User talk:Hoo Bot

From Wikidata
Jump to navigation Jump to search
[edit]

What was that? Infovarius (talk) 05:23, 20 March 2014 (UTC)[reply]

The page was deleted at that time, so the action was valid ([1]). Cheers, Hoo man (talk) 13:51, 20 March 2014 (UTC)[reply]

Can you make it detect more renames instead of just deleting links such as [2]? EmausBot catches some of these, see en:WT:Wikidata. Fayenatic london (talk) 15:48, 31 March 2014 (UTC)[reply]

Tough one (given the current architecture of my bot)... as the bot is only supposed to be a stop gap solution until Wikibase itself does all / most of this, I don't think that I'll implement something like that. The bot already acts on redirects (which are the semantically most correct way to express this). Cheers, Hoo man (talk) 00:15, 1 April 2014 (UTC)[reply]
Does Hoo Bot do things that EmausBot does not do? If not, would you consider stopping it? Today it has been removing links to "Category:Sports festivals in Foo" instead of renaming them. I've started reinstating them manually, and only just realised that a bot had removed them. Fayenatic london (talk) 11:31, 23 April 2014 (UTC)[reply]
As far as I know EmausBot doesn't unlink deleted pages, which is the crucial task of my bot. If you want item links to be preserved for categories, then move them using Special:MovePage or at least create a redirect from the old name to the new one, that's the semantically right way to express page moves (and both should work fine). - Hoo man (talk) 11:36, 23 April 2014 (UTC)[reply]
Here's a similar one that EmausBot caught and dealt with correctly, changing the link rather than removing it as Hoo Bot did on the one above. Please compare these.
Categories cannot be moved using the "move page" function. The bots that rename categories on enwiki do not create category redirects at the old names, and we would only rarely want those to exist. Would it be possible to exclude categories from Hoo Bot's scope? Fayenatic london (talk) 11:57, 23 April 2014 (UTC)[reply]
Who would then unlink categories? EmausBot doesn't do that. - Hoo man (talk) 12:00, 23 April 2014 (UTC)[reply]
Also this issue will reoccur once Wikibase automatically unlinks pages on page deletions (and that will happen). Thus I suggest that the proper way to solve this is to fix EmausBot so that it can fix category links after their deletion. Disabling my bot would only hide this issue temporary. - Hoo man (talk) 12:04, 23 April 2014 (UTC)[reply]
Thanks, I have requested this at User talk:Emaus. Fayenatic london (talk) 12:15, 23 April 2014 (UTC)[reply]

Again. --Infovarius (talk) 15:16, 28 August 2014 (UTC)[reply]

Good catch... I'm not sure how that could happen, will investigate. Cheers, Hoo man (talk) 02:22, 29 August 2014 (UTC)[reply]
I've just hardened the code that makes sure that a page really has been deleted, so that this shouldn't be happening again. Cheers, Hoo man (talk) 02:50, 29 August 2014 (UTC)[reply]

Emptying a page

[edit]

Hi, when Hoo Bot removes the only link that there was on a page, such as Q15877606, could it also nominate it for deletion?

Background: nl:Kurt Van Damme was created, and BotMultichill created the Wikidata item. The article was then deleted. HooBot removed the link, leaving an empty item.

If there is another bot that sweeps up empty data items, then never mind. Fayenatic london (talk) 11:21, 25 April 2014 (UTC)[reply]

I sweep up after my bot every now and then. I'll probably do that again soon, depending on when I'll find the time. - Hoo man (talk) 15:14, 25 April 2014 (UTC)[reply]
@Fayenatic london: fwiw, it looks like the sweeping up may be automated soon: Wikidata:Requests for permissions/Bot/Dexbot 6. \o/ John Vandenberg (talk) 03:31, 14 May 2014 (UTC)[reply]
[edit]

this is clearly incorrect, and does not appear to be within the scope of the bot task at Wikidata:Requests for permissions/Bot/Hoo Bot. John Vandenberg (talk) 16:28, 11 May 2014 (UTC)[reply]

Please tell me why this is incorrect, it's just the way we currently handle redirects on Wikidata and within Wikibase - Hoo man (talk) 18:43, 12 May 2014 (UTC)[reply]
A merge means joining two concepts into one, which Wikipedia does for notability reasons (in this case by user:Headbomb ), but that is not appropriate on Wikidata. If you actually looked at Ornis Hungarica (Q14506800), you will see that your bot changed the sitelink to link to a new target page w:Hungarian Ornithological and Nature Conservation Society (which was created two days before the merge, but five months before your bot edit), whereas that new page should have had its own item. Before leaving you a message, I created Hungarian Ornithological and Nature Conservation Society (Q16859791) and added it to Q14506800 as the publisher (P123).
Are you saying the Wikibase software does exactly the same thing as your bot did? John Vandenberg (talk) 02:45, 13 May 2014 (UTC)[reply]
Answering my own question, I have found one instance where the software did exactly this! See here. :-/ John Vandenberg (talk) 07:54, 15 May 2014 (UTC)[reply]

I see that it is doing the 'right' thing at Electrowavez (Q16751650), as there is only one sitelink present (maybe it should automatically update the label if it is the same as the old sitelink?). Maybe when there are two sitelinks, and one is changed, something different needs to happen. Spawn a new item? Or maybe put it on a public page of exceptions, for humans to process. John Vandenberg (talk) 08:29, 14 May 2014 (UTC)[reply]

Merges?

[edit]

Hello! Your bot seems to make many wrong merges, with close but not identical subjects. For example this when the bot merged with location in Harry Potter (Q208361) when it should have been with Forbidden Forest (Q2481438). Or this when it should have been this. Or this, or this, or [3]… And that was just clicking at random on your contributions.

I don't know how you decide on merges but it need more control! We merge identical entries, not close ones. --Harmonia Amanda (talk) 04:57, 21 May 2015 (UTC)[reply]

Thanks for letting me know... I changed my bot to no longer follow redirects if they link to a section (have a fragment), that should catch those cases. Despite of that, this is mostly doing things we generally did in the past and still do: Adding the titles of redirects pointing to a page as aliases. I stopped the bot yesterday, because the wbmergeitems api module doesn't respect the bot flag and will only start my bot again after the fix for that has been deployed. Cheers, Hoo man (talk) 15:04, 21 May 2015 (UTC)[reply]
After poking at this some more, I think it would be better to just delete the items that aren't notable for some other reason. I will do so tomorrow or in the next few days. - Hoo man (talk) 06:21, 22 May 2015 (UTC)[reply]
Actually, it's more complex than that. Many of these redirections point to general articles but exists as a full article on others Wikipedias. So it's wrong to merge them with the general articles, they should be merged with the precise one. But there is no direct link between these redirections and these precise articles, so it need an human intervention to do that. We should probably start a page with these kind of item (no interwiki link save for a redirect one) so we can treat them. For now the interface doesn't accept redirects but the community wanted to accept them. Deleting these probably won't help in the long run as others bots are creating these items… There are probably some we should delete but from what I have seen of your contributions, they are few and far between. --Harmonia Amanda (talk) 13:02, 22 May 2015 (UTC)[reply]
For example, you deleted Lightfoot (Q6546286) instead of adding the statements it needed. Or it was necessary to have this item to complete Transformers: Super-God Masterforce (Q2985264) (notability point 3). Like I said, I'm really not sure deleting is the right thing to do… --Harmonia Amanda (talk) 16:06, 22 May 2015 (UTC)[reply]
That item Lightfoot (Q6546286) was necessary to complete Transformers: Super-God Masterforce (Q2985264), but that wasn't clear from the state of the item (or from Tsutomu Kashiwakura (Q498624)) before you added more data to it. Due to the fact that these items don't have much (if any) use, can hardly be maintained or enhanced (as they are not linked to anything nor do they have statements) and since it's very easy to recreate them (should there be a use), I think they should be deleted. - Hoo man (talk) 22:33, 22 May 2015 (UTC)[reply]

Namespace category on ext.wikipedia

[edit]

@Hoo man: On edits such Special:Diff/255671298 your bot have removed the category page for ext.wikipedia. Apparently the namespace translation was changed from Categoría to Categoria (try opening the page ext:Category:Austrália). Is possible to re-add those pages within the new prefix? Lugusto (talk) 14:52, 9 October 2015 (UTC)[reply]

Supported. See [4]. Lymantria (talk) 11:08, 11 October 2015 (UTC)[reply]
Very weird… Usually these old names should have become namespace aliases which my bot can handle (it would have changed the sitelinks then to have the new namespace name). But for some reason that didn't happen (as you can see here), that also means that the old pages are no longer accessible at all (using their old names). I will compile a list of all links removed and re-add them today. Cheers, Hoo man (talk) 19:50, 11 October 2015 (UTC)[reply]
@555, Lymantria: All links have been restored (using the new namespace name of course). Thanks for letting me know. Cheers, Hoo man (talk) 23:13, 11 October 2015 (UTC)[reply]
Thank you! Lymantria (talk) 05:21, 12 October 2015 (UTC)[reply]

User pages

[edit]

Hi. I saw this edit by your bot. I think you should tell your bot that user pages are not allowed as site links. 188.99.140.133 12:38, 4 October 2016 (UTC)[reply]

Redirect to disambiguation

[edit]

Do you think it's possible avoid this kind of modify in disambiguation item? --ValterVB (talk) 17:25, 4 October 2016 (UTC)[reply]