36C3 Wikipaka WG: Software tools for wikis beyond MediaWiki and its extensions

36c3 preroll music

Herald: So now we present our next
speaker, who is Andre, and he will talk

about some cool tools related to the
Wikimedia stuff that you, as, maybe,

Wikimedia users could use and do cool
stuff with it. So let's have some applause

for Andre, please.

Andre: Batterien? Oh, it works, right?
Perfect. Sorry for that. Thanks for

coming. I'm Andre. I work for Wikimedia. I
was even wondering whether I should put

the logo here, because this actually has
nothing to do with the Wikimedia

Foundation itself. So this is all about
volunteer work, volunteer software,

because it's always a mix between several
entities like Wikimedia Deutschland,

Wikimedia Foundation, Sweden, also lots of
other companies, for example.

And I decided to give this talk because there is
if you are on a Wikimedia website, for

example, Wikipedia, there is some obvious
software there. Of course, the wiki

software itself, which allows you to view
and edit pages. There are a lot of

extensions, about 130 deployed on
Wikimedia servers, but there's also lots

of software around, which pretty often
isn't very visible.

Wikimedia Tech world is pretty complex, it's all free and
open source software, and some areas are

actually fully covered by volunteers, and
especially I'm going to be in the bots,

gadgets, user scripts, tools, and, a bit,
mobile apps area today. We have many

communities, many languages, for example,
there's already more than 300 Wikipedias

when it comes to different languages. So
there's a lot of diverse interests, use

cases, technical needs. You can probably
imagine a few things technically already

when it comes to different scripts used in
different cultures or right-to-left, left-

to-right. Many other examples like this.
And as everything is Free and Open Source,

a lot of volunteers experiment with new
ideas they have and also bridge some

workflow gaps that might exist for these
communities and maybe other communities

are not even aware of that. So pretty
often it happens, that some

community members come up with some ideas
and over time they evolve. Sometimes they

even become a code repository or a gadget
that also gets copied to another Wikimedia

site like a different language Wikipedia.
These kinds of things. And earlier this

year, some people decided that it would be
beautiful to appreciate the work and

create a showcase of the most impressive
software solutions which were implemented

outside of the Wikimedia Core Code
repository and the extensions.

Both to celebrate the software solutions and also
the people behind the tools. Because this

is a lot about ideas, about passion, about
skills. Pretty often also finding maybe

somebody who has more experience or
knowledge in a certain culture. If you

cannot create something yourself, and
teaming up. So, early this year, the idea

came up to celebrate such great pieces of
software by creating an award. And there

was a Wikimedia conference, I think, in
August in Stockholm, it's called

Wikimania, which is not only about
technical aspects, it's really about

everything related to the Wikimedia
communities. And we, beforehand, there was

there were a few people who came together
and tried to find categories and, for

tools to give an award to in these
categories. So I'm basically remixing this

award session from earlier this year here
without giving out awards. And my hope is

that you might see some great stuff,
might find some stuff interesting.

It's not necessarily if you run your own
Mediawiki installation on your server,

some stuff might be too Wikimedia-specific
use cases, but maybe you might get some

ideas or also stuff you weren't aware of
and might want to use. Because there are a

lot of tools out there, as I said, and
sometimes it's really hard to discover

them because they're. soundproblems Oh, thank you,
because they might be on on separate

wikis. All right, so the first category
was or is experience, and it was won by

the locator tool by Simon04. It's a tool
that helps you adding the geocode, the

exact position to existing images,
especially on Wikimedia Commons, which is

the place to share free media images,
videos, things like this. And why it

received this prize is because it's really
intuitive and easy to use. You can add

coordinates to one or more files. You can
find it in user preferences. So it's a bit

easier to discover. It's available in many
languages, it had great tutorials.

Actively maintained and it's been used
already a lot. So this is the tool.

I wonder if I should zoom in a bit. It's
called locator tool and you can enter a

category name here. For example, I have
one hand less than usual I realize.

"Coolest tool award"… In theory it should
also autocomplete. Let's try. Showcases.

Maybe if I… let's try something
else, then, I mean, that's what

autocomplete is for. And let's load. So in
theory, you get the map here with a

pointer on it. Or pointers of the files in
this category. In practice, I probably

chose a bad example, and the Wi-Fi isn't
that fast. Or maybe none of the images in

that category already has a location. That
might also be the case here. It's not the

category I tried beforehand when I tried
to prepare this session. Sorry for that.

I'll go back to the screenshot, where
you can hopefully imagine how things

should look like. The next one would
be HotCat, which is a pretty tiny

codebase, actually, but used a lot.
And "cat" in this case stands for

Categories, because that is one way to
organize, for example, files on Wikimedia

Commons, but also articles on
Wikipedias. So this is a screenshot

from a file on Wikimedia Commons. And at
the bottom you can see the categories,

and you can easily add categories via this
tool and also remove, change, add

categories. And it's also pretty
discoverable via the user preferences. So

to compare this, how much should I zoom
in? This here, down here you can see the

categories, how it usually looks.
Basically just the names, and you can

click the categories to get to the
overview page. If you've enabled the

gadgets, you see a few more buttons here,
which are added by JavaScript so you can

easily remove a category or add a category
by clicking the plus at the very end.

And then you could also type-ahead and add a
new category. It works on almost all

wikis, it actually has the highest number
when it comes to users. And yeah, as

usual, code is public. Several people
contributed.

"Impact". There is Internet Archive Bot by cyberpower, you
probably can guess a little bit from the

name what it's supposed to do. We are not
running an archive service. We're not

archive.org. But pretty often,
Internet websites or pages go down or get

removed or get moved. And as especially
Wikipedia articles have a lot of

references, then suddenly you cannot check
for references anymore. Or if that

statement is actually true because that
website got down. But there is the

Internet Archive and they archive
regularly websites and Web pages by

crawling the Internet. And then this
little Bot replaces those links and, or

references in, for example, Wikipedia
articles at the bottom by the link to

Internet Archive. So you can still
actually reach the Page, that was

referenced a while ago when that page
still existed. And the great thing about

this is that it automates work that
usually would be very cumbersome and very

tiresome to do, and the configuration also
depends on local wiki needs. As an example

I won't show you now running Internet Archive Bot
on some page,

but you can see here, I basically took the last
edits, a totally random one on English

Wikipedia. And you can see here that this
is a history of that article called

Gilberto Hernandez Ortega. And this is the
last edit that Internet Archive Bot made

on English Wikipedia by replacing this
obviously dead link here that you can see

on the left by a link to web.archive.org.
So if you go to that article on English

Wikipedia and you want to go to that
reference, you actually see the reference

and not a dead link, that's what it does.
Then we had a "reusable" category.

If you wonder where we are, if you get tired,
this is the fourth out of 10. That's page

views by MusicAnimal, Kaldari, Marcel Ruiz
Forns. It does what it says.

It's basically getting an idea how often does a
certain page on one of the Wikimedia sites

get accessed. So it's a pretty simple
graph, but that can be pretty useful when

when you want to have statistics, maybe
not necessarily about… well, also, some

people want to find out if… which articles
are the most popular ones on some

Wikipedias. Some people want to find that
out. But for me, for example, it's pretty

useful when when there are, when it comes
to technical documentation on

mediawiki.org and I wonder, OK, these two
pages kind of overlap when it comes to

their content and I would like to merge
them. But which one is more popular and

which way should I merge it. So these
things can be pretty useful.

You can include all wikis. You can also change
the time frame. You can get statistics

over a year now, that was recently
implemented. Before, it was per month.

And in life this, these are two pages
I'm comparing on meta.wikimedia.org.

You can see that I'm looking at the daily
statistics and in a certain time frame,

which you can change here, and I'm
comparing these two pages called the

Coolest Tool Award page and a page called
Requests for New Languages. And so here

you can see like on which day, how many
times those two pages were accessed.

Then there is quick statements by Magnus.
That's true, that's true. I tried to

access that earlier, and it somehow didn't
work for me when preparing this. So in

theory, it's a powerful editor for
Wikidata. You can use statements,

labels, descriptions and aliases to add
and remove them, via rather simple

text commands and you can see simply by
the numbers on Wikidata that it's

pretty popular. As I said, I wasn't able
to play with that yet myself. So I can

only read this text for you right now. So,
you can prepare things already in a

spreadsheet or a text editor to to run
several commands in a row.

Batch edits, basically, semi-automatically. And there's
also other tools like OpenRefine, the

Disambiguater, which also use this tool.
So as it was down, I could only go to its

help page and looked a little bit at the
statements down here. I hope that one

day I'm going to find time to try this
myself. Let's see.

Then, for developers, an award was given to MediaWiki Code
Search by Legoktm. Because, once upon a

time, there were, there was, for example,
a service by Google to do search

explicitly, like, public code, source code
repositories, and we wanted to have that,

especially for Wikimedia code. So
everything that's in WikiMedia git,

Gerrit, I don't think it supports stuff
that… all our Wikimedia code repositories

that are in GitHub or somewhere else. It's
a pretty simple interface. You can see on

the top you can filter by categories, in
which code bases you're looking for a

certain expression. Gerrit, it says
here at the bottom. And this is super easy

to use. Well, at least, if you know a
little bit of regular expressions or if

you just want to enter the name of a
function, for example, because one very

or, a great use case we actually have is,
when some function gets deprecated or even

later on even removed in the MediaWiki
core code base. Of course, somebody needs

to find and update all the extensions out
there, which might rely on that very

function in the MediaWiki core code base.
And this makes it way easier.

Of course you could also locally check out all the
extension repositories and then grep and

try to find that. But this makes it
especially, for those, or most people,

I guess, who don't have a complete check out
of all extensions and code repositories on

their own computer, to quickly use it on
the Internet online. I guess I don't need

to show you how to enter a search string
here. Still, if I, for example, enter,

getText, which would be a function name,
you'd then get the results listed by

repository. And then you could filter on
the top. If the server or the Internet is

fast enough. I might get back to you
later.

Seventh one out of ten we awarded is the "Mobile" area, there is a
Commons mobile app which is also entirely

run, managed, worked on by volunteers like
Josephine, Yuvi, Neshlihan, Vivek.

It allows you to upload photos to Wikimedia
Commons directly from your mobile phone or

from your smartphone, and you can
also, of course, add categories, or view

nearby missing images. So if you use your
GPS, if you know your location, that can

be helpful to find out which articles, for
example, on Wikipedia, still lack

images and view your contributions to
Commons in its own gallery. Those numbers

are probably now outdated. But what is
impressive to me is simply the large

number of different people who have
already contributed to the code base.

Still no results. OK. I don't think
I'm going to play that YouTube video for

for you now. Plus, I haven't sorted out
the sound beforehand, I realize. But you

can… oh, this just shows some of the
images uploaded via it, but it's a pretty

intuitive user interface. It's also
interesting to see that of course, it

also makes uploading content a bit easier
that might not be suitable for Wikimedia

Commons, like, for example, your selfies
of you and your friends.

But I think that's also being worked on and better
filter nowadays, for example, by

categorizing if this is a completely new
user and these kinds of things on Commons.

Then the category "Newcomer" is called NOA
Upload Tool by HappeJ, Sohmen. So this

takes scientific Open Access articles out
there, and fetches the images included in

them, and then anybody can help deciding
if this is suitable when it comes to the

content of, I mean, the license is already
pretty clear. But the content, if this

could be helpful on Wikimedia Commons. So
you go to the website, basically you get a

random image and you can help. Could or
should this be uploaded to Wikimedia

Commons to make it broader available to
make it more discoverable? It's beautiful

because it also gets a bit more into Open
Science. It's probably the most simple

user interface in this collection here,
and it does the attribution correctly.

Randomized. That's probably also
something. So I went to that tool.

And as you can see, you get a random image.
So the caption would be taken over and you

can click "Mark for upload" or "next
image". To actually upload it, you will…

It's probably a bit small… you would have
to log in, of course, first and authenticate.

Then when it comes to
outreach, more to social activities, there

is a programs and events dashboard by Sage
and others. So this is a bit more when it

comes to the teaching part or running
workshops for, for example, editors,

writers, uploading media. Or, for example,
Wikiversity, which is another Wikimedia

site which has courses. And this helps you
to get an idea how much outcome, how much

effect your program has. And it's pretty
useful and actually used by quite some

event organizers out there, for example,
by the Wikimedia chapters in quite some

countries. You can create and manage
education programs, you, as I said, track

some metrics and it's been used for more
than 100, or, with more than 100 000

students and editors, over a million
articles since 2010.

So this is the default view you would end up with on the
programs and event dashboard.

And you can see here the campaigns on the left, for
example, this was the Art and Feminism

2018 campaign. Related programs that were
run. And, for example, the number of

articles created, edited, and the number
of editors here, in the very end.

So you get an idea how much, actually, outcome
you have.

And, last but not least, probably Eggbeater doesn't tell you
anything. That was the logo we chose for

the award, and it's basically the
special or lifetime award or something

like that. That's probably how you would
translate it. Twinkle. It's also a

JavaScript gadget by AzaToth, Ioeth,
Amathea, atlight, MusikAnimal, AmoryMeltzer.

And this is when, when
you're a bit more of an experienced,

for example, Wikipedia user, it helps you
a lot with maintenance tasks like dealing

reverting vandalism, unscontract–
unconstructive edits. Which makes

administrative tasks way easier. It's
been around for 15 years, it has pretty

good help. And to give you a simple idea,
this would be the normal or nearly normal

view, I think I also have some some custom
gadgets enabled, on on a Wikipedia page

when you're logged in and you can see up
here, read, edit, view history.

The watch list, star button. When I enable
Twinkle, you see there's another drop

down, which leads to, for example, the
first link request, speedy deletion.

According to CSD, I should probably
know what that means. Speedy deletion?

And a few other options like "show most recent
diff", "unlink backlinks", a lot of

functionality that is way easier to access
and more common, especially if you try to

revert vandalism and watch pages.

This is a photo of, at the end of that actual award ceremony, of all the
people being around. There were also definitely a few

maintainers, developers, stewards of these
code bases around. So not everybody was

present at that conference, but we could
actually hand out some awards which were eggbeaters.

And what was this was
basically about was, as I said earlier,

appreciating, all the code in between,
that might not be obvious to you.

Sometimes, it might not even be obvious to
you that this is custom code or a gadget,

not in the core functionality, added by a
volunteer, because it is enabled by

default. And you just expect like, OK,
this is probably part of the core

software, but it's actually not. And, of
course, also, thanks a lot to all the

users of these tools. So if anything
was interesting here or if you have more

curiosity, you will find links on
meta.wikimedia.org, on the page

Coolest Tool Award. If you are generally
interested in the technical parts of it,

not necessarily only as a user using these
tools, I would recommend how to contribute

on mediawiki.org, which both covers how to
get technically involved, but also other areas.

For example, of course,
editing, but also, design, local user

groups, outreach or other things. So these
are probably only the credits.

So I'm done. Thank you.

applause

Herald: Hello, hello? Yeah. Thank you,
Andre! So do we have any questions in the

audience? If you do raise your hands and I
will hand you my microphone?

Andre: Basically, I would even say feel
free to ask anything. I mean, I might not

know the answer. It's not that I'm
actively working in all of these tools or

anything, but I can try to find out.
Q: I thank you for your speech.

Do you have a favorite tool for locations of
articles, how to add them, or edit them?

Andre: A favorite tool for locations of
articles, how to edit them?

Q: You can have locations for the images,
but also add locations to articles on the

other side? And I find it unhandy to
always copy paste a code with geolocation

and all that stuff. And I would also
prefer there to have a tool where I could

click on a map and say, OK, it's there.
Maybe it's existing.

Andre: That that is a good question. So,
yeah, so you basically go to an article

and you hope for some button, which
probably opens a map and then you say it's

here on this map. And then you edit with
one click the coordinates to the article,

I guess. Right? I'm not sure myself. I
would have to try to find out. Let me come

back to you later, please.
Herald: Any other questions? I don't think

I see any, but so then again, thank you,
have some applause for Andre please!

Subtitles created by c3subtitles.de
in the year 2021. Join, and help us!