(t | c)} 09:07, 6 January 2009 (EST)

Now got MediaWiki::API sorted out (having upgraded to Hardy Heron on the advice of Frohlich) and it seems to be working, now just got to get my head round the API....such fun :) Not wanting to tread on toes, I just find this sort of thing to be fun (Sad, but true). I had a look at R, and that may really take some time to get my head around, so Excel it is I'm afraid... :) Worm ^{(t | c)} 09:28, 20 January 2009 (EST)

I do care, in fact, I'm quite interested in your proceedings! Using MediaWiki::API seems to be a much more elegant way to gather the data than using my perl scripts! I've never looked into it before - but it sounds promising --LArron 09:52, 20 January 2009 (EST)

More elegant perhaps, but probably due to my ineptness, I have ended up following my usual cycle when it comes to coding a new project..

• Get excited and find out it is theoretically possible
• Download and install a few things
• Find a couple of examples and try them out
• Adapt things to my own purposes
• Scream and shout when things go all Pete Tong

At the minute I am banging my head against a module that appears to want to return 9 users in an array that calls for 10 users (but only on some wikis). grrr. Having to loop all the time because you can't get more than 500 items is also a pain (although understandable).... Maybe I'll need to roll my own modules, but I sometimes think it's a bit pointless. Worm ^{(t | c)} 16:45, 20 January 2009 (EST)

Managed to pull something together, although I'm not quite happy with it - simply based on the number of blocks for the top three blockers...er..if that makes sense. Still, somewhere to start. Need to find some other stats that are deeply interesting to dig into. Worm ^{(t | c)} 09:22, 21 January 2009 (EST)

I like it very much - thanks for posting it. I'm very curious what stats you dig out! LArron 09:39, 21 January 2009 (EST)

Blocked vs. Unblocked at CP

Date	Blocked	Not Blocked	Sum	Blocked Quota
2008-09-18	11,476	11,643	23,119	49.6%
2008-09-25	11,613	11,693	23,306	49.8%
2008-09-30	11,681	11,729	23,410	49.9%
2008-10-11	12,017	11,882	23,899	50.3%
2008-10-16	12,327	11,935	24,262	50.8%
2008-11-01	12,563	12,103	24,666	50.9%
2008-12-01	13,338	12,544	25,882	51.5%
2009-01-04	14,085	12,729	26,814	52.5%
2009-01-12	14,255	12,854	27,109	52.6%

Just noticed...

You liked it when people appreciated your graphics. i ceratinly do, so here's a...

		Barnstar
		For your tireless contributions to RationalWiki, I award you this barnstar.

Yours truly^{Dear Sir} 06:43, 21 January 2009 (EST)

Good move... but shouldn't that be called a Goatstar? lol. Refugee^{talk page} 04:11, 24 February 2009 (EST)

?

What software do you graph with? Sterile^{walkie-talkie} 09:23, 11 February 2009 (EST)

I used R. --larron^{sicut fur in nocte} 09:37, 11 February 2009 (EST)

A long time ago (wiki-wise, at least), I tried to calculate a blocking half-life at CP. Plotting as many time from new-user to blocked, you get a rough exponential decay, from which you can get a half-life. Unfortunately, it's a data-mining nightmare with say Excel macros (my tool o' choice), but you might be better at than I.Sterile^{walkie-talkie} 10:40, 14 February 2009 (EST)

I haven't looked at this (yet :-) - but I'll do so, after I'll have updated Conservapedia:Active users --larron^{sicut fur in nocte} 18:13, 14 February 2009 (EST)

FYI: This was my original plot:

Sterile^{walkie-talkie} 12:27, 15 February 2009 (EST)

Interesting - but I don't understand it completely: Does it really takes only minutes to get blocked? larron^{sicut fur in nocte} 05:50, 19 February 2009 (EST)

Yep. I took the new user log and the block log, matched them up the best I could with an Excel macros, and took the difference in time: ie, the time from sign-up to block. This is basically a histogram of those times, with an exponential fit as the red line. Sterile^{walkie-talkie} 08:08, 19 February 2009 (EST)

absolutely fascinating, I just looked into the days between the creation of the account and the first block, I never thought about smaller time intervals :-) larron^{sicut fur in nocte} 08:47, 19 February 2009 (EST)

Ah, your innocence corrupted your view of the data! So, from what that graph looks like, the half-life of a (blocked) username on CP is roughly 6 minutes? How efficient! ħuman 01:54, 24 February 2009 (EST)

Hi LArron. I'm quite a fan of your work producing data from CP logs. I have some experience with data mining and I'd love to get involved if possible. I'd love to know how you extract the data from the CP logs? JoeDuffy 05:38, 19 February 2009 (EST)

For this, I don't use any fancy stuff, I'm afraid, just plain old perl larron^{sicut fur in nocte} 05:50, 19 February 2009 (EST)

Where do you run the script, and what sort of queries do you send the db? ħuman 01:54, 24 February 2009 (EST)

I run my scripts from my PC at home. In the early days, a used CGI requests via index.php (quasi looking up the special pages), nowadays (see above- thank you, Worm) via api: for instance, you can get a list of all users via the allusers quest, and for each of them who, you can look up the edits using usercontribs. As CP is very unfriendly against scripts - even when the follow their robots.txt, this takes quite a while :-) larron^{sicut fur in nocte} 03:37, 24 February 2009 (EST)

Thanks

Thanks for the tips. CPWebmaster 21:40, 23 February 2009 (EST)

Be careful CPWM--TK's gonna block you now for fraternizing with the enemy. TheoryOfPractice 21:42, 23 February 2009 (EST)

You're welcome - maybe they are helpful if there is ever a crises in a future project of yours. BTW, nice touch to separate your editing personality from your administrative one. larron^{sicut fur in nocte} 03:07, 24 February 2009 (EST)

Con tu permiso

I've stolen your signature idea. I hope you don't mind.-- Kriss Akabusi^{AAAWOOOGAAAR!!1} 05:48, 24 February 2009 (EST)

No, of course not, I'm flattered. But you should try this: it prevents to much clutter in your comments... larron^{sicut fur in nocte} 06:03, 24 February 2009 (EST)

I see your point - let's hope this works.-- Kriss Akabusi^{AAAWOOOGAAAR!!1} 06:51, 24 February 2009 (EST)

Nevermind

You're the go-to guy for server stat .png's of CP. I was hoping to get a picture of the bottomed out CP activity because of the busted server, just because I liked seeing the week long or so zero in January when Andy had sex with broke it. But it's up and running so it won't be nearly as impressive. The Foxhole Atheist 21:39, 26 February 2009 (EST)

You are right, it is fairly noticeable. Have a look at the edit history:

1. (diff) (hist) . . North Pole‎; 15:22 . . (-19) . . CGoodwin (Talk | contribs) (rephrased sentence to smooth it)
2. (diff) (hist) . . m Caleb‎; 15:17 . . (+4) . . WesleyS (Talk | contribs) (link - deadend page)
3. (diff) (hist) . . m Conductor‎; 13:47 . . (0) . . CGoodwin (Talk | contribs)
4. (diff) (hist) . . Conductor‎; 13:47 . . (+118) . . CGoodwin (Talk | contribs) (added responsibilities of conductor.)

So, the outage didn't last for more than 1 1/2 hours. When they forget to switch off the Night Edit Mode, they generally suffer from a greater lack of edits. larron^{sicut fur in nocte} 03:10, 27 February 2009 (EST)

Wait, that's wrong. There was a one week gap in editing. Did they manage to spork the lost edits back in to the database? ħuman 04:19, 27 February 2009 (EST)

The discussion above was just about yesterday's outage - which lasted ~90 minutes. The graphs account creation at CP and daily edits at CP both show that that the week gap is still existing. Any attempt to retrieve the data seemed to have failed. I presume that Andy's position will be a classic Beeblebrox: I do not remember any losses of memory.... larron^{sicut fur in nocte} 04:35, 27 February 2009 (EST)

Oh, ok. Yes, the endless 1 hour periods on talk wigo CP where people think it's "gone for good" because they can't load it for an hour. Nevermind ;) ħuman 04:45, 27 February 2009 (EST)

Parsing the Conservapedia API data

So, I was working on my little script to get the list of universities blocked at Conservapedia, and I hit a bit of a snag. Perl's XML::Parser bitches about the XML not being well formed, and it's right. The root cause is their server move, when JM's name got all mangled due to some charset issue. Now it's echoing malformed UTF-8 out in to the document, and screwing up my parsing. I've "fixed" this problem by substituting out all the "by" clauses in the document before parsing it, but that hardly seems satisfactory.

What are you using to parse the document, and have you encountered the same problem? --JeevesMkII ^{The gentleman's gentleman at the other site} 04:12, 14 March 2009 (EDT)

I'm using perl, too - and I had the same problems! I fall back on basic regex, but I didn't want to invest to much time into the thing... larron^{sicut fur in nocte} 05:17, 14 March 2009 (EDT)

Stats gathering and mediawiki

How do you gather the stats for RationalWiki:Active users? You use http://rationalwiki.com/editcount.php, right? Does the api in newer MediaWiki versions allow you to do that? --  Nx/talk  10:56, 29 March 2009 (EDT)

Honestly, I never got the API at RationalWiki to work for me properly, I don't get all the information I'd like to have. So, I my perl programs use the CGI, i.e., question index.php : I basically look up the list of the users and user-groups, and then their entries...

Nothing elegant here, I'm afraid.larron^{sicut fur in nocte} 11:30, 29 March 2009 (EDT)

Hm, I thought you used Trent's editcount.php. Anyway, would this help? --  Nx/talk  12:15, 29 March 2009 (EDT)

Indeed, that's very nice! OTOH, to make all my little pics, I need

• an enumeration of all users, their blockstatus and the user-groups they belong to
• a list of all edits, with their dates and flags (namespace, minor, etc.)

When I have this list - and can make my graphs - the table of the active users is just a by-product... --larron^{sicut fur in nocte} 14:16, 29 March 2009 (EDT)

Take a look at the api documentation for MW 1.14. action=query&list=allusers and action=query&list=usercontribs should give you what you want, ucprop and auprop allow you to select extra info. You can only query 500 (5000 if you're a sysop) at once, so you have to loop through it. If there's something else you'd like just tell me. --  Nx/talk  15:39, 29 March 2009 (EDT)

That would be great, especially, as because it allows auprop=blockinfo|groups|editcount|registration, while http://rationalwiki.com/wiki/api.php only has auprop=groups|editcount... larron^{sicut fur in nocte} 06:31, 30 March 2009 (EDT)

Emulating Ed Poor

LArron, I have an Ed Poor style project for you. Could you run up a histogram for article size on CP? Just a fiendly suggestion you understand, there is absolutely no compulsion involved. :) Генгис^{Our ignorance is God; what we know is science.} 12:21, 29 March 2009 (EDT)

You call that emulating Ed Poor? Where are the crap redlinks? ) Totnesmartin 11:18, 3 April 2009 (EDT)

I hate working assignments :-) But I'll see what I can do... larron^{sicut fur in nocte} 11:20, 3 April 2009 (EDT)

Daily edits?

How about a graph of total edits per day over the last quarter? DogP 10:33, 3 April 2009 (EDT)

Why not, I'll see into it larron^{sicut fur in nocte} 11:20, 3 April 2009 (EDT)

You are teh amazing. Wait, that doesn't really work. DogP 11:24, 3 April 2009 (EDT)

Something like this: ? larron^{sicut fur in nocte} 17:46, 4 April 2009 (EDT)

Fantastic, LArron. A clear trend here, on both the 7-day and 100-day moving average. He Who Must Be Ignored may well be interested. DogP 02:27, 5 April 2009 (EDT)

Bureacratship

Why has this user not been cratted? He has excellent contribs, been here a while, and, most importantly, has created tons of useful charts and graphs, and has in general been incredibly useful to RW. I, for one, move that this user be bureacratized immediatley. The Emperor_{Kneel before Zod!} 14:47, 5 April 2009 (EDT)

But I've never been involved in administrative proceedings - I didn't sysop anyone, and made three measly blocks... larron^{sicut fur in nocte} 14:51, 5 April 2009 (EDT)

You can't give a sysopship until you are a bureaucrat. The only differnce is that burecrats can change user rights and also rename users. The Emperor_{Kneel before Zod!} 14:59, 5 April 2009 (EDT)

See? :-) I'm not worthy ;-) larron^{sicut fur in nocte} 16:09, 5 April 2009 (EDT)

Four month later, I feel worthy ;-) I'd like to fool around with a bot & a 'cratship would come handy. larron^{sicut fur in nocte} 12:54, 5 August 2009 (UTC)

Congratulations! You are now a bureaucrat. Nx (talk) 13:14, 5 August 2009 (UTC)

Not before time :) Congrats. Worm ^{(t | c)} 13:46, 5 August 2009 (UTC)

I never noticed you weren't one. Enjoy the power that you will just never use! ^postate 13:55, 5 August 2009 (UTC)

Wow, that went fast - and smoothly! Thanks, altogether! larron^{sicut fur in nocte} 14:25, 5 August 2009 (UTC)

Suggestion

I'm starting to feel bad that we just keep asking you to do things for us all the time. But, you're clearly brilliant at it, so.... My latest thesis is that Andy can't be a good educator to his students, or a good family man, because he spends so much goddam time online responding to socks, trolls, parodists and wandals. So - would it be interesting to have a look at the average number of hours (minutes?) Andy spends attentively online at CP, over, say, the last year? DogP 11:25, 7 April 2009 (EDT)

I won't go there. Clearly, Aschlafly spends lot of his time at CP, but it's his job and his past-time. Here's a pic of his posting behaviour (and that of some other CPns): . The man gets his sleep - other than our restless RJJensen. larron^{sicut fur in nocte} 13:05, 7 April 2009 (EDT)

I hope you approve of the way I used it in the Andy article. Let me know if you'd prefer it used or described differently. DogP 16:47, 7 April 2009 (EDT)

The article is great, and of course, you can use the pic as you see fit. Just to put it into context, I made this one this one. larron^{sicut fur in nocte} 17:46, 7 April 2009 (EDT)

Thanks Mr.Stats. If editors were weapons, we would win instantly just by pointing Human at CP. DogP 17:50, 7 April 2009 (EDT)

Range Blocks

Number of Active Range Blocks at CP on April 30th, 2009

Range	/16	/17	/18	/19	/20	/21	/22	/23	/24	/25	/26	/27	/28	/29	/30	/31	/32
IPs in Range	65536	32768	16384	8192	4096	2048	1024	512	256	128	64	32	16	8	4	2	1
#Blocks	228	15	34	57	59	10	130	15	181	3	2	3	2	7	0	0	864

larron^{sicut fur in nocte} 06:28, 15 April 2009 (EDT)

Do I assume correctly that multiplying the "IPs in range" by "# of blocks" you get the total number of devices blocked? ArmondikoV^bomination 13:00, 15 April 2009 (EDT)

Theoretically, yes. In fact, it's just an upper bound, as some of the blocks are nested or overlapping. For this pic, I'm taking this effect into account... larron^{sicut fur in nocte} 13:06, 15 April 2009 (EDT)

(The difference isn't that big: 14,428,755 vs 14,096,166 IPs, or 2,4%) larron^{sicut fur in nocte} 13:17, 15 April 2009 (EDT)

Traffic at aSK

Date	Mar 21	Mar 22	Mar 23	Mar 24	Mar 25	Mar 26	Mar 27	Mar 28	Mar 29	Mar 30	Mar 31	Apr 1	Apr 2	Apr 3	Apr 4	Apr 5	Apr 6	Apr 7	Apr 8	Apr 9	Apr 10	Apr 11	Apr 12	Apr 13	Apr 14	Apr 15	Apr 16	Apr 17	Apr 18	Apr 19	Apr 20	Apr 21	Apr 22	Apr 23	Apr 24	Apr 25	Apr 26	Apr 27	Apr 28	Apr 29	Apr 30	May 1	May 2	May 3	May 4	May 5	May 6	May 7
Accounts Created	4	1	2	0	79	35	18	5	8	6	3	2	3	2	4	0	1	2	0	0	3	2	5	0	1	0	0	0	1	2	0	1	0	1	1	3	1	2	1	2	1	1	1	1	0	0	1	1
Posts per Day	65	92	69	67	1403	1451	1022	831	868	639	589	464	380	366	240	241	221	253	224	208	187	190	494	195	179	220	102	183	341	166	252	167	129	103	114	121	119	90	76	116	81	65	77	175	133	168	100	84
Individual Editors per Day	4	5	5	5	76	86	65	46	52	52	51	46	40	39	33	20	30	30	26	22	16	19	23	19	26	25	22	17	20	24	32	26	18	24	19	26	21	21	13	14	15	14	14	17	19	26	15	23

|}

larron^{sicut fur in nocte} 12:41, 15 April 2009 (EDT)

I.e., it's died a quick and painless death? ArmondikoV^bomination 13:01, 15 April 2009 (EDT)

Oh no, I don't think so - it can drag on forever with this rate (~200 edits per day by ~20 individuals) ... larron^{sicut fur in nocte} 13:12, 15 April 2009 (EDT)

Not when you realize an enormous quantity of those are Wesley categorizing and FernoKlump wikilinking. 13:23, 15 April 2009 (EDT)

Present

A little present for you, you can choose between the gallery (fixed width) and table (expands), I think the gallery version is better --  Nx/talk  12:52, 17 April 2009 (EDT)

Shiny... but so. much. data. can't. take. it. in... ArmondikoV^bomination 13:04, 17 April 2009 (EDT)

Little present? That's great - and downright elegant. Thank you so much! larron^{sicut fur in nocte} 16:49, 17 April 2009 (EDT)

You're welcome :) --  Nx/talk  17:08, 17 April 2009 (EDT)

Suggestions/Requests

Could you do some graphs of some of the one trick pony obsessors?

I'd like to see visual representation of the growth of Daniel1212's homosex articles or the frequency with which FOIA edits Alger Hiss or number of times Ken

edits a page without adding anything.

I don't know if anyone else would be interested, I just thought it would be kinda fun to see. Neveruse513 17:21, 30 April 2009 (UTC)

I'll see what I can do. May take a while, though --larron^{sicut fur in nocte} 17:27, 30 April 2009 (UTC)

Hey, whatever. Thanks for even considering it. Neveruse513 17:33, 30 April 2009 (UTC)

Editors and Edits at CP

equidistant scale	logarithmic scale
logartihmic scale, sheared	protected articles

It's amusing to look how often an article was edited - and by how many editors. We all know Ken

's obsession, so, at one of his pet-articles, we'd expect few contributors, but many edits. In fact, even in the first pics - where most articles are muddled together - a few of his hobby horses are identifiable: cp:Homosecuality,cp:Evolution, and cp:Atheism. Other identifiable articles are cp:Talk:Mainpage, cp:User Talk:Aschlafly and - of course - cp:Barack Hussein Obama.

The blue line indicates the average number of edits per editor, i.e., ~1.98....

Going to a doubly-logarithmic scale, there are clearer tendencies even for articles with few editors to be seen: on user-pages, mainly the respective user is posting - and, perhaps, some vandals (Today, user-pages of new editors will be often be pestered by TK's de-redding.)

On talk pages, you find the back-and-forth of vivid discussion, a couple of editors contributing many edits each, while most of the debate seems to happen on user-talk-pages.

Shearing the image just helps to use the available space completely, and the articles can be seen better.

I added the names of the protected articles to the data points, though, these are illegible for articles with few editors and a small ration of edits per editor.

larron^{sicut fur in nocte} 14:24, 3 May 2009 (UTC)

Protected Articles at Conservapedia

table of the protected articles at CP larron^{sicut fur in nocte} 20:26, 3 May 2009 (UTC)

Gentleman! Another project!

Some day, maybe some day, you might scan this discussion Human and I were having recently. The issue being, that as the population round here has grown so too have the votes for Best of CP. So classic bonkers stuff from two years ago or whatever barely registers and recent asinine WIGO's are pretenders to the crown of Best Of's. We were wondering if there was a way to scale the vote tallies by the RW population of active editors over time, in an attempt to name the Top Top Best of Conservapedia's of all time, adjusted. There you go, that should keep you busy for about, oh, ten minutes. DogP 22:43, 3 May 2009 (UTC)

Or scale them by a factor relating to "how high" the highest contemporary wigos are tending to go? ħuman 23:44, 3 May 2009 (UTC)

I'll have a look into it... larron^{sicut fur in nocte} 08:26, 4 May 2009 (UTC)

There is of course no definite answer to your question. I made a little table to toy with (here). The effects of different ways of weighting the wigos can be seen in the pic. larron^{sicut fur in nocte} 13:21, 7 May 2009 (UTC)

Awesome. I like the 4th one (weighted by maximum of previous wigos), it "seems" about right - the Lenski and server crash events look like the highest scorers. Weighting by number of editors probably doesn't help, we don't know how voters are lurkers vs. members. One interesting thing that occurs to me is that there might be a snowball effect when WIGO is very interesting - when there's a lot of excitement, some people might tend to vote more on other less interesting items, because they are paying more attention. When the big excitement goes away, they drift off until the next big thing, whatever that may be. ħuman 22:31, 7 May 2009 (UTC)

mmm

You defended yourself somehow. This was unexpected. User:Mei

WIGO

Thanks - I couldn't figure out what was wrong ^{This message brought to you by: respond}_and honey 15:05, 28 May 2009 (UTC)

cached dpl

You can cache the image gallery dpl on your userpage to make it load a bit faster (most of the time), but it won't be updated automatically when you upload an image or make a link to one, so you'll have to manually purge it (or wait for the cache expiry, 24 hours). You can see the improved speed here. Copy this code if you'd like to use it:

{{#dpl:
|createdby = LArron
|namespace = File
|ordermethod=lastedit
|order=descending
|addpagecounter = true
|format=«gallery perrow="4" widths="160"»,\n%PAGE%¦%COUNT% views «br/» ²{Imageused¦%PAGE%¦4=true}², \n,«/gallery»
|allowcachedresults = true
}}

--  Nx/talk  12:31, 10 June 2009 (UTC)

Great! Thanks, I'll try it... larron^{sicut fur in nocte} 15:13, 10 June 2009 (UTC)

And thanks for the fix - so much to learn :-) larron^{sicut fur in nocte} 17:30, 10 June 2009 (UTC)

No problem. --  Nx/talk  17:32, 10 June 2009 (UTC)

Ah, are you still adding all your images to your user page? I foresaw that getting "clumsy" someday, like now... Why not put a bunch each in sub-pages and just link to the sub-pages? You could do it by category, or by date period, or any other scheme that makes sense to you? By date might even be automatable... ħuman 21:07, 10 June 2009 (UTC)

It is automatic already. The dpl code above does it. --  Nx/talk  21:19, 10 June 2009 (UTC)

Yeah, but by loading them all on his user page (which is what it does, right?) it takes forever to load. Breaking them up into chunks makes it easier to admire his handiwork (IMHO anyway). Just a suggestion. ħuman 23:50, 10 June 2009 (UTC)

Ah, I think I see now. I think ;) ħuman 23:17, 13 June 2009 (UTC)

Frankly, I don't expect that many visitors of my userpage - it's more a handy tool to find a pic if I want to make an up-date. larron^{sicut fur in nocte} 13:17, 16 June 2009 (UTC)

Ken edits

Arron, would you please do an analysis of Ken's recent edits per hour over the last month or so? User:Nutty Roux/sigtalk 04:49, 14 June 2009 (UTC)

I'm afraid that is infeasible: due to Ken's moronic way of editing (hundreds of tries to get a comment right) and his constant deletion of articles (which erases the history of the article, and therefore the lengthy list of Ken's edits to it), any analysis of Ken's data suffers from a severe undercount. larron^{sicut fur in nocte} 09:11, 14 June 2009 (UTC)

Ahh, you're right. The edit history, even from refreshing recent changes, disappears when he frequently deletes. You are smart. User:Nutty Roux/sigtalk 15:46, 14 June 2009 (UTC)

Captain Obvious, at your service. larron^{sicut fur in nocte} 15:52, 14 June 2009 (UTC)

Block Map

Fear Leads to Anger

You seem to know a lot about this intarwebz thing there... is there a way to translate IP addresses to geographic coordinates, automatically and in bulk? I thought it'd be cool if we had an actual map with a pictogram of a nuke over every spot TK has blown out of his queer little parallel universe. A small nuke for every /24, a medium nuke for every /18 or so, and a big nuke for every /16, or something. They could fade out over time, new nukes being superimposed over them as The Fist of Andy continues to roam his world, visiting death and destruction upon wherever he should chance to stride. Mountain Blue 06:51, 14 June 2009 (UTC)

I know of no easy way to get link the IP with its geographic position - or at least the country it was used in. I would love to have an elegant way to - for instance - enumerate the companies a certain \16 range is assigned to, but all I have are whois inquiries.

The map of the web thing is nice as xkcd did the heavy lifting, i.e., linked IPs and companies/places. In my earlier attempts, I just differentiated by the various RIRs.

larron^{sicut fur in nocte} 09:44, 14 June 2009 (UTC)

Aye. Thanks anyway. Mountain Blue 09:52, 14 June 2009 (UTC)

I have a database with the IP address assignments mapped to countries. I suspect more specific mappings may just use whois directly, which is somewhat tedious. --JeevesMkII ^{The gentleman's gentleman at the other site} 16:00, 14 June 2009 (UTC)

I'd like to have such a thing - did you make it yourself? How? Or where did you get it from? larron^{sicut fur in nocte} 16:03, 14 June 2009 (UTC)

Ah, even better. A moment with google has uncovered this, which is free and kinda sorta accurate. I suspect you'll get lots of "London" results for UK IP addresses, for example. The paid version seems more accurate still. However, it has lat-long coordinates, so you could make a pretty map if you so desired. --JeevesMkII ^{The gentleman's gentleman at the other site} 16:12, 14 June 2009 (UTC)

That seems to be great - seems to be exactly what Mountain Blue wanted :-) Thanks larron^{sicut fur in nocte} 16:21, 14 June 2009 (UTC)

You people totally fucking rule. Now I just need to find the time... Mountain Blue 16:31, 14 June 2009 (UTC)

Unfortunately, the free version seems not to support the latitudes/longitudes - but one gets the cities - not bad! larron^{sicut fur in nocte} 22:22, 14 June 2009 (UTC)

Close enough. Mapping city names onto long/lat is reasonably easy, even for me. Mountain Blue 23:10, 14 June 2009 (UTC)

Anger Leads to Hate

Heck, they even provide a list of the cities in their database, including the coordinates. But when I use the perl-module, $record->longitude and $record->latitude are plain nonsense... larron^{sicut fur in nocte} 06:32, 15 June 2009 (UTC)

The raw numbers I get following your link seem nothing less than completely reasonable at first glance. Maybe you should just regexp them out of there. Mountain Blue 06:45, 15 June 2009 (UTC)

The numbers in the list make sense, but when I use the api, i.e.,

use Geo::IP;
my $gi = Geo::IP->new(GEOIP_STANDARD);
my $gi = Geo::IP->open("/usr/local/share/GeoIP/GeoIPCity.dat", GEOIP_STANDARD);
my $record = $gi->record_by_name("24.24.24.24");

only the fields

       $record->country_code,
       $record->country_code3,
       $record->country_name,
       $record->region,
       $record->region_name,
       $record->city,
       $record->postal_code

are filled with sensible information, while

       $record->latitude,
       $record->longitude

give obviously wrong values. larron^{sicut fur in nocte} 07:10, 15 June 2009 (UTC)

Don't have access to Perl at work, but can fiddle when I get home. FWIW I got the php module running without much hassle.

       <?php
       
       include("geoipcity.inc");
       include("geoipregionvars.php");
       
       $gi = geoip_open("GeoLiteCity.dat",GEOIP_STANDARD);
       
       $record = geoip_record_by_addr($gi,"24.24.24.24");
       print $record->country_code . " " . $record->country_code3 . " " . $record->country_name . "\n";
       print $record->region . " " . $GEOIP_REGION_NAME[$record->country_code][$record->region] . "\n";
       print $record->city . "\n";
       print $record->postal_code . "\n";
       print $record->latitude . "\n";
       print $record->longitude . "\n";
       print $record->dma_code . "\n";
       print $record->area_code . "\n";
       
       geoip_close($gi);
       
       ?>

gives

       US USA United States NY New York Jamaica 11434 40.6763 -73.7752 501 718 

Worm ^{(t | c)} 09:08, 15 June 2009 (UTC)

Strange, the perl module yields:

       US USA United States NY Jamaica -1.#QNB 3.0678

But - with the list provided by MaxMind - I got my data :-) larron^{sicut fur in nocte} 21:01, 15 June 2009 (UTC)

Hate Leads to Suffering

Just a first try:

v0.0

larron^{sicut fur in nocte} 13:22, 15 June 2009 (UTC)

Wicked cool. I'm one of the red dots in New Zealand and one of the yellow dots in Australia. Mountain Blue 17:31, 15 June 2009 (UTC)

The map is taken from the map library of R. It's quite out of date: though it doesn't include the GDR anymore, it lacks Estonia etc. But it's good enough for an overview larron^{sicut fur in nocte} 20:34, 15 June 2009 (UTC)

Now, we have to look for the little nukes you proposed earlier :-) BTW, for these pics, I used perl and R. larron^{sicut fur in nocte} 21:03, 15 June 2009 (UTC)

Just a byproduct:

cp
wp

Though there are quite a few single IPs blocked at wikipedia, conservapedia is the real master of blocking ranges: there is no state without a rangeblock larron^{sicut fur in nocte} 21:54, 15 June 2009 (UTC)

Very interesting. /32 is a single IP, right? Since there is no red dot in NH, did TK decide to take out a few counties of comcast service to make sure I can't re-join CP using my real name? ħuman 00:12, 16 June 2009 (UTC)

Can it be done that we have a separate picture on Just CP blocks within the New Jersey (WP Not needed in this case for obvious reasons)? [[User:K61824|]][[User_talk:K61824|]] 03:39, 16 June 2009 (UTC)

A caveat: MaxMind state at their web-site: Accuracy Over 99.5% on a country level and 79% on a city level for the US within a 25 mile radius.

So, some blocks will not be displayed correctly, though 4/5th are (at least in the US)...

larron^{sicut fur in nocte} 05:56, 16 June 2009 (UTC)

I don't think anyone cares. The accuracy is perfectly adequate as is it. This is about total number and mass, not about individual data points. I liked the first version of your world map, the one where you can see the blockfest all but outline entire continents. Mountain Blue 06:41, 16 June 2009 (UTC)

wp 0.0

Then, you may enjoy this pic of the blocks at WP, too --larron^{sicut fur in nocte} 08:43, 16 June 2009 (UTC)

Thanks for the New Jersey map(was it flipped?). Apparently where Assfly lives isn't getting enough rangeblocks. [[User:K61824|]][[User_talk:K61824|]] 16:58, 16 June 2009 (UTC)

I'm afraid, I accidentally flipped it. Sorry. larron^{sicut fur in nocte} 17:50, 16 June 2009 (UTC)

Article using maps

Mind throwing the maps into an CP namespace article with a brief explanation of what we are seeing? There was some outside interest in your work comparing the block rates and I want to go back to promote this work you have done here as well. tmtoulouse 18:35, 16 June 2009 (UTC)

I've pondered about an article a pictured guide to conservapedia... larron^{sicut fur in nocte} 18:40, 16 June 2009 (UTC)

I am down for whatever you would like to put up, one of the wikipedia "newletter" things picked up the analysis you did on block rates between CP and WP and that in turn got picked up by a few blogs. These maps are excellent work, and not just for the CP/WP comparison, even just the WP data mapped out like that is interesting. I really think it could get a lot of "play" and would like to mess around with some Ken style promotion. Just need an article to push! tmtoulouse 18:46, 16 June 2009 (UTC)

That's great! I didn't know the WP's article was picked up - I hope it garnered some interest for us at RW. ^{and off he goes to write the article}larron^{sicut fur in nocte} 18:52, 16 June 2009 (UTC)

I second that emotion. These maps are impressive. That last iteration you just did, the one with the big dots for the big blocks, demonstrates the insanity at work here like nothing done before. You definitely need to find some place with a stable URL for them. Mountain Blue 19:00, 16 June 2009 (UTC)

"A pictorial guide to wiki block ranges"? Could lead with cool stuff about WP, then add comparos with CP and RW? ħuman 19:32, 16 June 2009 (UTC)

Conservapedia:An illustrated guide: have a look! larron^{sicut fur in nocte} 20:26, 16 June 2009 (UTC)

You, sir, deserve a medal. Mountain Blue 21:43, 16 June 2009 (UTC)

I deredded my title suggestion, now off to check out the new article! ħuman 23:43, 16 June 2009 (UTC)

dered < shudder > :-) I'm afraid that the language of the article is a little bit arcane, me being Jonny Foreigner ... larron^{sicut fur in nocte} 23:53, 16 June 2009 (UTC)

I fixed a few typos and a word or two. Some headers might be nice to break it up a bit? ħuman 00:02, 17 June 2009 (UTC)

(Edit conflict.) There's some things there that are unorthodox, yes. Not wrong, just idiomatically eccentric. I could suggest fixes if you'd like me to but I'd have to get some sleep first. Mountain Blue 00:04, 17 June 2009 (UTC)

I tend to idiosyncratic formulations - thought I try to avoid with regards to. Please, remove the crass unorthodoxies. And feel free to insert headers. larron^{sicut fur in nocte} 00:24, 17 June 2009 (UTC)

(undent) I may have gotten somewhat carried away there... please feel free to revert, rework, or otherwise undo as needed. Me being me, I probably added quite a number of typos. Mountain Blue 01:10, 17 June 2009 (UTC)

blockety blockety block

This might interest you: IP Geolocation SQL Database. Mountain Blue 21:51, 20 June 2009 (UTC)

Interesting: makes it easy to stay up-to-date and allows for another way of questioning the data base. Thanks larron^{sicut fur in nocte} 12:08, 21 June 2009 (UTC)

Revision time indexing

I got a big favour to ask. (perhaps because you got the chart "number of edits per day at CP") Since the wiki revision numbers are sequential, is it possible to compile a list of revision milestones with time, like indexing the time/dates of revision number, perhaps in linear increments of 10000(number is arbitrary, please adjust as you see fit)? This will help in finding stuff in archived talk pages (if it exists, usually between a certain time frame) in help of attempting to fix links to talk pages (Since revision numbers are in most of the difflinks, time spend finding which archive they are in would be way quicker if the time in indexed) Thank you for your time. [[User:K61824|]][[User_talk:K61824|]] 11:50, 21 June 2009 (UTC)

To get the number of articles created each day, I looked up the first revision of every article in main space. So, I could provide you with a list of the time/dates of these ~30,000 revisions. Would that be sufficient? larron^{sicut fur in nocte} 12:24, 21 June 2009 (UTC)

That should do it. The important part is the revision number of the article instead of the name I guess. A CSV or table somewhere would be most helpful. Thank you very much. [[User:K61824|]][[User_talk:K61824|]] 15:44, 21 June 2009 (UTC)

Here you go (tab separated list of number of revision, number of article, timestamp, year, month, day, hour, minute, second) larron^{sicut fur in nocte} 08:55, 23 June 2009 (UTC)

Copyright

Hey LArron, just a note, you have released much of your work under GFDL specifically, would you be willing to transition it over to CC-BY-SA just to ensure coherent copyright structure? tmtoulouse 04:57, 17 July 2009 (UTC)

No problem - any graph which I created may be put under CC-BY-SA. larron^{sicut fur in nocte} 06:14, 17 July 2009 (UTC)

Thanks for the smooth transition! larron^{sicut fur in nocte} 20:09, 17 July 2009 (UTC)

A special request

I have a request for your expertise, Larry Sanger, the founder of Citizendium made some predictions in October 2007 about the growth of his project: see here. I would love to see your analysis of his predictions vs. reality almost two year later. tmtoulouse 18:41, 28 July 2009 (UTC)

Interesting read: I suppose Nov 2007 was the happiest month in Sanger's live as his prophecy seemed to become real - it's the month with the highest rate of creation of articles! But after that....

I'll look into the details later on, I smell some geometric growth :-) --larron^{sicut fur in nocte} 19:12, 28 July 2009 (UTC)

By the way, I have been meaning to show you this, and I don't remember if I ever showed you this. Your images do a good job sparking interest/discussion. It is great work, that does get noticed around the net. tmtoulouse 20:03, 28 July 2009 (UTC)

Thanks for the links! Have you looked at Citizendium's own human-made statistics page? Nice pics, but no exponential growth to be seen there... larron^{sicut fur in nocte} 09:19, 29 July 2009 (UTC)

Congrats on the recognition by WP! Now, why isn't our Citizendium article funny enough? Генгис^{You have the right to be offended; and I have the right to offend you.} 09:42, 29 July 2009 (UTC)

Sanger's prognosis

History of CZ

And here it is - not much of an analysis, but - as usual - just a visualization. Though doubling the number of articles each year doesn't seem to be impossible, I suppose the reality is muddling on creating 14 articles a day. That's not to bad, but it won't meet Sanger's expectations. larron^{sicut fur in nocte} 18:07, 29 July 2009 (UTC)

Looks to me as if we are approaching a deviation in the path between "doubling" and the more linear growth. Will be interesting to see which path it follows. tmtoulouse 18:19, 29 July 2009 (UTC)

But though both models coincide at this time, there is nothing in the history which hints to exponential growth. I'd stick with muddling. larron^{sicut fur in nocte} 18:51, 29 July 2009 (UTC)

Interesting stuff. It's good to know you got a mention on WP and get some guys on Citizendium to have a cry. I mean, really, our article on the place isn't exactly wrong as they say, it's just focused on how CZ is really fucked up in relation to topics that RW is interested in. ^postate 18:59, 29 July 2009 (UTC)

Trent, if you're still reading this thread - did we get much traffic from the link? Генгис^{You have the right to be offended; and I have the right to offend you.} 19:22, 29 July 2009 (UTC)

Armond, link(s) to the WP mention and CZ crying if you have them handy please? Oh, and Larron, can you slap the "reality" line on the "prognosis" graph easily? I'm too lazy to "analyze" the scales on the vertical axes... ħuman 19:36, 29 July 2009 (UTC)

Human, are you just too fucking lazy to click on Trent's links above for the WP & CZ mentions? 21:00, 29 July 2009 (UTC)

Pictures at Teflpedia

Hi LArron, you uploaded a few pictures to teflpedia ([1]), these weren't moved to RW during the import. I don't want to upload them myself because then then they won't show up on you userpage. -- Nx / talk 08:04, 7 September 2009 (UTC)

Very considerate - I'm planning an article on W. Dembski's and R. Marks's paper Conservation of Information in Search. For this, I'll have to change some of the pics...

And thanks to you - and π I suppose - for all your work on the last few weeks! larron^{sicut fur in nocte} 09:25, 7 September 2009 (UTC)

Would people please stop acting like I do anything. - π 09:35, 7 September 2009 (UTC)

At least they're not accusing you of doing something bad this time :) -- Nx / talk 09:45, 7 September 2009 (UTC)

Lovely work

As ever. :) Totnesmartin 17:17, 1 November 2009 (UTC)

Thanks! larron^{sicut fur in nocte} 09:21, 2 November 2009 (UTC)

Categorizing images

I'm going over the images in Conservapedia related images and am recategorizing the graphs into Category:Conservapedia statistics (I'm not sure this is the best name, but once I'm done, the name can be easily changed by a bot). I'd like to ask you to use this cat in the future. Also, what cat did you use for images that were not Conservapedia related? (e.g. the RationalWiki graphs) -- Nx / talk 09:22, 14 November 2009 (UTC)

Ouch, I'm feeling downright lazy: I never added any categories to my pics. Sorry! larron^{sicut fur in nocte} 09:46, 14 November 2009 (UTC)

It's not like we had a category system for you to put them into :) I've created Category:Statistics images for all the non-CP related ones. -- Nx / talk 09:58, 14 November 2009 (UTC)

Monumental effort

Thanks for doing the licensing on all your images. - π 01:13, 29 November 2009 (UTC)

Welcome back!

I was just thinking the other day "I hope LArron hasn't gone for good, 'cos who will do all our graphs?" and whup! there you are. Welcome back. --PsygremlinSnakk! 17:13, 2 February 2010 (UTC)

Thank you very much! larron^{sicut fur in nocte} 09:30, 3 February 2010 (UTC)