This is an archive of past discussions with User:Legobot. 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.
As seen here, some RfC listing entries are seriously indented, others are not. I don't see a logical reason or purpose for this. There's something to be said for neatness and consistency, and in the 21st century how hard can it be for the software to prevent this indentation?
If the "solution" is simply for RfC creators to do something different in their formatting, I think that's a unreasonable expectation; correct RfC formatting is already too complicated for many editors. ―Mandruss☎19:36, 4 January 2018 (UTC)
I now see that the relevant software is the {{Rfcquote}} template. I'll cross-post there, but as I'm creating its talk page I don't know how much attention it will receive there. Consider this thread a notice of that discussion. ―Mandruss☎19:45, 4 January 2018 (UTC)
@Bellezzasolo: You were selected for this FRS notification because the RfC is listed in Wikipedia:Requests for comment/Maths, science, and technology, although that is not linked from the RfC as it presently stands. However, at the time that the RfC was initiated, it was in |sci. When Thewolfchild (talk·contribs) made this edit, they failed to remove the |rfcid=8B62314 parameter, so Legobot didn't fully update its data tables - it added the RfC to the style list and to the politics list, but did not remove it from the science list. When category parameters are added to an ongoing {{rfc}} template, the addition is handled fine by Legobot; but it is a known bug (no fix is presently expected) that when one or more are removed, Legobot ignores the removal unless|rfcid= is removed at the same time. This removal causes Legobot to rebuild its data tables for that RfC. --Redrose64 🌹 (talk) 08:52, 4 February 2018 (UTC)
Here's another one. Changing the date in the "Old business" section to January 1, 1970 with this edit on March 6. Thanks. --StarcheerspeaksnewslostwarsTalk to me15:32, 13 March 2018 (UTC)
This is obviously silly, so I figured I'd let you know. Probably a good idea to exclude blocked users from the FRS notices. Headbomb {t · c · p · b}14:01, 22 March 2018 (UTC)
Hello Legoktm. I doubt many people publish a RfC about RfCs, so you probably wont see this behavior often. Nevertheless, I thought I'd let you know. Legobot twice linked the example given in the proposal as if it were an actual RfC. The first time the example was between <syntaxhighlight lang=""> tags,[3] and the second time the example was between <nowiki> tags.[4] I removed the example and substituted another to prevent a third occurrence. Best regards.--John Cline (talk) 17:18, 26 March 2018 (UTC)
(talk page watcher) @John Cline: This is a known problem. Legobot looks for two opening braces followed by the three letters "rfc" (case-insensitive), and upon encountering that five-character sequence assumes that it is the start of an open RfC. Legobot doesn't care if these five characters are inside HTML comment tags <!--...--> or enclosed by other common methods for hiding code, such as <nowiki>...</nowiki>, <pre>...</pre>, <source>...</source> - such markup is ignored. The only proven methods of defeating that behaviour involve making sure that these five characters don't occur together in the page source. Four ways that I have used successfully are: (i) use a template-linking-template such as {{tlx|rfc}}; (ii) go through a redirect that doesn't begin with the letters "rfc" - this is why {{closed rfc top}} and {{closed rfc bottom}} exist; (iii) use numeric character entities { as in {{rfc; (iv) remove the sequence entirely. --Redrose64 🌹 (talk) 23:26, 26 March 2018 (UTC)
Thank you Redrose64. I appreciate that information; I had thought it was the level 2 header followed by rfc that caused the problem. would {{((}}rfc|hist{{))}} provide a work-around or would Legobot consider }}rfc as two template brackets followed by rfc? I know that you said two opening brackets but wonder why this wouldn't be a fifth work-around.--John Cline (talk) 02:08, 27 March 2018 (UTC)
It might work; I didn't mention it because I've never tried it. It seems an awfully long-winded way when one of the template-linking templates will do the necessary task with up to six fewer characters; Template:T is the shortest - {{rfc}}. --Redrose64 🌹 (talk) 14:42, 27 March 2018 (UTC)
It looks like the Talk:Joss Whedon page's archive index hasn't been updated since June 2017. Is there any way I can trigger a re-indexing? I verified the page code against the instructions--the page code had been <!-- HBC Archive Indexerbot can blank this --> , so I changed it to <!-- Legobot can blank this --> per the instructions in case that was the problem, but it's been four days and the index still isn't updated. Am I doing something wrong? — Narsil (talk) 19:01, 27 March 2018 (UTC)
Is this page monitored? The Joss Whedon archive index still isn't updating, and I'd love to know if (a) I made a mistake we can fix, (b) The indexing is delayed but will come, or (c) Indexing don't happen no more and I should delete the page... — Narsil (talk) 20:42, 10 April 2018 (UTC)
In our Talkback template left for new users when we've answered their questions at the Teahouse, we offer them a link to our archive of past questions. Nobody seems to have spotted that this ceased functioning in 2015. Is it a simple task to reactivate it? See Wikipedia:Teahouse/Questions/Archive Index. We're now on archive #750, but the page we (unhelpfully!) offer to newcomers stopped at #347. I'll post a TB template below this post. Regards from the Teahouse, Nick Moyes (talk) 01:24, 12 April 2018 (UTC)
The article does not display the GA icon, and it's just occurred to me that it probably won't so long as Legobot thinks it failed GAN. Factotem (talk) 10:59, 16 April 2018 (UTC)
GAN marked as on review when review page has been deleted
Hi Legoktm, WugBot's been running into a problem creating the GAN Report because the nomination for SethBling is being marked as on review (literally, ``exists=yes``) despite the review page not existing. It seems to be a reflex of the fact that a review was started by an IP and then deleted. I tried to fix the problem manunally but Legobot undid my edit. This behavior doesn't seem ideal, and I'm wondering why it's happening and if there's a fix. It's something that WugBot can handle, but if it's an easy fix on your end that would probably be better (since all WugBot would do is put it in malformed noms, but it can't be fixed since Legobot undoes the manual fix). If you could take a look and let me know, I'd be very appreciative. Wugapodes[thɔk][ˈkan.ˌʧɻɪbz]16:14, 3 May 2018 (UTC)
If you think this page should not be deleted for this reason, you may contest the nomination by visiting the page and clicking the button labelled "Contest this speedy deletion". This will give you the opportunity to explain why you believe the page should not be deleted. However, be aware that once a page is tagged for speedy deletion, it may be deleted without delay. Please do not remove the speedy deletion tag from the page yourself, but do not hesitate to add information in line with Wikipedia's policies and guidelines. —swpbTgo beyond16:17, 16 May 2018 (UTC)
Tyw7 I'm not sure you noticed at the top of Legotkm's talk page that it says they are on vacation until 7 July. I expect that ordinarily they are very responsive, but it sure is nice to leave work and stay away for an extended period of time. — jmcgnh(talk)(contribs)07:31, 16 June 2018 (UTC)
@Tyw7: (adding a {{ping}} template only pings if you also add a signature on the same edit) I'm afraid I don't have anything to add. There's been no response to our various queries, which gives an indication of how thinly supported this particular feature is. — jmcgnh(talk)(contribs)12:10, 18 June 2018 (UTC)
@Tyw7: Looking at the source code for the bot, it appears that the mask= parameter can start with a leading slash, so that it behaves just like the target= parameter. This looks like something that might need to be fixed in the documentation. But it may not help if Legobot isn't looking in user talk pages. I'm also intrigued that the bot has an undocumented alternative that it would use for pages with year/month-labeled archives but only for matches with full month names, not for month abbreviations or month numbers (like mine). — jmcgnh(talk)(contribs)07:33, 15 June 2018 (UTC)
@Tyw7: As far back as I looked in the month of June, Legobot has only touched archive indexes for Talk:Anatolia and Talk:Astrology. Now my experience is that the archive search box still does a full-text search of the archives without there needing to be an index, so it may be that there's no real need to index these archives. — jmcgnh(talk)(contribs)08:21, 15 June 2018 (UTC)
@Tyw7: I know. I'm not in any position of authority on this, I'm just reading the published docs and source code as best I can. If I make enough pesky comments, maybe we'll hear from someone who knows more. — jmcgnh(talk)(contribs)08:42, 15 June 2018 (UTC)
@Jmcgnh: Your pesky comments are much appreciated, and that warrants saying, so everybody sees it, and knows about it, and recognizes your effort. Bravo. So, thank you, jmcgnh! I'll also take this opportunity to urge everybody to use that "Thank" link in the Revision History more often: when you see something that makes you smile, or makes you realize that someone took the time to make the encyclopedia a little better, whether it be something like this Bot config that had me completely stumped, or a typo that needed fixing. I've been using the WP:THANK feature more lately, partly because I wonder if anyone cares, and I want them to know I saw their effort, and I recognize it. So, pay it forward, and thank someone today! Mathglot (talk) 08:58, 15 June 2018 (UTC)
@Tyw7: All the evidence indicates that Legobot is not looking at User talk space for archive index opt ins. While I can see the code that I think Legobot is running for this function, I haven't figured out its overall run configuration. We can see that it works for Talk: space, but have not found it to work for User talk: space. — jmcgnh(talk)(contribs)04:35, 16 June 2018 (UTC)
@Tyw7: Still nothing? I saw Legoktm on IRC recently, so they're apparently still alive, but also apparently still on vacation from dealing with this issue. The bot updated the index for Talk:Astrology/Archive index earlier today, so I think it's simply not configured to look for OptIn configs in User talk: space. It's not a matter of having the wrong coding, it's just not configured to look there currently. I suspect there's a file somewhere where this could be verified, but in the vast edifice that is Wikipedia, one sometimes needs a Virgil to say which bits are the ones actually running and which are just artifacts from some past version. — jmcgnh(talk)(contribs)07:01, 29 June 2018 (UTC)
@Tyw7: You're right, something's still screwy about that. I tried finding the Legobot logs, not having seen any responses from Legoktm, but no luck. Someone on #wikimedia-cloud also checked and did not see any published logs. I did find some more indexing script that supports the mask1, mask2 syntax. But I'm still lacking context on how Legobot looks for OptIns in various namespaces. The on-wiki configuration and log files don't appear to have been touched since 2012 or 2013. Without access to the current configuration and logs, I don't see how to determine whether there's an error that prevents Legobot from indexing these files or if it is simply not configured to discover that they've requested to be indexed. — jmcgnh(talk)(contribs)15:55, 29 June 2018 (UTC)
Hi Legobot: I am a GA-co-nominator for Samuel May Williams. I apologize for using the templates and the nomination page incorrectly, but I think I now understand the process better. Samuel May Williams is deceased, so I will reclassify the nomination under the World History subcategory. Thank you, Oldsanfelipe (talk) 19:36, 23 September 2018 (UTC)
RockMagnetist, this has been a known bug for years, which occurs when a talk page has a previous FailedGA template on it along with the new GA template; the bot for some reason does not go for the newest, but rather for the earlier FailedGA. The bot is not actively being supported, and various attempts to find a new operator have failed. Unfortunately, it's unlikely to be solved anytime soon; I have added the GA icon to the article, which the bot normally does when it correctly identifies a new GA. Sorry about the confusion. BlueMoonset (talk) 06:13, 4 October 2018 (UTC)
Duplicate notifications for feedback service
Not sure if this is a bug or if anything can be done about it, but the Legobot notified me twice for the feedback request service for the same item, once on Sept. 7 and once on Oct. 9th, and I'd already commented. Is it supposed to keep track of who has already been notified? Here are the alerts. [5] and [6]. TimTempleton(talk)(cont)16:49, 10 October 2018 (UTC)
@Timtempleton: If you look carefully at your two diffs, the seven characters after Talk:Sephora#rfc_ are different: this is a giveaway that different |rfcid= values are involved. Looking at these three edits to the talk page, we find that the original RfC had reached its thirty-day limit, so Legobot (talk·contribs) removed the {{rfc|econ|rfcid=8E50BAD}} template; then Cunard (talk·contribs) re-added the {{rfc|econ}} template with a fresh timestamp in order to allow a further thirty days; but instead of reusing the original |rfcid=8E50BAD parameter, they left it off. The absence of a |rfcid= has caused Legobot to assume that it was a new RfC and so assigned a fresh value |rfcid=A860981 with the result that anybody who had been notified about the RfC when it was 8E50BAD is liable to be renotified about it as A860981. --Redrose64 🌹 (talk) 19:54, 10 October 2018 (UTC)
@Redrose64: Ah - I see - thanks for researching. This happens so rarely and I suppose seeing it relisted isn't that big a deal - perhaps there might be new comments that I could comment on. But I am also interested in a point I saw raised on the feedback request service about how it seems that we are pinged for mature discussions, when things have already been hashed out. Did you see this? Wikipedia_talk:Feedback_request_service#Late_invitationsTimTempleton(talk)(cont)21:03, 10 October 2018 (UTC)
I hadn't seen that, but I was aware of the issue: Maproom makes (unsigned) notes on their talk page (which is on my watchlist) concerning how many days have elapsed from an RfC being raised to the WP:FRS notification being sent. See for example User talk:Maproom#Please comment on Talk:Collapse of the World Trade Center which indicates that 9 hours had elapsed from initiation to notification. Some of the notifications have queries about the selection, which I have occasionally answered.
On the FRS subscription list, different people have different RfC categories, and various monthly limits - which need not be the same for each category. For example, Maproom subscribes to three RfC categories, with monthly limits varying from 3 to 12. It's described at WP:FRS that Legobot picks users at random; what it doesn't say is that Legobot also picks RfCs at random, subject to certain restrictions, including: that the RfC belongs to a subscribed category (obviously); that it is still an ongoing RfC (i.e., it still bears a {{rfc}} template); that the user has not already been notified about this RfC. The last one is controlled by the |rfcid= parameter, which if altered, can have the consequences that we discussed to begin with.
Legobot sends out FRS notifications once a day, beginning at approximately 04:20 (UTC): if it only did this for the newest RfC in each of the sixteen categories, two problems arise. First, there is the possibility that more than one RfC has been started in a given category in the 24 hours that have elapsed since the previous FRS run, so only one of those RfCs will be publicised by FRS - those that were started earlier will be passed over entirely. Second, each FRS category has more than one subscriber, so those users who were randomly selected will all be informed about the same RfC.
By allowing random ages for the RfCs, each user gets a fair chance of being informed about each RfC, and each RfC gets roughly the same amount of publicity. --Redrose64 🌹 (talk) 22:54, 10 October 2018 (UTC)
Thanks for the detailed explanation. I'll start making a note of the time lapse between when the rfc template is added and I'm notified by the service, and see if it is indeed random. I've been subscribing for a while, and my subscription is just one per week, but it seems I'm never among the first few people arriving. TimTempleton(talk)(cont)02:12, 11 October 2018 (UTC)
I think there is a bug with the GA task. The bot keeps increasing the number of reviews made by a user (Vami IV) without the user making any additional reviews. --DannyS712 (talk) 06:41, 26 November 2018 (UTC)
It does that occasionally, it's been happening for years. When it happens, it's a sign that something is misformatted somewhere. I never did track down exactly what the triggers are. --Redrose64 🌹 (talk) 21:27, 26 November 2018 (UTC)
Good day Legobot! I have been working on cleaning up Linter errors, and the following pages could use a bit of cleanup, replacing {{collapse bottom}} with {{end}}. I made the necessary changes, but the bot overwrote them with code that sticks extra </div> tags into each page.
Please see the edits I had made to each of these pages for details on what to replace. It would be great if you could adjust the bot's code accordingly. Thanks! – Jonesey95 (talk) 18:38, 27 November 2018 (UTC)