User talk:Cyberpower678/Archive 39

This is an archive of past discussions about User:Cyberpower678. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 35

←

Archive 37

→

Senior Editor II

User appearance

Hi, I would like to know how have you changed the appearance of your user (User:cyberpower678) on the user page, for instance. --2212ca (talk) 12:27, 2 October 2016 (UTC)

{{DISPLAYTITLE}} will do it for you.—^cyberpower_Chat:Online 12:32, 2 October 2016 (UTC)

Thank you! Can I use it on my Hebrew page? --2212ca (talk) 12:48, 2 October 2016 (UTC)

If the template exists there, yes, otherwise just use the magic word.—^cyberpower_Chat:Online 12:49, 2 October 2016 (UTC)

Issuance of warning to ARUNEEK

I mistakenly pressed a wrong criterion in reason for deletion of Doomscrewed while using TWINKLE.Moments later I realized the eror and deleted it and again put up a correct one.What's the problem?Aru@baska^{❯❯❯ Vanguard} 15:48, 3 October 2016 (UTC)

Internet archive bot

I think this edit was quite problematic. The dead urls in question had already been archived using Webcite. The modifications made broke the links to Webcite, thus actually going against the bot's purpose. Is it possible to teach it to avoid changing links to Webcite and other archiving services? — Chris Woodrich (talk) 02:07, 1 October 2016 (UTC)
Nothing broke there. The URLs were converted to long-form as mandated by a recent community consensus.—^cyberpower_Chat:Online 12:17, 1 October 2016 (UTC)
That's odd. When I clicked on it the other day, the long form wasn't working. Now it is. Alright, NVM. — Chris Woodrich (talk) 14:45, 3 October 2016 (UTC)
There was probably a temporary WebCite outage when you clicked it.—^cyberpower_Chat:Online 14:47, 3 October 2016 (UTC)
Perhaps. The short form did work, hence my post here. Anywho, thanks. — Chris Woodrich (talk) 01:41, 4 October 2016 (UTC)

InternetArchiveBot – AFHRA

I see your bot has been rescuing citations that have been orphaned by the recent move of the entries on the Air Force Historical Research Agency to a new platform, because I received alerts for pages on my Watchlist. I believe there may be more on articles I don't watch, though. Since InternetArchiveBot likes a human to look after its work after the fact, is there a location where its recent changes are listed? I know that some of its dead link markings are incorrect, probab ly because they are to searchable databases, and to use them, agreements with TOS are required. --Lineagegeek (talk) 22:00, 3 October 2016 (UTC)

(talk page stalker) @Lineagegeek: I guess Special:Contributions/InternetArchiveBot will work? nyuszika7h (talk) 22:11, 3 October 2016 (UTC)

(talk page stalker) See also Category:Articles with unchecked bot-modified external links Optimist on the run (talk) 12:29, 4 October 2016 (UTC)

Talk:Administrative Office of the United States Courts

Dead link: http://findfederalagency.com/administrative-office-united-states-courts

This is to notify you of the above dead link, as stated in you bots revision of the above Article.

Birdymckee (talk)Birdymckee Birdymckee (talk)5 October, 2016 (0218 hrs.) —Preceding undated comment added 09:18, 5 October 2016 (UTC)

Okay, what about it?—^cyberpower_Chat:Online 14:28, 5 October 2016 (UTC)

Chart listing archives.

I checked these and was just wondering why the bot grabbed them? Reverted for the moment as the links are to the most up-to-date listings in the charts. Apologies if I overlooked something really obvious. Karst (talk) 09:36, 5 October 2016 (UTC)

I'm wondering what on earth happened there. Clearly a bug.—^cyberpower_Chat:Online 14:29, 5 October 2016 (UTC)

InternetArchiveBot: add or replace?

Thanks for addressing dead links with the wayback machine. In two recent edits I have seen, this one replaces the dead link with the archive.org copy, while this one leaves the (still dead) link in place and supplements it using a specific archive link, using {{wayback}}. Is there a rationale for the difference? David Brooks (talk) 14:32, 5 October 2016 (UTC)

Yes. The former happens for non-references and the latter for references.—^cyberpower_Chat:Offline 22:45, 5 October 2016 (UTC)

IABot is a bit confused

For this edit (note also edit summary), IABot left this Talk page notification. Note the claimed number of links rescued vs. actual changes in the diff; and the second link (www.heathamhouse.org.uk) in the notification. Also, is explicit mention of "0 dead links" (vs. omitting it when not relevant) the intended behaviour? --Xover (talk) 17:56, 5 October 2016 (UTC)

FAQ.—^cyberpower_Chat:Offline 22:45, 5 October 2016 (UTC)

Bot oops

It's been a long time since I've seen the bot do an error. The bot messed up on Adaptive hypermedia and Adrian Gunnell. I've fixed the both. Both articles have the same issue that caused the bot to do the same oops on both. Bgwhite (talk) 05:11, 5 October 2016 (UTC)

Are you referring to the references tag being converted to a ref? Or am I missing something else?—^cyberpower_Chat:Online 14:27, 5 October 2016 (UTC)

I'm usually confused and missing my brain. Take a look at the article's reference section after the bot was done. You will notice it doesn't look normal. One thing the bot did was add an extra <ref> tag that wasn't needed. Bgwhite (talk) 19:13, 5 October 2016 (UTC)

Here's] another fun-filled error Bgwhite (talk) 06:34, 7 October 2016 (UTC)

Question re: protocol

Hi had a question. If a URL doesn't contain a protocol will IABot process it? For example in an infobox |website=www.example.com vs. |website=http://www.example.com .. will IABot attempt to rescue in both cases or just the later? -- GreenC 15:21, 6 October 2016 (UTC)

No it won't. I somehow doubt it renders as a hyperlink on the software unless it's surrounded by brackets in the template.—^cyberpower_Chat:Offline 00:36, 7 October 2016 (UTC)

Ok that's good. Trying to determine what should be done with official URL's in infobox's. The solution may be to simply say, it should not be a hot link in the first place, only a printed URL "www.example.com". But there will be various opinions and various solutions. If consensus is for a hot link, it probably exceeds the scope of IABot to do anything different than it already does, and perhaps something like WM follows behind and fixes (whatever that fix is). Somewhat complicated problem. Similar to the {{official}} problem but different. -- GreenC 13:44, 7 October 2016 (UTC)

InternetArchiveBot removing archive urls

See [1]. The bot removed existing archive urls with the edit summary "Rescuing 8 sources and tagging 0 as dead. #IABot (v1.2.4)" - Evad37 [talk] 14:18, 7 October 2016 (UTC)

Note I've started a TfD for {{cite archives}} at Wikipedia:Templates_for_discussion/Log/2016_October_7#Template:Cite_additional_archived_pages. Rationale given there. -- GreenC 15:05, 7 October 2016 (UTC)

Badly behaved bot

I am feeling a bit frustrated about IABs edits. First your bot should not be editing {{Query web archive}} nor replacing the url value with an archive link (see diff showing added Wayback links). Second your bot can not seem to tell when a link is dead or unfit (see diff marking the dead ADV Films link as live (ie. |deadurl=no) and diff 2 the PopCultureShock.com links that now redirect to what appears to be a spam site). If you can't fix this soon please stop your bot until you can. Regards. – Allen4names (contributions) 20:32, 7 October 2016 (UTC)

@Trappist the monk: For CS templates that have the archiveurl and archive date fields set, how does the template behave when the deadurl field is omitted? My understanding is that omitting it is that same as deadurl=no. I may have that wrong. The bot is very capable of detecting dead links but it doesn't immediately acknowledge the site as dead unless also tagged as dead.

@Allen4names: As for the templates, I'm working on an update to ignore URLs found inside unrecognized templates. It will be part of the v1.3 release.—^cyberpower_{Chat:Limited Access} 20:41, 7 October 2016 (UTC)

—Trappist the monk (talk) 21:03, 7 October 2016 (UTC)

I think it preferable that the url value be a canonical link element in any case. – Allen4names (contributions) 21:15, 7 October 2016 (UTC)

Report

Tracked in Phabricator
Task T147691

[2]. Saw perhaps 50 like this reported in Category:Pages with wayback template errors. I manually fixed them, and assumed they were an old bug. But still happening as of 1.2.4 -- GreenC 17:40, 6 October 2016 (UTC)

So let's see what we have here. We have a bracketed URL detected as such, but still converted to a cite template inside those brackets, a wayback tag attached outside the brackets correctly, but a URL inside the Wayback template got converted to a cite template inside. :/ Lovely. Quite the cascading error.—^cyberpower_Chat:Offline 00:40, 7 October 2016 (UTC)

Most of the earlier ones were made July 3-6 when there was a definite bug since fixed. The tracking cat is helpful to monitor when new problems show up. BTW as bad as this mangled ref looks, it's in a predictable order (having manually fixed so many). -- GreenC 13:36, 7 October 2016 (UTC)

Do me a favor? Can you file all the bugs you've found so far onto seperate Phab tickets and include the InternetArchiveBot project in them?—^cyberpower_Chat:Offline 00:44, 7 October 2016 (UTC)

Good idea, they are all over the place. -- GreenC 13:36, 7 October 2016 (UTC)

For the Phab Subscribers field, is there a list-name that includes everyone, or otherwise a recommended list of people to include? For tags do you want InternetArchiveBot v1.2 or v1.3? -- GreenC 23:25, 7 October 2016 (UTC)

Put them in v1.3 and also add the project for the community tech team, that should automatically subscribe the necessary people. 1.2, despite bugs, is stable enough and isn't causing too many errors.—^cyberpower_Chat:Offline 23:34, 7 October 2016 (UTC)

I added "Community_Tech_bot" to subscribers. There's no project field, but there are some community-tech options for the tags field also. -- GreenC 23:54, 7 October 2016 (UTC)

Project and tags are almost the same thing on Phab.—^cyberpower_Chat:Offline 00:45, 8 October 2016 (UTC)

A barnstar for you!

	The Technical Barnstar
	For your well-programmed bot which has fixed countless sources and improved reliability everywhere Jjjjjjdddddd (talk) 00:50, 10 October 2016 (UTC)

Current AfD list

This very useful sortable listing of current AfDs, made by Cyberbot I, has not been updated since 29 August. May we have it back please?: Noyster (talk), 16:28, 8 October 2016 (UTC)

I'll into it when I get the chance.—^cyberpower_Chat:Online 14:18, 10 October 2016 (UTC)

Dead link fix?

This edit by Cyberbot II, with Edit summary 'Rescuing 1 sources, flagging 0 as dead, and archiving 0 sources. #IABot' doesn't seem to have provided anything of incremental use for the citation. The wayback machine url goes nowhere; to 404; and looks identical to the dead link url; and the date looks scrambled at best too. Background: I had reestablished the dead link template already, today, when I looked back at an earlier edit and came upon this Cyberbot edit; finding in the process that I was the one who did the last dead link template a year ago. Thanks in advance. Swliv (talk) 01:15, 9 October 2016 (UTC)

This is way too old to act on. Please report bugs from the current version, being 1.2.4.—^cyberpower_Chat:Online 14:19, 10 October 2016 (UTC)

IABot is archiving links to a non-dead PDF

Several times in the last few days, InternetArchiveBot has archived links to this non-dead PDF. The website that the file is on does not appear to have problems (note that the bot does not modify other links to the same domain). See edits here, here, and here. I would appreciate if you could fix whatever bug is causing these unnecessary edits. Thanks, Pi.1415926535 (talk) 05:30, 10 October 2016 (UTC)

There live_states were set to the dead state. Once there it requires manual intervention to reset.—^cyberpower_Chat:Online 14:36, 10 October 2016 (UTC)

Ok. So when this happens, is there any way to fix it other than informing you here? Pi.1415926535 (talk) 15:00, 10 October 2016 (UTC)

I am working on that solution as we speak. This will allow users to directly manage IABot's memory themselves. :-)—^cyberpower_Chat:Online 15:06, 10 October 2016 (UTC)

InternetArchiveBot

Please could you tell me why the IAB is changing the short "WebCite" URLs with the long version? Also why is it changing date formats to be inconsistent? What is the empty parameter "df =" it is adding? Why is it altering the layout of refs? It was brought to my attention by this edit. SagaciousPhil - Chat 06:01, 10 October 2016 (UTC)

For this edit [3] See Wikipedia:Using_WebCite#Use_within_Wikipedia and the RfC. -- GreenC 14:25, 10 October 2016 (UTC)

Evidently my mind reading powers are not up to scratch then? I always add an archive URL when inserting a ref to a website so will in future add the long WebCite URL instead as now very recently dictated by about a dozen people. However, that does not answer why the date format and layout were changed; I also cannot see what the empty "df=" parameter is supposed to be. SagaciousPhil - Chat 14:56, 10 October 2016 (UTC)

The df parameter is a date format parameter. You can set it to dmy-all or mdy-all to have the cite template format the dates in the output. It gets filled by the bot when there is a date format template on the page. If the bot detects a newline in the wiki-markup if the segment it's modifying, it switches to multiline formatting. On a last note, I'm not appreciative of your seemingly sarcastic tone on my talk page. I appreciate a sense of humor, and frustration, but not sarcasm.—^cyberpower_Chat:Online 15:14, 10 October 2016 (UTC)

Thanks, I do not appreciate being spoken to in the tone you are assuming either, especially when I am simply trying to ask about errors your bot has introduced. SagaciousPhil - Chat 15:24, 10 October 2016 (UTC)

Evidently my mind reading powers are not up to scratch then?

Struck as sarcasm to me. I have no issues with questions and am happy to answer them.—^cyberpower_Chat:Online 16:03, 10 October 2016 (UTC)

SORTKEY

Hi. I think this is a misunderstanding. SORTKEY does not allow "&" as a character. It is replaced by "and" in all other pages. This does not affect the title displayed. It was only to do with the categorisarion order. -- Magioladitis (talk) 14:49, 12 October 2016 (UTC)

Thank you for the clarification.—^cyberpower_Chat:Online 16:46, 12 October 2016 (UTC)

InternetArchiveBot creating error messages

When the Bot is archiving references it creates "Check |url= value" error messages (see Category:Pages with URL errors). Is it possible to modify the Bot to prevent this from happening, for example by having it add just the deadlink, archiveurl and archivedate parameters rather than the whole cite web template. EdwardUK (talk) 17:48, 8 October 2016 (UTC)

Looks like a bug.—^cyberpower_Chat:Online 14:18, 10 October 2016 (UTC)

Here's another one. And another one. – Jonesey95 (talk) 04:57, 13 October 2016 (UTC)

And another one, after you were notified of this problem. Please either check and revert the bot's errant edits or stop the bot and debug. Thanks. – Jonesey95 (talk) 04:47, 14 October 2016 (UTC)

It's finally

Fixed. The templates were not being recognized as such. It took a bit before I found the culprit.—^cyberpower_Chat:Online 11:44, 14 October 2016 (UTC)

InternetArchiveBot

Hi, I think the bot should skip adding archive urls to http://www.iucnredlist.org in citations to the IUCN Redlist from species pages (such as here: Ameerega pulchripecta). The IUCN front page contains only links to some news, nothing that is relevant for the citation. The relevant link is the link to the species page, e.g. [4] in this case. However, this is also getting redundant as IUCN has started to use DOI's. So in this particular case, adding the archive url only adds unnecessary burden. Otherwise, keep up the good work! — Micromesistius (talk) 14:22, 11 October 2016 (UTC)

The bot isn't programmed to skip archiving of certain links, and adding that would make it needlessly complex. Sorry.—^cyberpower_Chat:Online 11:48, 14 October 2016 (UTC)

Your InternetArchiveBot marked a valid link "dead"

InternetArchiveBot, in this edit [5], added a "deadurl=yes" parameter within a <ref> tag, but the URL in this reference ([6]) works perfectly well. Maybe it was down at the moment the bot check it? or perhaps the bot can't handle URLs with question marks in them? Anyway, the bot also added an archiveurl value that seems totally gratuitous, unless your goal is to duplicate every single link on Wikipedia with an archiveurl pointing to the URL on the date it was added. I've reverted the change, but I thought you might like to know. — Lawrence King ^(talk) 00:27, 12 October 2016 (UTC)

It did the same here, but instead a reference and in an external link. But the link is not dead. Hawkeye7 (talk) 02:17, 13 October 2016 (UTC)

There's an interface in the works for this, so regular users can submit changes to IABot's memory core.—^cyberpower_Chat:Online 11:50, 14 October 2016 (UTC)

Marvel Cinematic Universe

Various sources in articles about films set in the MCU have been archived using WebCite, but don't seem to be loading. I request you to look into this. Kailash29792 (talk) 07:43, 13 October 2016 (UTC)

I can't do anything about that sorry. Contact WebCite.—^cyberpower_Chat:Online 11:51, 14 October 2016 (UTC)

You no longer consider yourself active in xTools

Hello Cyberpower678, I have just read your message that you no longer consider yourself active in xTools on this talk page. May I point you to the fact that on [7] (at the bottom) you are still listed as the first contact person and even in bold, somehow indicating that you are the one to be contacted first. I am not complaining, just trying to be helpful to you and others when I suggest you to change your name to non-bold (or remove it, if you prefer). --85.181.49.137 (talk) 12:43, 13 October 2016 (UTC)

Thanks for that. It should be unbolded.—^cyberpower_Chat:Online 11:52, 14 October 2016 (UTC)

Question about archive URL formats

Hey Max, I know there was consensus for InternetArchiveBot to use long form archive URLs (rather than hash URLs) for services like archive.is and webcitation. But I wasn't aware of any consensus for InternetArchiveBot to replace existing short archive URLs with long ones (like this). Was this accidental or on purpose? IMO, it doesn't seem like a useful thing for the bot to be doing. If a citation already has both a url and an archiveurl, shouldn't we just skip it? Kaldari (talk) 01:40, 15 October 2016 (UTC)

There is consensus for both.—^cyberpower_Chat:Offline 02:06, 15 October 2016 (UTC)

fixed/replaced tick box?

I have just fixed some of the references flagged up at Alice Cooper. It included three links that the bot could not retrieve. I replaced one and managed to fix two manually. I tagged this in brackets and italics after it, to show other editors that I have done this. I am just wondering if a formatted option with a tick box or radio button next to it would be an option. Just a suggestion. Karst (talk) 10:44, 13 October 2016 (UTC)

I don't quite understand what you mean.—^cyberpower_Chat:Online 11:52, 14 October 2016 (UTC)

Apologies for the confusion. Let me use a different example. Here the bot highlighted a dead link to a blog. I checked it, the link is indeed irretrievable through archive.org. And as it is a blog, not a desirable reference to use anyway. So I removed it from the article (diff). Now this poses the following issue. When I set the review to true it highlights: "Archived sources have been checked to be working". And while this is correct for the archived links that the bot (rightly) changed, there is no option to show that the ^{[dead link]} problem has been dealt with. For the time being I have put (removed) in italics next to it. What I was suggesting is perhaps a more elegant solution similar to the true/false option. But if you think this works fine, I am happy to continue with it. Just a suggestion. Karst (talk) 11:37, 15 October 2016 (UTC)

Its really up to the community. The bot doesn't make use of it, it's just there to prevent other users from repeating the same work someone else just did. I'm working on an update that allows users to directly interact with IABot to make its database even better.—^cyberpower_Chat:Online 13:27, 15 October 2016 (UTC)

Sounds great! Thanks. Karst (talk) 13:32, 15 October 2016 (UTC)

Odd InternetArchiveBot edit to Nobel Prize broke nobelprize.org link

A recent edit by InternetArchiveBot broke the info box link to nobelprize.org on Nobel Prize in Literature. I've fixed this, but is something you want to look at to see why the bot made this edit? —C45207 | Talk 21:41, 13 October 2016 (UTC)

I see nothing wrong with the edit itself. Nothing appears to have broken.—^cyberpower_Chat:Online 11:53, 14 October 2016 (UTC)

When I said "broke the info box link", I was trying to convey that it no longer linked to a site controlled by the Nobel Prize, not that the rendered output looked different. It pointed to http://www.webcitation.org/5EsureXKk: this is not a site controlled by the Nobel Prize, and images on it don't render--at least for me. I assume this change was made because the site was deemed to be down because it was having availability problems due to the recent announcements. nobelprize.org is working now. Would the bot eventually have changed the link back to nobelprize.org? —C45207 | Talk 06:32, 16 October 2016 (UTC)

In an infobox, the |website= should ideally display a printer-friendly host name like "www.nobelprize.org" (see {{official URL}} created for this). However there is no current mechanism or consensus for dealing with dead links in the |website= field. There are opinions, like Codename Lisa's comment here who says it should not have any archive added and left alone even if dead. I suggested doing something like this: {{webarchive |url=https://web.archive.org/web/../.. |title=www.nobelprize.org }} to preserve a printer-friendly host display. And there are editors who don't agree (or know) that a printer-friendly URL is the right solution. Probably the least disruptive solution is simply skip processing |website= entirely until there is consensus (my recommendation). The second least disruptive would be preserve the original URL and text as given with a trailing {{wayback}} that has no title arg. The third would be replace the url and text with a {{wayback}} using |title=www.nobelprize.org option above. -- GreenC 17:03, 14 October 2016 (UTC)

The rendered output did not change, just the link.—^cyberpower_Chat:Online 17:20, 14 October 2016 (UTC)

True. There have been cases of bare URL links where the rendered shows the archive link. There are also some wrapped in templates like {{URL}} and {{official URL}}. -- GreenC 18:25, 14 October 2016 (UTC)

InternetArchiveBot error?

Hi! Does your bot have a whitelist for non-dead urls? If so, would you kindly add this url to it? Or could you kindly specifically exclude that url (which is used in hundreds of articles) from its activities in some way? The url works fine, so there's no need for it to be tagged as dead (as the bot did, for example, here). Many thanks, Justlettersandnumbers (talk) 19:49, 16 October 2016 (UTC)

Problem with <references> with IAbot

I've reported this before and it is still not fixed. When an article has:

</references>

IAbot turns it into:

<ref>

</references>

Examples are Area 25 (Nevada National Security Site) and Area 27 (Nevada National Security Site). Bgwhite (talk) 05:30, 18 October 2016 (UTC)

bot problem with comment tag

I see this around 5 times a day. in Arguments for and against drug prohibition, IAbot turns this:

<ref>{{cite news| last = Goldacre| first = Ben |date = June 2008| title = Cocaine study that got up the nose of the US| url = http://www.guardian.co.uk/commentisfree/2009/jun/13/bad-science-cocaine-study| work = Bad Science| publisher = ''The Guardian'' | location=London}} - re. [http://www.tdpf.org.uk/WHOleaked.pdf International study on cocaine executed by the World Health Organization].</ref>

Into:

<ref>{{cite news |last=Goldacre |first=Ben |date=June 2008 |title=Cocaine study that got up the nose of the US |url=http://www.guardian.co.uk/commentisfree/2009/jun/13/bad-science-cocaine-study |work=Bad Science |publisher=''The Guardian'' <!-- |archiveurl=http://www.webcitation.org/5hYgOP6zc?url=http%3A%2F%2Fwww.guardian.co.uk%2Fcommentisfree%2F2009%2Fjun%2F13%2Fbad-science-cocaine-study |archivedate= |location=London |deadurl=no |df= }} - re. [http://www.tdpf.org.uk/WHOleaked.pdf International study on cocaine executed by the World Health Organization]. </ref>

1) Without the closing comment tag, everything after the opening comment tag is blank. In this case, both the archive link and the second URL outside the cite template do not show up in the article.
2) IAbot leaves |archivedate= blank.
Bgwhite (talk) 05:58, 18 October 2016 (UTC)

Correct link replaced by archive link

In Talk:Hat/Archives/2022/March#External_links_modified your bot has failed to identify the dead link correctly and has replaced a working link instead. I have manually checked and corrected all links now, so currently there is (as far as I can say) no problem with this article any more. But you might wish to check what was going wrong to further improve the bot, which is very useful in the majority of cases. --78.53.236.134 (talk) 10:31, 18 October 2016 (UTC)

Vicious cycles

It appears as though Cyberbot II is having an edit war with itselfon The Gents (British band). On another note, whenever the aforementioned bot tags a link as dead, AnomieBOT adds the date a few minutes later; perhaps this is something your bot could do itself? --Highly Я!d¡cüłoʉ$ ^chat?_oops… 01:22, 19 October 2016 (UTC)

Asclepias curtissii

I noticed the bot marked a dead link with this edit. The EOL link works fine. Some plant database sites present temporary issues such as site down. I'm wondering if the bot distinguishes these from a true 404. Thanks, Declangi (talk) 21:01, 19 October 2016 (UTC)

BOT bug report

Hi, in this edit the link was indicated it was dead where the link is not dead. The organisation split into 2 last year and the details need updating but the original link still finds the correct page. I have updated to use the {{IoE}} template which gives the updated organisation details. I think you will find that all of the links to this domain work OK and do not need archiving. Keith D (talk) 16:48, 20 October 2016 (UTC)

IABot thinks a live link is dead

In this edit IABot changed a reference to http://www.hyperdictionary.com/dict-e/p-44.html to an archive link despite the site working fine for me. I've been noticing this happen a lot lately, I think the bot's "dead link" detection code has a flaw of some kind. —RP88 ^(talk) 03:10, 21 October 2016 (UTC)

Album chart

Just a quick question – does InternetArchiveBot support {{Album chart}}? A lot of old links to capif.org.ar will need archiving because they redesigned their site, moved from rankings.aspx to rankings, and rankings before September 2016 are not available at the new location. nyuszika7h (talk) 11:49, 21 October 2016 (UTC)

Actually it only needs to support the manual referencing mode of the template for now, as the automatically generated links point to an even older version of the site. nyuszika7h (talk) 11:52, 21 October 2016 (UTC)

InternetArchiveBot incorrectly flagged a live link as a dead link

In this edit, I think InternetArchiveBot incorrectly flagged a live link as a dead link. I will attempt to correct that edit, but I would be grateful if you could double-check InternetArchiveBot's dead link detection. Thanks! zazpot (talk) 18:30, 21 October 2016 (UTC)

Links incorrectly considered as dead

Your bot added |archive-url= for live urls, see Talk:Avishai_Cohen_(bassist)#External_links_modified. Thanks for your work otherwise! − Pintoch (talk) 16:31, 22 October 2016 (UTC)

Preemptively archiving URLs is good practice as long as you add |dead-url=no at the same time. If the bot did not do that in this case, it is possible that the page was not accessible temporarily. – Jonesey95 (talk) 19:29, 22 October 2016 (UTC)

InternetArchiveBot error?

Hi! Does your bot have a whitelist for non-dead urls? If so, would you kindly add this url to it? Or could you kindly specifically exclude that url (which is used in hundreds of articles) from its activities in some way? The url works fine, so there's no need for it to be tagged as dead (as the bot did, for example, here). Many thanks, Justlettersandnumbers (talk)

And here, and in a number of other places too. Would you kindly stop it from doing this? Thank you, Justlettersandnumbers (talk) 09:19, 24 October 2016 (UTC)

Inter-Parliamentary Union

I'm fairly sure I've raised this before, but the InternetArchiveBot is still flagging live links to www.ipu.org as dead. Can you put the website on whatever list it uses to override reports as it's getting a bit tiresome having to revert it. Thanks, Number 5 7 11:40, 24 October 2016 (UTC)

WebCite

Does InternetArchiveBot check WebCite for possible archives when attempting a rescue? SpinningSpark 13:39, 24 October 2016 (UTC)

It does not, there API is way to slow to be used reliably. It does however support WebCite, and absorbs them from existing sources.—^cyberpower_{Trick or Treat:Online} 13:45, 24 October 2016 (UTC)

Web archive templates merger proposal

Wikipedia:Templates_for_discussion/Log/2016_October_24#Web_archive_templates -- GreenC 16:53, 24 October 2016 (UTC)

Buggy IABot edit - something related to comments in cite templates, probably

Check this one, please. Here's a similar one.– Jonesey95 (talk) 02:52, 25 October 2016 (UTC)

Level 1 warning received in error

I recieved a warning that I removed an AfD template, from Cymbal (app), I closed an AfD, was reverted by Light2021 and I have still recieved the error. Thanks Nordic Nightfury 14:05, 25 October 2016 (UTC)

InternetArchiveBot error

Tracked in Phabricator
Task T147691

InternetArchiveBot destroyed a reference.(diff) It made a link to an archived version of the domain name, but ignored the actual reference and produced a confused tangle of a reference. Hopefully this behaviour has been fixed now.--Auric talk 12:46, 26 October 2016 (UTC)

Dead links: National Heritage List for England

InternetArchiveBot is tagging links as dead when they're not. Please see Wikipedia talk:WikiProject England#Dead links: National Heritage List for England. --Redrose64 (talk) 18:55, 27 October 2016 (UTC)

I already saw it and reset the domain in the DB. The algorithm behind it doesn't see it as dead, so it should only be a problem is someone else is tagging the site as dead.—^cyberpower_{Trick or Treat:Limited Access} 18:57, 27 October 2016 (UTC)

IABot thinks a live link is dead

In this edit IABot marked links to http://www.aoml.noaa.gov/phod/goos/xbtscience/reports/2013.pdf and http://www.aoml.noaa.gov/phod/goos/meetings/2008/XBT/index.php as dead despite both links working fine for me. This is the third time I've reported a bug of this nature to you, each time regarding a different site. Since you didn't respond to my previous notes, I don't know if you are looking into this issue. While I am sure the vast majority of the bots edits are fine, the fact that I've noticed errors from the bot in my watchlist three times, despite my relatively small watch list suggests that these errors are more frequent than they probably should be. Please investigate your bot's "dead link" detection code, it really looks like it has a flaw of some kind. —RP88 ^(talk) 11:09, 28 October 2016 (UTC)

I am. I'm not ignoring you. But if I responded to every dead link report, I wouldn't get anything done. :-)—^cyberpower_{Trick or Treat:Online} 13:56, 28 October 2016 (UTC)

It would be great if the bot exposed a log somewhere that listed when and why it determined certain links are dead. In the cases I've reported the bot thought the links were dead but I found them to live not that much later. I am happy to help investigate if you can provide some details. —RP88 ^(talk) 14:28, 28 October 2016 (UTC)

The algorithm is not at fault here. I tested the algo and it returns alive. This leads to 2 possibilities, a lingering false positive that simply needed resetting, or someone is tagging the site as dead somewhere, and the bot is picking up on that.—^cyberpower_{Trick or Treat:Online} 14:30, 28 October 2016 (UTC)

Can't you check your database to see which is the case? If not, I'd argue that the bot's database should keep enough metadata about links to determine exactly why it thinks a link is dead. I find it very unlikely that the bot would have previously encountered both of the above links on some other previously processed page. Also, are you saying that if a link is erroneously marked dead by a user this bot will propagate that error to potentially hundreds of other articles containing a reference to the same URL? I think it is fine for the bot to try to add an archive URL to a link that has been manually marked dead in an article, but I do not think it is OK for it to mark all other uses of that URL in other articles as dead without actually verifying that it is dead. —RP88 ^(talk) 15:05, 28 October 2016 (UTC)

According to the DB, it only exists on a single page as far as IABot is aware. IABot has quite a bit of metadata. It's DB is over 15GB in size. The bot has a configuration option called "tag_override" on it's configuration page. During trials it was deemed necessary to be switched on. Tag_override forces the bot to accept the tag and apply the dead state to that URL. When switched off, it will add it's own judgement to the tag, and either remove it if found alive, or agree with it if found dead and accept the URL as dead. This will of course have cons of it's own. IABot cannot detect soft 404s.—^cyberpower_{Trick or Treat:Online} 15:13, 28 October 2016 (UTC)

I have no problem with that behavior (although I think it should not be propagated to other articles). But that doesn't explain what happened here. When you made your comment I just in the process of mentioning that an external link search shows that Bathythermograph is the only article that links to http://www.aoml.noaa.gov/phod/goos/xbtscience/reports/2013.pdf . So we know that this link could not have been determined to be dead due to someone tagging the URL as dead somewhere else on wikipedia and this was the first edit by IABot to that page, so that would seem to also rule out a false positive from a previously encountered reference. It looks like neither of the two cases you propose can be the cause of this error. —RP88 ^(talk) 15:26, 28 October 2016 (UTC)

How does a lingering false positive from an old algorithm not fit the case? The algorithm could have reported it as dead at one point, but a bug fix to it recently, may have fixed it, and so the URL needed only to be reset in the DB.—^cyberpower_{Trick or Treat:Online} 15:29, 28 October 2016 (UTC)

Thanks for the explanation. Let me restate this so that I can make sure I understand you correctly. Are you saying that at some point in the past you found an error in the dead link detection code. Links marked as dead in the database due to actual attempts by bot to verify the URL prior to this point in time may be erroneous. This bug presumably does not apply to links detected to be dead by the bot noticing a manual "dead link" annotation in an article. If that is true, can't you fix a whole bunch of these issues in one fell swoop by just removing all links marked as dead due to URL verification that were added to the database prior to the date of the bug fix? Of course that would decrease performance for a bit as the bot would have to reverify some of those old links if it encounters them again instead of being able to rely on the cached state from the database, but that would seem to be a small price to pay to get rid of a bunch of false positives caused by a lingering bug that you've already fixed. —RP88 ^(talk) 15:44, 28 October 2016 (UTC)

Well that would be the entire DB. I'm not sure how happy Labs would be when I start hogging most of their bandwidth, again, to go crawling other pages, when it could be avoided. I have an easier solution, of simply reseting the entire domain in the DB, when someone brings up a false positive. This is of course after verifying the algorithm isn't the cause of the false positive. IABot's built around speed, consequently, while IABot's speed isn't impacted by making the checks as much, Labs will definitely notice the spike in bandwidth usage. I try to keep checks and other uses of heavy functions to a minimal. An additional reason to not constantly go crawling around the web again, is that certain sites will end up blocking the bot, which defeats the purpose of the anti-bot mask the algorithm uses to evade bot detection. This will ultimately create an even higher false positive rate.—^cyberpower_{Trick or Treat:Online} 15:53, 28 October 2016 (UTC)

Did you crawl the majority of WP's external links on some previous occasion to prepopulate your database or are the majority of links encountered being processed for the first time now? If it is the later, I just extracted the external links from 20 random articles, and the overlap of URLs between articles would seem to be quite low, so the bandwidth expansion would be relatively minor since most links are being processed for the first time anyways. Also you woudn't have to reset the entire DB, links learned to be dead by user annotation wouldn't have be removed and any link determined to be dead by the bot via direct inspection after the date of the bug fix wouldn't have to removed. In any case, I wanted to mention that I appreciate you taking the time to talk out these issues with me and explaining your process of handling false positives. —RP88 ^(talk) 16:13, 28 October 2016 (UTC)

They were all already crawled before. It took a lot of resources to get the job done. I'm working on a GUI that allows users to have more direct control over IABot.—^cyberpower_{Trick or Treat:Online} 16:20, 28 October 2016 (UTC)

OK, thanks for the update. If I see future "live" links that are marked as dead by IABot I'll just report them here until your user-facing GUI is available. —RP88 ^(talk) 16:22, 28 October 2016 (UTC)

Vicious cycles

It appears as though Cyberbot II is having an edit war with itself on The Gents (British band). On another note, whenever the aforementioned bot tags a link as dead, AnomieBOT adds the date a few minutes later; perhaps this is something your bot could do itself? Thanks, Highly Я!d¡cüłoʉ$ ^chat?_oops… 20:25, 28 October 2016 (UTC)

That task is old and should be disabled.—^cyberpower_{Trick or Treat:Limited Access} 21:50, 28 October 2016 (UTC)

Sometimes the site is available and just the page has changed position

I wonder if there is a way to detect these instead of sending things to the web.archive. Example. -- Magioladitis (talk) 22:43, 28 October 2016 (UTC)

Acting on this post I made this edit. Of the two links, I'm certain of the second, but not so sure of the first, not knowing what the page looked like before. --Redrose64 (talk) 22:50, 28 October 2016 (UTC)

Nope. Considering they don't redirect to the new URL, the bot can't look for a new one.—^cyberpower_{Trick or Treat:Offline} 22:59, 28 October 2016 (UTC)

InternetArchiveBot error?

Hi! Does your bot have a whitelist for non-dead urls? If so, would you kindly add this url to it? Or could you kindly specifically exclude that url (which is used in hundreds of articles) from its activities in some way? The url works fine, so there's no need for it to be tagged as dead (as the bot did, for example, here). Many thanks, Justlettersandnumbers (talk)

And here, and in a number of other places too. Would you kindly stop it from doing this? Thank you, Justlettersandnumbers (talk)

And here too. Please understand that this becomes more and more annoying as the bot continues to make the edit and the operators continue to ignore the complaint. Cyberpower678, Kaldari (assuming I've got that right), would you kindly stop the bot from marking that particular live link as dead and irrecoverable? I'd be grateful. Justlettersandnumbers (talk)

@Cyberpower678: This looks like it might be another case which was likely fixed by your patch from a few days ago. Would it be possible to purge all URLs from the deadlink cache that have %20 in them so that they can be rechecked? Kaldari (talk) 22:42, 28 October 2016 (UTC)

Again, it's not a cache, it's a memory core. It's been built on the fact that once a site goes down, it tends to never come back. As such the stored values stay at 0 once it hits 0. I'm doing entire domain resets when they get reported, and I hope to have my web interface for this up soon.—^cyberpower_{Trick or Treat:Offline} 23:13, 28 October 2016 (UTC)

Checklinks

Hi CP, the Checklinks tool by Dispenser mirrors some of the functionality of IABot, except it's user run-on-demand. It adds {{dead link}}, replaces with archive links (including other archives besides Wayback). However it has problems that need to be fixed and is breaking the {{wayack}} template and other stuff. I spoke with Dispenser and he says it needs a major rewrite and he doesn't have time and the source is closed (may contain some copyright code). Usage is not heavy thankfully but still creates problems with {{wayback}}. The intersection between Checklinks and {{wayback}} increases daily due to IABot adding so many. Do you think it would it be relatively easy to adapt a version of IABot to a single-article run-on-demand triggered by something? That could replace Checklinks. Checklinks recent changes. -- GreenC 00:36, 29 October 2016 (UTC)

That's a planned release for the webapp.—^cyberpower_{Trick or Treat:Offline} 00:46, 29 October 2016 (UTC)

Oh great. Very good. That solves a lot. Is that feature in development or in a future version? How's that going BTW is it turning into a big project? -- GreenC 00:55, 29 October 2016 (UTC)

How to document that I've found a replacement URL

Here, I found that the Seattle Times had moved the content to a new URL, so I just reversed the IAB edit and put in the new URL. I assume that's fine, but how should I record that on the talk page? Neither true, false, nor failed results in the correct text. Can you give advice in the FAQ? David Brooks (talk) 20:51, 29 October 2016 (UTC)

Just set it to true, or leave a note. It's only there so other editors don't repeat your work needlessly.—^cyberpower_{Trick or Treat:Limited Access} 20:55, 29 October 2016 (UTC)

When will IAbot be fixed?

You never commented on my bug reports and the problems are still happening. I filled the reports a week ago. My bug reports were archived. The bot is still destroying references. Bgwhite (talk) 22:55, 24 October 2016 (UTC)

[8], [9]. etc. -- Magioladitis (talk) 23:01, 24 October 2016 (UTC)

Sorry, I never meant to ignore your report. I'll work on a fix and have it hopefully tomorrow.—^cyberpower_{Trick or Treat:Offline} 01:56, 25 October 2016 (UTC)

Fixed in v1.2.6.—^cyberpower_{Trick or Treat:Offline} 22:23, 26 October 2016 (UTC)

Archive Links on Bodacious Bull

Hi, I'm a pretty new editor around a month. But long time tech writer outside of here. I am hoping you can help me, please? Anyway, want to start doing some write in a topic you add two archive links to, the Bodacious_(bull). In the talk page, back in March 2016, you left a note. No one paid much attention to it, other to move the links from external to internal. The checked parameter was never added. And for some reason the Title parameter is missed and there's an error on both links because of it. I checked the archive links and they are both working and the source material is good (I know alot about bullriding). I got the title to work once, but from then on nothing I tried will work. I went to the citation template topic and it really looks like I'm using it correctly. I can't spend all day on this. Also not sure where the checked parameter goes but I feel it can be added with the True value. Thanks for any response. Dawnleelynn (talk) 17:54, 29 October 2016 (UTC)

I'm a little confused of what you're asking of me. Btw, my bot made those edits you mentioned.—^cyberpower_{Trick or Treat:Limited Access} 20:54, 29 October 2016 (UTC)

Thanks for answering! Okay, a bot did it, I didn't realize that, good to know. I'm sure your bot did not cause those links to be missing the Title parameter, which is what I was having trouble adding. Anyway, I just tried to do the edits again and it worked. I think it was user error-using the wrong key for the | key. Feel so stupid. Anyway, sorry I wasn't clear but it's working now and all is okay. I appreciate that you answered so quickly. Have a great week! Dawnleelynn (talk) 02:14, 30 October 2016 (UTC)