User talk:Green Cardamom

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Template editing[edit]

Wikipedia Template editor.svg

Your account has been granted the "template editor" user permission, allowing you to edit templates and modules that have been protected with template protection. It also allows you to bypass the title blacklist, giving you the ability to create and edit edit notices.

You can use this user right to perform maintenance, answer edit requests, and make any other simple and generally uncontroversial edits to templates, modules, and edit notices. You can also use it to enact more complex or controversial edits, after those edits are first made to a test sandbox, and their technical reliability as well as their consensus among other informed editors has been established.

Before you use this user right, please read Wikipedia:Template editor and make sure you understand its contents. In particular, you should read the section on wise template editing and the criteria for revocation. This user right gives you access to some of Wikipedia's most important templates and modules; it is critical that you edit them wisely and that you only make edits that are backed up by consensus. It is also very important that no one else be allowed to access your account, so you should consider taking a few moments to secure your password.

If you do not want this user right, you may ask any administrator to remove it for you at any time.

Useful links:

Happy template editing! — Martin (MSGJ · talk) 14:40, 14 December 2015 (UTC)

  • Just wondering ...
    1. Could you archive your talk page?
    2. Have you considered adminship? — Martin (MSGJ · talk) 12:32, 15 December 2015 (UTC)
  1. @MSGJ:. Done
  2. I'll look into it. I know nothing about requirements. Thanks. -- GreenC 14:21, 16 December 2015 (UTC)
    Wikipedia:Guide to requests for adminship is perhaps a good place to start. Regards — Martin (MSGJ · talk) 22:43, 16 December 2015 (UTC)

Possible AWB Bug[edit]

Hey just a notice that you somehow duplicated the contents of Champ Clark twice. In the process of reverting to the last good version, I removed <ref name="Allan, Chantal page 17">, but I don't think that caused any harm and it seemed to also fix a ref error. Opencooper (talk) 05:20, 17 December 2015 (UTC)

Hi thanks for the notice. Never seen that before. I don't know if the problem is with my script or AWB itself. Since I don't know what caused it, what I'll do is add a sanity check if the edit size is > 50% of the original article it halts with a warning. I might also retry AWB on Champ Clark and see if it can replicate the problem. -- GreenC 14:19, 17 December 2015 (UTC)

Season's Greetings![edit]

Use {{subst:Season's Greetings}} to send this message

Merry Christmas and Happy New Year![edit]

Warmest Wishes for Health, Wealth and Wisdom through the Holidays and the Coming Year! Lingzhi ♦ (talk) 22:43, 25 December 2015 (UTC)

Administrators' noticeboard: red herring[edit]

Information icon There is a response discussion at Wikipedia:Administrators' noticeboard/Incidents regarding an issue you created. Clepsydrae (talk) 19:27, 27 December 2015 (UTC)

As for your comment, "See WP:3RR before you revert again at red herring. You may be blocked from editing Wikipedia," I only reverted the article twice. I edited it multiple times this morning, possibly immediately after a revert, which undoubtedly appeared to you as a "revert." I have no intention, desire, or time to sustain a revert war. I had hoped that you would follow the references, have an "ah-ha..." moment and leave a good, scholarly edit be. Clepsydrae (talk) 19:27, 27 December 2015 (UTC)

2015 Thalys train attack[edit]

Hi, through no fault of your own, a recent revert I made at 2015 Thalys train attack has undone a change of yours made on 14 November. This is due to an improper implementation of an agreed-upon Rfc by Tough sailor ouch on 12 November which I subsequently reverted, leaving your edit nowhere to go for the time being. Once the Rfc change is properly implemented (by him or another editor) your edit can be reapplied. I will be happy to apply that edit for you, upon request. Please wait for the re-implementation of the Rfc first.

For details, please see Talk:2015_Thalys_train_attack#Implementation_of_Rfc. If there's anything you can think of to explain it to him better than I was able to, I'd sure appreciate it, cuz I'm not sure I'm getting the point across clearly enough. Cordially, Mathglot (talk) 10:16, 29 December 2015 (UTC)

2016[edit]

Happy New Year .jpg
Happy New Year 2016!
Did you know ... that back in 1885, Wikipedia editors wrote Good Articles with axes, hammers and chisels?

Thank you for your contributions to this encyclopedia using 21st century technology. I hope you don't get any unneccessary blisters.
Cullen328 Let's discuss it 22:46, 30 December 2015 (UTC)    –

@Cullen328:, hey thanks! Happy New Year. Cool banner. -- GreenC 01:39, 31 December 2015 (UTC)

Happy New Year, Green Cardamom![edit]

Send New Year cheer by adding {{subst:Happy New Year fireworks}} to user talk pages.
@Northamerica1000:, thank you, North America, Happy New Year! -- GreenC 16:08, 2 January 2016 (UTC)

Public domain movies[edit]

Hello,

Thank you for removing The Klansman on the list of films in the public domain in the United States. I have added another film on the list that I believe to be in the public domain by the name of Born to Win, a 1971 movie released thru United Artists that features Robert De Niro in an early appearance. If the source I used isn't reliable, then by all means please remove it from the list. Thank you. Hitcher vs. Candyman (talk) 22:11, 17 January 2016 (UTC)

Lemelson[edit]

Sorry about the edit conflict on Emmanuel Lemelson. Coincidentally I was in the middle of a more drastic rewrite of that para., which I've added. Hope you think it's better overall - less promotional. If you want to make other changes to the article, let me know, and I'll work on Lemelson Capital Management for a while instead.  —SMALLJIM  14:33, 23 January 2016 (UTC)

No problem, your edit is a big improvement. I probably won't edit more for the moment. -- GreenC 14:49, 23 January 2016 (UTC)
OK. Thanks for your other edit there.  —SMALLJIM  15:12, 23 January 2016 (UTC)

hold[edit]

Dickens worked on David Copperfield for two years between 1848 and 1850. Seven novels proceed it, and seven novels would come after it, Copperfield being the mid-point novel.

Tolstoy regarded Dickens as the best of all English novelists, and considered Copperfield to be his finest work, ranking the "Tempest" chapter (chapter 55,LV - the story of Ham and the storm and the shipwreck) the standard by which the worlds great fiction should be judged. Henry James remembered hiding under a small table as a boy to hear installments read by his mother. Dostoevsky read it enthralled in a Siberian prison camp. Franz Kafka called his last book Amerika a "sheer imitation". James Joyce paid it relevance through parody in Ulysses. Virginia Woolf, who normally had little regard for Dickens, confessed the durability of this one novel, belonging to "the memories and myths of life".

The story is told almost entirely from the position of the first person narrative, through the voice of David Copperfield himself, and was the first Dickens novel to do so. It is considered a Bildungsroman and would be influential in the genre such as Dickens own Great Expectations (1861), Thomas Hardy's Jude the Obscure, Samuel Butler's The Way of All Flesh, H. G. Wells's Tono-Bungay, D. H. Lawrence's Sons and Lovers, James Joyce's Portrait of the Artist as a Young Man.

As a bildungsroman, it has one major theme throughout, the disciplining of the hero's emotional and moral life. We learn to go against "the first mistaken impulse of [the]] undisciplined heart", a theme which is repeated throughout all the relationships and characters in the book.

Talkback[edit]

Nuvola apps edu languages.svg
Hello, Green Cardamom. You have new messages at Cyberpower678's talk page.
Message added 04:58, 17 February 2016 (UTC). You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

cyberpowerChat:Offline 04:58, 17 February 2016 (UTC)

Works not by[edit]

Just so you know, it seems that the "Works By" project is linking to irrelevant works. For example, in the link to the Internet Archive you added to Brayton C. Ives labeled "Works by or about Brayton C. Ives", we find one relevant work (repeated 6 times), which already appeared among the external links in the article, and then 19 links to various volumes of the novels and tales of Robert Louis Stevenson, followed by more relevant links. I don't know why the RLS works are showing up, other than having a table of contents that includes "St. Ives". So maybe you need to evaluate your search strings or something similar. I suspect that searching with initials instead of the name actually used by the author (in this case, Brayton Ives or General Brayton Ives) is causing this. - Nunh-huh 05:56, 22 February 2016 (UTC)

Actually, I've moved the page because there's no source referring to him with a middle initial.... - Nunh-huh 06:08, 22 February 2016 (UTC)
It's because "Stevenson, Fanny Van de Grift, 1840-1914" (same birth-death) and "St. Ives" are being matched by the IA search engine to the search string ("1840-1914" AND Ives). This is intentional as other authors require this to avoid false negatives. Most of the time it works OK but some cases like this produce false positives. It's all a trade off between avoiding false negatives while producing as few false positives as possible. I check each one before adding and felt in this case it wasn't excessive though it did appear to be about 50% I usually aim for 40% or less before adding a custom search (which has its own trade offs thus not done often) but if most cases it's much less and often 0%. -- GreenC 15:02, 22 February 2016 (UTC)
Well, my feeling is that a link to works that may or may not be by or about the subject of the article doesn't really enhance the article. If the current method produces such links, it ought to be changed. It will probably require personal curation—you know, an actual personal assessment of the works linked to—if it's going to be done right. Thanks for fixing this one. - Nunh-huh 17:55, 22 February 2016 (UTC)
Thank you for bringing it to my attention. I do manually check each search (~93% are rejected on average) and sometimes make a mistake as the data can be hard to interpret why the IA search engine included a certain book. The template documentation {{Internet Archive author}} has more info on how anyone can fine tune the results, such as done here with the custom search. Unfortunately the goal of 0% false positives is unworkable for a number of reasons. The solution just implemented for Brayton Ives is called a custom search (see the docs) however it has trade offs. It's brittle, prone to future problems and lack of maintenance as changes occur at Internet Archive. That's why the template exists to address these problems. Sometimes the trade off makes sense, other times not. I make custom searches when I feel it is justifiable, and sometimes miss some. -- GreenC 19:55, 22 February 2016 (UTC)

Internet Archive author template[edit]

I noticed that you added {{Internet Archive author |sname=James Eustace Bagnall}} to James Eustace Bagnall. Following the link does find some of his publications, but also one by a very different James Bagnall. Not being familiar with {{Internet Archive author}} I don't know if this can be prevented. Peter coxhead (talk) 15:48, 24 February 2016 (UTC)

sopt=t worked for this case. It means if any future books are uploaded using only "James Bagnall" they won't be seen but if you put more emphasis on false positives than false negatives sopt=t will work. My preference is to allow in a few false positives so as to not cause false negatives in the future but it's a personal call on an article by article basis and what the results are. -- GreenC 15:57, 24 February 2016 (UTC)
Ok, thanks; I understand. Peter coxhead (talk) 17:16, 24 February 2016 (UTC)

re Croses Criquet[edit]

Thanks for removing spamy link. I tried, but failed, and got tied up in knots when making enquiries about how to resolve the matter! Thanks for putting me out of my misery! Regards --Observer6 (talk) 16:36, 28 February 2016 (UTC)

@Observer6: no problem. When using the template {{cbignore}} just type the letters {{cbignore}} into the article (without the tlx part, which is for displaying the name of the template without actually using the template). -- GreenC 17:01, 28 February 2016 (UTC)

Thx, GreenC. First, I couldn't see how 'cbignore' would solve the problem. Second, I didn't know that 'tlx' needed to be omitted if I was going to use it. Thirdly, I still couldn't understand how changing the bot flag from 'false' to 'true' (or was it the other way around?) was going to solve the problem of a spamy link! But 'failed' was brilliant! Glad to get a third party response and succinct solution to the fundamental problem! --Observer6 (talk) 17:23, 28 February 2016 (UTC)

Yeah if you don't know how the moving parts work it can be mysterious. There probably needs to be better documentation with clear examples. Maybe I can work on that. -- GreenC 17:30, 28 February 2016 (UTC)

Disambiguation link notification for March 4[edit]

Hi. Thank you for your recent edits. Wikipedia appreciates your help. We noticed though that when you edited Aaron Swartz, you added a link pointing to the disambiguation page Scribner (check to confirm | fix with Dab solver). Such links are almost always unintended, since a disambiguation page is merely a list of "Did you mean..." article titles. Read the FAQ • Join us at the DPL WikiProject.

It's OK to remove this message. Also, to stop receiving these messages, follow these opt-out instructions. Thanks, DPL bot (talk) 11:31, 4 March 2016 (UTC)

IA API bug[edit]

Are you getting anymore bad archives, as much as before. IA tech team reported to me in an email last night, "An earlier version from last week also fixed a problem in which I was returning the HTTP status of the CDX call rather than the underlying archived HTTP status code." which seems to describe the bug of bad archives with a non-200 OK being mistaken for a good archive with a 200 OK response.—cyberpowerChat:Limited Access 16:25, 8 March 2016 (UTC)

@C678: Good to hear, hope that is it. They are having challenges. Short outages daily. Connection timeouts. Also saw a case where it reported a match in wayback, then tried again and it reported none found, then tried again and matched - intermittent result. I started today running WaybackMedic which looks for and corrects 4 problems. In the first 250 articles checked (from the 90k+ CB has edited since Dec) it reported problems in 30 articles (about 12%). Will get a bigger sample tomorrow after I manually verify the 30 did the right thing -- GreenC 01:47, 9 March 2016 (UTC)
I also just deployed a massive update to Cyberbot that should fix a zillion bugs that were reported on the talk page. The scope of the change may have broken something else that I haven't seen, so please let me know if Cyberbot starts acting wierd as of the last hour of posting this. :P—cyberpowerChat:Limited Access 16:28, 9 March 2016 (UTC)
Ok great. Will mark this date as a terminus for the 4 problems listed at WaybackMedic (although the 404 problem might continue depending on IA). -- GreenC 17:06, 9 March 2016 (UTC)

Talk[edit]

Hello i just joined.How do i upload pictures? EnchantingZucchini (talk) 16:09, 24 March 2016 (UTC)

That's an easy sounding that can be complicated. It depends if the picture is Fair Use or Public Domain (or Creative Commons). If Fair Use then it is uploaded directly to this website (left side of page, box says "Tools", "Upload file"). However you will need to read up on what constitutes Fair Use and provide appropriate rationales etc.. if it is a public domain then it is uploaded to http://commons.wikipedia.org and they have their own set of procedures and rules there. -- GreenC 16:17, 24 March 2016 (UTC)

Edit summaries[edit]

Hi. Will you please use edit summaries? Cheers Face-smile.svg fredgandt 14:57, 29 March 2016 (UTC)

URGENT..._ _ _... morse ...---... ...___… Hi listen your summary https://en.m.wikipedia.org/wiki/Positron_emission_tomography you can't just end it like that you only gave one example right there it is not correct you need to at least do some research you know a little deeper than you already do might as well since you dug that far on ,if you are as smart asi think you are which I'll assume for now that you are, is a very unintellectual subject so just because I have OCD sorry for the caps I humbly ask you to just put another possibility that it may be used for thank you SubtleAlpha246 (talk) 09:42, 21 June 2016 (UTC)

Oh yeah sorry about that urgent OCD. Sir I am so sorry. SubtleAlpha246 (talk) 09:43, 21 June 2016 (UTC)

AfD[edit]

Hi Green Cardamom: A recent edit you performed at AfD has been reverted. You may want to check it out. North America1000 03:34, 1 April 2016 (UTC)

@Northamerica1000: - Clever. Thanks for the laugh - and not having to deal with a real dispute! -- GreenC 13:24, 1 April 2016 (UTC)
You're welcome? Cheers! North America1000 14:00, 1 April 2016 (UTC)

Disambiguation link notification for May 4[edit]

Hi. Thank you for your recent edits. Wikipedia appreciates your help. We noticed though that when you edited Best Translated Book Award, you added a link pointing to the disambiguation page Liu Xia (check to confirm | fix with Dab solver). Such links are almost always unintended, since a disambiguation page is merely a list of "Did you mean..." article titles. Read the FAQ • Join us at the DPL WikiProject.

It's OK to remove this message. Also, to stop receiving these messages, follow these opt-out instructions. Thanks, DPL bot (talk) 10:45, 4 May 2016 (UTC)

Edit to Reading Half Marathon by WaybackMedic[edit]

Hi. Could you take a look at the recent edit by your bot (WaybackMedic) to Reading Half Marathon. I don't know which archiveurl= parameter it was trying to fix, but it appears to have replaced all the newlines in the article with ***!!***, thus rendered the whole article pretty nigh unreadable. I've reverted the change, so you will need to look back in the history to oldid:722269132. -- chris_j_wood (talk) 12:12, 27 May 2016 (UTC)

Yes bug fixed. thanks for the notice. -- GreenC 17:04, 27 May 2016 (UTC)

WaybackMedic and talk pages[edit]

Thank you for this edit at Tomb of Absalom; I agree that there is no valid archived link for that citation.

Would it be possible for the bot to also update the {{sourcecheck}} template on the talk page? In this case I have updated the parameter at Talk:Tomb of Absalom. – Fayenatic London 13:08, 27 May 2016 (UTC)

There are problems with the sourcecheck system. If there is more than 1 link there is no clear way to designate a "fail" or "pass" for each link much less to do it with a bot since the formatting is somewhat free-form. It would take a lot of work to prevent messing up the talk page and not sure if it would be big gain because it doesn't impact anything other than a notification, and WM is operating on archive.org links besides those added by Cyberbot, and my guess is only impacting about 5-10% of those. I'm also uncomfortable doing by bot what was meant to be a manual check since there can be soft-404s that bots believe are valid but only a human can verify; it would still work for the deletions, but the problems mentioned above. -- GreenC 17:27, 27 May 2016 (UTC)
OK, thanks. I'll watch out for others that need checking manually. – Fayenatic London 11:20, 28 May 2016 (UTC)

Cucuteni-Trypillian culture edits[edit]

Could you please explain in plain English what is going on with this article? The last two edits are making me very concerned. Please reply to my own talk page. Thanks. --Saukkomies talk 20:59, 27 May 2016 (UTC)

@Saukkomies: It's the same software bug noted above, it impacted 10 out of 1500 articles edited, or about half of one percent. The bug is fixed, the impacted articles were corrected, and the bot will not be editing those articles again. (Keeping on this page since multiple threads on the same topic). -- GreenC 21:30, 27 May 2016 (UTC)
Thanks for the explanation. --Saukkomies talk 16:41, 30 May 2016 (UTC)

Re: Webcite[edit]

It doesn't redirect but is actually a way back archive page that I webcited lol. I noticed a while back that web archive links rot/_expire eventually but webcite seemed more lasting. Which is why I objected to the bot's changes Dan56 (talk) 23:48, 27 May 2016 (UTC) Dan56 (talk) 23:48, 27 May 2016 (UTC)

@Dan56: That is true, good idea as robots.txt can make pages disappear from Wayback, or even disappear for unknown reason. Rather then battling with the bot, there is a {{cbignore}} template ("Cyberbot Ignore"). -- GreenC 00:31, 28 May 2016 (UTC)
Cyberbot is not adding a badlink. Cyberbot is moving an archive URL to the proper field. I have no idea, why there is an archive of an archive, but the URL parameter needs to have the original URL of the page being displayed, especially a broken WebCite makes it impossible to figure out what the original URL was supposed to be.—cyberpowerChat:Online 01:42, 28 May 2016 (UTC)
This is true, the url= field should be the original URL not an archive URL (archive.org or webcite). Confusing when there are two archive URLs in a template plus the IA API provides a third option. I would probably go with what is known working, the API result 200, which I believe is what happened. -- GreenC 02:03, 28 May 2016 (UTC)
@Cyberpower678:, none of this requires removing the webcite link, which always shows at the top the link that's been archived. Dan56 (talk) 07:53, 28 May 2016 (UTC)
My point isn't about the bot removing the webcite archives, my point is the URL field needs the original URL field so people can easily see what the archive link was supposed to be archiving. Sure it's printed on the top of the snapshot, but what if WebCite goes down, crashes, their data centers have a fire, flood, or something that permanently destroys the snapshot. Then no one knows what that WebCite URL was holding in place of the original.
Well, you would if you'd gone back to the revision dated whenever the BOT made the changes I've been reverting lol Dan56 (talk) 23:57, 28 May 2016 (UTC)

2016 Wikimedia Foundation Executive Director Search Community Survey[edit]

The Board of Trustees of the Wikimedia Foundation has appointed a committee to lead the search for the foundation’s next Executive Director. One of our first tasks is to write the job description of the executive director position, and we are asking for input from the Wikimedia community. Please take a few minutes and complete this survey to help us better understand community and staff expectations for the Wikimedia Foundation Executive Director.

  • Survey, (hosted by Qualtrics)

Thank you, The Wikimedia Foundation Executive Director Search Steering Committee via MediaWiki message delivery (talk) 21:48, 1 June 2016 (UTC)

WayBack Medic[edit]

Hi. I'd like to learn how to use WayBack Medic. I've read User:Green Cardamom/WaybackMedic, but I'm still confused. I think I'd like to test it. Would you please walk me through it?Zigzig20s (talk) 13:44, 4 June 2016 (UTC)

It's a bot not an end-user tool. It will hopefully eventually process all articles that contain wayback links. -- GreenC 13:50, 4 June 2016 (UTC)
OK, makes sense.Zigzig20s (talk) 13:53, 4 June 2016 (UTC)

List of things creating Internet Archive search queries[edit]

Is there a list somewhere of people/bots who are creating Internet Archive search queries? We're looking into what raw Lucene syntax we need to support in our search interface, and it appears that there are several people/bots creating such searches. Thanks! Greg (talk) 00:20, 5 June 2016 (UTC)

@Greg Lindahl: .. sending reply via email. -- GreenC 01:29, 5 June 2016 (UTC)

Your dead link bot - idea for improvement[edit]

Hi Green Cardamom, I came across the useful work your GreenCbot is doing with dead links. Is it possible to automatically assign a tag on the top of the page and/or (hidden but visible for maintenance workers) category to it? Because now it appears very small in the list of references, e.g. Cesar Department, where I saw it, but the chance it will be picked up by someone and resolved is minute. Also because in the edit summary only appears "WaybackMedic" and no reference to "This page has dead links, please have a look at them" or something along those lines. In general I hate those tags at the top and they are useless in many cases as they stay on the pages for many years, but in this case I think it's useful, especially when linked to the Wikiprojects on the talk page. What do you think? Cheers, Tisquesusa (talk) 08:46, 5 June 2016 (UTC)

The WP:Link rot problem is pretty immense, and growing daily, this bot is not the only one trying to solve it. I'm not sure there would be consensus to add a top hat for dead links, partly because they are often unsolvable (no archive exists) and how common they are. All dead links are invisible tagged by inclusion in a category which other bots/people can use to find them. The edit summary I agree is not good; the bot uses AWB which currently has no feature to dynamically modify the edit summary - I've asked for this feature twice. I should be using the direct API instead, but AWB is how the bot was designed when it started out. It's on my list of things to change in future versions. -- GreenC 14:01, 5 June 2016 (UTC)

Edits claiming "not dead"...[edit]

This edit [1] removed templates showing a couple of links as dead: it removed two strings like this:

{ { dead link|date=June 2016|bot=medic } } { { cbignore|bot=medic } } (sorry, can't remember proper way to untemplate)

These links just "helpfully" redirect to a generic front page, but they are dead in terms of the relevant content. I'm not sure what to do: I would have thought that cbignore instructed the bot to leave the dead link template, but it doesn't seem to. Imaginatorium (talk) 03:59, 7 June 2016 (UTC)

Imaginatorium: Yeah. The WaybackMedic bot made some errors in about 300 articles so I wrote a patch to return and fix them but the patch created new problems by undoing dead links that were legit in some cases. There was no way around it, so I choose the path of least damage. I've been going though manually but if you see anything, please help restore the dead link and cbignore templates for any url's that seem to deserve it. If there was a cbignore before it should be restored (including the |bot=medic). -- GreenC 04:06, 7 June 2016 (UTC)

WP:BOWN[edit]

Information icon This message is being sent to inform you that there is currently a discussion at Wikipedia:Bot owners' noticeboard regarding an issue with which you may have been involved. Thank you. — xaosflux Talk 10:57, 7 June 2016 (UTC)

Question about Gutenberg list[edit]

Hey there, I see you put 99% for project Gutenberg, are you suggesting all the red links on guttenberg need a redirect as there are already articles for them? Because I noticed on the talk page you said there are 9,000~ gutenberg tags on wikipedia and the list has about 8,000? I ask because I notice a lot of redlinks on the gutenberg list still. Kind regards, Calaka (talk) 09:46, 8 June 2016 (UTC)

I may have confused things feel free to revert. My intention was 99% complete in terms of adding the template to available articles, not in creating new articles. -- GreenC 12:13, 8 June 2016 (UTC)

Help:Using the Wayback Machine[edit]

Hi! I see that you're doing some cleanup of archive.org links in articles. Over at Help:Using the Wayback Machine, those sections are specifically about asking archive.org to create an archive. Is your edit there correct, are or those URLs with /save/ a special case that should have been left alone? -- John of Reading (talk) 19:16, 14 June 2016 (UTC)

/save/ creates an archive of the page. It's a command that triggers the Wayback Machine to begin archiving the page. It only needs to be done once. Once the archive is created, linking to an archive page for citation purposes should be a 14-digit date. Since I don't know what the correct date is, replacing with "*" leads to the archive index page and hoping someone will fill in the correct archival date from the index. -- GreenC 19:46, 14 June 2016 (UTC)
Thank you. That sounds as if the /save/ URLs in the help page are correct, so I will undo your edit there. -- John of Reading (talk) 19:50, 14 June 2016 (UTC)
John of Reading: I see what you are saying, sorry. Yes that was a bad edit. I have a list of articles to process and I thought they were all mainspace. I see some WP mixed in. I'll remove those before proceeding. Thanks for the notice. -- GreenC 20:11, 14 June 2016 (UTC)

A barnstar for you![edit]

Peace Barnstar Hires.png The Barnstar of Diplomacy
Many thanks to you Good Sir, may the WaybackMedic find continuous blessings and wellness ongoing! SoS was answered and appreciated. "Dutch" Publican Farmer (talk) 23:35, 15 June 2016 (UTC)

Wikipedia:Bots/Requests for approval/GreenC bot 2[edit]

Your task has been approved for trial. — xaosflux Talk 03:26, 25 June 2016 (UTC)

Do you mean case 2 mentioned here or case 2 mentioned here or case 2 mentioned here here :) I think what your asking is a way for links to be added to the Wayback Machine in an automated fashion. If so, that is already done for all external link added to Wikipedia. As I understand it. It should be easy to test to make sure its working. Find an external link that is not yet in the Wayback or Wikipedia, add it to an article, check a few days later and see if it's now in the Wayback. -- GreenC 21:46, 28 June 2016 (UTC)
Sorry, I meant case two on C678's page. (Usually when I test links in the Wayback it automatically archives the page, so I'd have to experiment to try that test.) BUT regardless of whether IA is doing that in the background, my concern is the security of having |archiveurl= filled out automatically in a citation, as your second example did on C678's page. Is that something your bot is prepared to handle? If so, when will it add the archiveurls and is there a way to make sure it gets to a page within 24 hours, etc. (I imagine there might be some pushback if it added archiveurl params to all URLs sitewide, but I'm talking about cases in which editors specifically want the bot to do the work of archiving their links for them.) czar 21:51, 28 June 2016 (UTC)
WM doesn't create new archiveurl links. It deletes or modifies existing links that have problems. Thus "Medic". New additions are done by CB. -- GreenC 22:55, 28 June 2016 (UTC)
"cases in which editors specifically want the bot to do the work of archiving their links for them" .. you'll need to be more explicit what you mean. If you want a dead link to be archived in the template, add a {{dead link}} template. If you want a link archived on Wayback, then it will be done automatically by background processes. If you want a link that is otherwise working and not dead to be added as an archiveurl to a template, there is no auto mechanism for that. -- GreenC 23:06, 28 June 2016 (UTC)
Your last example is what I'm after. I thought that's what was happening in the second replacement in [2] but apparently not. Oh well, thanks anyway czar 23:56, 28 June 2016 (UTC)
...returning to this—don't you think it would be valuable to have such an "auto mechanism" for that? I imagine it'd be easier to repurpose one of the bots already doing the work than for me to write my own script/code czar 19:03, 8 July 2016 (UTC)
Medic is a 1-pass bot, to do what your suggesting would require a continually running bot and that is CB. If you are serious about making a bot.. take a look at medicapi.nim (Python-like) in Wayback Medic's GitHub account, I figure development > 1000 man-hrs at this point (development + running), a lot of hard-won lessons on how the Wayback API (doesn't) work. Or the CB source (PHP) which is even more complex. From a policy view not sure it's the best idea for a couple reasons. Ideally the snapshot date would be chosen by a human to ensure it contains the text cited and there is nothing wrong with the page (soft-404 etc). We're using bots to deal with link-rot for already-down links because there are too many to fix manually. This process is not error-free though, thus CB's talk page messages to manually check the links. If an editor is taking the time to add a cite template flag like you suggested, they really should be adding the archive link directly it doesn't take much longer. Otherwise why not have an automated process that adds archive links for every cite template added. That may eventually happen but not sure there is consensus for that level of bot intervention due to the problems of accuracy. It's one thing to fix existing dead links, another to fully automate the process for still live links. DASHBot did that for a while and honestly it had problems that WaybackMedic is currently fixing (not that WM is perfect either). If you decide to explore further I would suggest trial balloon in Village Pump and notify the talk page of cite web template as certain users like Trappist the Monk and Cyberbot would be key. -- GreenC 19:35, 8 July 2016 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Very helpful to hear the backstory—thank you. I was thinking more of a single-pass run, and less of a bot than a userscript that would save myself a few minutes on each article I write. I used to spend the time archiving links, but now when I write an article with 30–60 links, it's a significant chunk of time doing a relatively mundane task, so I don't. I'd rather handle only the errors/cases that do not archive properly with IA (usually the same sites are repeat offenders, so WebCitation would be better in those cases) but the idea is that I'd like to call such a script on specific pages in specific requests. The other policy stuff could be interesting/useful down the line and it might be useful to prepare for that, but I think of my own writing and the hours a month such a script would save me. (That is, if I still was bothering to archive my links, which I should.) By the way, you recommended that I check whether Wayback was automatically archiving new links added to articles. I added several bare URL refs to Nadia Kaabi-Linke a week ago. If you're curious, they either haven't been archived or haven't been archived in 2016. Same for another link that was added with {{cite web}} format: [3]. Basically, I don't feel that I can trust the silent archive mechanism but I see value in being able to run it myself and see the results. czar 20:28, 8 July 2016 (UTC)

I understand better what you're saying and actually that would be a useful tool since it would be run manually and you could check the results. Let's see.. if your comfortable with AWB and unix, WaybackMedic could do this with some modification I would be happy to provide. What environment are you working in with AWB, Windows? Do you run a VirtualBox Linux? I think Linux will be required, it's not a script but a compiled binary.
Interesting about the archive test failing.. I'll bring this up with IA, that was understood to be working. -- GreenC 02:01, 9 July 2016 (UTC)
I'm comfortable with AWB and Unix. I prefer to edit in OS X, so I have run AWB in Windows when necessary (though I'm open to other ways of running AWB). But the OS X should be fine with the Unix bit. This sounds really great if you could help. Also thanks for coordinating with IA czar 06:08, 9 July 2016 (UTC)
Ok. You will need to install Nim, in order to compile the program, I don't have access to an OS X system. Nim installs in a single directory not invasive of the system. Recommend these instruction (the first section "Install"). Also will need GNU Awk version 4+ if not already. Ping me when these are working. -- GreenC 13:22, 9 July 2016 (UTC)
@Green Cardamom: all good, I have Nim and gawk czar 18:30, 9 July 2016 (UTC)
Ok great. Nim recently upgraded to 0.14 and I am on 0.13 .. it should be OK, but as a test could you download *.nim here into a local directory and run "nim c medic.nim" and make sure it compiles without error (might have some warnings). -- GreenC 19:18, 9 July 2016 (UTC)

Looking at this more closely, the program is designed to fix existing archive.org links in an article. It's not designed to add links where none exist since that is what CB does. So it would be a significant rework to change the scope. However, if the article contained "best guess" archive.org links, it could automatically verify and fix those. There would be almost no coding changes needed. For example say there is a link http://www.moma.org/calendar/exhibitions/1317 with an access-date of March 1 2015. You would set a "best guess" of archiveurl=https://web.archive.org/web/20150301000000/http://www.moma.org/calendar/exhibitions/1317 and archivedate=March 1, 2015. Then run the program and if this link doesn't exist at that snapshot, it will be deleted -- or if there is a better snapshot date such as March 15 2015 it will change to that (it works by picking the closest available). -- GreenC 19:55, 9 July 2016 (UTC)

Ah, but it won't save a copy of the live site, if it's still up? CB does that, right? What would you recommend as the best way forward for this? I see a few options: (1) A variation on tagging live links as {{dead link}}s just to get a pass from the bot, (2) Adding a dummy archive.org link like you suggested, and hope that an archived version already exists, and I suppose run this process manually, (3) Ask Citoid team to build it IA auto-archive into their Citoid API, (4) Make a request for a bot and pray for help, or (5) write a custom script that uses the IA API to save a site and add it in my wikitext (likely most time consuming but all I'm missing in my script is the IA part right now). You're familiar with my use case, so what would you suggest? I'd prefer to have the tool available to everyone, but I'm really looking for the quickest way of getting IA links for live sites into my citations. (My citations are formatted through Zotero export and the Citoid API, usually.) Also the medic.nim compile is missing "docopt"—if you think I should continue with that route, should that be part of another package or is it a separate dependency? Appreciate your help, czar 20:50, 9 July 2016 (UTC)
No it doesn't trigger saves on working sites but it could easily, just a simple GET command. The easiest way right now would be (2) since the code is working. For (2) the program will take care of things automatically - it will change the dummy snapshot date to the working snapshot date, or remove the dummy archive link entirely (if no working snapshot exists). It's not far off from the bot flagging idea of (1) except the flag is an archive.org URL. Docopt can be installed using nimble, the package manager, if not already installed the Nim Install instructions above has the steps for nimble. I think we are close to testing it out. After it compiles download the rest of the files and follow steps 2-6 in the 0INSTALL text shouldn't take very long. Then run ./project -c -p main to initiate a project. Then ./driver -p main -n "article name" using any article name. The new article text will be in ~/data/wm-XXXX if any changes were made.-- GreenC 21:40, 9 July 2016 (UTC)
All right, I'll give it a go—thanks for your help! Getting this error on the nim c medic.nim compile: medic.nim(2, 18) Error: undeclared identifier: 'randomize' I have math installed and nimble install extmath didn't solve it czar 21:52, 9 July 2016 (UTC)
Great! Random is related to the change from .13 to .14 .. in line 2 of medic.nim change from math import randomize, random to import random -- GreenC 22:57, 9 July 2016 (UTC)
Had to modify a few things in project.cfg (should that be added to 0INSTALL?) but I'm stuck on Unable to find .../WaybackMedic/meta/main.auth - Unable to create .../WaybackMedic/meta/main/auth It doesn't appear to be a folder permissions issue czar 23:32, 9 July 2016 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Ahh, skip the ./project -c -p main step and manually create two directories ~/WaybackMedic/meta/main and ~/WaybackMedic/data/main then modify project.cfg

default.id = main

default.data = /path/WaybackMedic/data/

default.meta = /path/WaybackMedic/meta/

main.data = /path/WaybackMedic/data/main/

main.meta = /path/WaybackMedic/meta/main/

Shouldn't need to mess with projects again once this is working, it will always be project "main" .. it's a feature when dealing with 100s of thousands of articles to break it up into batches.

-- GreenC 23:53, 9 July 2016 (UTC)

Okay—got it running without errors, but the output in the data folder doesn't appear to reflect any changes. I ran it on User:Czar/drafts/WaybackMedic test with and without |deadurl= set. The two target refs are the first two in the list-defined {{reflist}} czar 00:38, 10 July 2016 (UTC)
I ran it here and got the correct result. Need to see some debug output. First change driver.awk as its currently setup for GNU Parallel which you are not using. Edit driver.awk and find the commented line that says "Create index.temp entry (re-assemble when GNU Parallel" .. below that are two lines, comment them out; and below that are 6 commented out lines, uncomment them. Now re-run ./driver -p main -n "article name" -- then run ./bug -p main -n "article name" -v (view info) which will show useful information. The third line starts with ./medic .. .. run that command and it will re-run medic with debugging output using the data cached in the ~/data/main/wm-XXX directory which was just created by driver. The second line of the "./bug -v" output should show Data: cd .. then cd to that directory and look for a file called "article.waybackmedic.txt" .. this is updated article (article.txt is the original). -- GreenC 01:46, 10 July 2016 (UTC)
Hm... still not getting it. Here is a link to my installation (removed all but the last attempt). I recompiled after changing those eight lines in driver.awk and I'm not getting "article.waybackmedic.txt" in the wm-XXX directory. Any ideas? I didn't get any error messages or special output. (You can see in the link that it downloads the article but doesn't change the archiveurl from the dummy to the updated URL.) Appreciate your help with all this czar 02:31, 10 July 2016 (UTC)
The dropbox is helpful. It doesn't look like medic ran, only the driver front-end which downloads the article.txt and creates the data directories. Driver then calls medic. Try this to run medic manually with debug output ("-d y"): ./medic -p "main" -n "User:Czar/drafts/WaybackMedic test" -s "/path/WaybackMecic/data/main/wm-0709212754N/article.txt" -d y (set "/path" to actual path). -- GreenC 02:48, 10 July 2016 (UTC)
When I run medic I get "Usage: medic" and then the bunch of flags but still no debug output in the data folder. I was playing around with it a bit more and changed the executable links in init.awk to be direct, in case that was the issue, and now I get this when running driver:
$ ./driver -p main -n "User:Czar/drafts/WaybackMedic test"
File "<string>", line 1
import urllib, sys; print urllib.quote(sys.argv[1])
(Caret pointing to the b in "urllib") Also I'm not even getting the right output anymore. I can revert if need be, but I wonder whether a badly linked executable was responsible for the silent lack of output in the first place czar 03:17, 10 July 2016 (UTC)
driver looked ok and seemed to be working the problem is with medic I would revert the changes to init.awk. If medic gave a "Usage: medic" result it's saying it doesn't understand the arguments or missing args. If not already, use cut and paste for the ./medic command above, because the order of args is significant. Beyond that I really don't know why (and I need to sign off for tonight). -- GreenC 04:15, 10 July 2016 (UTC)


User:Czar -- On auto archival, I reported your test links and the response from IA: "We have discovered, and are fixing, a breakdown on our end that was effecting the timely archiving of new links from Wikipedia." Thanks for the discovery of the problem. -- GreenC 18:02, 10 July 2016 (UTC)

Good, I'm glad—happy to help. I've been getting "Usage: medic" ever since I first ran medic, if that helps. I thought it was part of the output. I was directly copy/pasting the command from the bug output and from your suggestion with the aforementioned results. Not sure if we've hit a dead end. I don't have time to write it right now, but I poked around with the basic IA API and found a basic command for saving to IA last night so I might just try that down the line czar
I suspect it's an OS X issue with the command line parsing library (Docopt). You gave it a good go and almost there I think. If there was an OS X shell account I could ssh into I would be glad to try, but understand if you want to try other options. To save a page to Wayback: https://web.archive.org/save/_embed/$URL Might take a few minutes to show up. The archive said they plan on doing it automatically for all links on Wikipedia. -- GreenC 19:23, 10 July 2016 (UTC)

Greetings[edit]

Per wp:lead I can't say I agree with this edit summary, but I am fine with your changes, for what it's worth. wp:lead asks for it to include "any prominent controversies". Best. Biosthmors (talk) pls notify me (i.e. {{U}}) while signing a reply, thx 19:42, 21 July 2016 (UTC)

Well the truth is there is no lead section because there are no sections. The lead section repeats/summarizes the body of the article's key points, a mini-article one can copy to elsewhere (such as a sub-section on McPherson in the article on near term extinction if it existed). To do that we would need to create sub-sections. I don't really care and am fine how it is given limited content. It just didn't seem right to put so much weight on comments critical in the first paragraph while other substantial parts of his life such as long academic career get a passing mention of one or two sentences ie. wp:weight / @Biosthmors: -- GreenC 00:39, 22 July 2016 (UTC)