User talk:Citation bot/Archive 15
 | This is an archive of past discussions. 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 10 | ← | Archive 13 | Archive 14 | Archive 15 | Archive 16 | Archive 17 | → | Archive 20 |
Unclosed templates lead to parsing errors and the bot to do very strange things
- What happens
- Two headings were moved to one line (open the link and search for 'With varying CO').
- Relevant diffs/links
- https://en.wikipedia.org/w/index.php?title=Calcium_carbonate&type=revision&diff=887674726&oldid=885707535
- We can't proceed until
- Feedback from maintainers
I will try to figure out this. But it seems to be big but rare GIGO. AManWithNoPlan (talk) 13:09, 15 March 2019 (UTC)
- It's a case of unclosed template, as the previous greedy replace. Nemo 14:07, 15 March 2019 (UTC)
- This will catch most of those (basically a <ref> inside as CS1/CS2 template => bomb out): https://github.com/ms609/citation-bot/pull/1509 AManWithNoPlan (talk) 22:16, 17 March 2019 (UTC)
Caps: BJOG
Instead of Bjog. Headbomb {t · c · p · b} 12:49, 18 March 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1518 AManWithNoPlan (talk) 15:37, 19 March 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 14:08, 20 March 2019 (UTC)
HDL, DOI cleanup
- What should happen
- [1]
- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot/pull/1517 DOIs AManWithNoPlan (talk) 15:35, 19 March 2019 (UTC)
https://github.com/ms609/citation-bot/pull/1516 HDLs AManWithNoPlan (talk) 15:35, 19 March 2019 (UTC)
Added incorrect pmid, etc.
- Relevant diffs/links
- https://en.wikipedia.org/w/index.php?title=User%3AJosve05a%2Fcite-sandbox&diff=prev&oldid=887460997
- We can't proceed until
- Feedback from maintainers
The pubmed search does not allow you to search by title. You search by title keywords. On rare occasions we get false positives. It is very rare, I will look into doing a follow on search. AManWithNoPlan (talk) 23:23, 12 March 2019 (UTC)
Garbage DOI in, garbage all over
- Status
- mostly {{fixed}}
- Reported by
- ♦ J. Johnson (JJ) (talk) 21:49, 14 March 2019 (UTC)
- What happens
- Changes the source.
- What should happen
- Not change the source
- Relevant diffs/links
- diff
- We can't proceed until
- Feedback from maintainers
CitationBot has egregiously damaged a citation in this edit (bottom), where it relied on an incorrect doi to "fix" a citation by adding data from a different source. The assumption that the doi is always correct is unwarranted; there should be some kind of check or comparison between the original citation and what the doi pulls up. ♦ J. Johnson (JJ) (talk) 21:49, 14 March 2019 (UTC)
- People complain that we need to use DOIs to fix citations, then they complain that we need to trust the text. It is not a simple task to verify when titles are slightly different because of typos, etc. It can be quite circular at time. AManWithNoPlan (talk) 21:57, 14 March 2019 (UTC)
- Perhaps the bot should not be so aggressive in trying to "fix" things? I don't think we should tolerate bot-aided screwups just because they are bot-aided. One approach might be to display the data from the citation alongside the data from the doi source, letting the operator determine if they match. Another approach would be to do such changes in individual edits (and on a sectional basis) to facilitate reversion of specific edits. ♦ J. Johnson (JJ) (talk) 21:04, 15 March 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1482 AManWithNoPlan (talk) 04:03, 16 March 2019 (UTC)
- Not an ideal edit, but it was a positive one in that it helped identify a wrong DOI. Nemo 05:54, 16 March 2019 (UTC)
- The concept of screwing up something so badly that some else comes along to fix it, and incidentally notices something else to fix, is hardly a credit for the bot, and not a practice to be recommended. The same result could be had without corrupting the citation by simply trashing the DOI alone and tagging it for someone to check. ♦ J. Johnson (JJ) (talk) 19:45, 18 March 2019 (UTC)
- very hard to deal with since people often put series/chapter/journal/etc in title. AManWithNoPlan (talk) 18:46, 21 March 2019 (UTC)
e0152973
- What happens
- Bot changed the pages in the first citation to "e0152973".
- What should happen
- Not that.
- Relevant diffs/links
- https://en.wikipedia.org/w/index.php?diff=887968443&oldid=887428542&title=Sustainable_agriculture
- We can't proceed until
- Feedback from maintainers
Looks like source-data GIGO based on the citation generator, but I think it would be safe to say that's GIGO that we can look for. --Izno (talk) 17:25, 21 March 2019 (UTC)
- Is it even garbage at all? If you go to the doi and look at the recommended citation, it is "Barbier EB, Hochard JP (2016) Does Land Degradation Increase Poverty in Developing Countries? PLoS ONE 11(5): e0152973." So they appear to be saying to use e0152973 where the pages should go. —David Eppstein (talk) 17:35, 21 March 2019 (UTC)
- This citation is correct to list the pages as e0152973. A lot of journals including PLOS use something like the "article number" as the page number for citations. Each issue of the journal is not printed as one big book with continually increasing page numbers. —Chris Capoccia (talk) 19:17, 21 March 2019 (UTC)
{{wontfix}} these crazy page numbers are nuts. AManWithNoPlan (talk) 02:55, 23 March 2019 (UTC)
Untitled_new_bug
- What happens
- Deal with
|url=https://www.ncbi.nlm.nih.gov/pmc/articles/pmid/27036656/ - We can't proceed until
- Feedback from maintainers
Request: detect Publisher/journals that only differ by 'the'
- What should happen
- [2]
- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot AManWithNoPlan (talk) 22:47, 23 March 2019 (UTC)
Request: be consistent about which page number choice to make
Strange. It cannot make up its mind which database to believe. Both answers are correct values. AManWithNoPlan (talk) 15:50, 14 March 2019 (UTC)
- Page ranges that start with one are consider suspect. Pages are considered inferior to page ranges. It is a battle royal. AManWithNoPlan (talk) 15:53, 14 March 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1549 AManWithNoPlan (talk) 22:48, 23 March 2019 (UTC)
Japan Times
- What happens
- Added
|newspaper=The Japan Times Online - What should happen
- Also remove
|publisher=The Japan Times - Relevant diffs/links
- https://en.wikipedia.org/w/index.php?title=Caulerpa_lentillifera&diff=888848926&oldid=886083655
- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot AManWithNoPlan (talk) 22:48, 23 March 2019 (UTC)
Caps: NRC
Instead of Nrc, but see also User talk:Citation bot/Archive_15#Request: Smart acronym capitalization. Headbomb {t · c · p · b} 13:30, 23 March 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1554 AManWithNoPlan (talk) 15:07, 24 March 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 17:30, 3 April 2019 (UTC)
Caps: B/gcvs instead of B/Gcvs
To handle "VizieR On-line Data Catalog: B/gcvs" which should remain like that. Headbomb {t · c · p · b} 10:26, 24 March 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1554 AManWithNoPlan (talk) 15:06, 24 March 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 17:30, 3 April 2019 (UTC)
PMID url cleanup
- What should happen
- [5]
- We can't proceed until
- Feedback from maintainers
I had to remove this |url= manually before running the bot. (Also |journal=, but that's something the bot couldn't be expected to handle.) Headbomb {t · c · p · b} 11:02, 24 March 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1553 AManWithNoPlan (talk) 15:04, 24 March 2019 (UTC)
additional hdl resolver
- What should happen
- [6]
- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot/pull/1552 AManWithNoPlan (talk) 14:03, 24 March 2019 (UTC)
Fails to expand doi
- What should happen
- [7]
- We can't proceed until
- Feedback from maintainers
Facinating. The CrossRef database has no title. That naturaly fails to match old title. The patch after it is applied will not try to verify empty titles. https://github.com/ms609/citation-bot/pull/1558 AManWithNoPlan (talk) 21:54, 24 March 2019 (UTC)
- What kind of superhuman code does Citoid (Zotero) use to identify the title of that doi? Aren't there any public API we also can take advantage of? (t) Josve05a (c) 22:02, 24 March 2019 (UTC)
- Since we get a succesful cross-ref search, we do not use Zotero. That's a GIGO prevention policy. AManWithNoPlan (talk) 00:20, 25 March 2019 (UTC)
Caps: AAPOS instead of Aapos
Self-explanatory. Headbomb {t · c · p · b} 21:15, 25 March 2019 (UTC)
https://github.com/ms609/citation-bot/pull/1568 AManWithNoPlan (talk) 16:23, 27 March 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 17:28, 3 April 2019 (UTC)
Caps: IEEE S&P instead of IEEE S&p
Self-explanatory. Headbomb {t · c · p · b} 13:51, 26 March 2019 (UTC)
https://github.com/ms609/citation-bot/pull/1568 AManWithNoPlan (talk) 16:22, 27 March 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 17:28, 3 April 2019 (UTC)
Caps: PloS One, PloS one, Plos One → PLOS One
Self-explanatory. Headbomb {t · c · p · b} 02:51, 27 March 2019 (UTC)
They go by PLOS ONE now. AManWithNoPlan (talk) 04:49, 27 March 2019 (UTC)
- the journal named by Eustace from Narnia. It is “PLOS ONE”, but the “NE” is actually “ne” but in a font that lacks under-case characters. 🙄 AManWithNoPlan (talk) 12:50, 27 March 2019 (UTC)
https://github.com/ms609/citation-bot/pull/1568 AManWithNoPlan (talk) 16:22, 27 March 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 17:28, 3 April 2019 (UTC)
Caps: Angew Chem Int ed → Angew Chem Int Ed
Self explanatory. Headbomb {t · c · p · b} 03:19, 27 March 2019 (UTC)
https://github.com/ms609/citation-bot/pull/1568 AManWithNoPlan (talk) 16:22, 27 March 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 17:28, 3 April 2019 (UTC)
No longer seems to add bibcode?
- What should happen
- [8]
- We can't proceed until
- Feedback from maintainers
That is correct. Because of a coding oddity, the gadget API does not load the API keys. The fix is in the source code - I noticed this last week; but it is not deployed yet. AManWithNoPlan (talk) 15:57, 27 March 2019 (UTC)
Caps: Merip → MERIP
Self explanatory. Headbomb {t · c · p · b} 20:44, 29 March 2019 (UTC)
https://github.com/ms609/citation-bot/pull/1568 AManWithNoPlan (talk) 21:17, 29 March 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 15:38, 3 April 2019 (UTC)
Cite encyclopedia issue
- What happens
- https://en.wikipedia.org/w/index.php?title=User:Josve05a/cite-sandbox&diff=890285453&oldid=890285438
- We can't proceed until
- Feedback from maintainers
What is the exact issue? It might just be that I am tired too. AManWithNoPlan (talk) 17:08, 2 April 2019 (UTC)
- Adding
|title=already (kinda) present in|chapter=(and|encyclopedia=) (t) Josve05a (c) 19:18, 2 April 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1580 AManWithNoPlan (talk) 03:14, 4 April 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1579 AManWithNoPlan (talk) 03:14, 4 April 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1580 AManWithNoPlan (talk) 03:14, 4 April 2019 (UTC)
respect date format request templates
- What happens
- Does not respect {{Use mdy dates}} Inserted: "date=2019-03-25"
- What should happen
- Insert: "date=March 25, 2019"
- Relevant diffs/links
- [9]
- We can't proceed until
- Feedback from maintainers
- A proposed modification to how the cs1|2 templates work will make it unnecessary for citation bot to worry about how dates are formatted in wikitext. See the discussion.
- —Trappist the monk (talk) 15:07, 3 April 2019 (UTC)
- Our code does not allow dates in those templates. This should fix it once implemented: https://github.com/ms609/citation-bot/pull/1578 AManWithNoPlan (talk) 18:41, 3 April 2019 (UTC)
Request: Use author-link as we do with title-link
- What should happen
- Replace
|author=[[Example|Author1]]with|author=Author1|author-link=Exampleetc. - We can't proceed until
- Feedback from maintainers
The order should be |author=, |author-link=, not |author-link=, |author=. Headbomb {t · c · p · b} 21:27, 6 March 2019 (UTC)
- Adjusted above. (t) Josve05a (c) 10:44, 9 March 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1590 AManWithNoPlan (talk) 00:33, 8 April 2019 (UTC)
Do not add journal that matches existing title
- What happens
- Bot adds
|journal=The Building of Economics at Adelafor a book, which already has|title=The Building of Economics at Adela - What should happen
- (1) Don't add
|journal=to {{cite book}}. (2) Don't add|journal=where the same content is present in the|title= - Relevant diffs/links
- https://en.wikipedia.org/w/index.php?title=User:Josve05a/cite-sandbox&diff=889374931&oldid=889374918
- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot/pull/1585 AManWithNoPlan (talk) 15:16, 5 April 2019 (UTC)
Misunderstanding of DMY and MDY
- What happens
- Adds invalid ISO format dates
|date=09-23-2011 - What should happen
- Add dates in valid format, preferably in accordance with the appropriate {{use_mdy_dates}}.
- Relevant diffs/links
- https://en.wikipedia.org/w/index.php?title=Brandin_Cooks&diff=891262034&oldid=891261687
- We can't proceed until
- Feedback from maintainers
Interesting. MDY is not mm-dd-yyyy. I will change the code. AManWithNoPlan (talk) 22:08, 6 April 2019 (UTC)
- it looks like DMY is 13 June 2018 and MDY is June 13, 2018. Is that right? AManWithNoPlan (talk) 22:11, 6 April 2019 (UTC)
- Yes. (t) Josve05a (c) 22:26, 6 April 2019 (UTC)
- Once this pull is implemented. You will see "December 5, 2016" and "5 December 2016" date styles, when these templates are present. AManWithNoPlan (talk) 14:34, 7 April 2019 (UTC)
- Yes. (t) Josve05a (c) 22:26, 6 April 2019 (UTC)
Request: remove rogue title-links
Basically, if the link doesn't start the title, remove it. Headbomb {t · c · p · b} 17:41, 13 March 2019 (UTC)
- Current code removes links if more than one. Current code coverts to title-link if one. Current does not do what you want. I will leave this here for a week to see if anyone has a complaint against it. I think you are right. AManWithNoPlan (talk) 22:47, 13 March 2019 (UTC)
- The only objection I might have is if something is structured
|title=Series name: Specific title with many words here(t) Josve05a (c) 22:50, 13 March 2019 (UTC)- https://github.com/ms609/citation-bot/pull/1589 AManWithNoPlan (talk) 15:13, 7 April 2019 (UTC)
- The only objection I might have is if something is structured
Request: Smart acronym capitalization
In |journal=, when you've got a word that's only consonants, or only vowels, it's an acronym, and they should be in uppercase. Since Y can by either (most often a vowel), it shouldn't be considering in those.
- Bmj → BMJ; Prl → PRL, etc...
- Aaai Journal → AAAI Journal; Iau Proceedings → IAU Proceedings
I've done zillions of such cleanup, and I've yet to encounter a counterexample. If they exist, they are exceedingly rare, and can be bypassed manually. Headbomb {t · c · p · b} 15:59, 14 March 2019 (UTC)
- How about a list of acronyms, since I am always concerned about non-English words. AManWithNoPlan (talk) 14:40, 17 March 2019 (UTC)
- These lists would be utterly massive and near impossible to compile. Non-English words that are all consonants or all vowels would be close-to or just-as-rare as they would be in English, proportionally speaking, but would get cited orders of magnitude less. Those are easier to bypass when found than the other way around. Headbomb {t · c · p · b} 15:00, 18 March 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1587 AManWithNoPlan (talk) 18:08, 6 April 2019 (UTC)
- These lists would be utterly massive and near impossible to compile. Non-English words that are all consonants or all vowels would be close-to or just-as-rare as they would be in English, proportionally speaking, but would get cited orders of magnitude less. Those are easier to bypass when found than the other way around. Headbomb {t · c · p · b} 15:00, 18 March 2019 (UTC)
- How about a list of acronyms, since I am always concerned about non-English words. AManWithNoPlan (talk) 14:40, 17 March 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 14:40, 9 April 2019 (UTC)
Request: Please remove imbalanced brackets
https://github.com/ms609/citation-bot/pull/1586 AManWithNoPlan (talk) 23:39, 5 April 2019 (UTC)
Strip em-tags from new data
- Relevant diffs/links
- https://en.wikipedia.org/w/index.php?title=User:Josve05a/cite-sandbox&diff=889376487&oldid=889376457
- We can't proceed until
- Feedback from maintainers
They are used extensively in species names. AManWithNoPlan (talk) 13:53, 25 March 2019 (UTC)
- <i> (wikitext: '') is correct for species. The use of em inside titles should be rare. The bot should probably just strip styling/HTML in newly imported citations. --Izno (talk) 14:14, 25 March 2019 (UTC)
- Using
<em>...</em>for species names is improper markup. For species names in|title=, editors, and this bot, should use standard wiki italic markup:|title=article title about ''Spiecies name''. The same would apply to the example case where the journal article title is italicized for I-don't-know-what reason.<em>...</em>has a particular semantic meaning that is different from the mere rendering of text in italic font. - —Trappist the monk (talk) 14:26, 25 March 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1583 AManWithNoPlan (talk) 02:32, 5 April 2019 (UTC)
Citation bot
Why is the citation bot blocked? I find that tool very useful.BabbaQ (talk) 07:28, 9 April 2019 (UTC)
- See #March 2019 above. In the meantime, you can use the citation expander gadget. Headbomb {t · c · p · b} 09:58, 9 April 2019 (UTC)
{{Duplicate Issue}}
Request: do something with publisher/journal when wikilinked
- What should happen
- [14]
- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot/pull/1597 AManWithNoPlan (talk) 21:11, 9 April 2019 (UTC)
Publication place
- Status
- {{notabug}}
- Reported by
- Martin of Sheffield (talk) 09:41, 15 March 2019 (UTC)
- What happens
|publication-place=is incorrectly changed to|location=- Relevant diffs/links
- [ https://en.wikipedia.org/w/index.php?title=Safety_lamp&type=revision&diff=887842043&oldid=887224890 ]
- Replication instructions
- run automatically by Citation bot
- We can't proceed until
- Feedback from maintainers
The documentation for {{Citation}} states that |location= is an alias for |place=. This parameter is used for "news stories with a byline, that is, the location where the story was written". It does note that earlier usage was for the publication place and for compatibility "will be treated as the publication place if the publication-place parameter is absent". The damage is therefore minimised, but it is regrettable that Citation Bot is changing editors' correct work to current standards into a outdated form. Martin of Sheffield (talk) 09:41, 15 March 2019 (UTC)
- I wish CS1 and CS2 would agree on these things and not keep changing the documentation. AManWithNoPlan (talk) 13:08, 15 March 2019 (UTC)
- This documentation is consistent between CS1 and CS2 and has been for some time. See Template:Citation Style documentation#publisher, which under-the-hood refers to a single template that is the official documentation for the parameters in both CS1 and 2. That said,
|location=is by-far the most-used, so perhaps there should be a discussion at Help talk:CS1 about the parameters. Usage:- publication-place has about 14k uses.
- place has about 27k uses and then times out.
- location has about 300k uses and then times out.
- --Izno (talk) 13:29, 15 March 2019 (UTC)
- This documentation is consistent between CS1 and CS2 and has been for some time. See Template:Citation Style documentation#publisher, which under-the-hood refers to a single template that is the official documentation for the parameters in both CS1 and 2. That said,
- @Izno, is that not in part the historic use of location and also editors who "know how to" still using the old way? I would have expected the bot to be updating the old way to the new if it possible, not downgrading to fit in with the herd! If it is non-trivial, then the best thing would be to leave it be, the template makes a good fist of dealing with discrepancies. Martin of Sheffield (talk) 15:29, 15 March 2019 (UTC)
|publication-place=is not an 'upgrade' over the more concise and just as clear|location=. The solution is to update the CS1|2 documentation to use location by default, like people expect. Publication place should be fully deprecated. If, for some weird reason, people insist on specified where something was written, rather than published (which is was all style guides recommend), then you can have a specific parameter for that (e.g.|writing-place=or similar). Either way, the bot works fine as is. Headbomb {t · c · p · b} 15:45, 15 March 2019 (UTC)- @Headbomb, the place to raise this argument is at Template talk:Citation, not here. @Trappist the monk: may be able help with the rationale. Martin of Sheffield (talk) 14:30, 17 March 2019 (UTC)
- A better place to discuss this is probably Help talk:Citation Style 1. But, since I have been
{{noping}}ed into this conversation ...
- A better place to discuss this is probably Help talk:Citation Style 1. But, since I have been
-
- If I understand the rationale behind the simultaneous use of
|location=or|place=when the template also has|publication-place=, the latter is to specify the actual place of publication while either of the former specify the place of writing as one would find in the dateline of a newspaper article. This suggests a remedy.
- If I understand the rationale behind the simultaneous use of
-
- Because, per Editor Izno's researches,
|location=and|place=use far outnumbers|publication-place=use, perhaps we should:- deprecate
|publication-place= - create
|dateline=- for use only in
{{cite news}}or in{{citation}}when|newspaper=is set - this parameter would cause Module:Citation/CS1 to emit the 'Written at ...' text regardless of the presence of
|place=or|location=
- for use only in
- deprecate
- Of course, maybe we just don't care where something was written; the purpose of citations at en.wiki is to identify the source where we got some 'thing' that is claimed in an article; let the source say where it was written if the source's producers think it important. If this is the case, deprecate
|publication-place=and remove the code that emits 'Written at ...' We might also deprecate|place=because: why do we need two synonymous parameters especially when one of them is so obviously preferred over the other? - —Trappist the monk (talk) 15:04, 17 March 2019 (UTC)
- I'm happy with any solution. At the moment the documentation says "use publication-place", and as a fall back the earlier "location" will be interpreted if there is no publication-place. Citation bot reverses this. All I would ask is that either the documentation deprecates publication-place or else citation bot follows the documentation. What is annoying is when you carefully check the documentation and follow it, then citation bot comes along and does the opposite. The reason for mentioning you Trappist is that you seem to keep the citation templates under your wing, and would therefore be a court of ultimate appeal. Regards, Martin of Sheffield (talk) 15:35, 17 March 2019 (UTC)
- Because, per Editor Izno's researches,
- @Headbomb, the place to raise this argument is at Template talk:Citation, not here. @Trappist the monk: may be able help with the rationale. Martin of Sheffield (talk) 14:30, 17 March 2019 (UTC)
- @Izno, is that not in part the historic use of location and also editors who "know how to" still using the old way? I would have expected the bot to be updating the old way to the new if it possible, not downgrading to fit in with the herd! If it is non-trivial, then the best thing would be to leave it be, the template makes a good fist of dealing with discrepancies. Martin of Sheffield (talk) 15:29, 15 March 2019 (UTC)
- Is there something that the bot needs to change or does it keep on doing what it is doing? AManWithNoPlan (talk) 02:58, 4 April 2019 (UTC)
Check if blocked before trying to edit page
- What happens
- Citation bot runs through its entire wikitext editing routine and then tries to submit without checking if it is blocked first.
- What should happen
- Citation bot should check that it is not blocked before trying to make an edit. If it is blocked, the procedure should be aborted. Also, the Citation bot homepage should display a banner indicating that the bot is blocked whenever that is the case.
- Relevant diffs/links
- GitHub gist
- We can't proceed until
- Feedback from maintainers
Hope springs eternal. AManWithNoPlan (talk) 20:56, 9 April 2019 (UTC)
- @Qzekrom and AManWithNoPlan: Actually, wouldn't it make more sense to act as if the "commit edits" checkbox was unchecked in that case? (and, when the bot is blocked, show the homepage as usual with "commit edits" marked as disabled and a message explaining such). {{3x|p}}ery (talk) 23:28, 9 April 2019 (UTC)
- Yes. Although, it's generally unwise to spend development resources on a temporary situation such as this block. Nemo 06:21, 10 April 2019 (UTC)
Publisher/Journals that only differ by &
https://github.com/ms609/citation-bot/pull/1607 AManWithNoPlan (talk) 17:34, 12 April 2019 (UTC)
Request: clean-up Volume vs issue if the same number
If |volume=73 exists but no |issue= for a cite, but the bot adds |issue=73 based on e.g. doi, the bot should validate the |volume=. If it isn't 73 at source, a user most likley added it in the wrong parameter.
- In article: "Example". 73.
{{cite journal}}: Cite journal requires|journal=(help) - The bot adds issue: "Example". 73 (73).
{{cite journal}}: Cite journal requires|journal=(help)
In most cases volume and issues are different, and a check should be carried out to see if |volume=73 was ever true to begin with. (t) Josve05a (c) 15:24, 3 March 2019 (UTC)
- We have special code for journals that do not have issues or do not have volumes. Other than that, this is hard to catch reliably. AManWithNoPlan (talk) 18:37, 3 March 2019 (UTC)
- The word "issue" and "volume" are saddly on some occasions used as interchangeable, so I doubt other than adding specific journals, there is nothin that the bot can do. AManWithNoPlan (talk) 18:54, 3 March 2019 (UTC)
- Real example. In this case, where both issue and the volume ends up with the same value, it should clear the volume parameter and check the source if another value is available for
|volume=. If not, add back original value. (t) Josve05a (c) 19:02, 3 March 2019 (UTC)- There are some real issues with order within the code. I will have to ponder. AManWithNoPlan (talk) 19:16, 3 March 2019 (UTC)
- Looking at this now. AManWithNoPlan (talk) 13:31, 11 April 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1608 AManWithNoPlan (talk) 20:18, 12 April 2019 (UTC)
- Looking at this now. AManWithNoPlan (talk) 13:31, 11 April 2019 (UTC)
- There are some real issues with order within the code. I will have to ponder. AManWithNoPlan (talk) 19:16, 3 March 2019 (UTC)
- Real example. In this case, where both issue and the volume ends up with the same value, it should clear the volume parameter and check the source if another value is available for
- The word "issue" and "volume" are saddly on some occasions used as interchangeable, so I doubt other than adding specific journals, there is nothin that the bot can do. AManWithNoPlan (talk) 18:54, 3 March 2019 (UTC)
Marking as {{fixed}} since in GIT master now. Will go live at next update. I do not normally flag until live, but this is a feature and not a bug. AManWithNoPlan (talk) 14:29, 15 April 2019 (UTC)
Foreign capitalization
Since capitalization of foreign titles is often a matter of preference, these keywords will match a significant portion of these titles (French, German).
- -ische (i.e. ische at the end of any word)
- Abhandlungen
- Actes
- Annales
- Archiv
- Archives de
- Archives du
- Archives des
- Beiträge
- Berichten
- Blätter
- Bulletin de
- Bulletin des
- Bulletin du
- Cahiers
- Carnets
- Comptes rendus
- Die
- Fachberichte
- Jahrbuch
- Journal du
- Journal de
- Journal des
- Journal für
- L'
- La
- Le
- Mitteilungen
- Monatshefte
- Monatsschrift
- Mémoires
- Notizblatt
- Recueil
- Revue
- Travaux
- Studien
- Wochenblatt
- Wochenschrift
- Zeitschrift
- Études
What I mean by this is that if you have, e.g. |journal=Revue d'Histoire littéraire de la France or |journal=Revue d'Histoire Littéraire de la France or |journal=Revue d'histoire littéraire de la France, the bot should those alone. On those title, only the articles and preposition should be touched. E.g. |journal=Revue D'Histoire Littéraire de la France → |journal=Revue d'Histoire Littéraire de la France. I don't know how feasable that is, however. Headbomb {t · c · p · b} 23:39, 4 March 2019 (UTC)
- Let me see if I understand. If we find these words, then do not change existing capitalization? AManWithNoPlan (talk) 00:46, 5 March 2019 (UTC)
- that’s a pain: you want some cleanup; not all or nothing. Will think about. AManWithNoPlan (talk) 02:36, 13 April 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 16:11, 16 April 2019 (UTC)
Bare references
The number of bare references with DOIs has decreased by maybe 10 % from January to March, many persist: phabricator:P8395 (this also includes some pages in the Draft namespace). Should citation bot be more pro-active with those, for those who are using the gadget now? Nemo 13:41, 14 April 2019 (UTC)
- Sadly, WP:CITEVAR limits what is possible to do with automated runs. However, with manual runs, it's relatively easy to cleanup individual articles. Headbomb {t · c · p · b} 13:46, 14 April 2019 (UTC)
- I checked a few of these randomly and the articles containing them didn't seem to follow any citation style at all. Maybe I was lucky in my selection. Nemo 15:09, 14 April 2019 (UTC)
- If you point to a specific page or two, I can look at them. AManWithNoPlan (talk) 15:50, 14 April 2019 (UTC)
- The examples I find would be to much WP:CITEVAR, so {{notabug}}. AManWithNoPlan (talk) 23:20, 15 April 2019 (UTC)
- If you point to a specific page or two, I can look at them. AManWithNoPlan (talk) 15:50, 14 April 2019 (UTC)
- I checked a few of these randomly and the articles containing them didn't seem to follow any citation style at all. Maybe I was lucky in my selection. Nemo 15:09, 14 April 2019 (UTC)
Messes with italics in cite web titles
- What happens
- [17]
- What should happen
- leave italics alone here
- We can't proceed until
- Feedback from maintainers
- Might need to cover more than just {{cite web}}
|title=too. {{Cite journal}}|title=and others should be left alone as well. Headbomb {t · c · p · b} 15:18, 17 April 2019 (UTC)- https://github.com/ms609/citation-bot/pull/1614 AManWithNoPlan (talk) 04:03, 18 April 2019 (UTC)
Dtsch
- What happens
- [18]
- What should happen
- Keep 'Dtsch' as 'Dtsch'
- We can't proceed until
- Feedback from maintainers
The first violation of the auto-capitalization code. This list should be pretty short. AManWithNoPlan (talk) 14:01, 19 April 2019 (UTC)
Need to run twice, again
I do not see it with Guglielmino, A.; Olmi, M. (2010). "Revision of Nearctic species of Esagonatopus, with description of a new species from Florida (Hymenoptera, Dryinidae)". ZooKeys (70): 57–66. doi:10.3897/zookeys.70.764. ISSN 1313-2989. AManWithNoPlan (talk) 02:01, 21 April 2019 (UTC)
Better volume/issue/number handling?
- What should happen
- [22]
- We can't proceed until
- Feedback from maintainers
Not sure what's possible here, but this is worth investigating. Headbomb {t · c · p · b} 17:34, 19 April 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1619 AManWithNoPlan (talk) 17:11, 20 April 2019 (UTC)
Use ISSN to find journal name when missing
In a citation like
- Ahn, Kee-Jeong (1997). "A review of Liparocephalus Mäklin (Coleoptera: Staphylinidae: Aleocharinae) with descriptions of larvae". 73: 79–92. ISSN 0031-0603.
{{cite journal}}: Cite journal requires|journal=(help)
The bot doesn't do much. However, following the ISSN reveals that the name of the journal is |journal=Pan-Pacific Entomologist. The bot should be able to figure that
- Ahn, Kee-Jeong (1997). "A review of Liparocephalus Mäklin (Coleoptera: Staphylinidae: Aleocharinae) with descriptions of larvae". Pan-Pacific Entomologist. 73: 79–92. ISSN 0031-0603.
is better. (Likewise for {{cite magazine}}.) Headbomb {t · c · p · b} 20:15, 19 April 2019 (UTC)
https://github.com/ms609/citation-bot/pull/1617 AManWithNoPlan (talk) 01:29, 20 April 2019 (UTC)
- The name to be put into
|journal=must be the name of a journal at the time it was published and printed on the journal, not some later name. Can we be sure that the ISSN is always assigned to a single journal name only? What happens if the name of a journal changes (perhaps only slightly)? Is it mandantory for the ISSN to be changed as well then? - I'm asking because this isn't the case for ISBNs - I have seen cases, where different editions of a book where published under the same ISBN, so working from the ISBN alone it is not possible to reliably identify a particular edition of a book in all cases.
- If this would be the case for ISSNs as well, deriving a journal name from an ISSN (alone) will only give a clue, but will still need to be verified by a human, and should not automatically be inserted by a bot. --Matthiaspaul (talk) 15:20, 5 May 2019 (UTC)
- Journal names do change over time: even DOIs will almost always give the "current" name, etc. The first question is a matter of "net good": does a feature improved wikipedia on average. The second question is, would a human being actaully do any better, which the answer is that they would do the exact same search and insert the exact same current name. Even if they did go back and find the exact journal webpage for the exact issue, the new journal name would be emblazoned across the top of the website. The third question is, since references exist so that people can verify data, which name will help them find it better, new or old: unfortunatelly, from my own experiences, the newer wrong name is actaully better. Lastly, I should note that ISBN can in rare cases point to multiple books (that's a bad publisher!); so GIGO is a risk in all things. AManWithNoPlan (talk) 15:54, 5 May 2019 (UTC)
- The idea is not to rely on the ISSN first, but as a last resort. If you have a DOI and no journal, the journal gets added based on the doi. This is for the case where there's an ISSN, but no DOI nor journal. The citation was already extremely poor before, but this is a net improvement. Headbomb {t · c · p · b} 15:07, 6 May 2019 (UTC)
- the code will be so last resort that it might result in a situation where running the bot again results in more changes. AManWithNoPlan (talk) 13:49, 9 May 2019 (UTC)
- The idea is not to rely on the ISSN first, but as a last resort. If you have a DOI and no journal, the journal gets added based on the doi. This is for the case where there's an ISSN, but no DOI nor journal. The citation was already extremely poor before, but this is a net improvement. Headbomb {t · c · p · b} 15:07, 6 May 2019 (UTC)
- Journal names do change over time: even DOIs will almost always give the "current" name, etc. The first question is a matter of "net good": does a feature improved wikipedia on average. The second question is, would a human being actaully do any better, which the answer is that they would do the exact same search and insert the exact same current name. Even if they did go back and find the exact journal webpage for the exact issue, the new journal name would be emblazoned across the top of the website. The third question is, since references exist so that people can verify data, which name will help them find it better, new or old: unfortunatelly, from my own experiences, the newer wrong name is actaully better. Lastly, I should note that ISBN can in rare cases point to multiple books (that's a bad publisher!); so GIGO is a risk in all things. AManWithNoPlan (talk) 15:54, 5 May 2019 (UTC)
{{fixed}}
have to run it twice
I cannot reproduce this. AManWithNoPlan (talk) 22:59, 19 April 2019 (UTC)
- Yasushi Komori; Kohji Matsumoto; Hirofumi Tsumura (2010). "Hyperbolic-sine analogues of Eisenstein series, generalized Hurwitz numbers, and q-zeta functions".
{{cite journal}}: Cite journal requires|journal=(help)
Running on the above reproduces it for me. Running twice will produce the final output. Headbomb {t · c · p · b} 23:07, 19 April 2019 (UTC)
- There are some data integrity checks in the code that value certain data sources as more than others. Thus there can be some order dependence in the code. AManWithNoPlan (talk) 23:31, 19 April 2019 (UTC)
- code adds bibcode. Bibcode record contains arXiv record that gets added. It is too late at this point to use arXiv code. Will see about checking arXiv code a second time. AManWithNoPlan (talk) 23:33, 20 April 2019 (UTC)
- This situation is concerns mostly journals that don't get indexed by ADSABS, but that have preprints on the arxiv. This is a relatively common situation in mathematics and computer science. You could either parse the arxiv or ADSABS comments to search for a doi in a non-standard location, although in my experience, arxiv is more reliable there. Headbomb {t · c · p · b} 00:01, 21 April 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1623 AManWithNoPlan (talk) 02:10, 21 April 2019 (UTC)
- This situation is concerns mostly journals that don't get indexed by ADSABS, but that have preprints on the arxiv. This is a relatively common situation in mathematics and computer science. You could either parse the arxiv or ADSABS comments to search for a doi in a non-standard location, although in my experience, arxiv is more reliable there. Headbomb {t · c · p · b} 00:01, 21 April 2019 (UTC)
- code adds bibcode. Bibcode record contains arXiv record that gets added. It is too late at this point to use arXiv code. Will see about checking arXiv code a second time. AManWithNoPlan (talk) 23:33, 20 April 2019 (UTC)
Caps: bei
https://github.com/ms609/citation-bot/pull/1621 AManWithNoPlan (talk) 02:05, 21 April 2019 (UTC)
New handle repository
- What happens
|url=https://repository.si.edu/handle/10088/16510- What should happen
|hdl=10088/16510, see [27]- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot/pull/1622 AManWithNoPlan (talk) 02:04, 21 April 2019 (UTC)
Fails to expand because of GIGO filters
- What happens
- Fails to expand
- Pope, R. D. (1981). "Rhyzobius ventralis (Coleoptera: Coccinellidae), its constituent species, and their taxonomy and historical roles in biological control". 71: 19–31. doi:10.1017/S0007485300050999. ISSN 0007-4853.
{{cite journal}}: Cite journal requires|journal=(help)
- What should happen
- Expand to
- Pope, R. D. (1981). "Rhyzobius ventrails (Coleoptera: Coccinellidae), its constituent species, and their taxonomy and historical roles in biological control". Bulletin of Entomological Research. 71: 19–32. doi:10.1017/S0007485300050999. ISSN 0007-4853.
- We can't proceed until
- Feedback from maintainers
The title is wrong and so it gets rejected. Perhaps differing only by single quotes isn’t too different AManWithNoPlan (talk) 01:47, 21 April 2019 (UTC)
- Quotes only is a minor difference. But also here the DOI is already given, so it should just expand based on what's given. It's not like it's trying to find a DOI based on a close-but-not-quite match title. Headbomb {t · c · p · b} 02:20, 21 April 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1635 AManWithNoPlan (talk) 15:07, 2 May 2019 (UTC)
Crap issues=##(##)
- What should happen
- [28]
- We can't proceed until
- Feedback from maintainers
I achieved the above by replacing |issue= by |volume= and running the bot on that. Headbomb {t · c · p · b} 02:17, 24 April 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1619 I already thought of this. :-) AManWithNoPlan (talk) 13:57, 24 April 2019 (UTC)
Doi error
- What should happen
- [29]
- We can't proceed until
- Feedback from maintainers
Relatively common error (happens a few times a month) that should be easily handled by the bot. Headbomb {t · c · p · b} 14:09, 2 May 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1634 AManWithNoPlan (talk) 14:57, 2 May 2019 (UTC)
fails to change |url-access= to |chapter-url-access= when renaming |url= to |chapter-url=
- Status
- {{fixed}}
- Reported by
- Trappist the monk (talk) 18:44, 3 May 2019 (UTC)
- What happens
- this may have already been fixed (I didn't find mention of this particular issue in the archive): bot changes
|url=to|chapter-url=without also changing|url-access=to|chapter-url-access= - What should happen
- if
|url-access=has a value assigned to it, change parameter name to match the new url parameter name (|article-url=,|chapter-url=,|entry-url=,|section-url=); if|url-access=does no have an assigned value, remove it - Relevant diffs/links
- diff, diff, diff
- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot/pull/1636 AManWithNoPlan (talk) 21:49, 3 May 2019 (UTC)
Caps: JPN --> Jpn
- What should happen
- [30]
- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot/pull/1621 AManWithNoPlan (talk) 18:18, 7 May 2019 (UTC)
CAPS: SAE & AAUP
- What happens
|journal=Sae Transactions|journal=Aaup Bulletin- What should happen
|journal=SAE Transactions|journal=AAUP Bulletin- We can't proceed until
- Feedback from maintainers
Volume/issue cleanup
https://github.com/ms609/citation-bot/pull/1619 I think AManWithNoPlan (talk) 02:25, 10 May 2019 (UTC)
More handle links
- What should happen
- [33]
- We can't proceed until
- Feedback from maintainers
https://github.com/ms609/citation-bot/pull/1622 AManWithNoPlan (talk) 16:10, 11 May 2019 (UTC)
- thank you for mentioning that one; but I already added it to the pull having run across it myself. Must be common if we both saw them. AManWithNoPlan (talk) 13:15, 16 May 2019 (UTC)
Fails on every page
Don't really know what causes it, but I haven't been able to use the bot for several hours. Headbomb {t · c · p · b} 22:27, 13 May 2019 (UTC)
- still is.... AManWithNoPlan (talk) 16:34, 14 May 2019 (UTC)
- today i get 500 internal server error. — Chris Capoccia 💬 21:19, 14 May 2019 (UTC)
@Fhocutt, Kaldari, Mattsenate, Maximilianklein, and Smith609: I believe one of the people listed as running the bot needs kick it and restart it. AManWithNoPlan (talk) 02:14, 16 May 2019 (UTC)
{{fixed}} AManWithNoPlan (talk) 15:19, 17 May 2019 (UTC)
Request: Journal / publisher cleanup
More details: if journal and publisher are the same, but publisher is wiki linked then drop journal and rename publisher to journal. AManWithNoPlan (talk) 20:46, 17 May 2019 (UTC)
- https://github.com/ms609/citation-bot/pull/1648 AManWithNoPlan (talk) 21:15, 18 May 2019 (UTC)
add special case for journal with zeroes in issue numbers
- What happens
- [37]
- What should happen
- Don't strip 0s from issues. Or least don't strip them for Insecta Mundi
- We can't proceed until
- Feedback from maintainers
More hdl links
- What should happen
- [38]
- We can't proceed until
- Feedback from maintainers
Running a steamroller
I have run the Bot on several thousand pages and have made a couple of dozens of improvements based upon my experiences. Enjoy! AManWithNoPlan (talk) 22:49, 26 May 2019 (UTC)
Strip whitespace from hyphenated ISBNs
- What should happen
- [39]
- We can't proceed until
- Feedback from maintainers
Need to run twice
Volume cleanup
- What should happen
- [43]
- We can't proceed until
- Feedback from maintainers
Suggestion
I would like to suggest to temporarily add {{bots|deny=UTRS}} to the page to prevent getting messages on this page of these troll/fake unblock requests. If a real UTRS unblock request is made my the bot owner/operator (which seems unlikely per the discussion above) they should remove it again or do so when the bot gets unblocked. This should work per the explanation on User:UTRSBot. -- Redalert2fan (talk) 21:50, 27 May 2019 (UTC)
- An unblock request by the maintainers should be just as valid. WP:BUREAUCRACY and all that. In any case, this isn't trolling, or a fake request. Headbomb {t · c · p · b} 22:20, 27 May 2019 (UTC)
- Just to prevent any confusion, I didn't meant to exclude the maintainers of course. And further clarification, I meant that if a valid request is made {{bots|deny=UTRS}} should be removed by the maintainers/operators, not that the request should be removed by administrators. Redalert2fan (talk) 22:27, 27 May 2019 (UTC)
- On second thought, do you mean that a maintainer submitted the last 3 request that have been denied/dismissed as UTRS trolls? or am I misunderstanding? I only requested to add {{bots|deny=UTRS}} to prevent the talk page getting spammed with the understanding that these were fake/troll attempts, not to hide valid request if they are actually made by the operators/maintainer. Note that I am only talking about the UTRS request, this is all apart from the unblock discussion above. Redalert2fan (talk) 22:37, 27 May 2019 (UTC)
- I have no idea who submitted the UTRS tickets. I don't even know where to begin to see them, since there are no links, and the UTRS system is an inscrutable maze of confusion. Just saying that unblock requests from maintainers should be also accepted, and those would certainly not be trolling/fake requests, not just unblock request from the operator. Headbomb {t · c · p · b} 00:11, 28 May 2019 (UTC)
{{fixed}} seems to have stopped AManWithNoPlan (talk) 12:58, 29 May 2019 (UTC)
Caps exception: Srp Arh Celok Lek
https://github.com/ms609/citation-bot/pull/1681 AManWithNoPlan (talk) 03:36, 30 May 2019 (UTC)
Running twice
https://github.com/ms609/citation-bot/pull/1682 AManWithNoPlan (talk) 03:43, 30 May 2019 (UTC)
Helping the fakers?
I'm at a loss how to describe this. A very nasty vandal was faking cite book entries. In one article 36 of them. Searching for other uses of the faked google books url "XAb6ZH5xjH8C" I hit Islamic views on sin and was trying to figure out if I could just delete the added passages.
There were several small edits attempting fixes, and at least one person recognizing and deleting bogus text, but then I was confused by 'somebody' fixing the fake cite books. Um, what? (Search for "XAb6ZH5xjH8C')
Citebot 'fixed' ISBN to various values in 9 different entries without noticing they were all the same URL! Amazing... Shenme (talk) 23:37, 29 May 2019 (UTC)
- thank goodness the bot converted the urls to an ISBN that it is noticeable. There is always a balance between adding information and the bot had many checks but not infinite ones. The same url for multiple things is valid if each reference is a chapter of the same book. Lastly the bot treats each reference as a stand alone item for this reason. AManWithNoPlan (talk) 23:53, 29 May 2019 (UTC)
- I got some others too. Thank you for pointing this out. {{notabug}}, but a reminder to often go back and look at what you initiated. AManWithNoPlan (talk) 03:46, 30 May 2019 (UTC)
remove via when no url give, but doi/other identifiers present
- What should happen
- [49]
- We can't proceed until
- Feedback from maintainers
ISSN doesn't count, because it's a garbage identifier. Headbomb {t · c · p · b} 00:39, 31 May 2019 (UTC)
better cite article support
Looking into. We are leery of {{cite news}}. AManWithNoPlan (talk) 03:46, 31 May 2019 (UTC)
Fails to fill journal
In
- Butzkamm, Wolfgang (Winter 2003). "We only learn language once. The role of the mother tongue in FL classrooms: death of a dogma" (PDF). 28: 29–39. doi:10.1080/09571730385200181. Archived from the original (PDF) on December 14, 2010.
{{cite journal}}: Cite journal requires|journal=(help); Unknown parameter|deadurl=ignored (|url-status=suggested) (help)
The bot fails to add |journal=The Language Learning Journal. Possibly because of a title mismatch. However, there is a doi, and that should be good enough to bypass title mismatch 'safeguards'. Headbomb {t · c · p · b} 21:08, 1 June 2019 (UTC)
- doi shows “We only learn language once. The role of the mother tongue in FL classrooms: death of a dogma”. looking at. AManWithNoPlan (talk) 22:46, 1 June 2019 (UTC)
- works now. AManWithNoPlan (talk) 00:34, 2 June 2019 (UTC)
{{notabug}}
'User-activated'
- Status
- Not {{fixed}}, but flag for archive and will create new less verbose Request
- Reported by
- Headbomb {t · c · p · b} 19:23, 8 February 2019 (UTC)
- What happens
- Edit Summary says User-Activated
- What should happen
- The username should be reported.
- We can't proceed until
- Feedback from maintainers
Extended content
|
|---|
|
Link to the Issue on GitHub with links to documentation someone needs to read and use https://github.com/ms609/citation-bot/issues/948 AManWithNoPlan (talk) 19:45, 8 February 2019 (UTC)
. Headbomb {t · c · p · b} 19:49, 8 February 2019 (UTC)
Ok, see Wikipedia:Edit filter/Requested/Archive_12#CiteSeerX and Citation bot. —David Eppstein (talk) 20:45, 10 February 2019 (UTC)
It is good the bot is blocked. Now it can stay blocked until the ability of editors to edit wholly anonymously is removed. This is WP:TEACE, either by username or by revealing your IP; but citation bot currently has the facility to bypass both of those constraints. ——SerialNumber54129 16:39, 24 February 2019 (UTC)
A lot of progress could likely be easily made with by updating User:Smith609/citations.js and Wikipedia:Citation expander and other scripts to output the username (if they don't already, I notice the way User:Smith609/citations.js and MediaWiki:Gadget-citations.js gets usernames is different than User:Headbomb/citations.js, maybe something needs to be updated?), and make the Username box in [52] be mandatory. Those are really low hanging fruits. More could be identified if we knew how the bot was activated.Headbomb {t · c · p · b} 15:07, 19 March 2019 (UTC)
|
Discussion continues in block discussion
March 2019
{{unblock|reason=Your reason here ~~~~}}. — JJMC89 (T·C) 01:34, 17 March 2019 (UTC)- Thank you, JJMC89. I hope this can get resolved so that Citation bot can continue its other useful work soon, but this continued addition of copyvio links has been a big problem. —David Eppstein (talk) 01:57, 17 March 2019 (UTC)
- I tried to get more people involved at https://en.wikipedia.org/wiki/Wikipedia_talk:Copyrights but copyright seems to bore people. No one new showed up. AManWithNoPlan (talk) 02:02, 17 March 2019 (UTC)
- Once this goes into effect, the addition of CiteSeerX links will stop. https://github.com/ms609/citation-bot/pull/1500 AManWithNoPlan (talk) 03:56, 17 March 2019 (UTC)
- I too hope the bot can continue its useful work soon. – SJ + 04:30, 17 March 2019 (UTC)
- [53] ——SerialNumber54129 11:00, 17 March 2019 (UTC)
- As always, gadget mode still works and is 100% functional (that's the 'citations' button that appears next to 'save changes' when editing a page). AManWithNoPlan (talk) 14:38, 17 March 2019 (UTC)
- gadget is a good tool, but it doesn't work for all DOIs, and doesn't work for book chapters with DOIs. –Chris Capoccia (talk) 03:13, 28 March 2019 (UTC)
- As always, gadget mode still works and is 100% functional (that's the 'citations' button that appears next to 'save changes' when editing a page). AManWithNoPlan (talk) 14:38, 17 March 2019 (UTC)
- [53] ——SerialNumber54129 11:00, 17 March 2019 (UTC)
- I too hope the bot can continue its useful work soon. – SJ + 04:30, 17 March 2019 (UTC)
- Once this goes into effect, the addition of CiteSeerX links will stop. https://github.com/ms609/citation-bot/pull/1500 AManWithNoPlan (talk) 03:56, 17 March 2019 (UTC)
- I tried to get more people involved at https://en.wikipedia.org/wiki/Wikipedia_talk:Copyrights but copyright seems to bore people. No one new showed up. AManWithNoPlan (talk) 02:02, 17 March 2019 (UTC)
Citation bot (block log • active blocks • global blocks • contribs • deleted contribs • filter log • creation log • change block settings • unblock • checkuser (log))
Request reason:
The pull has been merged and pulled through; CiteSeerX links will no longer be added.
Decline reason:
There is currently no consensus to unblock this bot. Several editors have pointed out that it violates bot policy by not clearly identifying the editor who has triggered it. Some others have suggested that this could be accomplished fairly easily. It appears that the bot's operator is not very active on Wikipedia anymore. It might be better for an active editor to take it over and make it compliant with the bot policy. — Martin (MSGJ · talk) 19:57, 19 May 2019 (UTC)
If you want to make any further unblock requests, please read the guide to appealing blocks first, then use the {{unblock}} template again. If you make too many unconvincing or disruptive unblock requests, you may be prevented from editing this page until your block has expired. Do not remove this unblock review while you are blocked.
- @Smith609: How does the unblock request address the reason(s) for the block? ——SerialNumber54129 15:55, 18 March 2019 (UTC)
- I just saw the unblock request and came here to review it, but "The pull has been merged and pulled through; CiteSeerX links will no longer be added" is utterly meaningless to me (well, I think I understand the second part, but I haven't a clue about the "pulling"). Can you please explain what that means in English? Boing! said Zebedee (talk) 16:15, 18 March 2019 (UTC)
- @Boing! said Zebedee: The block was issued because the bot was adding CiteSeerX links. The bot will no longer do that, so the reason for the block no longer applies. Headbomb {t · c · p · b} 16:17, 18 March 2019 (UTC)
- There were two reasons for the block; that was only one. The other reason was "not disclosing the Wikipedia user directing each edit (see § 'User-activated'", which of the two is probably even more impactful. ——SerialNumber54129 16:19, 18 March 2019 (UTC)
- Right, I missed that part for some reason (a weird linebreak at my screen's width). Headbomb {t · c · p · b} 16:41, 18 March 2019 (UTC)
- the Bot was approved for anonymous editing before the guidance was changed to request the tracking of users. We have always been grandfathered in. I am not saying that times can change, but just giving the historical justification. AManWithNoPlan (talk) 19:50, 18 March 2019 (UTC)
- As Serial Number 54129 pointed out, I blocked the bot for two reasons. The PR only address the first one. No, the bot is not grandfathered in. All bots are expected comply with current policy. — JJMC89 (T·C) 03:46, 19 March 2019 (UTC)
- After a decade of being grandfathered in, it is not surprise. https://en.wikipedia.org/w/index.php?title=Wikipedia:Bot_policy&diff=next&oldid=204147936 AManWithNoPlan (talk) 04:25, 19 March 2019 (UTC)
- As far as I'm concerned the urgency of identifying user activations was so that we could track down the users responsible for adding bad CiteSeerX links and tell them not to do that. If the CiteSeerX links are not going to be added any more, then I don't think the other part is block-worthy. So I think it should be unblocked. But I was chided earlier when I suggested a block as being too involved to do the block myself, so by the same token I am not going to be the one to unblock. —David Eppstein (talk) 06:39, 19 March 2019 (UTC)
- After a decade of being grandfathered in, it is not surprise. https://en.wikipedia.org/w/index.php?title=Wikipedia:Bot_policy&diff=next&oldid=204147936 AManWithNoPlan (talk) 04:25, 19 March 2019 (UTC)
- As Serial Number 54129 pointed out, I blocked the bot for two reasons. The PR only address the first one. No, the bot is not grandfathered in. All bots are expected comply with current policy. — JJMC89 (T·C) 03:46, 19 March 2019 (UTC)
- the Bot was approved for anonymous editing before the guidance was changed to request the tracking of users. We have always been grandfathered in. I am not saying that times can change, but just giving the historical justification. AManWithNoPlan (talk) 19:50, 18 March 2019 (UTC)
- Right, I missed that part for some reason (a weird linebreak at my screen's width). Headbomb {t · c · p · b} 16:41, 18 March 2019 (UTC)
- Users to not direct the edits of Citation Bot. The bot makes the same edits whoever is activating it, and whether it is activated by a user or a CRON job. Users may only request that it visits a particular page first. Therefore I do not see that the policy that you reference is relevant here. Martin (Smith609 – Talk) 13:52, 19 March 2019 (UTC)
- you are correct, the bot does not allow any user control on configuration. That’s important. AManWithNoPlan (talk) 14:35, 19 March 2019 (UTC)
- @Smith609: That is in direct contravention of the claims made by several of the regulars at this page that the bot itself is not responsible for the edits made, because it is user activated and because it warns the activating user to check the edits. Absent your participation the position of these regulars in practice becomes the position of the bot's operator. I happen to agree with you on this: the responsibility for the bot's edits fall to you. Which was why I asked you to confirm whether you are in fact still the bot's operator (and to address WP:BOTACC, second para, and WP:BOTCOMM). Until the bot was blocked you had not edited this page in months (iirc), have not responded to multiple pings, and appear to have de facto delegated decisions regarding the bot's behaviour to AManWithNoPlan; including on issues that quite possibly could have led to sanctions for the bot's operator (that is, you, not the bot's account, could conceivably have ended up blocked or topic banned or...). The community needs to know who is responsible for the bot's edits, and thus also where to address concerns (which is why WP:BOTPOL requires responsiveness), and so far all I've seen is that nobody takes responsibility for Citation bot; its operator of record is non-responsive and its activating users are anonymous. --Xover (talk) 15:30, 19 March 2019 (UTC)
- The reason that I no longer regularly engage in discussions on Wikipedia talk pages is that I very often encounter a toxic and petty atmosphere that leaves me deeply discouraged. I have to be careful not to let this outweigh the pleasure that I derive from participating in the WP project. As such, I mainly interact with the bot at a technical level. I'm very grateful for AManWithNoPlan's voluntary work in maintaining the bot, without which the service would have stopped running a long time ago. Martin (Smith609 – Talk) 10:59, 2 April 2019 (UTC)
- @Smith609: I am sincerely sorry that your experience with your fellow Wikipedians has been so negative. I agree that "toxic and petty" are too often apt descriptions of interactions here, and that this is a significant threat (maybe even the significant threat) to editor recruitment and retention, and hence the long-term health of the project as a whole. Good communication is absolutely key on a wholly volunteer-run and collaborative project. Please take that as the backdrop when I ask again, for the third time, that you address WP:BOTACC, second para, and WP:BOTCOMM. Your desire to avoid negative interactions is entirely understandable (I sympathise completely), but I question whether that desire is compatible with operating a bot, and especially a bot like Citation bot. Other editors have a legitimate need to interact with the bot's operator, and to know who that actually is, and to be met by something other than "a toxic and petty atmosphere" when they raise concerns with the bot's edits. Discussion and consensus is not something one may "opt out" of on Wikipedia. --Xover (talk) 06:22, 3 April 2019 (UTC)
- @Xover: "quite possibly could have led to sanctions for the bot's operator" hardly. At worse, the bot could be blocked. Sanctions require active mischief. Headbomb {t · c · p · b} 15:39, 19 March 2019 (UTC)
- @Headbomb: If you actually think about it I'm sure you'll realise that that assertion is incorrect, even if it were not completely beside the point. Also, my comment was addressed specifically to the bot's operator and your constant interposition to "defend" Citation bot is not helpful (hint: it is not "under attack" and in need of being defended; there are just concerns being raised that are not being adequately addressed). Smith609 is presumably quite capable of speaking for themselves. --Xover (talk) 17:01, 19 March 2019 (UTC)
- @Xover: Point to me one time in the entire history of Wikipedia where a bot owner was sanctioned for simply being absent/uninvolved when a bot didn't edit according to consensus? Headbomb {t · c · p · b} 17:14, 19 March 2019 (UTC)
- Let's not debate a theoretical point that isn't even relevant anymore. What exactly needs to be done to get Citation bot unblocked? If the answer is "identify users who activate the bot", that involves some significant changes to how Citation bot works (assuming it needs to be implemented in way that can't easily be forged). Since it doesn't look like anyone has the bandwidth to do significant refactoring of Citation bot, I imagine this would effectively mean the death of Citation bot (at least the "bot" part). If that isn't the desired outcome, are there any other mitigations that could be made? Kaldari (talk) 20:04, 20 March 2019 (UTC)
- @Xover: Point to me one time in the entire history of Wikipedia where a bot owner was sanctioned for simply being absent/uninvolved when a bot didn't edit according to consensus? Headbomb {t · c · p · b} 17:14, 19 March 2019 (UTC)
- @Headbomb: If you actually think about it I'm sure you'll realise that that assertion is incorrect, even if it were not completely beside the point. Also, my comment was addressed specifically to the bot's operator and your constant interposition to "defend" Citation bot is not helpful (hint: it is not "under attack" and in need of being defended; there are just concerns being raised that are not being adequately addressed). Smith609 is presumably quite capable of speaking for themselves. --Xover (talk) 17:01, 19 March 2019 (UTC)
- The reason that I no longer regularly engage in discussions on Wikipedia talk pages is that I very often encounter a toxic and petty atmosphere that leaves me deeply discouraged. I have to be careful not to let this outweigh the pleasure that I derive from participating in the WP project. As such, I mainly interact with the bot at a technical level. I'm very grateful for AManWithNoPlan's voluntary work in maintaining the bot, without which the service would have stopped running a long time ago. Martin (Smith609 – Talk) 10:59, 2 April 2019 (UTC)
I'm keen to know whether this is a solution to a hypothetical problem, or a real problem. Has anyone ever used the edit summary to contact a user who was activating the bot? What was the outcome? Has anyone ever needed to contact an editor who operated the bot, but been unable to because they were anonymous? What was the outcome?
Headbomb: it sounds like your suggestions are
- Make .js scripts activated on Wikipedia send username to bot -- I believe that this is already the case
- Make manual entry of username mandatory on the web interface -- this could readily be done, or (as per WP) the user's IP could be used as a default.
Is there anything I've missed? Would these changes be sufficient for you to lift your opposition to the bot being permanently blocked?
Martin (Smith609 – Talk) 10:59, 2 April 2019 (UTC)
- I know it would for me. I'm fully aware there are some WP:BEANS-level concerns with those solutions, and that using some WP:OAuth thing would be more secure, but having the bot fail to edit when no username/IP is specified would go a long way to help trout people that use the bot carelessly (or help find people that suddenly took an interest in an article you've been trying to collaborate on). Headbomb {t · c · p · b} 16:05, 2 April 2019 (UTC)
- It is generally not sufficient to just log any old text string someone typed in: there has to be some form of authentication so you know that the user is who they say they are. A blocked user should also not be allowed to use Citation bot as a proxy. In practice this means either using OAuth so the actual edits are attributed to the authenticated user in the edit history, or implementing your own authentication solution for access to the bot (a separate user/password database) with some way to verify that the bot user account is identical to the Wikipedia user account (a per-user magic string on that user's Wikipedia user page for example). Not that these lesser mitigations will not be helpfull on their own (most users will not be seeking to avoid scrutiny, so just taking it on trust will be "good enough" for most normal situations), but they will not be sufficient on their own. In the mean time the bot's operator is responsible for its edits, including those initiated by bad-faith actors, when it is used to disrupt or edit war, etc. My strong recommendation would be to expend the effort (which is not insignificant, but neither is it insurmountable) to implement OAuth based authentication. --Xover (talk) 06:37, 3 April 2019 (UTC)
- @Xover: I think AManWithNoPlan has already done the OAuth coding work. {{3x|p}}ery (talk) 19:53, 3 April 2019 (UTC)
- I tested and needs to be seriously tested using the developer installation first. AManWithNoPlan (talk) 20:21, 3 April 2019 (UTC)
- @Xover: I think AManWithNoPlan has already done the OAuth coding work. {{3x|p}}ery (talk) 19:53, 3 April 2019 (UTC)
- I find the idea of the bot being used in an edit war to be odd and far fetched. How would one do that? AManWithNoPlan (talk) 15:27, 10 April 2019 (UTC)
- I agree. I have an uneasy feeling that some editors are trying to nuke the bot under whatever excuse comes to mind. Xover, can you please list a few examples when the bot was "used to disrupt or edit war"? Or is it just you are inventing arguments to keep mounting as many obstacles as you can? By the way, I see the bot has lately been recording user's IP address[54], so this problem seems to be resolved. Also, recording a user's IP has been introduced in MediaWiki in order to distinguish edits by different editors; on its own it has negligible impact on the site's security, given that a malicious actor can change their IP address at will using simple tools. — kashmīrī TALK 16:05, 27 April 2019 (UTC)
- Hi Kashmiri and welcome to Wikipedia. Since you have only been here for 13 years and 17k edits you have probably not yet had a chance to familiarize yourself with all the byzantine bureaucracy here, but since we're an entirely volunteer-driven project we have several policies and guidelines that are designed to create and maintain a collaborative and constructive environment. One of these is Assume Good Faith, which embodies the positive reinforcement of desired behaviour here. Its negative equivalent is WP:ASPERSIONS, an application of the policy No Personal Attacks. Since you're so new to the project you could of course not be expected to be aware of this, but you may want to keep it in mind for the future.PS. @AManWithNoPlan: I'm pretty sure leaking users IP addresses in edit summaries is against the WMF privacy policy. I haven't been paying attention lately, or looked at actual bot behaviour above looking at the diff above, so you may have it in hand already; but if by design or through a bug it's still leaking IPs that needs to be fixed ASAP (and the affected edits need to have their edit summaries hidden; WP:OVERSIGHT can help with that, and, I think, all admins can do it too). --Xover (talk) 18:45, 27 April 2019 (UTC)
- "I'm pretty sure leaking users IP addresses in edit summaries is against the WMF privacy policy" that's hilarious; your sarcasm just slays me. AManWithNoPlan (talk) 22:11, 27 April 2019 (UTC)
- OK, anything meaningful to add for a change, Xover? — kashmīrī TALK 23:07, 27 April 2019 (UTC)
- Verging on the aspersions yourself there, Kashmiri, be mindful. Cheers! ——SerialNumber54129 09:04, 28 April 2019 (UTC)
- Hi Kashmiri and welcome to Wikipedia. Since you have only been here for 13 years and 17k edits you have probably not yet had a chance to familiarize yourself with all the byzantine bureaucracy here, but since we're an entirely volunteer-driven project we have several policies and guidelines that are designed to create and maintain a collaborative and constructive environment. One of these is Assume Good Faith, which embodies the positive reinforcement of desired behaviour here. Its negative equivalent is WP:ASPERSIONS, an application of the policy No Personal Attacks. Since you're so new to the project you could of course not be expected to be aware of this, but you may want to keep it in mind for the future.PS. @AManWithNoPlan: I'm pretty sure leaking users IP addresses in edit summaries is against the WMF privacy policy. I haven't been paying attention lately, or looked at actual bot behaviour above looking at the diff above, so you may have it in hand already; but if by design or through a bug it's still leaking IPs that needs to be fixed ASAP (and the affected edits need to have their edit summaries hidden; WP:OVERSIGHT can help with that, and, I think, all admins can do it too). --Xover (talk) 18:45, 27 April 2019 (UTC)
- I agree. I have an uneasy feeling that some editors are trying to nuke the bot under whatever excuse comes to mind. Xover, can you please list a few examples when the bot was "used to disrupt or edit war"? Or is it just you are inventing arguments to keep mounting as many obstacles as you can? By the way, I see the bot has lately been recording user's IP address[54], so this problem seems to be resolved. Also, recording a user's IP has been introduced in MediaWiki in order to distinguish edits by different editors; on its own it has negligible impact on the site's security, given that a malicious actor can change their IP address at will using simple tools. — kashmīrī TALK 16:05, 27 April 2019 (UTC)
- I find the idea of the bot being used in an edit war to be odd and far fetched. How would one do that? AManWithNoPlan (talk) 15:27, 10 April 2019 (UTC)
I tried to be coy and avoid directly pointing it out, but anyone worried about the bot leaking IP addresses is saying clueless idiotic stuff. I will leave it as an exercise for the reader to think about why. AManWithNoPlan (talk) 13:11, 28 April 2019 (UTC)
- If it is obvious why this is not a problem, would it not be more constructive to simply explain why so that us clueless idiots can understand it? I have absolutely no problem being wrong or having my mistakes pointed out to me, and if there is no privacy issue here then so much the better. --Xover (talk) 13:48, 28 April 2019 (UTC)
- Maybe this would help you? When a user logs in with an IP address (not a named login ID), we call them "IP" users. The IP shows up in watchlists, diffs, signatures etc.. Since Citation bot has to credit someone for the edit, it credits the IP user. There is no "leaking". Citation bot does not have access to the IP's of named users. For example, it does not know what IP User:Xover is coming from. -- GreenC 14:00, 28 April 2019 (UTC)
- I never said anyone was clueless or idiotic, just that what was said was - that is different. The bot is not running so the idea that it is leaking anything is silly. Second of all users without accounts always leak their IP addresses: it’s Wikipedia policy, which is the exact opposite of the presupposition. AManWithNoPlan (talk) 14:02, 28 April 2019 (UTC)
- @GreenC: Ah, thank you. See? An easy straightforward answer: the bot does not, in fact, leak the IP address of logged-in users. Nothing could be better. --Xover (talk) 15:09, 28 April 2019 (UTC)
- @Headbomb: This bot still has an open unblock request. This is your jurisdiction, is the bot ready to be unblocked, or is it still pending the BAG all clear? If it's the latter, can I decline the unblock request with the understanding that BAG will unblock when the bot is ready to return to service? ~Swarm~ {sting} 23:01, 30 April 2019 (UTC)
- @Swarm: I'm involved enough here that the decision shouldn't fall on me. IMO, the issue is mostly one of convenience/annoyance which shouldn't be all that hard to overcome, or at least have some stopgap measures that minimize the annoyances (e.g. fail if no username is provided), rather than one hugely problematic behaviour. Others may disagree on this. Headbomb {t · c · p · b} 23:05, 30 April 2019 (UTC)
- I do disagree on that (well, the sufficiency, not necessarily the absolute scope), and do agree that you are sufficently involved to need to recuse from BAG discussion of this. But more to the point, is the BAG actually aware of this issue? I don't think this was a BAG-initiated block in any sense (BOTPOL, yes; BAG, no), so absent specific request or notification there's no particular reason to assume the BAG is doing anything active or even watching this. While I don't personally think unblock is warranted yet (I could be wrong; some mitigations have indeed been put in place and I haven't paid attention to details), it would be exceedingly counterproductive if the unblock was dependent on an "ok" from an instance (BAG) not aware that their opinion is solicited. --Xover (talk) 10:16, 1 May 2019 (UTC)
- I agree that with the removal of XCiteSeer links, the bot can resume normal operation. What remains to be done - having username in edit summary - is a minor annoyance at most, as Headbomb pointed out. Contrary to Xover's suggestions, we have not yet seen examples of the bot's misuse in this regard and in my view the advantages of having it up and running greatly outweigh any theoretical risks. — kashmīrī TALK 14:31, 1 May 2019 (UTC)
- As the blocking admin, if all of the issues that I blocked the bot for were resolved, I would have already unblocked. Only one of the two has been addressed. As such, I oppose unblocking until the other is addressed. — JJMC89 (T·C) 04:43, 2 May 2019 (UTC)
- @Swarm: I'm involved enough here that the decision shouldn't fall on me. IMO, the issue is mostly one of convenience/annoyance which shouldn't be all that hard to overcome, or at least have some stopgap measures that minimize the annoyances (e.g. fail if no username is provided), rather than one hugely problematic behaviour. Others may disagree on this. Headbomb {t · c · p · b} 23:05, 30 April 2019 (UTC)
- @Swarm: What Kashmiri casually dismisses as "a minor annoyance at most" the rest of us call policy; at the moment—and if unblocked without being addressed—the requirements that
the Wikipedia user directing any given edit must always be identified
andusers able to direct the bot to make edits must be positively identified to the bot at the time of edit,
are being flaunted. It matters not that the bot does other "useful work"—so have many now-indef'd editors over the years. Bot operation clearly does not fall within the scope of WP:IAR, and it is fallacious to demand examples of misuse when policy is being so flagrantly abused.Also pinging Xaosflux, who has been discussing this elswehere. Cheers, ——SerialNumber54129 11:58, 2 May 2019 (UTC)- @Serial Number 54129: thanks for the note, I'm trying to get a good understanding of the "bot side" of the issue here. One thing I'd like to see is: (a)An example of a few "bad edits" made by the bot, along with (b) which BRFA approved Task Number the operator says these edits are being made under. — xaosflux Talk 12:50, 2 May 2019 (UTC)
- @Xaosflux: Thanks for this. I'm less concerned with whether it's harmed the encyclopedia: if the "bot"—i.e. the editor controlling the bot at that point in time—has made a useful edit, I would like to be able to thank them for it; in the current situation, I cannot. Likewise, of course—and this is the nub—if the person behind the bot makes an error—possibly regarding WP:CITEVAR, for example—I would also welcome the opportunity to notify the controller. Again, at the moment, I cannot do so.On the matter of whether harm has been done, that's more subjective, but looking over just three of my recent featured articles, I see for example, changing a website to a journal (?!) [55], and multiple CITEVAR errors on the same articles—[56], [57], [58], [59], [60], and [61], [62], [63], [64], [65].Ultimately, I should have the opportunity to discuss these edits with the editor making them. WP:BOTCOMM, anyone? I find the situation honestly bizarre. ——SerialNumber54129 13:51, 2 May 2019 (UTC)
- @Serial Number 54129: regarding the "operating" portions, I'm not really buying that those editors have "operated" this bot, just "triggered it" (along the same lines that vandals don't "operate" antivandalism bots, or users don't operate signbot, etc). But ultimately, responsibility for the edits lays with the actual bot operator - and it is either making useful edits, or it is not. If the bot is set up to make bad edits based on certain triggers, the operator certainly needs to fix that. — xaosflux Talk 14:02, 2 May 2019 (UTC)
- Well, there may well be something the operator / maintainer can do to prevent the issue arising, but I disagree that the way forward is to semantically absolve editors from their responsibility to ensure their edit are not disruptive, whether we call that triggering or operating or whatever. I have to disagree with your analogy: editors using ("triggering"?) anti-vandal tools may not "operate" them, but we sure as hell hold them responsible for misuse, and the way CB is currently configured there is no similar way of holding those who trigger the bot responsible for their edits-by-proxy. And while I agree that the actual operator can and should be responsible for making sure the bot acts within policy, others should also; indeed, some responsibility should also be placed firmly at the door of—pace!—the various BAGgers over the years who appear to have repeatedly rubber-stamped its operations despite this lack of accountability. Basically, there's a long-standing issue with a concomitant amount of responsibility.In good news, though, this was a good edit with the bot...and I knw who to thank for it :) Just as it should be. ——SerialNumber54129 14:21, 2 May 2019 (UTC)
- @Serial Number 54129: regarding the "operating" portions, I'm not really buying that those editors have "operated" this bot, just "triggered it" (along the same lines that vandals don't "operate" antivandalism bots, or users don't operate signbot, etc). But ultimately, responsibility for the edits lays with the actual bot operator - and it is either making useful edits, or it is not. If the bot is set up to make bad edits based on certain triggers, the operator certainly needs to fix that. — xaosflux Talk 14:02, 2 May 2019 (UTC)
- @Xaosflux: Thanks for this. I'm less concerned with whether it's harmed the encyclopedia: if the "bot"—i.e. the editor controlling the bot at that point in time—has made a useful edit, I would like to be able to thank them for it; in the current situation, I cannot. Likewise, of course—and this is the nub—if the person behind the bot makes an error—possibly regarding WP:CITEVAR, for example—I would also welcome the opportunity to notify the controller. Again, at the moment, I cannot do so.On the matter of whether harm has been done, that's more subjective, but looking over just three of my recent featured articles, I see for example, changing a website to a journal (?!) [55], and multiple CITEVAR errors on the same articles—[56], [57], [58], [59], [60], and [61], [62], [63], [64], [65].Ultimately, I should have the opportunity to discuss these edits with the editor making them. WP:BOTCOMM, anyone? I find the situation honestly bizarre. ——SerialNumber54129 13:51, 2 May 2019 (UTC)
- @Serial Number 54129: Given that the bot was approved well before the policy came into force and that such operation has to-date not resulted in any significant problem, I stick by calling this non-compliance as "annoyance at most". Certainly, in the light of the bot's historical usage, it is not a sufficient reason for indeffing. — kashmīrī TALK 13:28, 2 May 2019 (UTC)
- @Serial Number 54129: thanks for the note, I'm trying to get a good understanding of the "bot side" of the issue here. One thing I'd like to see is: (a)An example of a few "bad edits" made by the bot, along with (b) which BRFA approved Task Number the operator says these edits are being made under. — xaosflux Talk 12:50, 2 May 2019 (UTC)
- Irrelevant. When policy changes, we change with the policy, we don't just ignore it because we were here before the policy. ——SerialNumber54129 13:51, 2 May 2019 (UTC)
- Agreed, but non-compliance is not a sufficient reason for an indefinite block in absence of actual damage to the project, especially when the non-compliance was pointed out only a few weeks earlier after 10+ years of safe operation. To put it simply, going around and blocking everything that's non-compliant with some policy will leave us with very little in this project, so I am always in favour of a fair dose of common sense. — kashmīrī TALK 17:39, 2 May 2019 (UTC)
I hate to dive into the middle of a discussion here but this is absolutely kafka-esque to read from the outside. Look, I see a few general points here that need to be summarized:
1. The blocking editor, User: JJMC89, wants both issues resolved--the XCiteSeer links AND the lack of identification of editors who use this bot. As I understand it, at this point only logged-in editors remain unidentified. IPs are tracked for editors who are not logged in. Someone please correct me if I'm wrong here, but it sounds like we are 1/2 + 1/4 of the way done here--that's 3/4! We are well on our way.
2. Some people are interested in if the bot is really doing or capable of doing harm. Some people are interested in what really necessitates an indefinite block. Still more people are worried about whether bots should follow this or that policy. But all of that is mind-numbingly detailed and misses the big picture.
3. The big picture is, forget about blocks for a second. It would be a darn nice feature to be able to see who is using the bot, wouldn't it? Then you could work with them, or give them a barnstar, or give them tips, or even learn from them if you are just learning about this bot like I am. Whether or not the block is in place, it would be really good to just collect this automatically from editors so that there is no hassle for them or for the rest of us checking up on edits. And if we were to solve this problem, we wouldn't have to worry about this or that policy or whether the bot is dangerous. The best possible scenario, as I see it, is having this feature and then having the bot unblocked. Remember, we are trying to make an encyclopedia with good references.
4. But all that's a bit of a hassle for User:AManWithNoPlan or anyone else who works on the code for this thing. So let's all put our heads together and ask. Is that possible to set up? What are the challenges of doing so? What's the timeframe? And, dare I say it, what do you need in order to get this thing done? Is there any way for us to help you? After all, that's the point of this entire exercise. This bot is clearly a useful resource. So let's get it up and running, safely, for all our sakes. And let's please discuss how to do that rather than who is right or wrong about this or that academic issue. When we are done with all that jazz, why don't we then pop on over to the relevant policy talk pages if we want to talk about if this or that makes sense or is fair--perhaps with a link to this discussion. While these are weird circumstances, and while I have my own comments, it does seem like continued discussion of that sort doesn't really belong on a user's talk page after an unblock request. So tell us, Man, how do we get this train rolling? Prometheus720 (talk) 04:10, 3 May 2019 (UTC)
- So we're just stuck on this user identification mechanism? Are there actually any real cases where someone has incorrectly recorded their edit as from some other user? Or is this another fake disaster where we've turned off Citation bot for a theoretical problem.Chris Capoccia (talk) 20:32, 6 May 2019 (UTC)
- Theoretical mostly. You can technically impersonate people if you know how (although, why the hell would you?), and I believe AManWithNoPlan did one such edit posing as me as a proof of concept (I could be wrong here), although I couldn't tell you which it is. OAuth is better and more secure, yes. But right now, we could just implement an edit filter blocking the anonymous edits, or have the bot fail if no username/IP is specified. That such a simple fix still isn't in after 2+ months is rather mind boggling. Headbomb {t · c · p · b} 20:45, 6 May 2019 (UTC)
- @Chris Capoccia: Yes, and the most vigorous opposition is by editors who don't seem to have been ever using the bot[66][67][68] or contributing to its development. — kashmīrī TALK 21:51, 6 May 2019 (UTC)
- Which is completely irrelevant; those that are at the sharp end of the bot's misuse might be writing featured content instead; albeit featured content which some bot-misuser(s) can subsequently come along and fuck up anonymously (passim). ——SerialNumber54129 15:11, 12 May 2019 (UTC)
- again, are there any actual examples of this? or is this only theoretical? — Chris Capoccia 💬 16:02, 12 May 2019 (UTC)
- again, 13:51, 2 May 2019. ——SerialNumber54129 16:14, 12 May 2019 (UTC)
- Fuck up the featured content??? Are you sure you know the function of this bot? — kashmīrī TALK 18:06, 12 May 2019 (UTC)
- As much, I imagine, as you would claim to know about featured content :) the point is that consistent citations and MOSCITE adhereance is mandatory for FAs, and bot operators—not the bot, its anonoperators—f-u-c-k it up occasionally. And one occasion is too often. ——SerialNumber54129 18:20, 12 May 2019 (UTC)
- "And one occasion is too often" that statement is not just false, it is laughably false. Remember we are talking about website that allows anyone to edit anything with anonymity. A few pages do require you to set up a sock-puppet (ooops, I mean account) before editing them; but most do not. AManWithNoPlan (talk) 18:46, 12 May 2019 (UTC)
- I think you do not understand anonymity, here at least. Not "anonymous=real life identity unknown"; rather, "anonymous=still indentifiable". We are all anonymous, if we choose to be: me with my serial number, you with your lack of a plan, for instance; but we are all identifiable on this site. As long as everyone who fires up Citation Bot is identifiable in the same way, there'll be no problem. At the moment, however, the only place on the Emglish Wikipedia where truely anonymous editing is possible...is via this tool. That is the issue. ——SerialNumber54129 19:01, 12 May 2019 (UTC)
- "the only place on the Emglish Wikipedia where truely anonymous editing is possible...is via this tool." This is false. The edit is fully attributed to the bot. You may not know who activated the bot, but the bot is the one that made the edit. Headbomb {t · c · p · b} 00:57, 13 May 2019 (UTC)
- I still fail to comprehend how being identifiable protects FA content. No, I can't really get this entire argument about fuckup defense - that a username or IP address listed next to an edit makes fuck-ups less of an issue. Any proof please? Before you reply, let me also add that I have been long active and very active in reverting damage, SPI, etc. — kashmīrī TALK 19:15, 12 May 2019 (UTC)
- can we get a concrete real example of what we're talking about instead of this theoretical stuff? — Chris Capoccia 💬 00:49, 13 May 2019 (UTC)
- I think you do not understand anonymity, here at least. Not "anonymous=real life identity unknown"; rather, "anonymous=still indentifiable". We are all anonymous, if we choose to be: me with my serial number, you with your lack of a plan, for instance; but we are all identifiable on this site. As long as everyone who fires up Citation Bot is identifiable in the same way, there'll be no problem. At the moment, however, the only place on the Emglish Wikipedia where truely anonymous editing is possible...is via this tool. That is the issue. ——SerialNumber54129 19:01, 12 May 2019 (UTC)
- "And one occasion is too often" that statement is not just false, it is laughably false. Remember we are talking about website that allows anyone to edit anything with anonymity. A few pages do require you to set up a sock-puppet (ooops, I mean account) before editing them; but most do not. AManWithNoPlan (talk) 18:46, 12 May 2019 (UTC)
- As much, I imagine, as you would claim to know about featured content :) the point is that consistent citations and MOSCITE adhereance is mandatory for FAs, and bot operators—not the bot, its anonoperators—f-u-c-k it up occasionally. And one occasion is too often. ——SerialNumber54129 18:20, 12 May 2019 (UTC)
- again, are there any actual examples of this? or is this only theoretical? — Chris Capoccia 💬 16:02, 12 May 2019 (UTC)
- Which is completely irrelevant; those that are at the sharp end of the bot's misuse might be writing featured content instead; albeit featured content which some bot-misuser(s) can subsequently come along and fuck up anonymously (passim). ——SerialNumber54129 15:11, 12 May 2019 (UTC)
- What's the weather like at Cyberdyne Systems today?! 🌞 To take your points in turn (When this discussion moves to AN it will probably deserve individual responses).Headbong: It's certainly creative thinking: the bot wakes up and decides to make certain edits. AKA, it clearly does not: the editor does so, and is responsible for the edits. The bot, being code, cannot be individually responsible for anything. Kashmiri: congratulations on your previous efforts at "reverting damage", etc: now is the time to prove yourself. But if you don't understand how identifying an editor (not the bot they fire up) who makes poor edits is useful to the encyclopedia...it's page one. Chris: I have done so, above ^^^ with the timestamp. Although if you'd actually read the discussion, I would not have had to point it out in the first place. "Concrete examples" you got em. On a lighter note, the bot remains blocked; this is, of course, a Good Thing. Cheers all! ——SerialNumber54129 06:05, 13 May 2019 (UTC)
- My username is Headbomb, not Headbong. Also, the bot and its operators are responsible for its edits, even if someone else activates it. Headbomb {t · c · p · b} 19:27, 13 May 2019 (UTC)
- OK, so imagine you noticed that an unidentified user has run this bot on a FA which malformed some citations, and they did not revert. How on earth will your knowledge of the user's identity help you to prevent further damage? Are you perhaps aware that unless the article has been protected, anyone can do exactly the same damage manually, without resorting to the bot, and you will have precisely zero way of defending against it? After such a long time on Wikipedia you should already know that the only way to protect articles from malicious editing is to set appropriate protection level.
- Also, this is all pure theory, because you have still not offered any evidence that anyone has ever used the citation bot to damage Wikipedia. While me and others are telling you again and again that the continuous block of this bot has brought qualitative damage to this project, as 99.999999% of the bot's edits were improvements.
- Eh, it is so hard to argue with someone who tries to defend the humanity against imaginary threats! Please, please read WP:COMMONSENSE! — kashmīrī TALK 07:28, 13 May 2019 (UTC)
- I was meaning with a diff to an actual damaging edit. I haven't seen that, but there is a lot of awfully dense text and I could have missed it. And featured articles have nonsense in them all the time. For example, today's featured article Interstate 94 in Michigan someone added "HAD A BLACK CAR ON IT EW" at the very top diff. I ran Citation Bot on the page manually. You can see from this diff that this is nothing "damaging". Looking forward to some real concrete examples. — Chris Capoccia 💬 13:04, 13 May 2019 (UTC)
- You clearly do not know what you are talking about. But thanks for identifying yoursel as using CB in your edit: When everyone does that, the bot can be unblocked :D ——SerialNumber54129 13:12, 13 May 2019 (UTC)
- You're just dodging the question, name calling & failing to even know how citation bot has been leaving names in edits for a long time. — Chris Capoccia 💬 13:35, 13 May 2019 (UTC)
- "Dodging the question" is being provided examples of misuse and ignoring them. Pray continue your WP:VAGUEWAVEs though. Cheers! ——SerialNumber54129 13:39, 13 May 2019 (UTC)
- I don't have a horse in this race, but, it does seem rather silly to me to change:
|author = [[Michigan Legislature]]
- to:
|author = Michigan Legislatureplus|author-link = Michigan Legislature
- There is no reason for this sort of duplication. If you want to do something good for that article, remove all of the
{{open access}}templates. Newspapers.com is not open access. The content of the articles hosted there may be free-to-read, but is, most certainly, not free-to-be-reused. Suggested style for cs1|2 templates linking to readable clippings at Newspapers.com is at Wikipedia:Newspapers.com. There was discussion about the use of{{open access}}and{{free access}}templates with these citations on the talk page. - —Trappist the monk (talk) 13:36, 13 May 2019 (UTC)
- @Trappist: I mostly edit medical and science articles, so highways are well outside of the norm for me. I just wanted to show a concrete example of what Citation bot would do with a featured article. I looked at what Citation bot was proposing before I saved it, but the changes are only the automatic ones. Just trying to show that it's not making some disaster out of high-profile Good Article pages like SerialNumber seems to think. I see there are a couple of diffs, and Citation Bot did have problems with the fake DOI URL at oxforddnb, but it's hardly a citevar issue when the correct way to cite that article would be to use the correct DOI and not a URL that could change and an archive URL. If If the only problem left is eliminating the "User-activated" thing, there have been a couple ideas for how to reasonably accomplish that for all but a beans hacker whatever that is. — Chris Capoccia 💬 22:22, 13 May 2019 (UTC)
- @Trappist the monk: The bot does't invent cite parametres but just uses them per template. Your complaint should go to the editors at Template:Cite who expressly request not to use wikilinks in author field. My guess is this particular form has been adopted to avoid pipe links, but if you want to be sure, please head to Template talk:Cite. Cheers, — kashmīrī TALK 06:44, 16 May 2019 (UTC)
- That do-not-wiki-link text applies to
|last=and to|first=because when those parameters are used, cs1|2 renders authors names in last-first order while that same author's article at en.wiki is titled in first-last order. For editors who want both of an author's last and first names wikilinked,|author-linkn=is required when the author's name is split between|last=and|first=. Because|authorn=is the whole author's name, there is no need to require|author-linkn=. The cs1|2 documentation can always be made better, but at least for this case, is consistent; when a parameter should not be wikilinked, the documentation so states. You can wikilink both|first=and|last=but that sort of makes for nonsense: - —Trappist the monk (talk) 11:05, 16 May 2019 (UTC)
- That do-not-wiki-link text applies to
So, let me get this straight: This extremely useful bot, which is so valuable that it even has a link in the "tools" menu on the left side of the screen, has been blocked for two months because someone is upset that it doesn't report who clicked the "expand citations" button and triggered the bot's completely automated edit to an article. This is ridiculous, and a perfect example of someone disrupting Wikipedia to make a point.--Srleffler (talk) 17:34, 19 May 2019 (UTC)
- That link is produced by a gadget you have installed. It is not on by default. --Izno (talk) 18:02, 19 May 2019 (UTC)
The current block is not well-founded on the policy
The current block of this bot is based on a misunderstanding of the policy. The outstanding issue is whether the bot is obligated to report the identity of the user who triggered it. The policy cited is WP:BOTMULTIOP, "Bots operated by multiple users". This section of the policy is about a bot that is shared between several operators. Users who merely trigger a bot are not "bot operators". The relevant part of the policy is WP:BOTCONFIG, where the fourth bullet begins "Providing some mechanism which allows contributors other than the bot's operator to control the bot's operation is useful in some circumstances..." [em. added]. The policy explicitly allows bots to be controlled by individuals who are not "operators" of that bot. That is exactly the situation with citation bot. There is no outstanding reason why the current block should be continued. --Srleffler (talk) 02:04, 20 May 2019 (UTC)
- I should note that the reason to unblock it given was not founded in policy either ‘no consensus to unblock’ is rubbish since you don’t need consensus just a decision on is there a policy violation. Thank you for tracking this link down, since multiple people made this point and knew it to be the case, but no one knew the link. Considering when and how the multiple operator policy was adopted and the text given, it is pretty obvious it does not refer to this bot (ie no user configurable controls). AManWithNoPlan (talk) 17:04, 20 May 2019 (UTC)
Citation bot (block log • active blocks • global blocks • contribs • deleted contribs • filter log • creation log • change block settings • unblock • checkuser (log))
Request reason:
the policy referenced does not apply AManWithNoPlan (talk) 02:36, 21 May 2019 (UTC)
Decline reason:
I am marking this declined in an effort to clean up the queue of unblock requests from blocked users. I have no opinion to share on what should be done with the bot. However, the unblock request process is not a suitable one for resolution of this matter. UninvitedCompany 22:16, 22 May 2019 (UTC)
If you want to make any further unblock requests, please read the guide to appealing blocks first, then use the {{unblock}} template again. If you make too many unconvincing or disruptive unblock requests, you may be prevented from editing this page until your block has expired. Do not remove this unblock review while you are blocked.
- Previous unblock request was rejected because of “meh, no consensus” AManWithNoPlan (talk) 02:58, 21 May 2019 (UTC)
- There was no consensus to block in the first place. Demanding consensus in order to unblock is absurd. Bots are not blocked or unblocked based on consensus but on policy-based arguments and risk assessment which were all extensively discussed above. — kashmīrī TALK 09:34, 21 May 2019 (UTC)
- I support the request to unblock, for the reasons I gave above. Despite discussions that assume otherwise, the policy explicitly allows bots to be controlled by users who are not operators. Keeping this bot blocked on the grounds that it doesn't identify the user who triggered it is not in accord with the policy. --Srleffler (talk) 01:55, 22 May 2019 (UTC)
- @Smith609: as this bot's task is contentious, would you be open to seeking reapproval via Wikipedia:Bot requests as I feel this would be a better forum to discuss the bot's functions and how they fit with the bot policy. — Martin (MSGJ · talk) 07:56, 22 May 2019 (UTC)
- Oppose unblock notwithstanding semantic wikilawyering: please desist from promoting anonymous editing. Thank you. ——SerialNumber54129 08:05, 22 May 2019 (UTC)
- Anonymous editing was aprroved in https://en.wikipedia.org/wiki/Wikipedia:Bots/Requests_for_approval/Citation_bot_6 In approval number 6 "Automatic, in response to user input" was added. AManWithNoPlan (talk) 15:48, 22 May 2019 (UTC)
- "wikilawyering" what else do you expect people to do when discussing the details of what words and policies mean? This is one case where wikilawyering is actually what almost everyone is doing on this thread and correctly so. AManWithNoPlan (talk) 15:48, 22 May 2019 (UTC)
- The court calls @Headbomb: to the Talk Page. Where were you on the night of 21 May 2019 when the second unblock request was made? Can you state your IP address for the record and please include any IP addresses within your subnet that can collaborate you whereabouts? Also, it is true that the alias Headbong has been use for you? AManWithNoPlan (talk) 15:52, 22 May 2019 (UTC)
- Uh? Not sure if this is a joke or not, but I was asleep, and I'm traveling, so I couldn't know what my IP was even if I wanted to. Either way, I'm of the opinion that the bot should be updated to disallow anonymous activation / required usernames to be given (which should have been done ages ago, and certainly when it was requested months ago), but that this isn't a block-worthy issue. Headbomb {t · c · p · b} 17:07, 22 May 2019 (UTC)
- it was a joke reference to being a lawyer. Happy travels. AManWithNoPlan (talk) 17:25, 22 May 2019 (UTC)
- Uh? Not sure if this is a joke or not, but I was asleep, and I'm traveling, so I couldn't know what my IP was even if I wanted to. Either way, I'm of the opinion that the bot should be updated to disallow anonymous activation / required usernames to be given (which should have been done ages ago, and certainly when it was requested months ago), but that this isn't a block-worthy issue. Headbomb {t · c · p · b} 17:07, 22 May 2019 (UTC)
- The court calls @Headbomb: to the Talk Page. Where were you on the night of 21 May 2019 when the second unblock request was made? Can you state your IP address for the record and please include any IP addresses within your subnet that can collaborate you whereabouts? Also, it is true that the alias Headbong has been use for you? AManWithNoPlan (talk) 15:52, 22 May 2019 (UTC)
- @Serial Number 54129: Maybe you wanted to add a meaningful comment about the core issue here? I guess you didn't mean it, but you admonishing other GF editors sounded plainly rude. — kashmīrī TALK 20:21, 22 May 2019 (UTC)
- considering that user directed editing by bots is allowed (they remain unnamed and within the witness protection program) and given that user-directed editing was called out in an approved bot request; I ask the question “why is the bot account blocked?” AManWithNoPlan (talk) 21:09, 28 May 2019 (UTC)
For any watchers not yet aware, there is a discussion at WP:AN#User:Citation bot unblock discussion is stalled. — JJMC89 (T·C) 03:22, 3 June 2019 (UTC)
{{fixed}} flag for archive
Request: Support Template:Retracted/Template:Erratum + their redirects
The only support needed is for identifiers. This would expand, e.g.
- (Retracted, see doi:10.1136/heartjnl-2012-302337ret)
to
- (Retracted, see doi:10.1136/heartjnl-2012-302337ret, PMID 23640983)
Headbomb {t · c · p · b} 15:21, 22 April 2019 (UTC)
- not a high priority but doable with more work than one would think. AManWithNoPlan (talk) 01:41, 23 April 2019 (UTC)
Flagging as {{wontfix}}, since the level of complexity is too high after careful investigation AManWithNoPlan (talk) 21:23, 5 June 2019 (UTC)
Running a steamroller again
I removing citation bot blocking templates from Wikipedia. And fixing all ‘bad’ behavior as I go. Do not upgrade ISBN to ISBN13 unless year is 2007 or newer. Special code for Oxford Online Encyclopedia, since the URLs are DOIs, just the wrong ones! I hope this will fix all unreported bugs. AManWithNoPlan (talk) 00:21, 4 June 2019 (UTC)
- over a hundred pages done. AManWithNoPlan (talk) 02:44, 4 June 2019 (UTC)
- @AManWithNoPlan: slowdown here. If you remove a do not run template, you should run the bot/citation helper on the page to ensure nothing actually breaks. Headbomb {t · c · p · b} 12:16, 4 June 2019 (UTC)
- I am running it on the pages. On most of them, the bot does nothing now. On some it does bad things, so I don’t remove the template. My goal is to eventually remove them all—with all bugs fixed of course. AManWithNoPlan (talk) 12:25, 4 June 2019 (UTC)
- except for a few pages with significant style differences. AManWithNoPlan (talk) 21:45, 4 June 2019 (UTC)
- also all pages with invalid dois from 2018. AManWithNoPlan (talk) 17:25, 5 June 2019 (UTC)
- except for a few pages with significant style differences. AManWithNoPlan (talk) 21:45, 4 June 2019 (UTC)
- I am running it on the pages. On most of them, the bot does nothing now. On some it does bad things, so I don’t remove the template. My goal is to eventually remove them all—with all bugs fixed of course. AManWithNoPlan (talk) 12:25, 4 June 2019 (UTC)
- @AManWithNoPlan: slowdown here. If you remove a do not run template, you should run the bot/citation helper on the page to ensure nothing actually breaks. Headbomb {t · c · p · b} 12:16, 4 June 2019 (UTC)
{{notabug}} flag for archive
Caps: DTSCH./JPN. --> Dtsch./Jpn.
Note the dots. Headbomb {t · c · p · b} 18:16, 5 June 2019 (UTC)