User talk:Johnuniq/Archive 14

Convert error category

I was just thinking that Category:Convert error could get full of rubbish quite quickly once the template goes live. One thing that could help in reducing that is to not categorise talk pages. The reason for the category is to track errors and get them fixed by editing the page with the issue, but for issues on talk pages, it's not work fixing them most of the time so it's probably not worth tracking them. Another thing that could be added is a |nocat=true parameter which would then be used to manually suppress the category on the page with the error. -- WOSlinker (talk) 12:05, 17 September 2013 (UTC)

Good idea. It looks as if it would be easy for the module to get the namespace number of the current page. One simple thing would be to output a category if and only if namespace == 0. If necessary, that could be tweaked later, but do you think it's likely a category would be needed in any of the others at WP:Namespace?

If we use the namespace == 0 test, would we also want a nocat parameter? Parameters are easy to add, but I think we should avoid them unless really needed because too many leads to confusion.

Please check the section just above this because I've added a new comment. Johnuniq (talk) 09:32, 18 September 2013 (UTC)

Yes, just namespace 0 would be fine, and no need for nocat. -- WOSlinker (talk) 09:45, 18 September 2013 (UTC)

Worth including template space too? That way errors inside infoboxes &tc. are listed by their source page, not hidden in articles. But this can include many controlled errors too (like empty infobox input in the template). -DePiep (talk) 14:26, 19 September 2013 (UTC)

@WOSlinker: I have implemented the above suggestion, thanks. It defaults to include categories only in main and template namespaces, and can easily be adjusted. While I have your attention, I recreated {{convert/sandboxlua2}} to include "warnings = true" so that template will show warnings. Johnuniq (talk) 10:17, 21 September 2013 (UTC)

Category:Convert invalid option

Your overview subpage says:

• Category:Convert invalid option ∙ can only occur if the template invoking the module includes "|warnings=true"

I suggest it is switched on right away always. That way we can find and clean up all these errors. Unless there is a pre-Lua convention that relies on this setting, but I do not see |warnings= in template code or documentation. -DePiep (talk) 14:40, 19 September 2013 (UTC)

I'm not sure, but am happy with whatever is wanted. It's reasonably common for convert templates to contain some junk (for example, "abr=on" instead of "abbr=on"), and I felt there should be a configuration setting (an option in the template) so warnings could be turned off if they generated too much noise. I thought it would be desirable to start with warnings off so more important error messages could be handled first (I'm expecting quite a few "unknown unit" problems), with fewer ugly messages visible in articles where there is probably no actual problem—that is, the displayed text is probably correct, even if not precisely what the editor intended. I have scraped examples from various articles, and just ran a test on a local computer with 8,000 converts. The were no warning messages, and I think I haven't cleaned problems from the converts, so perhaps warnings are sufficiently rare that having warnings on would not be a problem.

I don't think I've documented the following issue: Superfluous positional parameters are ignored, for example {{convert|12|m|ft|junk}} gives the same output, namely "12 metres (39 ft)", with and without warnings. Could think about some extra code to check for them later. Johnuniq (talk) 04:39, 20 September 2013 (UTC)

Ugh! Bit of turmoil going on here, and my test was broken. When done properly, the 7,540 converts (not 8,000) generate 31 warnings:

2 of "abr=on", 3 of "bbr=on", "abb=on", 1 of "abbrev=on" (should be "abbr=on")

1 of "abbrev=off" (should be "abbr=off")

1 of "sigfigs=1" (should be "sigfig=1")

1 of "dispp=flip" (should be "disp=flip")

1 of "link=on" (should be "lk=on")

1 of "sep=s" (possibly wanted "disp=s")

1 of "sp=eu", 3 of "sp=uk", 12 of "sp=UK" (should omit)

2 of "spell=us" (should be "sp=us")

1 of "lbabbr=on" (occurs in {{convert|20000|to|25000|kg|lbabbr=on}} which should be {{convert|20000|to|25000|kg|lb|abbr=on}})

Still 31 warnings in 7,540 converts is a pretty small number, so that sample suggests having warnings on by default would be ok. Johnuniq (talk) 05:59, 20 September 2013 (UTC)

Category names

The module uses four cetegories:

• Category:Convert error
• Category:Convert dimension mismatch
• Category:Convert invalid option
• Category:Convert unknown unit

Suggestions:

1. Make category names plural by general wp style.
2. Be more descriptive: like "Pages with convert dimension mismatch"
3. If I read code well, only the first message is shown. In that case, the generic (least precise) Category:Convert error should be added last, not first.

I will put all four in a parent template. -DePiep (talk) 15:04, 19 September 2013 (UTC)

Re #1: Yes, that's fine, although I'll hold off until there is more discussion on #2.

Re #2: It's just personal opinion, but the redundancy in a category like "Pages with X" is distasteful to me. A category is a list of pages so it's pointless putting "Pages with" in the name. For articles, we say Category:Living people or Category:1961 births. It's not Category:Articles about living people or Category:Articles about people born in 1961. However, I'm pretty flexible and am happy with the outcome of a discussion, although I hope we could attract a few opinions. The good news is it would be very easy to change the category names (see "all_categories" in Module:Convert/text).

Re #3: Yes, only one message is displayed no matter how many errors are in a convert, but each message has it own category so there is no realistic way of putting "Convert error" last (see "all_messages" in Module:Convert/text for the full story). Looking at the wikitext here shows the category for each message. Some expected errors have a dedicated category with a fairly obvious meaning: "Convert dimension mismatch", "Convert invalid option", "Convert unknown unit". Other miscellaneous messages are in the generic "Convert error" category. Johnuniq (talk) 09:55, 20 September 2013 (UTC)

Talk continued at Module_talk:Convert#Maintenance_category_names. -DePiep (talk) 12:17, 21 September 2013 (UTC)

A barnstar for you!

	The Barnstar of Diligence
	For your willingness to go way out of your way to help people from other wikis in technical matters. Siddhartha Ghai (talk) 17:44, 19 September 2013 (UTC)

Thanks very much! I was having a pretty mediocre day, and that was before looking at my watchlist (see below), so this message is most welcome. For future reference, this refers to a discussion above and at hi:User:Johnuniq/number. Johnuniq (talk) 02:26, 20 September 2013 (UTC)

sigfig error message

Just to let you know that I've changed the sigfig error from a critical one to a warning instead. -- WOSlinker (talk) 20:21, 21 September 2013 (UTC)

A "critical error" vs a "warning"? So far I only saw "errors" (orange, replacing the text, in one of three cats, all the same level of alarm) and "warnings" (pink, trailing the convert text, in a separate category). Does this edit mean there is a third, hybrid sort of message? At least we should give it its own category. But better we discuss its status & consequences. -DePiep (talk) 01:17, 22 September 2013 (UTC)

No, there's only two types. If there was |sigfig=-1 for example then previously it would have just displayed an error message and no conversion. I've changed that so that it now ignores the bad sigfig value and does the conversion with the default sigfig value instead and will display a warning instead (provided that warnings are on). -- WOSlinker (talk) 07:00, 22 September 2013 (UTC)

Thanks! In retrospect, it's clear that a sigfig problem is not fatal. For next time, you copied and edited the line:

add_warning(parms, 'cvt_empty_option', loc_name)

but loc_name is only suitable for cvt_empty_option (which reports that no value was given for loc_name), whereas cvt_bad_sigfig needs to be given loc_value because it reports that the given value is bad.

On looking at the code, I realized I had to do a little refactoring because the next section should not give a warning if the sigfig code already has done so. That's a benign problem currently because only the first warning is shown, but it's cleaner to get it right. Johnuniq (talk) 11:13, 22 September 2013 (UTC)

I've made Template:Convert/testcases/sigfig using {{convert/sandboxlua2}}. note that now we have warnings and errors in the same maintenance category. I would not mind adding categories for systematic message tracking. -DePiep (talk) 12:39, 22 September 2013 (UTC)

I'll look at the good work you're doing later (won't have time for at least 24 hours I'm afraid). However, I wouldn't worry about niceties such as whether errors and warnings belong in different categories. If we're going to output a message of any kind, it needs to be tracked so it can be fixed, and it doesn't matter what the category is called, or how optimized is the system of categories. Rather than have more categories, perhaps we should have fewer? Johnuniq (talk) 12:48, 22 September 2013 (UTC)

I've created a set of testcases at Template:Convert/testcases/rounding which involve combinations of the round_to, disp=5 and sigfig=number parameters. There are some differences between the current convert and the lua version which will need looking at. -- WOSlinker (talk) 13:09, 22 September 2013 (UTC)

Very nice, and I'm looking at it right now. I haven't finished, but so far haven't seen a single error from the module, although that may be prejudice. I'm not quite sure what the text at Template:Convert/testcases/rounding is saying. If a convert specifies round_to and sigfig, why wouldn't that be a case of "undefined behavior"? At any rate, what the module is doing looks reasonable to me. Johnuniq (talk) 13:20, 22 September 2013 (UTC)

Adding gross tonnage

I'd like to have an option to convert gross tonnage to cubic meters and back again. This involves a screwy formula which I have no idea how to invert, except numerically. [1] I don't want to intrude into your good code and screw things up, but I would propose adding something like this to line 1778 of the current Module:Convert:

        elseif in_builtin == 'gross tonnage' then
            -- gross_tonnage = (0.20 + 0.02 * log10(cubic_meters)) * cubic_meters
            -- inquiry at the Mathemetics Refdesk failed to obtain a closed form solution for cubic_meters
            if (invalue <= 0) {
            	return false, { 'cvt_bug_convert' }
            	}
            local gross_tonnage = invalue
            local k_factor = 0.20
            local cubic_meters = gross_tonnage / k_factor
            do {
            	k_factor = 0.20 + 0.02 * log10(cubic_meters)
            	local new_cubic_meters = gross_tonnage / k_factor
            	if (math.abs(new_cubic_meters - cubic_meters) < 0.1) {
            		return true, {
            			outvalue = new_cubic_meters
        			}
        		cubic_meters = new_cubic_meters
        	} while (iterations<100)
            return false, { 'cvt_bug_convert' }

The inverse is mathematically simpler; I assume you'd hook into it in the same place using out_builtin == 'gross tonnage' and set outvalue = invalue * (0.20 + 0.02 * log10(invalue)).

The main spot to invoke this one would be in Template:GT.

This is the kind of funny logic that might deserve to be exiled to a special module to prevent recalculating all the instances of math.convert... in any case, I'll leave it at this point in your capable hands. Wnt (talk) 22:30, 21 September 2013 (UTC)

Now this is definitely going to make my head explode! Let me focus on the trivial first. Are you saying that the unit code (the thing that has to be entered in a convert template) should be "gross tonnage"? What about "GT" as per Gross tonnage? That's a bit dubious because it could be confused with "Gt" (gigatonne), but I think there are a couple of similar cases where people have to be careful to get the unit code exactly correct (and at least GT would be volume while Gt is mass). I don't mind what the unit code is, but they are usually brief.

Actually, I'd better confirm that: is GT a volume? Should it be possible to convert, say, 1234 cubic feet to GT?

I guess the symbol is "GT" and the name is "gross tonnage"? But how are each written with a value? I guess there is no plural name?

What is the default output unit (in a convert like {{convert|120|GT}})?

Where are the testcases!? Any chance of a few sample convert templates with what the output should be?

This looks interesting! Johnuniq (talk) 11:49, 22 September 2013 (UTC)

Well, the default output is obvious: m3 (cubic meters). Johnuniq (talk) 13:31, 22 September 2013 (UTC)

I'm not sure about all the synonyms and how you'd like to organize them. I supposed "gross tonnage" was fairly safe, but there might be room for revision. GT is used widely on Wikipedia via the template; it could also be G.T. to be less ? confusing for the template. My thought was that if it got inserted into the template, someone at one of the articles would set me straight about any errors. :) GT should indeed be a volume, and you should be able to convert to and from other units such as cubic feet. The problem of course is that most units come from science, but this one was invented by bureaucrats!

If you search "114,000 gross tonnage" online you'll find many sources using the phrase (occasionally with a hyphen after the 114,000...) So for now, as someone with dubious competence to command a rowboat, I'll say that my assumption is that

{{convert|120|m3|GT}} --> 29 gross tons

With the huge caveat that I have not looked into your module deeply enough to see how you generally decide the issue of the level of precision to report. Wnt (talk) 19:05, 23 September 2013 (UTC)

@Wnt: Your code works well, although despite the fact that I have made it work in Lua, I still haven't actually read it! I don't think I have time at the moment to put this in Module:Convert because of the details that would need to be figured out. Apart from the question of whether the output should say "gross tons" or "gross tonnage", there is the issue of the precision. The module has to do something when a user enters (say) precision 8 (8 decimal places), yet clearly the fixed iteration exit test produces considerably less than that. Either the module has to limit any given precision to, say, 2 places (and similar for the more tricky sigfig=8), or the iteration has to do something more sophisticated (which seems silly since talking about 0.1 GT seems totally pointless). I've got a bunch of things I want to do to get the module ready for deployment, and would like to defer consideration of this for a couple of weeks.

Please experiment with Module:Sandbox/Johnuniq/grosstonnage (edit it there, or copy it) and see what it does. Following is a quick test:

• {{#invoke:sandbox/Johnuniq/grosstonnage | run_tests | 9000 | 11000}}

   Vol_input    Gross_ton    Vol_output  Iter
     9000.00      2511.76   9000.00280   4
    10000.00      2800.00   9999.99990   5
    11000.00      3089.11  10999.99989   5

Johnuniq (talk) 01:33, 25 September 2013 (UTC)

re: bed size table

...are you a wizard?! <3 — Preceding unsigned comment added by 193.63.174.211 (talk) 16:07, 5 October 2013 (UTC)

edit: *sees rest of talk thread only AFTER clicking the "send me a message" link on the main userpage and writing the above*

OHHHH I SEE NOW.

All the same, much thanks for fixing that. I'll try to follow that example next time something needs tweaking. 193.63.174.211 (talk) 16:23, 5 October 2013 (UTC)

This relates to a fix I made to how {{convert}} was used at Bed sizes. I've been working on Module:Convert for a long time and I'm glad there is benefit.

Is there a "send me a message" link on my user page? Some WMF wonder I assume. Johnuniq (talk) 09:30, 6 October 2013 (UTC)

Guns, Germs, and Steel

Johnuniq -

Your recent edits to this article were reverted by another editor, because they were unsourced. I think you made some very good points, and would like to see you put them back in, with sources. (What are the rules for citing a book when the article is about that book? I've never done it.) Just my 2 cents. Paulmlieberman (talk) 22:58, 7 October 2013 (UTC)

My recent edit was only to revert a very minor change that broke something in the text, and I did not write the section you mention. That section ("Response to criticism of Eurocentrism and determinism") can be seen in this permalink at 02:49, 27 September 2013. A quick look at the history suggests the removed text was present in the article four years ago, and is possibly older than that. I saw the edit which removed the section and decided to not challenge it because whereas the text was helpful to someone wanting to see Diamond's views, describing it as WP:OR is correct—an editor had essentially written a mini-essay extracting pieces from the book (aka cherry picking), and that's not suitable for Wikipedia. I'm afraid I don't have a recommendation about what to do because while it would be nice to present Diamond's views as he intended them, the only reasonable way to do that here is to find a reliable secondary source with an analysis of the book and its responses, then present a due synopsis of that information, with the source as a reference.

By the way, there is a problem with your signature as it is supposed to include a link to your user or talk page. There is something in your preferences where you possibly changed the default signature setting, and you might like to remove whatever you entered so the default occurs. Johnuniq (talk) 00:17, 8 October 2013 (UTC)

I try not to get into the politics around the GGS article, though I think it's a very important book, and I hate to see it trashed as Eurocentric. I think I fixed my signature. Here it is: Paulmlieberman (talk) 12:04, 8 October 2013 (UTC)

Yes, I agree. There has been a lot of nonsense written in response to the book. Signature is good. Johnuniq (talk) 23:15, 8 October 2013 (UTC)

Template editor role

Just to let you know that I've added the new template editor role to your account, which will come in useful when Module:Convert goes live & needs get protected, as it means that you'll still be able to edit it. -- WOSlinker (talk) 22:20, 17 October 2013 (UTC)

Thanks, nearly there!? Johnuniq (talk) 23:45, 17 October 2013 (UTC)

Titus Andronicus

Hey. Thought you might like to know I've opened a discussion about the non-free images in the Titus Andronicus article, here, if you wish to contribute. Cheers. Bertaut (talk) 20:42, 21 October 2013 (UTC)

Thanks, I noticed that and may comment later. I would suggest toning it down on the other editor regardless of the cluelessness of their behavior regarding collaboration. Johnuniq (talk) 21:57, 21 October 2013 (UTC)

Books and Bytes: The Wikipedia Library Newsletter

Books and Bytes

Volume 1, Issue 1, October 2013

by The Interior (talk · contribs), Ocaasi (talk · contribs)

Greetings Wikipedia Library members! Welcome to the inaugural edition of Books and Bytes, TWL’s monthly newsletter. We're sending you the first edition of this opt-in newsletter, because you signed up, or applied for a free research account: HighBeam, Credo, Questia, JSTOR, or Cochrane. To receive future updates of Books and Bytes, please add your name to the subscriber's list. There's lots of news this month for the Wikipedia Library, including new accounts, upcoming events, and new ways to get involved...

New positions: Sign up to be a Wikipedia Visiting Scholar, or a Volunteer Wikipedia Librarian

Wikipedia Loves Libraries: Off to a roaring start this fall in the United States: 29 events are planned or have been hosted.

New subscription donations: Cochrane round 2; HighBeam round 8; Questia round 4... Can we partner with NY Times and Lexis-Nexis??

New ideas: OCLC innovations in the works; VisualEditor Reference Dialog Workshop; a photo contest idea emerges

News from the library world: Wikipedian joins the National Archives full time; the Getty Museum releases 4,500 images; CERN goes CC-BY

Announcing WikiProject Open: WikiProject Open kicked off in October, with several brainstorming and co-working sessions

New ways to get involved: Visiting scholar requirements; subject guides; room for library expansion and exploration

Read the full newsletter

Thanks for reading! All future newsletters will be opt-in only. Have an item for the next issue? Leave a note for the editor on the Suggestions page. --The Interior 20:43, 27 October 2013 (UTC)

Convert/q needs Lua

Hi, Wikid77 here. I have created a Template:Convert/q, as a quick (ultra-fast) version of {Convert} and found that the Lua version is fast enough. So, I would like to use the Lua version, inside {convert/q} for limited main-space articles, to reduce 6-second conversions to below 1 second (where other templates are exceeding 30 seconds). I know you planned to release the Lua version as a system-wide upgrade, but a {convert/q} is needed soon, and the long-term testing for full-scale release of Lua is tedious, plus retro-fitting the dozen wp:wrapper templates: {Convert/2}, {Convert/3}, {Convert/4}, {Convert/flip2}, {Convert/flip3}, {Convert/show2}, {Convert/show3}, (etc.). Anyway, if {convert/q} will be a burden, then I can implement it through other means. Things are moving so fast, and people are complaining that Template:Jct is crashing pages (reformat: 58+ seconds!), where {Convert} is used 150-250 times (taking up to 6 seconds compared to Lua 0.80 second). They are planning a Lua version of {Jct} in December/January, but it has over 1,780 road-type formats to rework in Lua script. Meanwhile, {convert/q} with Lua would reduce those large pages by 10% faster, and so that would be excellent. -Wikid77 (talk) 22:42, 1 November 2013 (UTC)

Good! I would be very pleased to get some real-world usage as I'm hoping to deploy soon. I've been hit with some RL issues that mean I haven't had much energy or time for thinking recently, and I might be on reduced capacity for a while yet. What I wanted to do was finish devising some tests (for options) to be added to Module:Convert/sandbox/testcases, and do a few tweaks to the documentation that I have planned, then propose switching {{convert}} to use the module. I could go on testing and tweaking the module indefinitely, but I think I should just do the boring business of those couple of tasks I mentioned, then get the module running live, then spend a few weeks cleaning up problems which arise. That is, pull my finger out! I'm pretty weak on templates, so I don't know if you intentionally omitted the closing tag for <noinclude> in {{convert/q}} or whether it was an oversight. I'll have to look at Template:Jct because I hope to need a new project soon... Johnuniq (talk) 02:30, 2 November 2013 (UTC)

Okay, I'll start using {convertq}, to confirm the 6x-8x faster calculations. The "</noinclude>" tag has been optional for years, but I added it in case others wonder if it hurts the speed. Meanwhile, beware there are already people working on Lua {Jct}, but perhaps find new tasks under wp:Lua_requests, and just give the road-project people time to test the 1,780 {Jct} road-type codes (see: "Template_talk:Jct#Start Lua version" for status). -Wikid77 03:38, 2 November 2013 (UTC)

sheldrake NPOV tag-spam

While I agree it is more than fair to reward Vzaak for his efforts today, by taking down the spam-tag...

...it is a *little* bit going too far, however, for you to say "absolutely no justification" that it might still be left up.

It does finally say scientist again in the first sentence (and PhD as well as occupation:biochemist over in the infoboxen... not sure if that last bit is new or not).

But the three cites for the term 'scientist' are still the same old parapsychologist ones.  :-)

Which needs fixing. Anyways, I expect the remaining issues can be worked out on the talkpage, and I'll try to explain the same to Blippy. Thanks for improving wikipedia. 74.192.84.101 (talk) 13:21, 3 November 2013 (UTC)

Hi

Just wanted to say that I agree with your statement here.Tstrobaugh (talk) 16:39, 10 November 2013 (UTC)

Thanks. Another editor thanked me for that comment, so there are three of us who agree! Johnuniq (talk) 22:25, 10 November 2013 (UTC)

careful

You should be careful with removing reliably sourced content. You, of all people, should know that. I can remember that you have been involved in a number of reuest for comments where you complained about other editors usage of sources. Yet here you are not giving the slightest darn about sopurces right? hmmmm. Pass a Method talk 04:10, 17 November 2013 (UTC)

Also, i noticed you are following editors to articles you have never edited before. I hope you know that that would fall under the definition of wp:stalk. Since i have seen editors getting blocked for such behavior it would be in your best interest to stop. Consider this a warning. Pass a Method talk 04:23, 17 November 2013 (UTC)

You visit a lot of articles, so perhaps you are confusing me with someone else? May I suggest using article talk pages for content issues because user talk pages should not be used in an attempt to repel editors with a different view. I see that I mentioned something similar to you recently (diff). Johnuniq (talk) 05:18, 17 November 2013 (UTC)

It is clear why Johnuniq reverted you at the Faggot (slang) article. Interesting that you should speak of WP:HOUNDING/WP:STALKING, given recent comments with regard to your behavior, as shown here and here. Perhaps Johnuniq has simply taken an interest in topics that we both edit. Not to mention that he's had some of these topics and similar topics, such as Female genital mutilation, on his WP:Watchlist already. Furthermore, he is allowed to WP:STALK any editor who welcomes his stalking; if he wants to stalk my edits, for example, he is more than in his right to do so. He is also allowed to stalk clearly problematic editors. Flyer22 (talk) 05:29, 17 November 2013 (UTC)

From the comments of the OP: Yet here you are not giving the slightest darn about sopurces [sic] right? hmmmm. Again, as in the previous incident of reverting Athenean's edits at Europe and then giving Athenean an AGF-defying TW third-level blanking warning for which he was warned, this editor fails to consider the edit-summary that John left which explained his reversion. In addition the OP fails to acknowledge that he is the only editor supporting his edit and that Flyer22 also opposes it and acts, judging by their message here, as if he is on the up and up and John had absolutely no reason of reverting him. I find his approach highly selective to say the least. This looks as yet another incident of ignoring a user's good-faith explanation of their edit and the fact that John's edit is supported by another user and proceeding to give them stern, belligerent and serious but unwarranted warnings. Overall this new incident is showing little improvement in the OP's behaviour toward well-respected, established and good-faith users. Δρ.Κ. ^{λόγοςπράξις} 19:40, 17 November 2013 (UTC)

Re: Visual editor

Sorry and I've "fixed" all <nowiki>...</nowiki> added by me yesterday using {{tag}}. Yesterday I added these <nowiki>...</nowiki> manually, not automatically generated by VisualEditor. Because I thought it would be safer and quicker to use <nowiki>...</nowiki>, preventing some of the element names from yielding real HTML elements by MediaWiki. I don't have a list that tells which tags are rendered as real elements and which tags are not. And I thought it does no harm at all to add <nowiki>...</nowiki> even if it's not needed. So I did like that. --Tomchen1989 (talk) 16:26, 18 November 2013 (UTC)

Thanks. I've replied at your talk. Johnuniq (talk) 21:51, 18 November 2013 (UTC)

Hume

Hume's racialism is fully sourced in his own writings. See the Talk page for Hume, in Archive 1, where it is mentioned four times over. — Preceding unsigned comment added by 86.176.252.226 (talk) 12:19, 19 November 2013 (UTC)

Perhaps the above comment relates to my recent revert of this edit at David Hume. That edit could not stand because it is unsourced commentary injected into the lead of the article. There have been a lot of passers-by wanting to say that Hume was racist, and the matter has been well aired, with my comment appearing at Talk:David Hume/Archive 1#Racism (again). Any further discussion should be at Talk:David Hume, but I stand by my archived comment that it is not satisfactory for editors to apply 21st century norms to reach a conclusion that words written over 200 years ago indicate racism. Johnuniq (talk) 04:39, 20 November 2013 (UTC)

A big favour ...

Johnuniq, I'm currently talking to someone at EBSCO Publishing about the potential to get some accounts donated to the Wikipedia Library. I'm passing along to them this great page you put together for HighBeam usage. The representative mentioned EBSCO would be interested in seeing similar metrics for Credo, as they have a similar business model. How much work would it be to put that together? Best, The Interior (Talk) 19:22, 25 November 2013 (UTC)

Some results from 2010 to July 2012 are at User talk:Johnuniq/Archive 11#Credo Reference results. Those numbers are for articles only. Using LinkSearch shows there are currently 492 links in all namespaces.

I guess I could get more details from an analysis of the monthly downloads of external links (that is, more since July 2012). That will take a bit of time because the relevant files are on another computer that is half-dead. I've been meaning to pull out its hard disk for a couple of months, but haven't got around to it. That disk has the monthly files up to and including May 2013, so I would need to download those since then. There is probably a more clever way to do this (WP:LABS?), but it would be quicker for me to stick with what I know.

I will do it, but first please confirm what months you want, and what links to search for. See the second bullet point at the top of my first link in this reply. Johnuniq (talk) 22:14, 25 November 2013 (UTC)

Well, that data you have already is useful. What will be persuasive to them will be an illustration of increased usage comparing the pre-donation period, to present day. I believe the first logins were given out in late March 2010. You have data back to Jan. 2010 - is it possible to grab numbers from 2009, for instance? That might be difficult. Since Credo (or xRefer) has been around since before us, I'm assuming the year-to-year increase 2001-2010 was a lot lower than the period 2010-2012. As for results post-July 2012, which I believe was the last report, three-month intervals up to the present would be ideal, but I don't want to tax your resources. And a graph would be great :) But whatever you can realistically put together to give them an indication of usage will help. I am now going to plug EBSCO into that handy linksearch tool ... The Interior (Talk) 23:03, 25 November 2013 (UTC)

Looks like 3481 links to ebscohost.com, if I'm doing that right ... The Interior (Talk) 23:09, 25 November 2013 (UTC)

A tip: After doing the first linksearch (using *.ebscohost.com), put &limit=5000 at the end of the URL, then press Enter. That will show up to 5000 hits (the maximum allowed), so you get them all on one page.

Re 2009: That definitely needs the hard disk in the other computer, and that will take a bit of time. Is there a deadline!? I'm not talking weeks—what I mean is, would it be too late for your discussions if I don't get a result within, say, ten days? OTOH, if I get in the mood it might happen more quickly... The old data (before June 2011, which is the first month for which I have a copy of the external links database dump) is based on a somewhat fragile method outlined in the archived results. Extrapolating it back to 2009 would probably be reasonable, but there were only 75 links in articles in January 2010, so I wouldn't think 2009 will show much of interest. Still, if I find the files showing me how to do what I did, it's just a matter of starting it up and drinking coffee while it runs.

I need a confirmation per "what links to search for" above. Please check that those links are correct. Johnuniq (talk) 23:26, 25 November 2013 (UTC)

I've been looking through the credo urls, and they all seem to be www.credoreference.com/something, or with the .corp., or without the www. So your three prior search terms should catch them all as far as I can see. I'll let you know if I see anything different. As far as deadline, we're going to talk again in two weeks, so if we could get something to him before then, it might bear fruit. But I've told him we run on volunteer schedules, so he knows what he's working with. Having these data pages is also useful for future pitches to other providers, I think stats help convince the more reticent. Really appreciate you working on this, I wouldn't know where to start. (and thanks for the page limit tip too, that's useful for all sorts of things!) The Interior (Talk) 23:44, 25 November 2013 (UTC)

OK, I'll start and will {{ping}} you when I have something. Johnuniq (talk) 23:47, 25 November 2013 (UTC)

Cheers, The Interior (Talk) 00:00, 26 November 2013 (UTC)

Thanks - not

For not answering my question on Talk:Rupert Sheldrake and instead making it very difficult for me to understand what the hell was going on. Even your edit summary wasn't helpful when I eventually thought to look. Not a single mention of sockpuppetry on that page about the correct use of talk pages. I'm very tempted to move my question back to where it was in the first place where people can at least see it.

And another thing, why has the Talkpage vanished from my watchlist? Is it something to do with your edit? You are a techie, do you know??

My paws are hurting now from mashing them on this keyboard for the last three minutes.

I've calmed down now.  ;) --Roxy the dog (resonate) 00:35, 26 November 2013 (UTC)

Nothing on Wikipedia is enforceable. So long as an editor smiles (that is, uses reasonably polite edit summaries), they can do what they please, as often as they please, while taking care not to technically breach WP:3RR. That particularly applies to talk pages, and people try to abuse them all the time. I understand that you were asking a genuine question and were not abusing the page, but there is no good way to respond to such a question. I contemplated deleting it and replying on your talk, but I guess laziness gripped me, plus the thought that removing a comment on such a talk page would only provide ammunition to the POV pushers (I don't mean you) when the Sheldrake article inevitably ends up at WP:ANI.

To answer more directly: hiding or removing an off-topic discussion on an article talk page is quite common, but it is not enforceable. At a page like Talk:Barack Obama, trolls are reverted a couple of times per week, and because there are enough good editors at that page, the reversions generally stick. The reasons for hiding or removing off-topic posts vary—the most egregious cases are removed per WP:DENY, and the merely misguided are collapsed. The Sheldrake article has enough trouble without people chatting about whether a sockpuppet should have been indeffed. With so many SPAs active, it is necessary to focus everyone on policy-based improvements to the article, and discussions about other issues must be closed off to avoid the page turning into even more of a WP:NOTFORUM problem. Also, it is quite offensive for an article talk page to be used to question the removal of the sockpuppet of an indeffed user. Such questioning encourages the sockpuppet to continue pushing.

Re your watchlist: No, that's nothing to do with me—I cannot influence that. If you don't see the page at Special:Watchlist, something weird has happened. It is fairly easy to misclick while moving a mouse around, and a plausible explanation is that you accidentally unwatched the page. Other than that, you have seen a very rare bug (I've never heard of that problem). Happy editing! Johnuniq (talk) 01:03, 26 November 2013 (UTC)

Thanks so much for taking the time to respond to my grumpy comments above - I do appreciate it. I'll investigate the watchlist thing, though the mainspace Sheldrake page is still appearing. --Roxy the dog (resonate) 01:13, 26 November 2013 (UTC)

Postscript - The Shelly Talkpage has just come to life on my watchlist again; I thought you might be interested. --Roxy the dog (resonate) 18:31, 26 November 2013 (UTC)

John Prendergast Edits

Hello,

I have observed that many academics look to Wikipedia for bios of upcoming speakers to their universities, not understanding Wikipedia's restrictions, at least in the case of this bio; therefore, missing pertinent information.

If, for instance, you look at the page for Samantha Power, you will find an extensive history of her work and accomplishments. I don't understand why you feel it inappropriate to include more factual information, providing a more complete summary of a person's work, in this case.

I don't think that people naturally assume they should look to another source for more information.

Thank you for any enlightenment you can provide. 69.37.220.244 (talk) 23:14, 26 November 2013 (UTC)

It's edits like this (where an innocuous statement is changed to make yet another mention of this person's name more prominent) that eventually get up the nose of those editors who do not like seeing Wikipedia exploited by POV pushers of any persuasion. I do not know if you are aware of the long history in relation to John Prendergast (activist)—the subject is a very admirable person, and does not need fluffing up. Onlookers were uneasy but tolerant when someone started the drip-drip-drip addition of factoid after marvellous factoid. Eventually, enough-is-enough was declared, and after very time consuming discussions, the user was topic banned. They have not edited using the topic-banned account since that occurred in September 2011. In the last few weeks attention has returned. I am personally happy to overlook a moderate amount of over-icing because Prendergast really is talented and admirable, but "Give an inch and they'll take a mile" comes to mind. I have provided these thoughts in case they are of interest, but I'm not going to spend any time debating the issue here. Any suggestions or discussion regarding what should be in the article should be on the article talk: Talk:John Prendergast (activist).

Some extra thoughts on re-reading your message: The principle used here is that articles are not a substitute for the official website. If a company or person needs every detail mentioned, they can put it on their website. It's quite likely that other articles are abused, abuse which editors have not yet noticed. The principle used here is that "other stuff" is not relevant to what should occur in the article under discussion. We don't rely on precedent. Johnuniq (talk) 01:12, 27 November 2013 (UTC)

Well-stated and I agree. JohnInDC (talk) 03:26, 27 November 2013 (UTC)

John, John, I agree with y'all's edits on the topic of the other John. Dr. John from Montgomery (talk) 06:32, 27 November 2013 (UTC)

We're on a roll. I'm going to revert anyone not called John. Johnuniq (talk) 09:09, 27 November 2013 (UTC)

I think all Johns should be excommunniccatttede from wiki. Johns are horrible people. --Roxy the dog (resonate) 00:06, 28 November 2013 (UTC)

I'm impressed this is still going on. Nice to know some things stay the same ;) The Interior (Talk) 23:58, 3 December 2013 (UTC)

Credo Reference results

@The Interior: I have finished getting the Credo links requested at #A big favour ... above. All the results follow. The values from 2010-01-01 to 2012-07-02 inclusive are a repeat of the earlier results.

All links to credoreference.com/something in articles

Date	Unique articles	Unique links	Total links
2009-01-01	48	51	64
2009-03-01	50	53	66
2009-06-01	55	59	73
2009-09-01	56	59	73
2009-12-01	58	61	75
2010-01-01	58	61	75
2010-03-01	59	62	76
2010-06-01	71	77	93
2010-09-01	76	85	102
2010-12-01	88	100	119
2011-03-01	88	100	119
2011-06-20	129	137	163
2011-07-22	137	145	172
2011-09-01	145	152	180
2012-02-11	159	175	209
2012-03-07	158	175	209
2012-04-03	159	174	207
2012-05-02	160	174	208
2012-06-01	165	178	213
2012-07-02	166	181	215
2012-08-02	168	175	206
2012-09-02	167	174	205
2012-10-01	189	195	229
2012-11-01	214	226	260
2012-12-01	218	232	267
2013-01-02	221	235	270
2013-02-04	224	239	274
2013-03-04	222	238	273
2013-04-03	224	237	274
2013-05-03	224	236	273
2013-06-04	225	237	274
2013-07-08	225	238	274
2013-08-05	225	238	273
2013-09-04	226	238	273
2013-10-01	241	258	293
2013-11-04	248	267	302

Results for 2011-03-01 and before are from a "history" analysis performed by examining old revisions of the articles that contained Credo links in August 2012—any links added then removed before that date would not be detected. Results after 2011-03-01 were extracted from dumps of the external links table at the dates shown above.

The last row shows results from the external links dump for November 4, 2013. There were links in 248 different articles, and there were 267 different links, giving a total of 302 links because some links are used more than once.

As hinted in my last message, the computer I was using is deceased, and I haven't got around to installing the plotting tools that I occasionally use (I started, but there were some issues). I'm hoping that if you want a graph, a request at somewhere like WP:HELPDESK or perhaps WP:VPT would locate someone willing to use the table above for a plot. I will install the tools eventually, but I'm too occupied elsewhere to do that in the near future. Johnuniq (talk) 10:20, 2 December 2013 (UTC)

Johnuniq, thanks so much for taking the time to get these together. I will forward them on to the EBSCO person asap. Graph isn't necessary for the short term. When you get talking to these folks, metrics come up immediately. I realize that the Credo numbers are far from spectacular, but that reflects the general nature of its resources more than anything. Thank you kindly, and I'll ping you when I hear back from EBSCO, The Interior (Talk) 23:52, 3 December 2013 (UTC)

Browning Citori

The reason for the RfC is that one or more IP editors have been updating that article, changing to caption of the photo of Obama shooting skeet, to say that the photo is staged or fake. As I said in my own comment on the RfC, that's not supported by reliable references, and in fact is another Obama-related conspiracy theory that doesn't make any sense. But I was the only one reverting those edits, so I thought it would be helpful to get other editors to join the discussion, and maybe even help me patrol the article (which so far seems to be going well). — Mudwater ^(Talk) 01:40, 6 December 2013 (UTC)

Groan, alright I'll add Browning Citori to my watchlist. Thanks for the explanation. I suppose an RfC was a technique to get rational editors to look at the article, although a request for help at WP:BLPN or Talk:Barack Obama would be ok as we don't worry about a little canvassing when combating birthers. Johnuniq (talk) 02:21, 6 December 2013 (UTC)

Rational editors, forsooth! Your other ideas make sense too. And I might also request some level of edit protection for that article. Anyway, thanks! — Mudwater ^(Talk) 02:25, 6 December 2013 (UTC)

The Wikipedia Library Survey

As a subscriber to one of The Wikipedia Library's programs, we'd like to hear your thoughts about future donations and project activities in this brief survey. Thanks and cheers, Ocaasi ^{t | c} 15:13, 9 December 2013 (UTC)

Convert module is live

Ok, you're on the hook. Hope everything goes well!

—Trappist the monk (talk) 02:18, 11 December 2013 (UTC)

Yikes, should be interesting. Thanks. Johnuniq (talk) 02:20, 11 December 2013 (UTC)