Wikipedia

User talk:Δ/20110901

< User talk:Δ
Latest comment: 13 years ago by Hammersoft in topic A few comments


Another fix ref error

Latest comment: 13 years ago4 comments2 people in discussion

At Roman Opałka it gave the following error:

[code]

<type 'exceptions.ValueError'> Python 2.7.1: /usr/bin/python
Sun Aug 28 15:09:01 2011

A problem occurred in a Python script. Here is the sequence of function calls leading up to the error, in the order they occurred.

in '()
34 if __name__ == '__main__':
35 try:
=> 36 main()
37 finally:
38 pass
main = <function main>
in main()
28 page = wikipedia.Page(site,title)
29 text = page.get()
=> 30 text = fix_refs(text)
31 f = websave.put(page,text,'Fix references',cgi=True)
32 print(f.encode('UTF8'))
bgcolour = #6495ED\n| nam...:Roman Opalka]]\n[[sv:Roman Opa\u0142ka]]\n[[zh:\u7f85\u66fc\xb7\u6b50\u5e15\u5361]]', global fix_refs = <function fix_refs>
in fix_refs(text=u'{{Infobox artist\n| bgcolour = #6495ED\n| nam...:Roman Opalka]]\n[[sv:Roman Opa\u0142ka]]\n[[zh:\u7f85\u66fc\xb7\u6b50\u5e15\u5361]]')
1234 else:
1235 text += refs
=> 1236 text1,text2 = text.split('reflist')
<font color="#909090"><tt> <small> 1237</small> text2 = re.sub(re.compile(u'\n(\*(\s+)?)?<ref name=\"(.*?)\"/>',re.I),'\n',text2)<br /></tt></font>
1238 text = text1+u'reflist'+text2
bgcolour = #6495ED\n| nam...:Roman Opalka]]\n[[sv:Roman Opa\u0142ka]]\n[[zh:\u7f85\u66fc\xb7\u6b50\u5e15\u5361]]', text.split = <built-in method split of unicode object>

<type 'exceptions.ValueError'>: need more than 1 value to unpack args = ('need more than 1 value to unpack',) message = 'need more than 1 value to unpack'

[/code]

--Piotr Konieczny aka Prokonsul Piotrus| talk to me 15:11, 28 August 2011 (UTC)

thanks, looking now. ΔT The only constant
Can you give me a link to the offending revision that you attempted this on?. ΔT The only constant 18:28, 30 August 2011 (UTC)
Hmmm. Either this or one of the latter ones. --Piotr Konieczny aka Prokonsul Piotrus| talk to me 19:14, 1 September 2011 (UTC)
Unable to reproduce, so I think its fixed. ΔT The only constant 17:56, 2 September 2011 (UTC)

yet ref script error

Latest comment: 13 years ago7 comments2 people in discussion

As I run the script through a dozen or so larger articles, here are the errors (obviously, I self-rv after them): [1], [2] look similar to the past one; in [3] the bot stops looking very quickly? --Piotr Konieczny aka Prokonsul Piotrus| talk to me 17:32, 2 September 2011 (UTC)

All three of those cases have been fixed, its just situations where {{reflist}} is formatted in a unknown format for the script it just drops it at the bottom of the article. If you preview it before saving you can catch that, and just move it to the correct section. 17:42, 2 September 2011 (UTC)
Would it be possible to teach the script to recognize those formats and fix them if necessary? In the end, we want such tools to be as one-click as possible. Oh, and at History of Poland (1945–1989) I got a weird error: fix ref returned no changes, and cleanup, this:
--Piotr Konieczny aka Prokonsul Piotrus| talk to me 17:48, 2 September 2011 (UTC)
Looks like the webservers for webcite are having issues, Ive added code to handle this. As for teaching the script, thats what we are doing here :) ΔT The only constant 18:14, 2 September 2011 (UTC)
Troy H. Middleton is having the regular issue of the script not recognizing where the references are. "|3" ? Cannot it just find the text reflist, then modify it and leave whatever comes after, well, after? --Piotr Konieczny aka Prokonsul Piotrus| talk to me 04:40, 3 September 2011 (UTC)
Fixed, the issue is template parameter order matters, thus its not that simple. ΔT The only constant 10:41, 3 September 2011 (UTC)
You're the boss on that. If we are building an exception list, here's a new one: Smolensk War ({{reflist|colwidth=35em}}). Maybe we could just add all possible reflist parameters as an exception? --Piotr Konieczny aka Prokonsul Piotrus| talk to me 03:08, 4 September 2011 (UTC)
I just did a minor re-write of that part of the code, it should cut down on the number of these cases that we see. ΔT The only constant 03:31, 4 September 2011 (UTC)

Deltabot suggestion

Latest comment: 13 years ago3 comments2 people in discussion

Hey. I've noticed that when Deltabot updates the sandbox, it marks cases that have {{SPI case status|open}} as being 'unknown'. The case status template seems to handle a tag of open (in that it defaults to showing the open message) but would it be possible to get the bot to handle the open tag as well? — HelloAnnyong (say whaaat?!) 13:24, 5 September 2011 (UTC)

already done, people just need to tell me when they change things. The bot is not a smart AI, it cannot learn by itself. ΔT The only constant 21:02, 5 September 2011 (UTC)
No worries. The issue really only comes up when a non-SPI clerk reopens a case. But thanks for taking care of it. — HelloAnnyong (say whaaat?!) 22:59, 5 September 2011 (UTC)

The Signpost: 05 September 2011

Latest comment: 13 years ago1 comment1 person in discussion

American Boy Scouts

Latest comment: 13 years ago3 comments2 people in discussion

File:United Nations Fight for Freedom colored, white and Chinese Boy Scouts in front of Capitol, They help out by delivering poster to help the war effort (LOC) (2179085559).jpg is obviously an image of Boy Scouts from the BSA. The ABS was defunct well before 1941. ---— Gadget850 (Ed) talk 10:14, 6 September 2011 (UTC)

Sorry about that, was a little too tired and should have paid closer attention. ΔT The only constant 14:34, 6 September 2011 (UTC)
  ---— Gadget850 (Ed) talk 15:26, 6 September 2011 (UTC)

script testing update

Latest comment: 13 years ago9 comments2 people in discussion

The rewrite seems good, no more errors of the previous type returned so far. One problem - at [4], the script didn't move most references, only one. I wonder why? Also, I'd suggest changing the edit summary from "Fix references" to "Fix and organize references" or such, and perhaps link it to a doc page, so people who like it could start using it as well. I know it is still in testing, but the more people use it, the quicker we will weed out all the bugs. --Piotr Konieczny aka Prokonsul Piotrus| talk to me 17:09, 7 September 2011 (UTC)

Ive fixed the script, it as a bug related to spacing around the = sign and it not seeing it as a ref. (yeah its a stupid mistake). If you want to get a jump on the doc page User:Δ/Reffix is where it will end up. ΔT The only constant 21:07, 7 September 2011 (UTC)
I may. In the meantime, here's another color coded error report (from eBay stable rev link): --Piotr Konieczny aka Prokonsul Piotrus| talk to me 21:35, 7 September 2011 (UTC)


And here is the colorful error from JSTOR:--Piotr Konieczny aka Prokonsul Piotrus| talk to me 01:37, 8 September 2011 (UTC)

there is no need to post full traceback, just the article and revision the error occurred on. Both cases where related, and caused by a typo in a regex. Its been fixed. ΔT The only constant 03:08, 8 September 2011 (UTC)
All right. My recent observation is that the script is still missing a bunch of references that don't have a name. First, cite web seems not touched (ex. History of Poland (1989–present)): Second, ones without a cite template (ex. History of Poland (1945–1989)); I'd suggest either using the first word (but without {{), or something automatic, like reffix#. In cases where a google book link is present without a citation, could the script could try to run http://reftag.appspot.com perhaps? --Piotr Konieczny aka Prokonsul Piotrus| talk to me 17:25, 8 September 2011 (UTC)
Just took a look at the code and fixed/re-implemented a feature that I thought was already live. Due to my current location I cannot update the live script but I should be able to get that done in about 5 hours. It tries several guesses and then uses the longest word in the citation. As for your second idea, I looked into that, however its formatted using javascript in the user's browser, which makes using it in a script useless. ΔT The only constant 18:34, 8 September 2011 (UTC)
Ah. In that case, could it somehow notify the user(s) that reftag app could be used? Could drop a message on article's talk page, user's talk page, or add a template to the article, asking for the links to be transformed through the applet (if you prefer the latter one, I can draft such a template). --Piotr Konieczny aka Prokonsul Piotrus| talk to me 18:40, 8 September 2011 (UTC)
Ive sent request to the author, and take a look at [5] warning BIG diff, I think thats exactly what you are looking for. ΔT The only constant 20:08, 8 September 2011 (UTC)

The American Civil War Barnstar

Latest comment: 13 years ago4 comments2 people in discussion
  The American Civil War Barnstar
For your outstanding efforts at editing Arkansas Civil War Regimental Histories. Thanks! Aleutian06 (talk) 13:24, 7 September 2011 (UTC)

.

Uh, which pages would you be referring to?...... ΔT The only constant 21:09, 7 September 2011 (UTC)
On 6 September 2011 you cleaned up the following articles. There are about 44 more Regimental Histories on List of Arkansas Civil War Confederate units which could use your attention if you have time.


  Done I went ahead and cleaned up all pages linked from that page. ΔT The only constant 12:13, 9 September 2011 (UTC)
Outstanding! Thanks. Aleutian06 (talk) 17:54, 9 September 2011 (UTC)

1st Arkansas 30 Day Volunteer Regiment - image parameter

Latest comment: 13 years ago3 comments2 people in discussion

This is a fairly inconsequential point, but Template:Infobox Military Unit specifies "The image must be given in the form [[Image:Example.jpg|300px]]". Was there a reason you changed Image: to File: here? Nick Number (talk) 04:01, 9 September 2011 (UTC)

Image is the old out dated markup, Its now called file. ΔT The only constant 10:14, 9 September 2011 (UTC)
Ok, thanks for editing the template docs to reflect that. Nick Number (talk) 17:20, 9 September 2011 (UTC)

There is a mop reserved in your name

Latest comment: 13 years ago3 comments2 people in discussion
  I have observed some remarkable contributions from this account. I am curious, why are you not an administrator. Pardon that you have struck me as the kind of editor who could be a good one. and that you seem qualified by a cursory review. You exemplify the essence of an Administrator without tools! I hope you will consider serving in the fuller capacity.

My76Strat (talk) 04:18, 11 September 2011 (UTC)

  • With all respect and collegiality My76Strat, but this is absolutely hysterical. There is considerable history behind when Δ was an administrator, why he is not now, and the never ending group of people who want Δ's head on a platter. The compliment, regardless of lack of knowledge of the history, is wonderful. I'm sure Δ is pleased. --Hammersoft (talk) 16:46, 11 September 2011 (UTC)
    No one was in any danger. I know many would see this as a reason to mock. Δ helped me develop a useful script related to this essay, and the associated intent. It was added here as a gesture of thanks, and to link him to the essay, because I personally believe he is honorable. He knows, and I agree that a snowball has a longer life expectancy in hell, than he has at adminship. But the faults are by no means entirely his. My76Strat (talk) 04:32, 12 September 2011 (UTC)

The Signpost: 12 September 2011

Latest comment: 13 years ago1 comment1 person in discussion

Why moving references?

Latest comment: 13 years ago9 comments5 people in discussion

Hi. I noticed you have been moving references and notes out of the body of the text and piling them into a list at the bottom. May I ask why? It makes pages much more difficult to edit. Walrasiad (talk) 16:28, 9 September 2011 (UTC)

Its what is known as List Defined References, and it actually makes it a lot easier to edit the prose of the text for those who are not familiar with wiki-markup, and for many others. ΔT The only constant 16:33, 9 September 2011 (UTC)

And makes it inordinately more difficult for anybody else. It is much more difficult to edit the content of a page if you can't see what's in the footnotes at the appropriate place. And if editors don't read them, footnotes are liable to get pushed around to the wrong places. I really can't see the advantage. The wiki mark-up on footnotes are little different from the wiki-mark-up of the list notes you're making - indeed, the latter is more uninintuitive for those unfamiliar with mark-up because something like < ref = "googleconsentia86"/> in the middle of a phrase makes a lot less sense to a newcomer (and maybe more apt to be inadvertently deleted as gibberish) than < ref>Cite of book and its page</ref>, which anybody can quickly catch on is a footnote with a few seconds of reflection. Walrasiad (talk) 16:46, 9 September 2011 (UTC)

Actually its <ref name="nameofreference"/> which most people see as a valid reference, and is a fairly common format, take a look at History of Poland (1939–1945) ΔT The only constant 16:58, 9 September 2011 (UTC)

Oh, I know that. But a person who "is not familiar with wiki-mark-up" (the person you're suggesting would benefit from this) would not know that. List-defined references are of little or no benefit to newcomers, and a tremendous hassle for experienced editors. The only case I see for them is if the same reference is repeated multiple times (and even then with trepidation). It is just a lot of pain for no gain. Walrasiad (talk) 19:38, 9 September 2011 (UTC)

I have to agree that it is very unnecessary to move the references like that. There are a bunch of pros and cons, and it ends up being better to stay how 99% of Wikipedia articles do it. Like Walrasiad said, this makes the references not in the body of text, and then they can get pushed around and their area undefined. It sounds like an overall bad idea. Feel free to do it on your own articles, but please don't go around doing it to random ones. If you got a general consensus to do it to a certain area of articles, that is great, but I haven't heard of such thing. Blake (Talk·Edits) 03:10, 12 September 2011 (UTC)

Weather or not an article uses LDR's the chances of a reference being moved incorrectly is the same. My recent edits where to address a major issue with references and their numbering order. If you have exact questions you can poke My76Strat. ΔT The only constant 03:19, 12 September 2011 (UTC)
Yes I asked Δ to help develop a tool capable of doing these things. I also advised that it wasn't an issue we intended to force on an editor who preferred otherwise. It comes down to a style preference, and deference is given to the contributions of earlier attribute. In accord with WP:BRD, there is nothing wrong with being bold. And there is nothing wrong with reverting and discussing. This is how we build the encyclopedia. Δ holds a 1RR line with this tool, and I was interested in using it as a tool for review, because it identifies link rot, dead links, and reference strings that are out of chronology. It is valuable to me as a reviewer, but I don't save the changes. Only cite in the review that problems exist. I let the contributors fix it themselves, according to their desire. My76Strat (talk) 04:52, 12 September 2011 (UTC)
No, Blake, what 99% of wiki articles does is not always correct. Once, none had inline references, and I recall people saying we should not use them, because no articles use them, and reverting them because they are not "common". Moving refs to the bottom accomplishes one major goal: it reduces the amount of code in plain text, and makes the articles more friendly to edit to n00bies. As somebody who has introduced hundreds of new editors to Wikipedia, I feel confident in saying that this articles with refs at the bottom are significantly more n00b friendly, more likely to increase their enjoyment of editing and transform them into editors (rather than scare them). --Piotr Konieczny aka Prokonsul Piotrus| talk to me 17:34, 14 September 2011 (UTC)

Script issue (minor)

Latest comment: 13 years ago2 comments2 people in discussion

One thing that I've identified: the script should arranges the refs by alphabet, but only the ones it moves, it does not mix them with older ones. See this (N, W, W, W followed by an OLDER A). Other than that, no bugs for the past few days, we are getting close to a stable version (at least, with the current features). PS. I appreciate your attention to all the points I raise, some others are not so considerate :( --Piotr Konieczny aka Prokonsul Piotrus| talk to me 17:37, 14 September 2011 (UTC)

I'll take a look. is a test edit for an update to the ref tool to convert URLs to cite book's for google links. (I think everything went well can you give it a second set of eyes?) When Im sure that, that edit is success full, Ill update the live script, so please take a little more care in double checking the edits for a while. ΔT The only constant 18:36, 15 September 2011 (UTC)

Time zones

Latest comment: 13 years ago2 comments1 person in discussion

Hi there! Could you, please, elaborate on this? I didn't notice anything broken (and I did specifically check this morning). Thanks!—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); September 15, 2011; 21:09 (UTC)

P.S. Sorry about the edit summary. It's that damn Firefox trying to be too smart and offering a selection of old edit summaries to choose from :)—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); September 15, 2011; 21:10 (UTC)
it added 500+ pages to a tracking category and made them spew a huge red error message Help:Cite errors/Cite error refs without references ΔT The only constant 21:11, 15 September 2011 (UTC)
Hmm, didn't see any red error messages. Perhaps it's only on some pages? Could you let me know which one you saw it at, so I could investigate? Thanks again.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); September 15, 2011; 21:13 (UTC)
It spewed it on every article without a {{reflist}} or equivalent. ΔT The only constant 21:14, 15 September 2011 (UTC)
Ah, got it. There's nothing wrong with the template per se, though. The only way to fix the error is by actually going through those five hundred pages and adding a ref section to them (which is the very purpose of that tracking cat anyway). Do you perchance know a bot that can do it? If not, I'll restore the ref tomorrow and will go through the articles in the tracking cat myself. Cheers,—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); September 15, 2011; 21:18 (UTC)
Latest comment: 13 years ago2 comments2 people in discussion

File:Mesut Özil.jpg appears to be a copyrighted image and it was uploaded to a Russian site. --Walter Görlitz (talk) 05:14, 17 September 2011 (UTC)

It appears to be legit, the Russian site appears to have the original image and per an OTRS ticket everything looks like its not a copyvio. ΔT The only constant 13:48, 17 September 2011 (UTC)

The Signpost: 19 September 2011

Latest comment: 13 years ago1 comment1 person in discussion
Latest comment: 13 years ago6 comments3 people in discussion

I'm curious, does Google Books allow you to preview all the links you have added here? Many don't even have snippet view for me. So the links don't really help with the verification more than the ISBN alone does, from where I sit. (I have actual copies.) Have mörser, will travel (talk) 05:36, 21 September 2011 (UTC)

Also, I find dubious the exact date (day, month) that Google somehow conjures for book publication. The year is usually sufficient, and often the only such information in the book itself. Have mörser, will travel (talk) 05:38, 21 September 2011 (UTC)

It does not let me preview all of them, however I also fill in and standardize book references at the same time. Most of the publication dates are that of the initial release date, with a little bit of leg work is fairly easy to verify. ΔT The only constant 10:49, 21 September 2011 (UTC)
Release dates are not publication dates, change the bot to not give them; only the year of publication is cited in ALL referencing styles. FWiW Bzuk (talk) 13:04, 21 September 2011 (UTC).
Actually you are wrong. The cite template has a date parameter exactly for that reason. And published date is the date released, and refers to the date that the books where issued/first permitted to be sold. ΔT The only constant 13:16, 21 September 2011 (UTC)
No, actual year of publication is all that is required. FWiW Bzuk (talk) 16:07, 21 September 2011 (UTC).

Changing bibliographic information

Latest comment: 13 years ago3 comments3 people in discussion

Please do not change correctly formatted references into templates. FWiW Bzuk (talk) 13:02, 21 September 2011 (UTC).

I think standardized references is a good thing, and templates make that very easy. ΔT The only constant 13:12, 21 September 2011 (UTC)
Just as a caution (not in any way to be meant as administrative action caution, but more to avoiding anything down the road), changing from formatted references to templated references is really really frowned on (to the point of potential large scale edit wars), except when it is to create a consistency in reference format within the article. Particularly if this is even close to a semi-automated task. While template versions are nicer, it is first editor's choice and/or page consensus that determines the preferred format. See WP:CITEVAR for what's considered acceptable. --MASEM (t) 13:22, 21 September 2011 (UTC)

sockpuppet editing

Latest comment: 13 years ago1 comment1 person in discussion

There is an open WP:SPI case looking at sockpuppet editing primarily on the Johann Hari/ Talk page. As you edited the Johann Hari/Talk page between 2004 and 2011, your input is welcomed. Yonmei (talk) 19:18, 23 September 2011 (UTC)

Script error

Latest comment: 13 years ago4 comments2 people in discussion

Those are rare these days, but here is one. I guess some weird ref formatting, probably due to the split of notes from ref in a rare way? --Piotr Konieczny aka Prokonsul Piotrus| talk to me 20:26, 22 September 2011 (UTC)

Actually it was a typo in my code, I didnt have the proper code for an already LDR with the new variable column size code. but its been fixed. ΔT The only constant 22:45, 22 September 2011 (UTC)
It works now, but I see it still cannot order old and new refs alphabetically (it just adds the new ones before the old ones). It is a bit annoying, seeing as how we want to "organize" the refs, per script summary... --Piotr Konieczny aka Prokonsul Piotrus| talk to me 00:58, 24 September 2011 (UTC)
its something Im looking into, just really havent had enough time recently to really take a look at how to address the issue. ΔT The only constant 00:59, 24 September 2011 (UTC)

AN notice

Notice of noticeboard discussion

Latest comment: 13 years ago1 comment1 person in discussion

  There is currently a discussion at Wikipedia:Administrators' noticeboard regarding an issue with which you may have been involved. The thread is Δ (Betacommand) and community restrictions. Thank you.. Peachey88 (T · C) 07:15, 25 September 2011 (UTC)

Clean up at Paul Kelly (musician)

Latest comment: 13 years ago1 comment1 person in discussion

I reverted your recent 'Cleanup' at Paul Kelly: some of these changes disrupted ref templates, added unnecessary urls to Google books (no preview is available), added non-functional archiveurls (I archived those pages but did not use these as being unsatisfactory for verification of page content), adding dead link 'plates for (now) working links or incorrectly formatted refs (mostly dates, where dmy is required but mdy or ISO were supplied). I've subsequently gone through your changes and have re-added/modified the ones that I believe are appropriate.shaidar cuebiyar (talk) 00:43, 26 September 2011 (UTC)

The Signpost: 26 September 2011

Latest comment: 13 years ago1 comment1 person in discussion


Cleanup problems

Latest comment: 13 years ago14 comments5 people in discussion

In your cleanup script, besides a lot of helpful changes, you seem to make a lot of unnecessary ones as well (purely inconsequential ones which shouldn't be made as they seem to be imposing one accepted style over another for no benefit or reason), e.g.

  1. moving the "s" of plurals inside piped wikilinks instead of leaving them outside, e.g. here, changing "[[church (building)|church]]es to [[church (building)|churches]]
  2. changing piped section links to redirect links (i.e. changing a direct link to an indirect one, e.g. "American football positions#Offense|offensive line" to "offensive line")
  3. adding "" to named refs even when it isn't needed, e.g. here, changing <ref name=xinhua> to <ref name="xinhua">
  4. changing refs with one line per parameter to refs with the parameters one after another (which uses less space but is also less easily readable)
  5. changing parameter capitalization for no good reason (e.g. here you change "Date" to "date", even though the template uses "Date" with the capital).
  6. Adding extra blank lines before the defaultsort, e.g. here where you increase this from one to three lines

Any reason for these changes?

Edits that actually makes the articles worse are:

  1. the addition of the publisher and title to bare urls; even if the same title and/or publisher were added to the reference but outside of the url. After your change, the ref now has the same info twice, once as a link and immediately afterwards as plain text, e.g. here you change * [http://www.facebook.com/pages/Droban-Apherna/140585052673905] Droban-Apherna's Facebook to * [http://www.facebook.com/pages/Droban-Apherna/140585052673905 Droban-Apherna | Facebook] Droban-Apherna's Facebook
  2. Here you add a defaultsort, which has the result that all four cats sort this article in the same way, instead of the deliberate difference that originally existed between the first two cats and the last two ones.
  3. In this cleanup, you change a ref to a magazine, which links the magazine name, has an ISSN, and a quote, to a "cite book" template which loses the ISSN and the quote
  4. Here, you change the year in a number of citebooks to an exact date. I couldn't find any indication of where you got that date from.
  5. In the same change [6] you change google.books.ca links to google.books.com links, and at the same time change the link from leading to the actual page that is referenced to the general page about that book, thereby decreasing the actual usefulness of the link

Please correct me if any of these actually are improvements or have a consensus for them, and please remove the other ones from your cleanup script. Fram (talk) 07:44, 27 September 2011 (UTC)

(I went ahead and numbered your questions for easier responding, I hope you don't mind.) 1&2 are just there to make wikimarkup simpler. If the links piped there is no reason for text to be outside the brackets (as its considered part of the link anyway) If there is much opposition to 2 that can be stopped, however I dont see any good reason to leave the complex links that I fix in 1. 3 is just done out of standardization and correct markup. the correct markup includes ". 4 is done because I rebuild the citebook template with a combination of data from the page and from google books, (which addresses your other number 4) in order to fill in and correct information possibly left out of the original citation. 5 & 6 are just minor errors, (6 was just recently introduced and subsequently fixed). (crap this is taking longer than I thought to answer these questions and I have to run Ill finish up when I get a spare minute.) ΔT The only constant 09:40, 27 September 2011 (UTC)
No problem on the numbering, of course. Issue 2 gives strange results for the unexpecting reader, who sees that mosques isn't changed, but the one for churches is, for no apparent reason. It doesn't help anything, doesn't make things more standardized, and only adds clutter to already lenghty diffs. I wouldn't support the reverse change either (for none of the 6 first items in fact), they are just things better left alone IMO. For issue 3: WP:NAMEDREFS states that "The quotes are optional unless the name or group includes a space[...]". If something is optional, it shouldn't be added; both versions are correct markup. Finally, for issue 4, I don't see how this is a relpy to my "second" issue 4 (the error).
In general, if you could restrict your cleanup to things that are truly improvements (e.g. the addition of "deadlink"), and the removal of truly outdated stuff (like replacing "image" with "file"), then you have less risk of errors (like the issue 6 which you already corrected, thanks), and people looking at the diffs will more clearly see what the "cleanup" actually does, instead of a bunch of irrelevant stuff that didn't clean anything up.
No problem with waiting for a reply for the second bunch of issues of course, real life is just slightly more important than Wikipedia :-) Fram (talk) 10:00, 27 September 2011 (UTC)
"4) in order to fill in and correct information possibly left out of the original citation": I think he gets that info from Google metadata, which is of questionable accuracy and value. It was pointed out before on this page (now archived) that publication year is usually the only information in mainstream citation styles for books. Have mörser, will travel (talk) 12:10, 27 September 2011 (UTC)
I actually use google books API. As for your second set of issues, I just grab the contents of the <title> of said webpage and use that as the text, it works well 99% of the time. Numbers 3 and 5 Im looking into. ΔT The only constant 15:57, 27 September 2011 (UTC)
#3 fixed ΔT The only constant 16:14, 27 September 2011 (UTC)
#5 has been fixed. Anything else? ΔT The only constant 17:50, 27 September 2011 (UTC)
Um, yes. If 3 and 5 have been fixed, 1, 2, and 4 are still open. If there is doubt whether Google Books API is correct, it would be better to leave it out (e.g. here you change a 2001 year for a book to a March 1975 date, even though the actual text of the book says "original printing 2001", and the book references things happened after 1975 and the preface was finished in 2000: if this comes from Google API, please don't use it anymore). I doubt that for issue 1, it works well 99% of the time.

New issues:

  1. You change "Cite web" to "cite web", but "citenews" to "Cite news". Wouldn't it be better to let the capitalization be, or else at least to change it consistently?[7]
  2. You change the capitalisation of a book title here
  3. Here you add a date to a book, remove the year, and let the month stay
  4. In the same edit, same ref, you add "author" to a cite book template which already has first, last and authorlink. Note that "author" is tagged as "deprecated" in the cite book template, so adding it is not productive (it also changes nothing for the reader, only adds clutter to editors). You add "author" in quite a few places on other pages as well.
  5. here you removed two files, instead of noticing that they were misguided attempts to create wikilinks to e.g. Pálinka.
  6. Here you change "notes" to "references". Both are equally accepted, and you shouldn't be imposing your preference over that of others.
  7. Here you change identical refs (battles) to two different ones, and add the unnecessary quotes around the ref name (as discussed above). Again an example of unnecessarily imposing your own preference over other accepted methods.
  8. Here you change two different books to the same citebook template.
  9. Here you simply remove three references (two to Wikipedia, one to Commons) instead of converting them to Wikilinks.

There may be more, I stopped looking for more.

You seem to be going again on the path where you do a lot of semi-automated work, which includes a lot of unnecessary work, and don't check it thoroughly, which allows errors to get in repeatedly. We all know how this ended last time, please be more careful this time around. It's better to do less but to do it correctly, than to do a lot but with too many problems. Fram (talk) 08:22, 28 September 2011 (UTC)

2,3,4,8 are now moot, Ive stopped using Google books API. 1 has been fixed. 5 was just human error, I normally catch those, it didn't register as an attempt at wikilinks. 7 take a closer look those two refs have the same name but different content, yes its similar but still different which is why it was renamed. 9 yes I removed self references because links to Wikipedia and Commons do not pass WP:RS and thus where removed. ΔT The only constant 15:25, 28 September 2011 (UTC)
7 is the same for a human, but different for a bot. Please edit as a human. Fram (talk) 17:12, 28 September 2011 (UTC)
No its not, it may be for those who are partially blind or are vision impaired it may be. However to the rest of us its different content. If two refs have different content they shouldnt have the same name. ΔT The only constant 19:51, 29 September 2011 (UTC)
Actually, the idea that these kind of issues might be solved by implementing a simple CAPTCHA and time throttle in MediaWiki is far from unattractive. --Tristessa (talk) 20:39, 28 September 2011 (UTC)

A few comments

Latest comment: 13 years ago23 comments3 people in discussion

Delta, I am sorry to be so blunt with you, but you are persisting in driving everyone round the bend with your "semi-automated" editing, no matter how good your intentions.

The time spent by other Wikipedia editors watching you and correcting your mistakes is far in excess of what it would have taken for you to have made well-considered and properly reviewed edits to the articles in the first place. Worse is your proclivity to utterly ignore the fact that the community has not given you consensus for many of the en masse changes you have a penchant for making, and also seem determined to ignore that it has imposed sanctions on you.

Until such a time that mass changes are approved by a community discussion after proposal, you don't have a mandate from the Wikipedia community for making arbitrary changes to multiple articles using automated editing that have not received Wikipedia consensus. In some cases, your edits (amongst helpful ones) go explicitly against prevailing consensus, such as your mechanical-looking use of Google Books API data in rewriting references. I have no doubt that you are a well-meaning editor trying to do useful work for the encyclopaedia, but being told "no" by the community is not the same as being told "yes". I really don't know why you seem determined to ignore what the Wikipedia community thinks of your behaviour, or why you wish to edit right up to the margin of the boundaries it has set or ignore the processes you are meant to follow to gather input on whether changes are a good idea; I have no idea why you feel so inclined to act against the wishes of many editors in this way.

So, I'm going to plead with you. Please, please consider making only the changes to articles using your cleanup script that are absolutely necessary and have unambiguous consensus; your preferences are not relevant. Consider also, please, taking more time over each article you are running through the script. I want you to be able to continue to be the productive editor you are, rather than sliding back into this mode of getting everyone's backs up, but you make it very hard. If you carry on -- and this isn't a threat -- I feel sure that ultimately the community will actually make good on giving you no more chances. You are editing Wikipedia as part of the community, not as yourself. --Tristessa (talk) 09:15, 28 September 2011 (UTC)

  • Your analysis of Δ's "proclivity" and supposed ignorance of sanctions is false. Look at your own text. "Other Wikipedia editors watching". Surely you know at least some of the history here. With so many people watching what Δ does, if he violated the sanctions he would be blocked. Yet, of late, he hasn't been. That's because he IS paying attention to his sanctions. Δ is not restricted from making the changes he is making. There is no sanction in place that requires he get permission to make edits. Patterns of edits where a large number of articles are being affected, yes. Not edits. You are attempting to apply a sanction that doesn't exist; that he must seek approval for his editing here. He doesn't. Full stop. As for him acting against the wishes of many editors, Δ is in a virtually impossible position. Were I in his shoes, I would have gone bezerk a long time ago. If I hadn't, I certainly would be editing right up to the boundaries imposed on me by the community. What you're asking for is for him to be even more self restrictive than his sanctions call for. Yet, your demands are arbitrary.
  • If you have specific grievances with regards to edits Δ is making, then lay them out. You will note from the section above that when such specifics are provided, he is quite amenable to responding and adjusting as deemed needed. But, arbitrary "Won't you just be better?" requests rightfully get the /dev/null treatment, and well they should. --Hammersoft (talk) 13:09, 28 September 2011 (UTC)
Hammersoft: You write as though I am here about "specific grievances" of mine and no applicable sanction is relevant to his recent editing -- but there is a sanction that clearly and unequivocally requires he seek approval for editing of a kind he has recently returned to, (see his entry on WP:RESTRICT). I am in here in an attempt to help him call this situation of antagonistic editing to an end, as his recent editing falls under its criteria:
[...] Before undertaking any pattern of edits (such as a single task carried out on multiple pages) that affects more than 25 pages, Betacommand must propose the task on WP:VPR and wait at least 24 hours for community discussion. If there is any opposition, Betacommand must wait for a consensus supporting the request before he may begin. [...]
Either there's something I'm missing here, or you're simply wikilawyering arguing semantics over the meaning of the word "pattern" and the other wording of the sanction: he has recently applied a series of rapid edits to more than 25 articles involving changes that are obviously automatically generated even if they require interactive approval -- hence the various issues noted above with the unreviewed and blanket use of Google Books API data, amongst various other regex personal-preference changes to article syntax and attempts to rewrite external links in certain ways. Indeed there appears to be some level of agreement at AN/I that he has in fact violated his restrictions, but that blocking would be inappropriate because of a lack of demonstrated harm. I'm certainly not here with the intent of seeing him blocked and am an uninvolved admin who has never participated in the prior discussions, but I have reviewed them. Whilst he may well be amenable to specific technical concerns as you note above, dialogue and cooperation with other members of the Wikipedia community is more crucial, areas in which he appears to deliberately fail when I believe he could do better. You do him no service to encourage him to ignore dialogue from other users, as he will simply push the matter up the WP:DR chain by doing so -- no matter how much it is lawyered argued, he is exhausting the patience of a large segment of the community. --Tristessa (talk) 17:07, 28 September 2011 (UTC) Corrected. --Tristessa (talk) 18:07, 28 September 2011 (UTC)
  • Either I'm missing something, or there is no evidence being provided by you or by the recent thread at WP:AN that there is, in fact, a pattern going on here. I've not encouraged him to ignore the comments of others in any respect. I do not appreciate you labeling my comments as wikilawyering, or attributing actions on my part that never took place. I'm actually providing him considerable service in beating back the hordes that want his head on a platter. Either specific evidence needs to be supplied, or as User:Georgewilliamherbert noted, the matter should be closed. Your demands for him to do something are arbitrary; you're making no specific request, basing it on no specific diffs, and asking for nebulous change on his part. You should not be surprised at the result of such a request. Please take a cue from the "Cleanup problems" section above, as I previously noted. Thank you, --Hammersoft (talk) 17:41, 28 September 2011 (UTC)
And I don't want his head on a platter. That has never been my intent. As I said, I'm an uninvolved admin; I am not in a position to be providing evidence on the AN/I thread against him at all and I don't especially want to. I think it's quite clear that a "pattern" exists in his recent contribs performing a task, namely his general "cleanup" scripts, across over 25 articles without requesting on WP:VPR and hence being in violation of his restriction. I don't think you've really addressed my points, and I think you know I intended "wikilawyering" specifically with reference to your interpretation of the sanction and not as a general denigration of your comments. --Tristessa (talk) 17:59, 28 September 2011 (UTC)
  • I don't know that, though I accept your explanation. "Wikilawyering" tends to be a word that sets people off, me included. It shouldn't be used except when necessary to describe ill intent. As to the pattern, I took a look at 10 edits either side of the one diff provided by the person who brought the complaint. I found no pattern. Unless, you want to indicate a pattern being changing "C" to "c". If we're to make a mountain out of that hill of toothpicks, we might as well just ban him from the project. Of course, nothing would surprise me now after I was taken to task for removing an errant 't' and told I was in violation of wiki rules for doing so. --Hammersoft (talk) 19:47, 28 September 2011 (UTC)
To me, looking at Special:Contributions/Δ of the changes made over the last couple of days shows unanimously that 1) he is using an automated editing tool; that 2) he is making a specific pattern of edits; 3) these are applied to a range of over 25 articles (the bar in the original sanction); and 4) he is doing this without requesting consensus on WP:VPR. Ignore whatever diff was provided in the original AN/I complaint if it helps; just glance through the contributions from top to bottom, and you can see the pattern -- this is not an exhaustive list, but just first impressions of what one sees; the script does this:
  • Removes any HTML comment tags within articles -- this tends to only result in removing annotation comments in infoboxes that were from the original substituted templates; this is not a consensus based change
  • Capitalises citation templates from "cite" to "Cite", a personal preference change
  • Attempts to consolidate references together and replace Harvard style refs with <ref> tags but doesn't always do this right -- I've not kept track of the diffs where I saw it do it wrong unfortunately but I'll go back through the diffs and find it. This is also not a consensus change as Harvard referencing is permitted in the WP:MOS.
  • Appears to try to replace dead links with Wayback Machine references -- whilst this may be helpful in certain circumstances this makes me nervous also, as it is certainly again not a consensus matter that all dead links are replaced with nominally live ones by linking to the Wayback Machine
  • Does a few whitespace changes here and there which obviously are triggered by some sort of regexp, or alternatively this may be a change caused as a side effect of the other tasks
  • It used to muck around with ref data by replacing it with stuff from the Google Books API but he seems to have switched this off now mercifully -- thank you Delta for addressing this
  • etc, etc
These are all very specific patterns of changes being applied automatically to all articles that Delta is putting through his script, apparently with minimal or no review. If this is not a pattern of edits in the meaning of the sanction, I'm not sure what is. I'd like to clarify however that I am not saying many of these changes are necessarily unhelpful, but Delta's sanction specifically states that he must request community consensus for such mechanical changes before carrying them out; he has failed to do so. --Tristessa (talk) 00:55, 29 September 2011 (UTC)
  • So far as I'm aware, Δ is not prohibited from using editing scripts. I see nothing at WP:RESTRICT to indicate that he is so restricted. I did glance through 10 edits either side of the diff provided in the WP:AN thread. As I noted, I saw no pattern, unless changing the case of "c" counts as a pattern (which is frankly absurd). To your particular points, I want to address the idea of "pattern". If I replace a template with another template, and do so across a number of articles, that's a pattern. If I remove a particular image from a number of articles for failing WP:NFCC #10c, that's a pattern. If I remove image 'A' from article 'N', and some time later remove image 'B' from article 'X' for a similar reason, it isn't a pattern. I think you're going to have to get consensus on what "pattern" means before you can ascribe restriction violating behaviors to Δ. If Vincent van Gogh creates 30 portraits in a similar style, each is its own individual artwork. See this. Is that a pattern? Or is that applying a similar set of ideas across several works? What exactly is "pattern"? If Δ did 100 edits that replaced a template with another template, that's an obvious pattern case. If Δ did 100 edits adding content to 100 different articles from 100 different sources, that's an obvious non-pattern case. Within the middle of those two extremes is a considerable amount of grey area. It would behoove you to go look at the history underlying Δ's sanctions regarding patternistic edits to gain an understanding of why the restriction was put in place, and what the crafters of that restriction most likely saw as "pattern". Any one of us, me included, applying our own definition of pattern is rife with problems. It can be an exceptionally broad paintbrush. Using it as such creates a situation that makes Δ's editing here completely untenable. Of course, that is the intent of the lynch mob that is out for Δ. But, regardless of the lynch mob, Δ isn't banned from this site. Therefore, we must work productively to contrive a set of restrictions in which we can reasonably expect an editor to be able to move forward, as opposed to contriving a scenario where he's doomed to failure no matter what he does. Ascribing "pattern" to a broad swath of editing behaviors is precisely that sort of scenario, and entirely misses the point of his original sanctions. --Hammersoft (talk) 13:23, 29 September 2011 (UTC)
Hammersoft, I really don't mean to be nasty by saying this, but I'll be honest; such a rationale reads like a mix of non sequitur and the Chewbacca defense. I never said he was banned from using editing scripts; instead, he is banned from using them on a wide range of articles without asking for the community's consensus on what it does. Hair-splitting over the meaning of the word "pattern" doesn't modify this -- I'd consider it reasonably self-evident that the same cleanup script run on multiple articles that makes particular series of repeated changes, represents a "pattern". I believe a reasonable uninvolved editor would also consider this to be a "pattern". I think the community's sanction is perfectly clear, and Delta was (but is not, in fact, this particular 24 hour period) violating his sanction. --Tristessa (talk) 18:30, 29 September 2011 (UTC)
  • I'll just sidestep your first sentence. Sorry. I think you need to ask for Δ to be banned from the project. You're effectively saying all of his edits are a pattern. You're not offering any possible situation in which he can (even if he wanted to) comply with your wishes. --Hammersoft (talk) 18:34, 29 September 2011 (UTC)
No, I am not saying all of his edits are a pattern; I'll recap. The automated rule-based changes made by his cleanup script when run on a series of articles produces, in my opinion as an outside admin, a pattern of automated/semi-automated edits within the meaning of the sanction. I do not wish to ask for him to be banned. I believe my position is perfect clear in this regard. --Tristessa (talk) 18:41, 29 September 2011 (UTC)
  • And as I previously outlined, you're describing a grey area in nebulous terms. There's no possible way he can modify his behavior to comply with your wishes. You're giving him an impossible demand. --Hammersoft (talk) 18:44, 29 September 2011 (UTC)
I'm not giving him an impossible demand. All he has to do to comply with his sanction is file a description of the work the script is meant to do on WP:VPR and run it only if he gets that consensus. There would be no problem at all then and the whole point would be moot. It wouldn't have even got to AN/I again. --Tristessa (talk) 18:47, 29 September 2011 (UTC)
  • You're telling him that if he introduces a white space, it's a pattern. You're telling him that if he changes the case of a letter, it's a pattern. You're telling him that if he uses Google Books API, it's a pattern. What next, if he adds "~~~~" to the end of a talk page message of his that it's a pattern? This is impossible. --Hammersoft (talk) 18:49, 29 September 2011 (UTC)
Of course not. On the other hand, if he does all those things automatically to each article processed through a script, that is definitely a pattern. He just needs to seek consensus for pattern edits; that is all. --Tristessa (talk) 18:55, 29 September 2011 (UTC)
And I think that's a straw man. As you wish, Hammersoft; I think I've made my views quite clear, and done my best to discuss it with you. I'm sorry that you feel that way. --Tristessa (talk) 19:15, 29 September 2011 (UTC)
  • Refusing discussion on my points by labelling them "strawmen", "non sequitur", and "Chewbacca defense" is your choice. But, the point remains; you're asserting that his addition of white space constitutes a pattern, apparently simply because he uses an editing assist script to do so. That's flatly absurd. Hey, I'm a certified idiot, so I'm sure I completely missed your point. But, perhaps you'd do me the favor of answering why in holy hell we're discussing adding of whitespace as some sort of problem? Worse, that it's some kind of problem because there's an editing assist script involved? --Hammersoft (talk) 19:26, 29 September 2011 (UTC)
  • I would have to agree with Hammersoft. The "script" that I use to assist with cleanup is publicly available for usage. (Not sure that its the most up to date version but that's besides the point) and I use it like AWB's genfixes, alongside other edits. Its not really a pattern of edits. ΔT The only constant 19:47, 29 September 2011 (UTC)
Thanks for describing this, Delta. I'm grateful for contact with you. In which case, won't you please consider filing what you've just said (albeit with some more specific information on what exactly it does) to WP:VPR to stop this madness of the debate once and for all, so you can actually get consensus for running it? --Tristessa (talk) 01:11, 30 September 2011 (UTC)
  • "Dear VPR, please grant me permission to use an edit assisting script that tons of other people use to change the case of a letter and introduce whitespaces as format occasionally requires". Right. --Hammersoft (talk) 15:24, 30 September 2011 (UTC)
Hammersoft, I haven't refused discussion on your points and I apologise if it feels that I have done so; my remarks were not intended in such a vein. Whitespace itself was not my issue; this was a point that you, and not I, raised. Any series of repeated edits that occur as part of a general pattern, and in this case it is a series of different types of changes in a single edit, qualfiy as a pattern. Indeed, I hope I made it clear I didn't consider whitespace changes on their own to be a pattern, but of course if the same rule-based whitespace changes were made to 25 articles I'd definitely call that a pattern. Does that make sense? Maybe I really have failed to make myself understood, I really don't know, but this seems really quite an obvious point to me. --Tristessa (talk) 01:11, 30 September 2011 (UTC)
  • Do you read yourself? "A series of different types of changes in a single edit qualify as a pattern"? Wow. ANY edit he does would then qualify as a "pattern" under that criterion. --Hammersoft (talk) 15:24, 30 September 2011 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=User_talk:Δ/20110901&oldid=1089916845"