User talk:Jimbo Wales: Difference between revisions

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Content deleted Content added
Russavia (talk | contribs)
→‎Wrong: lighten up
Line 99: Line 99:
:It is their responsibility to not disrupt our project, and we ''will'' out things like IP addresses (at least suspected ones) and location (as a result) if necessary to prevent further disruption.[[User:Jasper Deng|Jasper Deng]] [[User talk:Jasper Deng|(talk)]] 22:13, 26 March 2012 (UTC)
:It is their responsibility to not disrupt our project, and we ''will'' out things like IP addresses (at least suspected ones) and location (as a result) if necessary to prevent further disruption.[[User:Jasper Deng|Jasper Deng]] [[User talk:Jasper Deng|(talk)]] 22:13, 26 March 2012 (UTC)
::I would have to look at the specific case before I could have an opinion about it.--[[User:Jimbo Wales|Jimbo Wales]] ([[User talk:Jimbo Wales#top|talk]]) 22:18, 26 March 2012 (UTC)
::I would have to look at the specific case before I could have an opinion about it.--[[User:Jimbo Wales|Jimbo Wales]] ([[User talk:Jimbo Wales#top|talk]]) 22:18, 26 March 2012 (UTC)
:I mean, seriously, a move was done, it was challenged by moving back, with a request to discuss as per [[WP:BRD]], and you ignored that and moved back without discussion. Please discuss. The hook was clearly done for April Fools DYK, and yes, yes we know we have to turn the project into Prudeapedia. Hence, why for the next Russian election, I am going to suggest to Zhirinovsky that he get a giant rooster, put a circular band on one of its feet, and then place 2 chickadees belonging to Ms XXX on either side. We should then have the hook,...DYK...Vladimir Zhirinovsky put a ring on his cock, and then placed it between Ms XXX's tits? Let's see how you guys would handle that one eh. Lighten up.....geeeez. [[User:Russavia|Russavia]] <sup>[[User talk:Russavia|ლ(ಠ益ಠლ)]]</sup> 22:21, 26 March 2012 (UTC)

Revision as of 22:21, 26 March 2012


(Manual archive list)

ARIN CEO asks us to go to IPv6

I cannot verify the authenticity of it, but apparently ARIN CEO John Curran has asked us to hasten the move to IPv6 to the point that the WMF can participate in World IPv6 Launch. It would be sad if the WMF could not take the leap with everyone else, but I realize that there are major outstanding issues that need to be resolved. Nevertheless, it seems that contact by Curran himself is a big motivator.Jasper Deng (talk) 01:12, 22 March 2012 (UTC)[reply]

  • IP Address: 12.174.51.2
  • ISP: AT&T Services
  • Organization: INTERCONTINENTAL HOTELS GROUP
  • Region: Austin (US)
    cyberpower ChatOffline 01:21, 22 March 2012 (UTC)[reply]
    • Who says he couldn't have been traveling? The topic, whether the message is real or not, is still a valid, pressing one for the WMF.Jasper Deng (talk) 01:23, 22 March 2012 (UTC)[reply]
      • I'm simply giving information to start the discussion with. I'm not trying to prove anything. I would suggest contacting ARIN's CEO by e-Mail or phone to verify that he indeed did write that.—cyberpower ChatOffline 01:24, 22 March 2012 (UTC)[reply]
SxSW just wrapped up this past weekend - lots of tech types were there, or so I understand. Then again, you'd think he'd use a more official channel. ​—DoRD (talk)​ 02:35, 22 March 2012 (UTC)[reply]
Well, we aren't too transparent about how to contact the WMF either...Jasper Deng (talk) 02:37, 22 March 2012 (UTC)[reply]
I've met him before, and I see no reason to doubt this request is genuine. Perhaps someone can help him find the right people to talk with!--Jimbo Wales (talk) 04:09, 22 March 2012 (UTC)[reply]
However, with it being a hotel IP, I don't believe there's really anything the community can do; the WMF has to do something, I think.Jasper Deng (talk) 04:20, 22 March 2012 (UTC)[reply]
There is a problem with quickly switching over to IPv6. Majority of systems out there are old style an run off of IPv4. Anything earlier than Windows vista will not be able to connect to IPv6. By switching over to IPv6, we may be kicking off millions of experienced editors including me.—cyberpower ChatLimited Access 10:06, 22 March 2012 (UTC)[reply]
Um, my Windows XP is IPv6 ready (talk→ BWilkins ←track) 10:21, 22 March 2012 (UTC)[reply]
It's not so by default, and I can't expect all users to know the command.Jasper Deng (talk) 17:23, 22 March 2012 (UTC)[reply]
It's expected that the IPv4 address's will be running out by the end of 2012. To be fair, we're in March. Testing and stuff could go underway, but I don't think there is a major rush to have it completed. I just did a quick search and there are also ways to communicate with IPv6 websites from an IPv4 address through a proxy, so it may not mean getting an upgrade in hardware straight away, however, I would assume, it will eventually be phased out entirely. Mrlittleirish 11:26, 22 March 2012 (UTC)[reply]
Isn't it possible to support connections from both IP versions? If WMF implemented support forth both, then they wouldn't force old users to have have to upgrade to new hardware to continue editing Wikipedia.—cyberpower ChatLimited Access 11:39, 22 March 2012 (UTC)[reply]
I don't think there is any reason why we shouldn't seamlessly support both. I have been told (but only randomly by random people) that we don't yet do IPV6 at all. This is not about kicking people off who use old hardware/software or making them use a proxy or anything - that would be very very wrong for us to do. This is about making sure that we are proactive about assisting with the move into the future. I personally have no real idea of the status of that right now.--Jimbo Wales (talk) 11:41, 22 March 2012 (UTC)[reply]
Well, we have some on-wiki issues too. We just talk to you as you're the head of the WMF; unfortunately we don't know how to reach the WMF network admins.
With that said, though, I think that we can resolve these in time if gadget and Toolserver developers hurry up.Jasper Deng (talk) 17:23, 22 March 2012 (UTC)[reply]
You can always write to answers@wikimedia.org. :) Currently I man this email address, but if I don't know, I'll ask around. Director of Technical Operations CT Woo tells me that, while there are a couple of tasks to be completed first (including particularly to do with Mediawiki and load balancer LVS), the WMF hopes to participate in the IPV6 Launch Day on 6/6/12. --Maggie Dennis (WMF) (talk) 13:56, 26 March 2012 (UTC)[reply]
Well, the WMF does not appear on the participants list, yet.Jasper Deng (talk) 22:11, 26 March 2012 (UTC)[reply]
Most things are working although there are a zillion little toolserver scripts and bots which don't know about ipv6 yet and may need to be upgraded. Also, there are still some bugs. In general, reading from ipv6 is ready for prime time, but editing might still be months to years out. 70.59.28.93 (talk) 17:00, 22 March 2012 (UTC)[reply]

I think you've been informed right Jimmy, there is a way to seemlessly support both, many sites have already implimented it. It would definately need to be something that gets implimented by tech staff. Most people, I would imagine, would be contacted by their ISP's about the transition. Mrlittleirish 14:32, 22 March 2012 (UTC)[reply]

  • Here's something of interest that would like a WMF comment: mw:Roadmap#April 2012 (search for "IPv6" in that section).Jasper Deng (talk) 06:34, 23 March 2012 (UTC)[reply]
    • Do you (Jimbo) think that if I start an active on-wiki movement to get ready, you can convince the WMF to follow suite and be ready ahead of World IPv6 Launch?Jasper Deng (talk) 04:42, 25 March 2012 (UTC)[reply]
  • Not entirely true, the problem with dual stack (as this is called), is that people who have IPv6 enabled, but incorrectly working (either due to older OS or due to ISPs being incompetent) would not even be able to reach the IPv4 website. That's why the IPv6 change is being done so carefully, even by large organizations like google. Just flipping the switch will loose you about between 1 to 4% of traffic. That was part of last years effort. To better get a feel about what would break, and make it possible for infrastructure organizations to better assess where they themselves had issues. But in general yes, dual stack is possible (actually it's what everyone is doing) and it's the organizations idea to go there as well. Last years effort was unable to succeed because it was forgotten that we would need to update the databases, which would be at least 1 week of work, before IPv6 can be enabled for wiki. Because of that we missed the window of opportunity. It would be good if someone could at least put that Db update on the map for the immediate future, before we again are too late. —TheDJ (talkcontribs) 14:31, 25 March 2012 (UTC)[reply]
    • IPv6 brokenness was listed as an issue on wikitech, but it's not directly related to the MediaWiki software and the webservers themselves.Jasper Deng (talk) 18:35, 25 March 2012 (UTC)[reply]

Wikipedia ranked by Google and Bing

A study by Search Engine Watch shows that Bing surpasses Google in giving prominence to Wikipedia pages.

Wavelength (talk) 18:59, 24 March 2012 (UTC)[reply]

  • Current tests show Google and Bing closer now: At least when comparing the results for Wikipedia among the top-ten results, on both Bing and Google, the search-results show very similar ranks, based on those 1,000 test-cases, which gives a large enough sample to ignore several questionable cases. However, when I have re-run some of the actual data from February 2012, many of the current search-results seem closer than what that report had stated. For example, the Google/Bing ranks are similar 18/19 for: 'united states life tables 2001'. Also, there are several math counting errors in the report, where the count of test-cases (in the report) totals to 1,095, but the actual data-file entries list only exactly 1,000 test-cases (errors: there are only 281 one-word searches, not 290; only 197 three-word, not 208; only 78 five-word, not 86; and 43 cases of six-word searches, not 86). Plus, there are some bizarre search terms among those 1,000, with some unusual misspelled words, such as "tigonal" rather than the more-common "Trigonal" or "mitronchondrial dna" rather than "mitochondrial dna" (see SEW data file: [1]). Again, with 1,000 total cases, then a few spelling errors will not affect the statistical averages; however, errors in the data (or counting) typically indicate that the data was not carefully checked for "tests of randomness". In a formal experiment, the test data should be given a "sanity check" as an independent step, to ensure that the basic data represents realistic scenarios and is not wildly skewed, as being a "false premise" to support false conclusions about that data. Another unusual issue IMHO, is listing 11 month names (except "April") but no weekdays or years, as I would have included: Tuesday, Thursday, 1492, 1493, 1520, rather than 11 months. However, overall, the list of 1,000 entries is a good start. Anyway, I think the best conclusion, from that study by Search Engine Watch, was that both Google and Bing typically list a Wikipedia article within the top ten search-results, but the exact rank might be slightly higher, as either Google or Bing can be higher for a specific search. Beware the report is inaccurate for many specific, current searches. -Wikid77 (talk) 14:13, 25 March 2012 (UTC)[reply]
A Google, Bing, Yahoo! and Ask search on "Jimmy Wales" will all bring up Jimmy Wales as the top result. This helps to explain why Wikipedia BLP articles are the subject of close scrutiny. This is to be expected for a search on the name of a living person.--♦IanMacM♦ (talk to me) 14:43, 25 March 2012 (UTC)[reply]
  • Results depend on type of subject: That is a good point about the high search-rank of BLP names, versus other search-phrases. Another issue which seems to limit that SEW report's accuracy is that, for one-word searches, Wiktionary is "Wikipedia" (rather than claiming as "no match"). So, rather than pseudo-calculating "86%" as the percentage where Wikipedia articles appear in the top ten results, I would pick "95.86%" as typical search-results which list a Wikipedia article in the top ten. It is not accurate to run a search for Wikipedia articles incorrectly 12% of the time, then claim "Wikipedia has a 14% failure" (not true) to match within 10 results, when the flawed experiment was the cause of most of those failures. The SEW report started with a "good idea" about using 1,000 search-phrases, but then the experiment went terribly wrong in multiple areas. Another example is the search for "Limerick school of art and design" which currently ranks #1 in both Google and Bing, but the report's data-table claims no match for that phrase, when the article has existed, with that title, since October 2006. Another error is claiming "car" matches WP at Google rank #13 when actually, "car" matches WP "Automobile" at Google #5 (in the top ten) and Bing #9. It looks like those Search Engine Watch people were asleep on their watch. The word "sun" matches WP in both Google/Bing at #2, and "Dell" matches as Bing #4 or Google #5. Both Google and Bing list similar results, and almost no test-cases (from February 2012) now fail to match within 50 results. The claim that 10% (of those cases) fail to match a Wikipedia article within 50 results is just not true. -Wikid77 (talk) 17:48, 25 March 2012 (UTC)[reply]

I wanted a little help with my wikia's wiki

When I saw that Uncyclopedia is a wikia's wiki, I found that a Wikia's wiki can have the wikipedia's layout.

Jimmy, help me, I was wondering how my wiki can have the wikipedia's layout?

The link of my wiki:: pt-br.weirdpedia.wikia.com

I thought, as co-founder of Wikia, you could help me do this.

A hug from your fan Renan Marcos (A Brazilian Wikipedian) — Preceding unsigned comment added by 187.15.121.227 (talk) 01:27, 26 March 2012 (UTC)[reply]

Hi, questions about Wikia are best directed to community@wikia.com!--Jimbo Wales (talk) 08:03, 26 March 2012 (UTC)[reply]

A DYK in the making

See Template:Did you know nominations/Zhirinovsky's ass.

I mean, seriously? --JN466 20:56, 26 March 2012 (UTC)[reply]

Jimbo, please, can you step in and sort out DYK? The Eastenders incident was horrific, and this looks as though it's heading into a similar way. I cannot believe these kinds of 'jokes' are being permitted for the front page. The shop window of Wikipedia is going to look like it's splattered in post-Friday night vomit. doktorb wordsdeeds 21:11, 26 March 2012 (UTC)[reply]

Wrong

Is appropriate for users to create a user page about an Ip that keeps coming back,about what he does,or personal info.Doesn't that break a Wikipedia rule,if not we should have a rule about it. 74.163.16.20 (talk) 22:10, 26 March 2012 (UTC)[reply]

It is their responsibility to not disrupt our project, and we will out things like IP addresses (at least suspected ones) and location (as a result) if necessary to prevent further disruption.Jasper Deng (talk) 22:13, 26 March 2012 (UTC)[reply]
I would have to look at the specific case before I could have an opinion about it.--Jimbo Wales (talk) 22:18, 26 March 2012 (UTC)[reply]
I mean, seriously, a move was done, it was challenged by moving back, with a request to discuss as per WP:BRD, and you ignored that and moved back without discussion. Please discuss. The hook was clearly done for April Fools DYK, and yes, yes we know we have to turn the project into Prudeapedia. Hence, why for the next Russian election, I am going to suggest to Zhirinovsky that he get a giant rooster, put a circular band on one of its feet, and then place 2 chickadees belonging to Ms XXX on either side. We should then have the hook,...DYK...Vladimir Zhirinovsky put a ring on his cock, and then placed it between Ms XXX's tits? Let's see how you guys would handle that one eh. Lighten up.....geeeez. Russavia ლ(ಠ益ಠლ) 22:21, 26 March 2012 (UTC)[reply]