The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section. The result of the discussion was Approved.
Some more info: Addbot processes this toolserver report, which in turn is based on WikiProject Orphanage criteria. When Addbot requests the data, a process is kicked off that verifies the orphan status of the 5000 articles in the list, so the orphan status is up to date as of the request. The list is organized by link count; no one-link orphan articles appear until all zero-link orphans have been tagged. All the work Addbot has done thus far has been on zero-link orphans. --JaGatalk 09:17, 16 February 2009 (UTC)[reply]
Comments I oppose this task for this bot. It is being used to tag organism species stubs, for which there may never be more than one link from the genus article, if even that. In some cases genera articles will not have the species listed, merely a link to a list of species, or the species list may be in the family article for plants or order article for animals as the primary grouping units for those particular organisms. The tag in no way aids the reader of the article, and, in fact, may confuse the reader of the article, especially when the tag is longer and larger than the stub, thereby trumping article space. In addition, as usual there is much confusion over what is an orphan, here, for example, is another definition of orphan:
Articles which should link to the orphaned article need expansion.
Parent articles were deleted, or their references (wikilinks) were removed.
The article was created in error: perhaps its name is mispelled or not a well known variation of the name.
The article is not noteworthy.
In the case of a single obscure species, the parent article, the genus or family or order, depending upon what type of organism, may not be in need of expansion, simply because the species article is a stub, these two criteria don't tie directly when dealing with plants, which are sorted by their hierarchial rank (Linnean or APG clade with plant article).
The parent article is probably the family article and plant family articles are not deleted on Wikipedia for any reason. Their contents may be deleted if they're copyvios, I suppose, but the plant family itself will still have an article, so this does not apply.
If the article was created in error, it should be deleted, not tagged orphan.
All named plant species and even a few un-named ones are considered notable by Wikipedia policies and practices already in case.
So, the bot is doing a task that is against guidelines.
Oh, and the policy on wikilinking says common sense prevails. I could make species list in all the genera articles, but these would just be eventually converted to list of genera articles, and then the article would still be an orphan.
You haven't served the reader. You haven't served the editor. The tag hogs the article when it comes to a species stub. The Wikiproject that Addbot's programmer defers to for guidelines doesn't edit the species articles, so you haven't done anything except tag the article and over-ride the point of the article, the most basic information about the species.
The bot is not just being used to tag organism species stubs, It is slowly tagging ALL orphans. The tag is not aimed to aid the reader but it is aimed to suggest things to be edited by those that go around adding links to pages to fix orphans. The criteria to become an orphan that I am using is very clear, WikiProject Orphanage criteria.
"If the article was created in error, it should be deleted, not tagged orphan." Yes and it probably would be deleted and not tagged as an orphan. The main scan by the toolserver page to find orphans does not happen very often, (maybe a few weeks). I do not see how the bot is doing a task that is against guidelines and you say the bot doesnt serve the editor but its does. The orpahnage will edit any article they need to to make an article non orphan, saying they dont edit the species articles could in time be wrong as they get around to the backlog of orphans. ·Add§hore·TalkToMe! 09:32, 16 February 2009 (UTC)[reply]
(edit conflict) Yes, the bot is tagging organism species stubs. Here are diffs.[1] Good god, I can't believe this time wasting. And, no, wikiproject orphanage editors are not editing organism species articles to make them non-orphans, in fact I could only find a few species article edits in the dozen edit histories of project members I looked at. It is doing a task that is against guidelines when it says use common sense, and I tell you that common sense says these articles may wind up just being orphans without three links, and you insist they have to have an over-powering orphan tag.
No, they're not editing species articles, they're not doing it. And the guidelines on species is that named organisms are notable. If you want to change that policy, go ahead and change that policy, but don't have your bot work against it. --KP Botany (talk) 09:47, 16 February 2009 (UTC)[reply]
The hope is, some day, an editor will see the tag and will build the article into the web. I always imagined navboxes that link flowers sharing some characteristic would do the trick. For instance, for Arabis kennedyae you could have a navbox that holds only flowers of that genus. Each Arabis article would have the navbox. Arabis kennedyae would then be linked to many articles, instead of being completely isolated as it is now, and user navigation would be much easier. Furthermore, a navbox would look more attractive than the ugly list currently on genera Arabis. Now, I'm not saying you have to do this. But if we put the tag in place, the article might get connected with someone willing to do the work. I don't see that it does any harm to the article and could help improve it vastly. That's the point of our entire project - build the web. --JaGatalk 10:00, 16 February 2009 (UTC)[reply]
What? So, we're going to link all red flowers together, all flowers with three petals together? I don't have a clue what you are talking about, this is original research. Plant species articles discuss the botanical, horticultural, evolutionary, and ethnobotanical story of a plant, they're not arbitrarily and randomly linked just so a wikiproject can have something to do, that they're not doing. --KP Botany (talk) 10:07, 16 February 2009 (UTC)[reply]
Well im getting bored of all these long messages, why dont we all stop fighting and come up with something sensible? Such as i change to bot so it will only edit pages over a certain length or it will ignore all stubs? ·Add§hore·TalkToMe! 09:37, 16 February 2009 (UTC)[reply]
One problem with that is it would be like having Special:Lonelypages all over again; the list would fill with 5000 stubs that would never be removed to make room for new articles. --JaGatalk 09:44, 16 February 2009 (UTC)[reply]
::::Why don't you ignore organism species articles for plants, and find out what unit should be ignored for other organisms, possibly a "higher" level in the hierarchy for other organisms? Seeing an orphan tag attached to a plant order article was very useful, because that should not be the case, ditto families. Why not ask at the organism wiki projects, since this is not a policy guideline, but the criteria of another wikiproject? In fact, why don't they just ask the other wikiprojects how they want their articles tagged? You know, consensus. You can program the bot to read the taxobox for the species. I'll ask at Wikiproject plants what they want. Also there's a notice at Wikipedia:AN about this. I'm not the only one who finds it distracting rather than useful.[2] --KP Botany (talk) 09:51, 16 February 2009 (UTC)[reply]
{{BAGAssistanceNeeded}} Lets get some bot input, Can someone from AG please assess the technical bot side to this request? ·Add§hore·TalkToMe! 09:57, 16 February 2009 (UTC)[reply]
I take it back. Now that Addshore is calling me "insanely specific"[3] for holding my arguments to only the plant species stubs, I want the bot to stop tagging all articles with orphan tags. If I can't argue for having it stop tagging for only the very specific reason I see it as being a problem, then I guess I'll argue that it simply shouldn't be tagging articles at all until a community consensus is reached about what precisely is an organ. --KP Botany (talk) 10:07, 16 February 2009 (UTC)[reply]
I raised an issue with this bot over the Orca Basin article. Just to make it clear that without the bot I wouldn't have got around to adding the links on other pages and the same is true of a group of articles that I'm watching that have recently been tagged (I'll be working on them over the next few days). IMO the tag is a very useful spur to get on and create links, I have no problem with it, I just thought that it might be malfunctioning in the particular case I mentioned here [4]. As to your suggestion Addshore, I would say yes to minimum article length as a criteria, but no to ignoring all stubs. I don't know if that would answer KPB's problems, which I entirely understand BTW, I'm just talking from my perspective. Mikenorton (talk) 10:19, 16 February 2009 (UTC)[reply]
As I replied to this on Addshore's talk page, I'm not a horse, and I don't need spurred on. If that's the purpose of the tag the wikiproject needs to be reminded that wikipedia is a group of volunteer editors, not personal slaves or draft animals. I dislike having other people set my volunteer time schedule to their desires. You could be spurred on by a note on your talk page. And, I think you tried to force the wikilinking, and it was not particularly useful. I reverted one of your links, and I think the Gulf of Mexico link needs more work, but I left it. However, yes, I found the bot's orphan tags useful on a couple of articles, but, Addshore won't let me say that, since he/she's now claiming I'm "insanely specific about plant stubs," the articles I had issues with concerning this tag. --KP Botany (talk) 10:24, 16 February 2009 (UTC)[reply]
Sorry but it is when you say things such as "This tag has not been used for years on the plant articles I've been editing." and "Link the bot approval page that shows where you got approval to add this tag to plant species stubs." I dont want this getting incivil so lets keep this to the point. The bot is a general bot and is not only tagging these plant stubs. They fit the criteria and are therefor tagged. ·Add§hore·TalkToMe! 10:31, 16 February 2009 (UTC)[reply]
You find it threatening or something that someone asks you to link the bot approval? You do understand that your bot runs by community consensus, don't you? And, fact of the matter is, I haven't seen a tag like this on the edit I've done on plant species stubs for a couple of years, so I said that. Now, you seem pissed off that I'm only bothered by the tag on the plant or species articles (I reverted it on another organism, also, a fruit fly I think), so I'll gladly protest the bot for everything.
Right now, since the bot owner is paranoid about being asked to link the bot approval for other editors to read, I protest the bot owner operating a bot. --KP Botany (talk) 10:38, 16 February 2009 (UTC)[reply]
I am not trying to avoide linking to the approval I was running under, It is located here. I have already said this on the discussion on my talk page. I understand that my bot runs from consensus and that I have gained from the BRFA linked. ·Add§hore·TalkToMe! 10:43, 16 February 2009 (UTC)[reply]
Then what's your problem with my requesting that you post it? You say above that it's an issue that I requested that you post it. You raised the issue, and you continue to discount my concerns, and also demand that I be concerned with things that don't bother me with the bot.
I'll be taking a break now, and maybe you can argue why I have to dislike the tags on all articles, rather than just the articles I think it is inappropriate to tag. However, I offer up, that if you think I should be obliged to argue against tagging all articles tagged by the bot with the orphan tag, you've kind of shot this request down without anyone else. --KP Botany (talk) 10:48, 16 February 2009 (UTC)[reply]
I have no probelm with requesting that I post it (as i just did) it was simply that you asked for a BRFA specificly for a "Link the bot approval page that shows where you got approval to add this tag to plant species stubs". I am not discounting your concerns but from my eyes there is nothing wrong with what the bot has been doing. I am not demanding that you be concerned with things that bother you but at a guess if if concerns you that a plant stub gets tagged with an Orphan tag then does it also concern you that maybe a vehicle stub be tagged with an Orphan tag?
Again why is it inappropriate to tag the plants stubs? They fit the criteria... ·Add§hore·TalkToMe! 11:00, 16 February 2009 (UTC)[reply]
I agree with KPB that some articles, such as the ones he describes, are difficult to link to and most such links would be entirely artificial (just keeping the bots at bay). Any way that you could exclude specific categories from the bots activities? There are special cases out there and bot designers need to be aware of them, despite the wish to make the application as general as possible, which I also understand. Mikenorton (talk) 11:10, 16 February 2009 (UTC)[reply]
Good god, are you going to listen, if I answer, since I explained it in all my posts, and now you're asking me why? It's inappropriate because obscure botanical species may not have any more links that to a list of genera, or, if it's been inappropriately created at the species level and should be a genus article, to the family. Not all genera lend themselves to a list within an article, some genera are huge, hundreds of species, meaning their parent article may be very large and full of information the makes adding a list of all of the genera pointless. In which case, there may be a separate page with a list of genera, and lists are specifically excluded from counting as the type of link that is considered a valid link. So, an obscure plant species may naturally only be mentioned in only one article. It's linked upwards to its genus, and its family, the latter being the primary taxonomic unit for plants. Family pages would be the natural place, then, to wikilink to species articles, except, again, some families are so huge that their species lists would be hundreds of thousands of bites, again, making the list an impossible asset for reader usage in the family article, and again, requiring a separate list only page, that would be excluded from the viable links that would forestall an "orphan" tag. So, at best, many obscure species of plants could have only one link, if it is in a genus with a only a few species, and that link would be from a list within the article. Other possible links would be artificially created solely for the purpose of wikilinking, as above, red flowers, three-petaled flowers. This would be original research, which is against policy, not just guidelines.
Now, the reason I am "insanely specific" about plant species (don't call people names, it's not civil), is that is only matters in plant species articles, and in animal species articles, and probably in a few other organisms, as a function of the criteria for notability for organism species articles: if it's got a scientific name, it's notable. The issue does not arise with higher order taxa in plant articles, as there is no reason that higher order taxa should be without links to the article because the higher order taxa are all extensively linked through species taxoboxes.
And, in addition, many plant species stubs and other organism species stubs will be very short articles, with less text than the tag. The tag itself links to a wikiprojects page, and, as such, should be on a talk page, not an article page, because it is not linked to information for the user, but rather, info for potential editors. Article space is not the proper location for recruiting editors for projects.
Well I think at the moment the best chance of resolving this is here. And I think this would be a change for the better. But then would this impact the other maintance tags e.t.c ·Add§hore·TalkToMe! 11:34, 16 February 2009 (UTC)[reply]
Note that proposals to move maintenance tags to talk pages have been routinely rejected at Wikipedia:VPP. So far, I see most of the discussion above consists of User:KP Botany repeating concerns at length, and User:Addshore and User:JaGa replying. The situation is a little better at the linked discussions, but not by much; the concern in all cases seems to involve species stubs. From a technical perspective, there's not really anything to comment on: no source is posted, and it's not even clear if anything is changing since the last BRFA (when the language above is listed as "PHP", does that mean this part of Addbot is being redone in PHP or is it referring to the language of the toolserver report?).
I suggest that User:KP Botany stop repeating the same concerns, stop misrepresenting others' statements, and generally calm down. We are abundantly clear on your position, there is no need to drown out all other discussion by repeating it. Then hopefully everyone can get on with trying to find a way to address the concerns. Possibilities include adding a {{not-an-orphan}} tag (possibly with no output), adjusting the orphanage criteria to require 1 link instead of 3 for certain classes of articles, tagging the articles (or certain classes) with a {{WikiProject Orphanage}} wikiproject banner on the talk page instead of {{orphan}}, or just accepting that these articles really are orphans by the definition. Anomie⚔ 12:59, 16 February 2009 (UTC)[reply]
I was asked to repeat myself.[5] I always ask myself on Wikipedia, how rude and inconsiderate can another editor be, then it's always topped. Either read and follow or don't post. --KP Botany (talk) 21:34, 16 February 2009 (UTC)[reply]
The bot itself is now PHP to speed things up rather than using AWB as it did before which turned out the be rather slow. As you have said the movment of templates to the talk page has been turned down then I dont think that is going to change. My next bet is seeing if the criteria are to change. ·Add§hore·TalkToMe! 17:02, 16 February 2009 (UTC)[reply]
Addbot was only tagging articles with zero links whatsoever, which is entirely within the BRFA. I don't think that anyone can logically dispute that an article with zero incoming links is an orphan, and Addbot was approved for tagging orphans. A debate over what is and what is not an orphan should be had, but not here. The question here is, is Addbot going beyond its BRFA, and the answer is no. --JaGatalk 19:13, 16 February 2009 (UTC)[reply]
It's not tagging articles "with zero links whatsoever," it's tagging articles that do not have incoming wikilinks from other articles, so, please, if I'm to be asked to repeat myself, then told to shut the fuck up and stop repeating myself, don't repeat yourself when you're not asked to, and stop misrepresenting the facts, so that I have to correct them. Oh, wait, no one is telling you to stop repeating yourself, only me, because others disagree with my viewpoint. Oh, wait, no, Anomie does not disagree with me, since Anomie didn't even read the discussion. So, go ahead and repeat yourself. But I'll repeat correcting you. --KP Botany (talk) 21:39, 16 February 2009 (UTC)[reply]
The bot needs a break. --KP Botany (talk) 22:09, 16 February 2009 (UTC)[reply]
That error was reported, and Addshore has said that he will fix the bug. And in effect, the bot is on a 'break', at least from orphan-tagging, as it has not done any tagging since the error was reported. Richard0612 22:17, 16 February 2009 (UTC)[reply]
Thanks for the information. --KP Botany (talk) 22:22, 16 February 2009 (UTC)[reply]
For those that care the fix is simply somethign along the lines of:
if(empty($text)){continue;}
For the page being totally empty (not acctually happening here, or
if$text=$orphantagcontinue;}
Or something similar.. I could also simply make sure the sive of $text is bigger after the adition of the tag. ·Add§hore·TalkToMe! 22:24, 16 February 2009 (UTC)[reply]
If you'd like me to give the code a once-over, just let me know (on-wiki, email, or IRC). Anomie⚔ 22:29, 16 February 2009 (UTC)[reply]
I don't see how the two codes above dealt with the issue of the bot removing all of the text from an article and replacing it with the orphan tag? Why is it removing any text when adding an orphan tag? It shouldn't. Yes, the little text afterward, that the text should be bigger after addition wouldn't hurt, but it also doesn't seem to deal with the issue of the two articles above which were deleted in their entirety. --KP Botany (talk) 22:30, 16 February 2009 (UTC)[reply]
The two code tweaks above fixes the bug, the bug fixes the problem even though the problem has already slightly impacted wikipedia. The examples shown have already been reverted and no doubt any other cases have also been reverted and rightly, I suspect these are the only cases as they have been the only cases reported and when a bot removes and article with a link saying "Report Errors" in the summary people generally do that.
Anomie what is your IRC nick? Get hold of me :) Currently Addshore_·Add§hore·TalkToMe! 22:33, 16 February 2009 (UTC)[reply]
Neither of these deal with the instances above, as they don't address what happened. The bot removed all of the code of an article, it won't be stopped from doing that by continuing when the article is empty text, that will fix the other bug, and it won't be stopped by continuing when the code is only an orphan tag, because neither of these deal with the issue, unless you're talking about the last being after the bot has already deleted the entire article. --KP Botany (talk) 22:40, 16 February 2009 (UTC)[reply]
$text is the variable of what the bot thinks the text is on the page, If $text = nothing dont continue. In the end the save is simply $wpi->post($page,$text,$summary,true); and $text = "$orphantag\n$text";. the only way for the above to occour is if $text doesnt equal anything. This will also fix the creation of articles containing only an orpahn tag. ·Add§hore·TalkToMe! 22:45, 16 February 2009 (UTC)[reply]
No, I'm not getting it. Please explain it on my talk page, as neither of these cases appear to be the case where the text was NOT nothing and it is changed. Okay, text is not nothing, don't continue to do what? It's adding orphan tags to articles that are not nothing, so where is the code that deals with this situation. You're not addressing the two links above with what the bot does that causes it to remove text. Why is the bot removing any text from an article when adding an orphan tag? This should not happen, as the code should simply not allow the removal of text to place a tag. --KP Botany (talk) 22:57, 16 February 2009 (UTC)[reply]
If $text = nothing the bot WILL NOTmake the edit, the $text comes before the tag is added, therefore it is what the bot thinks the page content is (requested from api). The bot is not made to remove content from an article and the code will not allow this to happen once i have added the fixes pasted above. The reason the content of the page was removed was more than likely an api bug or a drop in connection or something imposible virtually imposible to take accout for. -·Add§hore·TalkToMe! 23:05, 16 February 2009 (UTC)[reply]
I have to disagree that it was something impossible to take account for. This concerns me. --KP Botany (talk) 23:14, 16 February 2009 (UTC)[reply]
Changed to virtually imposible, but again with the fixes mentioned above the two example shown would not be able to happen. ·Add§hore·TalkToMe! 23:17, 16 February 2009 (UTC)[reply]
From what I understand, the second code block above would prevent the orphan tag replacing the text on the page (i.e. deleting everything) as it essentially means 'if the text of the page (I assume after editing) contains only the orphan tag, continue without editing' (personally, I would have the bot log the error, but that's neither here nor there). It may not be the most elegant solution to the blanking problem, but I think it would work. Richard0612 23:23, 16 February 2009 (UTC)[reply]
If the bot gets the pagetext from the API, it should (could) also be using the API to edit. In this case, just use the 'prependtext' parameter with action=edit, then it might not even have to get the pagetext, unless it does something with it besides adding the template. The proposed fix sounds like more of a hack than anything else. Mr.Z-man 23:40, 16 February 2009 (UTC)[reply]
The content of the page needs to be retreived from the API as the tag is not always placed at the top of the page.
If the page contains these then it should paste below e.t.c ·Add§hore·TalkToMe! 23:44, 16 February 2009 (UTC)[reply]
I should also note that accounting for the case where the API gives an incorrect result should be fairly easy, unless you expect that it should return nothing in some cases. Just check if the query result returned something, and if not, redo the query. Mr.Z-man 23:43, 16 February 2009 (UTC)[reply]
If th API does return nothing at all the bot will simply die (does not normally return nothing at all). Somthing else must be happening randomly but it makes little sense as it happens to only a few articles edited and I can not see any link between them. ·Add§hore·TalkToMe! 23:48, 16 February 2009 (UTC)[reply]
When writing my own bot framework using the API, I found this mailing list postbyUser:CBM rather helpful in explaining what needed to be accounted for. Mr.Z-man 02:23, 17 February 2009 (UTC)[reply]
Thanks for that, I have also been thinking about writing my own framework for my php bots, I am currently using my own modified Wikibot classes (cobi's classess). It would be nice to be able to pick an choose exactly what happens :). Thanks again for the link. ·Add§hore·TalkToMe! 08:40, 17 February 2009 (UTC)[reply]
{{BAGAssistanceNeeded}} Ok The list of articles that the bot gets its artiles to edit from here is now 100% only showing articles that have NO Incomming links from the article space. With this which fixes the problem of what is an orphan and the fixed bugs I see no reason for the bot not to be approved (unless I again have missed something). Would appreciate someone looking over this, ·Add§hore·TalkToMe! 17:41, 17 February 2009 (UTC)[reply]
Approved for trial (25 edits). Please provide a link to the relevant contributions and/or diffs when the trial is complete. Seeing as the list has been updated to only cover articles with no incoming links (which are quite unambiguously orphans) and the bug that led to pages being blanked has been fixed, I think a trial run is warranted. It should also help to cut through some of the confusion and get everything back on track. Richard0612 21:39, 17 February 2009 (UTC)[reply]
Every edit went as expected, no errors, I might increase the limit for posts to try to reduce the ammount of smaller articles tagged, does any one have any suggestions? ·Add§hore·TalkToMe! 22:32, 17 February 2009 (UTC)[reply]
You'll probably want to redo that database query before running fully (and redo it every few runs to make sure it doesn't get stale). The reason I say this is that the bot taggedAperture (botany), when the page has one incoming link from the article space. Richard0612 16:07, 18 February 2009 (UTC)[reply]
That article that you referred to has one incoming link but it is a disambiguation and is therefor ignored (disambiguation pages are not counted towards the incoming links). Also there is no need to refresh the list until the bot has completed a full sweep as if an article in the list is edited when it is in the list, it will be removed and will have to wait for the next scan till it can be re checked. ·Add§hore·TalkToMe! 16:11, 18 February 2009 (UTC)[reply]
I did wonder after I made the above comment whether DAB pages counted as 'incoming links'. In that case my comment about the database query is really irrelevant, and I see no reason not to mark this request as (for pages with no incoming links as defined by the toolserver query) Approved.Richard0612 16:18, 18 February 2009 (UTC)[reply]
The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.
What the heck is approved? You've rejected all arguments to the contrary and are just going to make the bot do the same thing it should not have been doing without offering one single contrary reason why it should tag species stubs? --KP Botany (talk) 03:50, 19 February 2009 (UTC)[reply]
Replied on talk page. Richard0612 11:37, 19 February 2009 (UTC)[reply]
And it's trial run appears to have consisted of reverting my edits. What the hell, I'll just revert the bot, since it seems that my human editing is not valued in the least bit. Oops, I'm supposed to shut the fuck up.
Oh, and one more thing: Wikipedia should not be allowing bots run by programmers who thing codes do "random" things. What's wrong with the bot? The programmer. Maybe you need a few non-programmers and non-bot owners on Wikipedia bots, since it seems you think computer programs behave "randomly." --KP Botany (talk) 08:45, 21 February 2009 (UTC)[reply]
Approved.
Approved for trial (25 edits). Please provide a link to the relevant contributions and/or diffs when the trial is complete. Seeing as the list has been updated to only cover articles with no incoming links (which are quite unambiguously orphans) and the bug that led to pages being blanked has been fixed, I think a trial run is warranted. It should also help to cut through some of the confusion and get everything back on track. Richard0612 21:39, 17 February 2009 (UTC)[reply]
Approved. Richard0612 16:18, 18 February 2009 (UTC)[reply]
