Wikipedia

User talk:Pbsouthwood/Archive 7

< User talk:Pbsouthwood
Latest comment: 6 years ago by Pbsouthwood in topic Interested in becoming a new page patroller?
Archive 1←Archive 5Archive 6Archive 7Archive 8Archive 9Archive 10

Outline of underwater diving

Latest comment: 6 years ago25 comments2 people in discussion

You wrote:

At what point should a draft outline be moved to mainspace? It seems that there are people opposed to incomplete outlines so I would prefer to avoid an attack of knee-jerk deletionism by getting Draft:Outline of underwater diving to an easily defensible condition before making the move. On the other hand, I do not want to delay unnecessarily, as the outline will be useful to WP:SCUBA and anyone interested in underwater diving.

I am making use of the recently created experimental short descriptions as annotations on WP:SCUBA articles, so will have annotations for most if not all of the links already in the list, however there are hundreds more article which will go into other sections, many of which do not yet have a short description. Cheers, · · · Peter (Southwood) (talk): 09:55, 11 February 2018 (UTC)

Nice job on the Outline of underwater diving. I look forward to seeing how it develops further.
It's way past the stub stage, so I wouldn't worry too much about knee-jerk deletion. The knee-jerking these days is from editors who aren't familiar with outlines and who don't look below the top of the page and think it is a copy of an article.
We've lost very few outlines to AfD.
I've taken the liberty of moving it to article space and have placed a construction tag on it, to show it's still being actively worked on.
Welcome to outline building. I'll post some more comments (and maybe some tips) soon. Keep up the good work.    The Transhumanist 10:27, 11 February 2018 (UTC)
That will probably give me an edit conflict, but thanks anyway. · · · Peter (Southwood) (talk): 10:30, 11 February 2018 (UTC)
You can overwrite it and then put the construction tag back, or leave the tag off. Either way. No worries. :)    The Transhumanist 10:33, 11 February 2018 (UTC)
Since I was expecting it I managed to recover without undue difficulty. That is the first time I have had an article disappear on me during an edit. It could have come as quite a shock. I would like your opinion on the utility of short descriptions for this purpose. Cheers · · · Peter (Southwood) (talk): 10:37, 11 February 2018 (UTC)
I don't understand your request. But, concerning edit conflicts, the program displays your version at the bottom of the edit conflict page, so your edits are not lost. You can either save them or cut and paste them, from that page. What did you mean by "short descriptions"? And for what purpose, exactly?    The Transhumanist 10:46, 11 February 2018 (UTC)
[ec] As I said, I recovered from the edit conflict without loss - pasted the section I had been working on over the equivalent on the moved article, saved, and all is fixed.
Short descriptions will be added to all main space Wikipedia articles (yes, eventually all 5.5 million of them - a non-trivial endeavour) to provide disambiguation of search results following several discussions and an RFC. I take it that you were not aware of the issue. I can provide links if you want the whole sordid story, but it started with problems in the quality of short descriptions WMF was drawing from Wikidata to describe Wikipedia articles on mobile view and in Visual Editor, which were a mixture of good, indifferent and bad, and included vandalism and BLP violations on Wikidata which Wikipedians were not interested in patrolling, and much of the time could not even see. This will change to a short description hosted on Wikipedia where Wikipedians can curate it along with the rest of the article. They are intended to provide enough information to the reader to identify what the article is about, while remaining small enough to work usefully on a mobile device. I ran a trial on WP:SCUBA, where I added about 470 short descriptions, which I am now copying over to the outline of underwater diving as annotations to the relevant links. If the annotations I am using appear suitable for purpose, then short descriptions should be a major resource for WP:Outlines, and may allow you to make greatly improved automated outlines. The remaining work will largely be sorting into the sections. Cheers, · · · Peter (Southwood) (talk): 11:08, 11 February 2018 (UTC)

I wrote a script in perl to do this several years ago in conjunction with WP:AWB's external script calling feature, lifting the first 2 sentences from each article's lead to insert as annotations, but ran into numerous problems, the main one being that it wasn't scalable, as it required thorough proofreading/copyediting because many of the leads included redundant phrases. For example, "is a species of shark" showing up again and again in a list of shark species, or "is a branch of" showing up in a "Branches of" section. There were also problems with redirects and disambiguation page links, resulting in gaps the script couldn't fill, requiring a human editor. I expect these challenges will still need to be overcome even with the short descriptions. Though those look like they'll make better annotations than the lead sentences.

By the way, what's the function of the magic word?    The Transhumanist 15:00, 11 February 2018 (UTC)

You may find the Phabricator ticket helps. · · · Peter (Southwood) (talk): 16:26, 11 February 2018 (UTC)

Initial question

I just realized that I didn't answer your initial question above. The answer is: when it is less likely to be deleted. A new outline might be nominated for deletion if it is full of red links, has mostly empty sections, or it is an exact duplication of a navigation footer (not that that is a valid reason for deletion, it isn't -- see WP:CLN). More than likely, though, a new page not ready for article space will simply be moved back to draft space. To avoid negative attention, it is helpful to develop it past those limitations. Annotations definitely help, as they are a feature not supported by navigation templates.

By they way, where are the "recently created experimental short descriptions" you mentioned in your initial post?    The Transhumanist 10:58, 11 February 2018 (UTC)

Thanks for the clarification.
They can be found on all articles in [[Category:Articles with short description]], usually as the first line of code. · · · Peter (Southwood) (talk): 11:11, 11 February 2018 (UTC)
Wow. That short description project could sure make gathering annotations easier for outlines. Don't know what to make of the magic word, though.
By the way, you can provide a link to a category, without it being processed as a category tag, by including a leading colon (before "Category"), like this: [[:Category:Articles with short description]], which shows up like this: Category:Articles with short description.
For other useful tips, see WP:TOTD. To display the daily tip, include {{totd}} somewhere on the page. If you put it inside ref tags, like this: <ref>{{totd}}</ref>, it will always appear at the bottom of the page.    The Transhumanist 14:30, 11 February 2018 (UTC)
P.S.: Sorry, thought you were new. Now I see you've been around for 9 years. Congrats, we're lucky to have you here. Cheers. -TT
No worries, we learn new things all the time. I didn't know about outline or index list articles until yesterday when I saw something you had posted somewhere, and as you say - 9 years, and not idle ones either. I do also tend to forget things I don't use often enough, like the leading colon, but the tip of the day ref tag trick is a new one. Strangely obvious though when you think about it. I wish I could see the obvious more often... I have been curating WP:SCUBA for longer than I care to think about, and when I saw that Outlines exist I knew we needed one, so here I am creating it. Looks like it is going to be big. I may have to split it later, but that is a problem for another day.
The magic word is WMF's solution, I don't know enough about them to know how valid it is, but the story goes that it is significantly faster to run than other possible solutions, which is a good reason if true. It may even have some advantages for outlines if you can use it to call up short descriptions from the source articles in run time which would make updating automatic. Maybe some template which uses the title as parameter, displays it as a link and calls up the short description as an annotation. The coding would be totally beyond me, but the concept looks possible. · · · Peter (Southwood) (talk): 15:07, 11 February 2018 (UTC)
I felt the same way when I found WP:JWB the other day. It's been around for years, but I was clueless.
In regards to our respective awareness of Wikipedia know-how, perhaps we should swap our favorite tricks. I'll start the first round of trading here: I've configured my account to show redirects in green, and links to disambiguation pages in orange. Your turn.
Speaking of big outlines, check out Outline of Buddhism. They win the prize for biggest outline. I had no idea Buddhism was that extensive.
JavaScript coding is pretty easy once you get the hang of it. I'm in the process of writing walk-throughs for each of my scripts (on their talk pages) that explain the source code line by line, so that others can learn JavaScript faster than I did. It's also so I can remember exactly what the code does so the language doesn't fade if I don't look at it for a few days. :)
Everything JavaScript (user scripts, gadgets, articles, etc.) is being organized by WP:JS. Since my main involvement with WP right now is writing scripts to develop outlines, I've taken on the curating of the the JavaScript WikiProject as well. If you are interested in learning JavaScript, I suggest you start with using user scripts (if you haven't already). Writing anything into your vector.js page makes you a JS programmer, because that page is a user script!
By the way, which user scripts do you use, and which gadgets do you have activated? (Most gadgets are user scripts, that is, JavaScript programs; the rest are .css snippets).
Anything you could tell me about the short descriptions project would be helpful.
Want to hear a slightly humorous story?
Outlines started out as "Basic topics lists". So, Outline of geology used to be List of basic geology topics. There were a set of these in the Wikipedia namespace from the very beginning of Wikipedia. They were used as article wish lists to gather topics for general subjects. Initially they were lists of redlinks. You would go to them to see what articles needed to be written in your area of interest.
Well, eventually the links all turned blue, and everybody forgot about these lists. After gathering dust for years, I stumbled across them and said "Hey, these are useful lists of links. I think I'll move them to article space." And so, I did. They became the Basic Topics List WikiProject.
There were already other lists called topics lists, like List of Ireland-related topics. The idea behind the basic topic lists were that they were basic lists, while the topics lists were more comprehensive.
But then something unexpected happened. There was no off-switch. The department was better supported, and became more visible. Editors wouldn't stop adding to them. Wikis are designed by their very nature to grow. There's no way to stop it. Pretty soon most of the basic lists were more comprehensive than their topic list counterparts. Some of the basic lists were several times larger than the topics lists on the same subjects. It was embarrassing.
So, through a major political struggle, the department was re-established as the Outline WikiProject, which shared the exact same scope as the topics lists (fortunately, nobody noticed that). Soon, the collection of outlines surpassed the number of topics lists, and the latter have mostly been absorbed into the outline collection and renamed to outlines.
Some outlines par excellence you might want to take a look at include:
Outline of canoeing and kayaking (fully annotated)
Outline of sailing
Outline of fishing (fully annotated)
Outline of forestry (extensively annotated)
Common errors in outlines include:
  • Breaking the hierarchical structure. Outlines are trees and should be parseable as such (a parser should be able to easily identify parents/offspring/offspring's_offspring/offspring's_offspring's_offspring/etc.
  • Prose (free-standing paragraphs) anywhere in the outline. Annotations are not considered paragraphs in this context. Lead paragraphs tend to creep in here and there in the outline system. It may be due to the influence of Featured List Criteria. That's fine for regular lists, but outlines are taxonomical lists.
  • Embedding topic navigation footers in or even at the end of the outline. They aren't part of the outline, and they aren't in outline format. They can't be annotated, they disrupt the outline's internal structure, they introduce redundancy, and render the make list feature in WP:AWB useless. The links from on-topic nav footers should be incorporated into the outline, and the nav footer removed. Off-topic or parent nav footers should definitely be removed. The only nav footer that should be on outlines is the outlines nav footer.
  • List section. This is a deprecated feature. Lists are a type of topic link, and therefore should be placed wherever they are relevant in the outline.
  • General concepts – this is a temporary holding bin for links until you know where they go. It is an in innocuous way of saying "miscellaneous", or "other". It should eventually be cleared out, all its links placed elsewhere in the outline, and removed completely. It becomes a problem in seasoned outlines, as the whole outline is a structured list of the overall subject's general concepts, and therefore this section is synonymous with the title of the outline.
  • "Miscellaneous" or "other" sections. The links should be placed in explicitly named topical sections. Every outline is a taxonomy of member topics and member topic classifications. "Miscellaneous" and "Other" are shoddy classifications. There are always better ways to classify the links in those sections. Unclassified links in a section generally go at the top of the section, before its subsections, where they are themselves default classifications. Once they grow significantly, they are typically given their own subheading.
  • Use of annotation separators other than en dashes. They are the standard, and make outlines more uniform and are thus the easiest to recognize and browse. The use of the standard facilitates easier script development. For example, the Annotation toggler script turns list item annotations off and on, based on the en dash.
  • Use of {{main}} links. All items within an outline's structure, that are not headings, are list items. List items are annotatable. Main link templates are not. They were included in early outline design, before any of us knew what we were doing.
  • Inclusion of tables. They are not outline format, and are a different format altogether. Tables are matrices, not tree structures.
I hope you found the history lesson interesting, and the editing guidance helpful. Feel free to ask any questions you might have. Now, I'm off to answer your question below...    The Transhumanist 12:58, 12 February 2018 (UTC)
Information overload! My brain is now full. The stack is so high that the bottom items are suffering gravitational collapse and disappearing into an alternate reality. It will take a little while to work through this. I have set the css to show redirects in green. That will be useful. Disambiguation pages are already orange. I can't remember how I got that. I have been using Remember the dot's code highlighter for some time, and like it. and I use vector because why not? I use my watchlist a lot but not recent changes - I hardly ever run out of more urgent things to do. I am firstly a content creator, secondly a curator of a Wikiproject. Those take up most of my available time and enthusiasm, but every now and then I can't avoid policy issues, but do not enjoy them so much as endure them for the greater good. I enjoy collaborating on articles with people who really do want to improve them, and have done a few GAs and an FA. I could probably speed up my productivity with some additional tools, but don't know which are worth the effort. My broadband is a bit slow and I do not use VE because it makes it even slower, and will not do what I want to do most of the time.
History anecdote appreciated. You get that sort of thing on Wikipedia, so mildly amusing but not surprising.
I have been using hyphens (the usual numeral line key) as annotation separators because there is a key and I need a lot of them. Is there a quick way to do en dashes? Even more useful would be how to do a find and replace. I feel there must be a way but have not found it yet. More later. Still on steep part of learning curve. I will work through your points above and fix what I have done wrong and try to avoid what I have not done wrong yet. · · · Peter (Southwood) (talk): 15:23, 12 February 2018 (UTC)
I know what you mean by information overload: the outline system as an effort to summarize all 5,500,000+ topics on Wikipedia, and how in the hell am I going to do that? I figure it will take about 10,000 outlines (existing lists do most of the rest), but the fastest I've been able to produce outlines is about 100 per year (when working full-time on them). Doing the math, it will take me 100 years to complete the project building them by-hand. And so, I decided to build programs to automate the process as much as possible, which is yet another learning curve. But using the resulting tools has turned out to be a lot of fun, and so it has been well worth the effort.
Having this stuff in print (like above) helps, because you know where to find it when you need it. Since it is going to be put to good use, I'll be happy to add to this resource (your talk page) with further helpful tips.
The orange link trick is a gadget that you probably check-marked in Preferences.
Thank you for the heads up on mw:User:Remember the dot/Syntax highlighter, I'll give it a try.
To make the watchlist more useful and easier on the eyes at the same time, I use the WatchlistSorter, which sorts your watchlist by namespace, making it much easier to browse.
Concerning tools, and other time-saving techniques, I have done extensive work covering them through the tip of the day project (which I used to curate), and now in curating the user script WikiProject (WP:JS). And I'm a bit of a power tools user (160,000+ edits). I am therefore in a position to be able to point out at least some tools that are worth your effort to learn.
I don't use Virtual Editor either (but am overdue to look at how it has progressed). You mentioned that it doesn't do what you want to do most of the time. What is that? I might know some relevant tools.
Concerning hyphens and search/replace, I'll cover that in a new section below.
As for trading tips, here's another one for you: buried deep in the special item on the sidebar menu, is a feature called Special:CategoryTree. It is great for browsing the category system. Unfortunately, it is set by default to "categories only" mode. You'll want to select the mode "pages except files", so that it shows the pages in the categories, and not just the categories.
That's all for now. Enjoy.    The Transhumanist 00:51, 13 February 2018 (UTC)

StripSearchSorted.js

I've created a user script, called StripSearchSorted.js, which has a switch (menu item in the side bar), so you can turn it on/off.

What it does is strips out all the extraneous descriptive stuff on the search results page, sorts the results, and adds link formatting for easy copy paste. It also removes those annoying redirected and category results, and includes a work around for the intitle bug. If you include intitle:"search term" (with your search term between the quotes), as your search string, the script will remove all titles that do not match. (WP's search engine fails to do that when the intitle search term includes common words like "the", "in", "of", etc.).

So, the script strips your search results to a bare list of links that are ready to be copied into a wiki editor.

And it remembers its on/off state between pages. It will stay on for all your searches until you turn it off.

I hope you find it useful.    The Transhumanist 11:24, 11 February 2018 (UTC)

I am a bit clueless with scripts, so will have to spend some time looking into this before I know enough to make a sensible comment. · · · Peter (Southwood) (talk):
The only way for you to comment then, I think, would be to install it and try it out. Here's a step-by-step, in case you are intrigued:
Step 1: If you use the Vector skin, proceed. If you don't know your skin, click on Preferences, then click on Appearance, and look at Skin at the top of the page. If it is Vector, then the script will work for you. (The script only works for users of the Vector skin).
Step 2: add this line to your vector.js page: 
importScript("User:The Transhumanist/StripSearchSorted.js");
(For an example of a vector.js page, see mine at User:The Transhumanist/vector.js)
Step 3: clear your cache. For Firefox, press Ctrl+R
Step 4: Go to the search page (click on the magnifying glass in the search box)
Step 5: Once there, type in this search term, exactly like this: intitle:"in Rome"
Step 6: In the sidebar menu, click SR sort (turn on)
Step 7: Click on 500
Step 8: Let me know what you think.
Note that the SR sort menu item only shows up on the side bar menu when you are on the search results page (and that shows up only after you've done a search).
If you do a search and everything disappears, add &limit=5000 to the url in your browser's address field. I do that anyways, because that is the maximum, and is way more than the usual maximum of 500 allowed by clicking the view number on screen.
If you don't understand any part of the above, let me know, and I'll be happy to clarify. The jargon is a bit thick. And the script is still in its alpha-test stage. ;)    The Transhumanist 09:19, 12 February 2018 (UTC)

Have you seen the movie They Live?

In that movie, there are these glasses, and if you wear them, you see things way differently, including aliens from another dimension walking amongst us, and subliminal messages everywhere.

Well, I'm working on a system for list and outline developers that will let them see Wikipedia differently than everyone else...

It will display all list items, wherever they appear, in list/outline wikiformat, ready to copy/paste, like the results in the search script above (with some further enhancements). Eventually, most it will be handled in a single script, with an on/off switch (menu item). But, while I'm developing it, I'm working on it page type by page type, in separate scripts, with no off switches.

So far, I've completed conversions for AllPagesWithPrefix, Books, and Category pages, with some other types partially working. These save time converting the links from those pages to wikilink format, because they are already in wikilink format!

I have them activated now on my account, and have gotten quite used to the view. I find this view very useful for gathering links. The one for nav boxes forces them all to the show state, which, combined with list formatting, is a mind blower.

Let me know if you are interested.    The Transhumanist 12:24, 11 February 2018 (UTC)

The thing about being clueless with scripts - same but more so :-/ · · · Peter (Southwood) (talk): 15:12, 11 February 2018 (UTC)
And no, haven't seen the movie, but it looks good in the article. · · · Peter (Southwood) (talk): 15:31, 11 February 2018 (UTC)
Okay, so there's this scene in the movie where the main character tries to get an acquaintance of his to put on a pair of the glasses. The guy refuses, because he thinks the star is crazy. But, the only way for the star to get the guy to believe him is to put them on. So he leaps on him and tries to force them on. And they fight. And fight. And fight some more. After they are both beat black and blue and nearly exhausted, the glasses finally go on... And the guy's view of the world is changed, forever.
In this case, the glasses are scripts that change what you see on the computer monitor. If you do this experiment, you'll be past the clueless stage:
Step 1: Add importScript("User:The Transhumanist/ViewAsOutline-Book.js"); to your vector.js page (if you don't use the vector skin, switch your skin to it in Preferences#Appearance)
Step 2: Clear your browser's cache
Step 3: Take a look at Book:Abraham Lincoln
Step 4: Copy and paste that page right off of the screen, to your sandbox
And at that very moment, I would love to see the expression on your face. :)    The Transhumanist 10:12, 12 February 2018 (UTC)

Dealing with articles which fit into more than one section/category

The Transhumanist, How does one handle articles which could logically fit in more than one place in the tree? · · · Peter (Southwood) (talk): 08:47, 12 February 2018 (UTC)

Keep in mind that any title could become a branch of the tree. You don't want 2 identical branches. Certainly not 2 identical big branches on the same page. So, for a duplicate branch, you include a section link instead. Make a value judgment as to which location is dominant, and put the section link in the less dominant location in the tree.
If it's just an entry (a leaf), and unlikely to grow into a branch of the tree, including it in multiple places is fine. Sometimes two such links will grow into parallel branches, at which time one should probably be pruned and section linked to the other.
Keep in mind that identical collections of links in an outline are identical branches, and one or the other should probably be pruned and section linked to the other.
Some links that will often be in an outline twice are the names of branches (fields). They'll be included in a "branches of" section, which only lists branches (fields, including subfields, as they are also fields) and then as the main topic of a branch of the outline further down the page, where the branch's (field's) subtopics are presented. The same thing applies to a types section and types. Such duplication is legitimate and normal.
Extremely large branches, regardless of whether they are duplicates or not, should probably be split off as a separate outline. You could have it as an end node (a leaf), or use a variation of the summary style, in which you provide a link to the branching outline, but still include some of its offspring beneath it (but be ready to prune it later, as all trees, and all branches of trees tend to grow - it's the nature of trees and of wikis).
Note that straight lists are also branches of outlines. The influential persons section, for instance. There may be an extensive list of them on its own page. In which case, you could provide a link to that list, and then in the spirit of summary style, include the most prominent people from that list in the outline's influential persons section. There's an example of this approach at Outline of film#Notable people from the film industry. If someone complains that the persons included are not the most famous, then you could explain they are just examples, and rename the section "Some notable directors", or whatever they are.
I hope the above answer helps.    The Transhumanist 14:05, 12 February 2018 (UTC)
Looks like it might, I will test it and see if I understand, · · · Peter (Southwood) (talk): 20:03, 12 February 2018 (UTC)

Hyphens, and search/replace

Latest comment: 6 years ago4 comments2 people in discussion

The code for en dash is &ndash ending in a semi-colon; which looks like this: –

You could type out the first one, and then copy/paste it in wherever else you need it.

Or, you could copy/paste an en dash from the screen and copy/paste that in wherever you need it.

Or you could use search/replace. Are you ready for a ride? Hold on to your chair...

Search/replace is available in 3 major places:

Text editors

Text editors, like Notepad++ (this one is very powerful yet easy to use). Cut and paste the article into there, and then do the search/replace, and cut/paste it back to WP. The things I like about Notepad++ the most are, that it supports regex, it can have many documents open at the same time, and it remembers your documents even if you didn't save them. They'll be in there the next time you use the program. An almost identical program for Linux is Notepadqq.

WikEd

Or you could use a better text editor on Wikipedia.   WP:WikEd replaces WP's standard wiki-editor. It's a gadget you can activate in preferences. It has a very good search replace feature, that even supports regex (there's a check box to activate it near the search box). Regex (regular expressions) is a powerful little language for representing searches and replaces. It was my stepping stone into programming in perl and JavaScript. Those weren't as alien to me after learning regex. Regex is especially useful, because you can include ↵ Enter (\n) into your searches for multi-line search/replaces, which is incredibly convenient for working on list items.

The nicest thing about WikEd's search feature, in my opinion, is that it can do search/replaces in text that you have selected, rather than for the whole page. It can do the whole page too, of course.

AWB and/or JWB

The killer apps for search/replace are WP:AWB and WP:JWB. These are semi-automatic editors / automatic page loaders, that will process a list of as many pages as you feed into them, and do all the search/replaces that you specify to those pages, including in regex. Instead of a mere search box, these programs can accept as many search/places as you want before you start the processing. The search/replaces are kept in a list that you can modify as needed.

These editors are semi-automatic, because you have to personally review the changes for each page and then press the save button yourself, after which they load the next page in the list automatically. Without you, they would be bots, which would make some errors because search/replacing doesn't catch all exceptions. So a human must remain in the loop. When you want to make changes to a batch of 20 or 20,000 pages, these two babies can get the job done.

Both AWB and JWB are capable of making lists of pages. After all, who can make lists of pages to process better than a program? Many options to choose from. AWB's make list feature is quite extensive.

The difference between WP:AWB and WP:JWB is that AWB has more bells and whistles but can only be installed on Windows, while JWB has fewer features but is a user script, and so it works across almost all platforms.

You'd use AWB or JWB if you wanted to look over or process all of the pages relevant to the underwater diving WikiProject, in one batch, for example.

AWB has a feature to make common fixes, while both AWB and JWB have a feature to fix standard typos (based on a list of many thousands of these). So you might want to run the diving pages through one of these just for that.

Another possible use is to feed them a list of all the talk pages for all the diving articles, and have them check those for the WikiProject's banner, placing it in there if it is missing. I use AWB's append feature to post notices to the talk pages of all outlines. I also contacted (posted messages to the talk pages of) all JavaScript programmers I could identify (by their user boxes), to tell them about the JavaScript WikiProject. But next time, I'll hone that list down to accounts with recent activity (some of the users hadn't logged in for years, but this will take use of a totally different tool).

You can also process the categories for each page, to replace a category, add a category, or remove a category.

And now for some more power tools...

AWB's database scanner

AWB has another search feature that goes way over the top, but the power of which would be hard to beat. It is the database scanner feature. You download the Wikipedia database, and use this to do searches straight from that, offline. I generally download both the articles-only version, and the full-blown version that include all namespaces. AWB can handle either. The search parameters for the database scanner are much more powerful than WP's search feature, and it also supports regex so you can fine-tune your searches to the nth-degree to find things not possible with other methods. You can search just titles, or just contents, or both. Both AWB and its database scanner have a filter as well, so you can modify results further after they are done searching. Searches with the database scanner take 20 minutes or more, so you would use this for important look ups, and have it running in the background while you work on something else in another window.

AWB's list compare tool

Another nice feature of AWB is its list compare tool. Find out which items are on only one list, or the other, or in both.

AWB's external processing feature

If AWB can't process the pages the way you want, you can use its external processing feature to have it call a script or program you've written, feed the pages from your list to it one by one, and show the results to you as a diff for your approval or rejection on each page before moving onto the next.

In conclusion...

There's more, but that's all I can think of off the top of my head.

I hope you've found the above info interesting, if not helpful.

Of course, if you have any questions, feel free to ask.

Cheers,    The Transhumanist 10:41, 13 February 2018 (UTC)

I have installed AWB and am now waiting for the permissions to come through. What you have told me convinces me that it will be worth learning to use. Steep learning curves ahead - my productivity will probably take a slight dive in the short term but then I hope it will soar to new heights. </hyperbole> Cheers, · · · Peter (Southwood) (talk): 10:44, 13 February 2018 (UTC)
Interesting, yes. Helpful? Time will tell. My guess is yes. · · · Peter (Southwood) (talk): 10:57, 13 February 2018 (UTC)
Congratulations!
By the way, while you are waiting, it may interest you that AWB's database scanner feature works without even being online, logged in, or registered (but you have to download a database first). The make list feature and the list comparer both work without being logged in or registered, but you have to be online so that AWB can gather the links you've told it to fetch.
So you can start playing around with AWB right away, if you want. I recommend getting familiar with the list features, because the power of the program lies in the quality of the lists of pages you feed into it. It can't work on what you don't tell it to work on.
A very obvious but subtle thing many people miss, is that you can copy and paste those lists into a WP page for many other uses, like creating list pages on WP, and creating lists of links to modify with WikEd's search/replace to have the make list feature read them back in to AWB. You can make a list of links, copy it to a WP page, and then have AWB's make list feature pull all the blue or red links out of it. The limitation with these approaches is that you have to add the link formatting yourself, but WikEd's regex search/replace handles this quite well like this:
Search: \n
Replace: ]]\n* [[
That will make a list like this:
Freediving
Scuba diving
Surface-supplied diving
Saturation diving
Atmospheric pressure diving
Unmanned diving
Look like this:
* [[Freediving]]
* [[Scuba diving]]
* [[Surface-supplied diving]]
* [[Saturation diving]]
* [[Atmospheric pressure diving]]
* [[Unmanned diving]]
To get a head start on regex, see AWB's regex help page.
There are plenty of regex help pages and tutorials online. Note that there are different "flavors" of regex. AWB is written in .NET and so its regex is compatible with "perl 5 regex flavor" (I think), while WikEd's and JWB's regex is "JavaScript flavor". The differences are not huge, but it is important to access the right documentation to support the regex activity you are engaged in.
I'm a somewhat advanced regex user, and so if you have any questions, I have a good chance of knowing the answer, or figuring it out. There is almost no end to the complexity of a regex you can write.
The learning curve for AWB is long but not particularly steep, except for a few esoteric poorly documented features (which I've yet to figure out), or when you start pushing the limits of the program...
The trickiest problem I've run into so far with AWB, was that its prepend feature pulls in the page title via the %%title%% key word, which can be very useful. But I was processing outlines, which all have "Outline of" in their titles. And rather than "Outline of subject", I needed just "subject". AWB's search/replace or even its regex won't modify the prepends or appends, and I didn't want to save the pages and re-edit them (that could be messy for so many pages). So after banging my brains out on occasion for weeks on end, I finally realized I could prepend %%title%% without using the prepend feature by using regex's ^ beginning-of-string anchor character, and then modify the first regex with a second regex and strip out the unwanted "Outline of"'s. And then I forgot how to do it later, and had to figure it out all over again. What a pain in the ass.
But don't let that deter you, there's plenty of power to work with even while disregarding the fringe cases. Most of it is menu-driven.
Have fun.    The Transhumanist 12:15, 13 February 2018 (UTC)

WikiProject X Newsletter • Issue 11

Latest comment: 6 years ago1 comment1 person in discussion
 
Newsletter • February 2018

Check out this month's issue of the WikiProject X newsletter, with plans to renew work with a followup grant proposal to support finalising the deployment of CollaborationKit!

-— Isarra ༆ 21:26, 14 February 2018 (UTC)

"Short description" mass-edits for disambiguation pages

Latest comment: 6 years ago6 comments3 people in discussion

Hi! Rather than add a new template to all of the bazillion disambig pages in existence, I think it should instead be added to Template:Disambiguation directly, so all articles transparently gain the description. Have you considered that? -- intgr [talk] 14:20, 22 February 2018 (UTC)

Hi intgr, Thanks for the suggestion, but I don't think it would be technically practicable, as it is not only disambiguation pages, but all article pages that will have to get a short description template, and the short description itself depends on the page topic. Dab pages are one of the few types where there are a relatively small number of possible descriptions, most pages will eventually get a unique short description, though as a start some may get one that fits a category, and be refined later. You can see more about this at Wikipedia:Short description and Wikipedia:WikiProject Short descriptions. Cheers, · · · Peter (Southwood) (talk): 14:55, 22 February 2018 (UTC)
  • Just in case this hasn't become clear from the discussion at Wikipedia talk:Disambiguation pages with links#Short descriptions, you should stop adding short descriptions to dab pages for the time being as it floods people's watchlists. You're welcome to recommence if after the discussion the community is happy with it and there is no easier way to achieve the same result. Thanks! – Uanfala (talk) 23:39, 22 February 2018 (UTC)
    • Uanfala, Sorry, it was not clear, as this is a thing that was discussed on a Village Pump RfC, which is about as wide an audience as one can get on Wikipedia, and WMF Reading Team have made it very clear that Wikipedians' happiness with it is not a thing they consider important. (For the record, I have made it as clear as I could to WMF that I am also not happy with the situation, but will do what is reasonably practicable to make the best of the situation as it is.) I have two choices, Go with the local preference to discuss again locally, which is unlikely to have any enforceable result, or go with reality, which is that we currently do not have amother useful option. All that stopping will achieve, is a longer period where WMF Reading Team inflict bad descriptions from Wikidata on our articles. I have now put disambiguations on hold, but as one of the fastest sets of short descriptions to create, this will slow down reaching the 2 million target that Danny Horn at Reading Team has unilaterally imposed on Wikipedia before they will turn off the Wikidata descriptions. I suggest you all complain to User talk:DannyH (WMF) about this if you think it is disruptive. I have already done so without noticeable effect through RfCs and other wide audience discussions. Cheers, · · · Peter (Southwood) (talk): 05:54, 23 February 2018 (UTC)
      • Sorry if it hasn't become clear enough in the discussion, but the objections are not against the use of short descriptions. The objections are against you adding them by hand on every single dab page. You know, there about one or two hundred thousand disambiguation pages, and each one of them has the {{dab}} template. If an identical string needs to be added to all those pages, then this can be achieved via a single edit to this template. If there are technical reasons for this not to work, then the placement of these short descriptions can be done with a bot, minimising disruption to watchlists. – Uanfala (talk) 13:47, 23 February 2018 (UTC)
        • Uanfala, If you can wangle up a better way of doing it I will be delighted. It is not in any way a fun job and I will cheerfully clean up the existing templates once something is working. Feel free to complain to Danny anyway. I don't think he has a clue of how much hassle he is causing us, or maybe just doesn't care... Cheers, · · · Peter (Southwood) (talk): 13:52, 23 February 2018 (UTC)

Outline image support

Latest comment: 6 years ago3 comments2 people in discussion

Here's a new outline I thought you'd be interested in looking at. Notice the use of images to support the list sections.    The Transhumanist 20:02, 24 February 2018 (UTC)

Where, The Transhumanist? Images for Outline of underwater diving is my next step, and I am looking for guidance. There is not much on Wikipedia:Outlines. · · · Peter (Southwood) (talk): 20:17, 24 February 2018 (UTC)
Oops. Here it is: Outline of Dresden. Enjoy.    The Transhumanist 20:30, 24 February 2018 (UTC)
Latest comment: 6 years ago1 comment1 person in discussion

An automated process has detected that when you recently edited Outline of underwater diving, you added a link pointing to the disambiguation page Cylinder valve (check to confirm | fix with Dab solver).

(Opt-out instructions.) --DPL bot (talk) 09:19, 1 March 2018 (UTC)

Administrators' newsletter – March 2018

Latest comment: 6 years ago1 comment1 person in discussion

News and updates for administrators from the past month (February 2018).

 

  Administrator changes

  Lourdes†
  AngelOfSadness • Bhadani • Chris 73 • Coren • Friday • Midom • Mike V
† Lourdes has requested that her admin rights be temporarily removed, pending her return from travel.

  Guideline and policy news

  • The autoconfirmed article creation trial (ACTRIAL) is scheduled to end on 14 March 2018. The results of the research collected can be read on Meta Wiki.
  • Community ban discussions must now stay open for at least 24 hours prior to being closed.
  • A change to the administrator inactivity policy has been proposed. Under the proposal, if an administrator has not used their admin tools for a period of five years and is subsequently desysopped for inactivity, the administrator would have to file a new RfA in order to regain the tools.
  • A change to the banning policy has been proposed which would specify conditions under which a repeat sockmaster may be considered de facto banned, reducing the need to start a community ban discussion for these users.

  Technical news

  • CheckUsers are now able to view private data such as IP addresses from the edit filter log, e.g. when the filter prevents a user from creating an account. Previously, this information was unavailable to CheckUsers because access to it could not be logged.
  • The edit filter has a new feature contains_all that edit filter managers may use to check if one or more strings are all contained in another given string.

  Miscellaneous

  Obituaries

  • Bhadani (Gangadhar Bhadani) passed away on 8 February 2018. Bhadani joined Wikipedia in March 2005 and became an administrator in September 2005. While he was active, Bhadani was regarded as one of the most prolific Wikipedians from India.
Sent by MediaWiki message delivery (talk) 03:00, 2 March 2018 (UTC)

Helium and oxygen

Latest comment: 6 years ago5 comments2 people in discussion

The helium balloon gas is supplied as 80 helium, and up to 15% air. [1] Although this can kill you, the small amount of air will cause the hypercapnic response, which is why Humphry says it won't work. Ratel (talk) 12:39, 2 March 2018 (UTC)

Ratel, 15% air in helium will give 3% oxygen. At normal atmospheric pressure this should not support consciousness long enough to notice hypercapnia. Once unconscious, hypercapnia is a moot point, as it is no longer possible to do anything about it. It may allow spontaneous recovery if the next breath is fresh air, but not if it is a 3% mix, regardless of carbon dioxide levels. This is an issue well known in rebreather diving. Is Humphry's claim supported by any experimental evidence? ·
The Worthington safety data sheet has some dubious information. It does not claim that the air is included for safety reasons, it is quite likely just to make balloon gas cheaper to produce. A 0 to 15% air content cannot be claimed to provide any reliable safety margin, as it may be 0%, and from a health and safety perspective must be considered anoxic. · · Peter (Southwood) (talk): 14:49, 2 March 2018 (UTC)
Moot point, because remember we are only reporting Humphry's opinions in that article, not scientific fact. As a peaceful death advocate, any chance of hypercapnia is anathema to him. Ratel (talk) 22:48, 2 March 2018 (UTC)
Ratel. Yes, but Humphry's opinions are not balanced by any indication that they may not be based on reality. They may be misleading to the reader. When there is reliable evidence available that an opinion reported in Wikipedia may be misleading and counterfactual, we should indicate this to be the case. The article about Humphry is an appropriate place to report his unsupported opinions, the article on suicide bags should be based on facts about suicide bags, not opinions of advocates and opponents, which should take a secondary place. · · · Peter (Southwood) (talk): 06:22, 3 March 2018 (UTC)
I have copied this discussion to Talk:Suicide bag, in case other editors wish to express an opinion.· · · Peter (Southwood) (talk): 06:28, 3 March 2018 (UTC)

Script to provide annotation viewing enhancement

Latest comment: 6 years ago9 comments2 people in discussion

That's an impressive job you've done on building and annotating the Outline of underwater diving.

Kudos.

It's now one of the best outlines.

As you know, lists (including outlines) have two main formats: bare and annotated.

Each has its advantages.

Bare is easier on the eyes, and is arguably more convenient to browse when you already know what the terms mean.

Annotated is highly useful when you need descriptions to help you decide what to click on, or to read the outline as a general overview to increase familiarity with the subject.

So, I've been developing a script to deliver the advantages of both, enabling the user to switch between the two modes at will.

That way, a user can browse the bare list, and when they come to a term they are unfamiliar with, press ⇧ Shift+Alt+a to see the descriptions. Then, after reading the desired description, press it again to continue browsing the bare list.

It's still under development, and doesn't reposition exactly the way it should, but it is developed enough to be useful, and as proof of concept.

I thought you might want to give it a test drive.

It is ViewAnnotationToggler.

The menu item is in the same tab menu as Move, in case you need to use that instead of the hot key.

If you give it a run, please let me know what you think of it, and point out any problems it may cause, ways to improve it, etc.

Sincerely,    The Transhumanist 17:11, 23 February 2018 (UTC)

P.S.: It works across all pages, and it is easy to forget when annotations are hidden, which may surprise you when you go into edit mode to add annotations, and discover that there are some already there. If you start to wonder why there are so few annotations around Wikipedia, it might be that you have them turned off.

Thanks The Transhumanist, I will take a look, hopefully soonish. Thanks for the reshuffle, It has let me look at the structure with a refreshed eye. I find it mildly ironic that the first thing you did was remove the under construction banner, just before a radical restructuring. Also slightly amusing is that I was still busy, and pasted in a new section without noticing your changes. So it goes.
I was lucky in that both my current projects overlap, so I got a whole bunch of short descriptions done which are now also annotations. When the short description template is in proper working order I want to see if it is possible to extract the short description and plug it into another article as an automatic annotation that is updated at the home article. I bit like transclusion, but possibly a bit more not like transclusion. The new magic word my help or not. · · · Peter (Southwood) (talk): 17:55, 23 February 2018 (UTC)
I'm glad you like the tree adjustments. Keep adjusting as you see fit. You know the subject way better than I do. I know it only as much as I have imagined: I dive in, look around, and scream when I see a shark swimming toward me. Thanks to Stephen Spielberg, I will probably never go swimming in saltwater again, let alone with diving gear.    The Transhumanist 18:18, 23 February 2018 (UTC)
Please explain to me how the magic word is supposed to work.    The Transhumanist 18:18, 23 February 2018 (UTC)
I wish I knew. All I can be reasonably sure of is that it is supposed to allow apps to access the short descriptions and display them along with the article titles. So it seems reasonable to assume that if they can be accessed from outside the server, it should be possible to do it from inside a mediawiki installation too. See Meta:Magic words, Meta:Variables and Meta:Parser functions, and the Phabricator ticket phab:T184000 which gives the syntax. · · · Peter (Southwood) (talk): 19:59, 23 February 2018 (UTC)
I added the install script for annotation toggler after reading some of the explanations on how it works, which were surprisingly understandable most places that I read, and it appears to work as described - so far, so good.
Have you had any luck working out how the magic word is supposed to work, if we will be able to call it, and how? It is tricky particularly since it does not exist yet, so experiment is not an option. · · · Peter (Southwood) (talk): 07:53, 26 February 2018 (UTC)

The Transhumanist, I had a problem that I tracked down to the annotation toggler script. It appears to disable the move page function, in that the page opens normally from the tab, but the dropdown menu for namespaces greys out and does not activate the menu when clicked. Removing the script and purging gets move page working again. I tried this twice and effects seem repeatable. I have had a few other anomalies on both my browsers on both my machines, so going to check if this might not have some other effects by leaving it off for a while. Cheers, · · · Peter (Southwood) (talk): 11:26, 17 March 2018 (UTC)

I recently discovered that it disables the script editor. The script editor shows, but then you can't get the cursor to show up in the window in order to edit. I fixed that by making the script ignore editing pages. I think I will have to do something similar to fix these other problems. Thank you for the heads up. I will let you know if I can replicate and fix these problems.     — The Transhumanist    00:43, 18 March 2018 (UTC)

Wikidata: I just don't get it

Latest comment: 6 years ago2 comments2 people in discussion

I've tried to understand Wikidata several times, but I still don't get it. All the help pages I've read so far seem to describe it in roundabout terms. Wikidata:Introduction doesn't give me any idea what it is for.

What, in essence, is it?

A database? Of what? That can do what?

What can it be used for now?

What will it be able to be used for in the future?

Frustratedly,     — The Transhumanist    10:06, 30 March 2018 (UTC)

You are asking me? I classify it among the things that are probably more trouble to understand than they are worth.
It is claimed to be a database that is freely available to anyone who wants to use the data in it. Now my knowledge of databases is superficial and sketchy on a good day. As I understand it, it is a system for storing structured data (mostly independent of language, or at least defined in all languages as the same thing, if that is possible) in a way that can be easily retrieved. I know enough of the user interface to be able to edit in Wikidata, and have added, deleted and modified data associated with some of the Wikipedia and Wikivoyage articles I work on. I understand that Wikidata is used to link articles on the same topic from Wikipedias in different languages, but how it is done is beyond me. There are people who claim that eventually it will be possible to automatically generate articles from data on Wikidata , but not on English Wikipedia in the foreseeable future. It is fairly easy to vandalise and there are not nearly enough editors or bots to keep the vandalism under control equivalent to Wikipedia. The current consensus on Wikipedia for how Wikidata may be used on Wikipedia is undecided, I think the RfC is still running. There are Wikipedians who think it is the best thing since sliced bread, and others who consider it the devil's work. I like to think I am more neutral, and think it may have potential, but is not very useful at present, because the data is insufficiently reliable, and vandalism is too easy and goes undetected for too long. I oppose using data from Wikidata on Wikipedia when it is unverifiable - ie. may be complete rubbish, perfectly accurate, or anything in between, and I can't easily check it. If short descriptions from Wikipedia were to be stored there as read only items, and remain editable on Wikpedia, it would make them relatively easily accessible to anyone, as RexxS suggests. This would have to be negotiated with the Wikdata people, as they are a separate project. I hope this helps, but suspect it may not. Cheers, · · · Peter (Southwood) (talk): 10:37, 30 March 2018 (UTC)

Administrators' newsletter – April 2018

Latest comment: 6 years ago1 comment1 person in discussion

News and updates for administrators from the past month (March 2018).

 

  Administrator changes

  331dot • Cordless Larry • ClueBot NG
  Gogo Dodo • Pb30 • Sebastiankessel • Seicer • SoLando

  Guideline and policy news

  • Administrators who have been desysopped due to inactivity are now required to have performed at least one (logged) administrative action in the past 5 years in order to qualify for a resysop without going through a new RfA.
  • Editors who have been found to have engaged in sockpuppetry on at least two occasions after an initial indefinite block, for whatever reason, are now automatically considered banned by the community without the need to start a ban discussion.
  • The notability guideline for organizations and companies has been substantially rewritten following the closure of this request for comment. Among the changes, the guideline more clearly defines the sourcing requirements needed for organizations and companies to be considered notable.
  • The six-month autoconfirmed article creation trial (ACTRIAL) ended on 14 March 2018. The post-trial research report has been published. A request for comment is now underway to determine whether the restrictions from ACTRIAL should be implemented permanently.

  Technical news

  Arbitration

  • The Arbitration Committee is considering a change to the discretionary sanctions procedures which would require an editor to appeal a sanction to the community at WP:AE or WP:AN prior to appealing directly to the Arbitration Committee at WP:ARCA.

  Miscellaneous

  • A discussion has closed which concluded that administrators are not required to enable email, though many editors suggested doing so as a matter of best practice.
  • The Foundations' Anti-Harassment Tools team has released the Interaction Timeline. This shows a chronologic history for two users on pages where they have both made edits, which may be helpful in identifying sockpuppetry and investigating editing disputes.
Sent by MediaWiki message delivery (talk) 01:23, 2 April 2018 (UTC)

Sister cities of Birmingham

Latest comment: 6 years ago2 comments2 people in discussion

I've now found the following:

Those are better, I think. --RexxS (talk) 19:39, 8 April 2018 (UTC)

Bare urls are linkrot waiting to happen, yet some people seem to think they constitute a proper reference. Better than nothing, but some of our long time editors, even admins, appear to consider that a bare url is all they need to provide. I must take you up on your offer to show me how to find archived sites. Cheers, · · · Peter (Southwood) (talk): 05:04, 9 April 2018 (UTC)

Interested in becoming a new page patroller?

Latest comment: 6 years ago5 comments3 people in discussion

User:Amorymeltzer/sandbox/npp/note ~ Amory (u • t • c) 19:20, 10 April 2018 (UTC)

Note that there is no obligation to actually review new articles. — Insertcleverphrasehere (or here) 06:38, 11 April 2018 (UTC)
Insertcleverphrasehere, is there any useful purpose to be served by becoming a new page patroller if one des not review new articles? It is something that I might occasionally do, but I am usually too busy with other things. Cheers, · · · Peter (Southwood) (talk): 06:47, 11 April 2018 (UTC)
Well if it is something you might occasionally do, then definitely apply, but even if not, we are trying to get as many AfC reviewers under the NPR user right as possible, to facilitate better coordination between the two systems, potentially bringing AfC under the NPR user-right umbrella in the future (there actually are not that many non-NPR AfC reviewers anyway). — Insertcleverphrasehere (or here) 06:51, 11 April 2018 (UTC)
Good enough reason for me. I will apply. New page review is a serious matter and an important part of quality assurance. Cheers, · · · Peter (Southwood) (talk): 07:19, 11 April 2018 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=User_talk:Pbsouthwood/Archive_7&oldid=864135592"