This past Wednesday marked a milestone in the evolution of Wikimedia’s Database infrastructure: the completion of the migration of the English and German Wikipedias, as well as Wikidata, to MariaDB 5.5.

For the last several years, we’ve been operating the Facebook fork of MySQL 5.1 with most of our production environment running a build of r3753. We’ve been pleased with its performance; Facebook’s MySQL team contains some of the finest database engineers in the industry and they’ve done much to advance the open source MySQL ecosystem.

That said, MariaDB’s optimizer enhancements, the feature set of Percona’s XtraDB (many overlap with the Facebook patch, but I particularly like add-ons such as the ability to save the buffer pool LRU list, avoiding costly warmups on new servers), and of Oracle’s MySQL 5.5 provide compelling reasons to consider upgrading. Equally important, as supporters of the free culture movement, the Wikimedia Foundation strongly prefers free software projects; that includes a preference for projects without bifurcated code bases between differently licensed free and enterprise editions. We welcome and support the MariaDB Foundation as a not-for-profit steward of the free and open MySQL related database community.

Preparing For Change

Major version upgrades of a production database are not to be made lightly. In fact, as late as 2011, some Wikipedia languages were still running a heavily patched version of MySQL 4.0 — the migration to 5.1 required both schema changes, and direct modifications of data dumps to alter the padding of binary-typed columns. MySQL 5.5 contains a variety of incompatibilities with prior versions, thanks in part to better compliance with SQL standards. Changes to the query optimizer between versions may also change the execution plan for common queries, sometimes for the better but historically, sometimes not. SQL behavior changes may result in replication breakage or data consistency issues, while performance regressions, whether from query plan or other changes, can cause site outages. This calls for a lot of testing.

Compatibility testing was accomplished by running MariaDB replicas outside of production, watching for replication errors, replaying production read queries and validating results. After identifying and fixing a couple of MediaWiki issues that surfaced as replication errors (along the lines of trying to set unsigned integer types to negative values which previously caused a wrap-around instead of an error) we replayed production read queries using pt-upgrade from Percona Toolkit. Pt-upgrade replays a query log against two servers, and compares the responses for variances or errors. Scripts originally developed for our recent datacenter migration to simultaneously warmup many standby databases from current production read traffic helped with rough load testing and benchmarking. Along the way, a pair of bugs in MariaDB 5.5.28 and 5.5.29 were identified, one of which was a rare but potentially severe performance regression related to a new query optimizer feature. The MariaDB team was very responsive and quick to offer solutions, complete with test cases.

Performance Testing In Production

As a read-heavy site, Wikipedia aggressively uses edge caching. Approximately 90% of pageviews are served entirely from the edge while at the application layer, we utilize both memcached and redis in addition to MySQL. Despite that, the MySQL databases serving English Wikipedia alone reach a daily peak of ~50k queries/second. Most are read queries served by load-balanced slaves, depending on consistency requirements. 80% of the English Wikipedia query load (up to 40k qps) are typically handled by just two database servers at any given time. Our most common query type (40% of all) has a median execution time of ~0.2ms and a 95th percentile time of ~50ms. To successfully use MariaDB in production, we need it to keep up with the level of performance obtained from Facebook’s MySQL fork, and to behave consistently as traffic patterns change.

Ishmael views of pt-query-digest data collected via tcpdump for the most common Wikipedia read queries (pdf). The first page of a query shows data from db1042, running 5.1fb-r3753, the second from db1043 over the same time period, running MariaDB 5.5.30.

Once confident that application compatibility issues were solved and comfortable with performance obtained under benchmark conditions, it was time to test in production. One of the production read slaves from the English Wikipedia shard was taken out of rotation, upgraded to MariaDB 5.5.30, and then returned for warmup. The load balancer weight was then gradually increased until it and a server still running MySQL 5.1-facebook-r3753 were equally weighted and receiving most of the query load.

Also from the Percona Toolkit, we use pt-query-digest across all database servers to collect query performance data which is then stored in a centralized database. Query data is collected from two sources per server and stored in separate buckets — from the slow query which only captures queries exceeding 450ms, and from periodic brief sampling of all queries obtained by tcpdump. Ishmael provides a convenient way to visualize and inspect query digest data over time. Using it, along with direct analysis of the raw data, allowed us to validate that every query continued to perform within acceptable bounds.

For our most common query type, 95th percentile times over an 8-hour period dropped from 56ms to 43ms and the average from 15.4ms to 12.7ms. 50th percentile times remained a bit better with the 5.1-facebook build over the sample period, 0.185ms vs. 0.194ms. Many query types were 4-15% faster with MariaDB 5.5.30 under production load, a few were 5% slower, and nothing appeared aberrant beyond those bounds.

From there, we upgraded the remaining slaves one by one, before finally rotating in a newer upgraded class of servers to act as masters. The switch was seamless and performance continues to look good. We’ll be completing the migration of shards covering the rest of our projects over the next month. Beyond that, we’re looking forward to the future release of MariaDB 10 (global transaction IDs!), and are continually assessing ways to improve our data storage infrastructure. If you’re interested in helping, the Wikimedia Foundation is hiring!

Asher Feldman, Site Architect

The Wikimedia Language Engineering team continued to develop additional features for Translate UX (TUX) and to fix bugs, including a number of critical bugs that affected Microsoft Internet Explorer. TUX has now been deployed on Wikimedia sites and on translatewiki.net. The team also launched the new outreach program, called Language Mavens, and conducted an office hour.

The view after the user logs in to the new home page of translatewiki.net. This is currently in development.

Redesigning the main page of translatewiki.net

As part of the original specifications for TUX, the translatewiki.net main page is being redesigned. The layout will now include a set of tiles for the project groups that will contain the projects’ logo and statistics. The links conveniently placed on the tiles can be used to access the project messages for translation and proofreading. Once logged in, users will be presented with summarized statistics of their recent activities.

Significant changes are also planned for the account sign-up form. Along with features that will allow newly signed-up users to familiarize themselves with the translation workflow, it’ll make the process of getting the translation permissions smoother. The page will also include artwork to depict cultural diversity from around the world. The main page is currently under preparation, but the special page can already be accessed on translatewiki.net.

Language Mavens: The new outreach program

Language Mavens, the new outreach program for the Language Engineering team, was launched to provide better support to the various language communities within Wikimedia projects. The program was announced during the monthly office-hour (read the logs). After a meeting with the initial group of participants, the pilot phase is now underway. The program aims to regularly connect with the language communities across the Wikimedia projects to get feedback about the user experience and any improvements needed in the language support features. To optimize the internationalization tools and workflows, the program will also facilitate participation in activities such as bug triages and testing days. The team will hold regular meetings with the Maven team members. The program is open for participation, and new members can sign up online. To get more details about the program, you can read the program’s description and write to runa at wikimedia dot org.

Up next

For the next development sprints, the team will continue to work on the projects mentioned above, and also devote significant efforts on redesigning the Universal Language Selector (ULS) interface based on the design review document. The Language Engineering team will host a community bug triage session on April 24th, 2013 at 17:00 UTC. The team also invites students participating in Google’s Summer of Code 2013, to work with them on several internationalization projects.

To know more about our projects and ways to participate, please find us on the IRC channel #mediawiki-i18n (Freenode) or write to runa at wikimedia dot org.

Runa Bhattacharjee, Outreach and QA coordinator, Language Engineering

In the Wikimedia movement, we have a vision statement that inspires many contributions to our endeavor: “Imagine a world in which every single human being can freely share in the sum of all knowledge. That’s our commitment.”

comScore traffic data to Wikimedia sites.

We’re still a long way from realizing that vision, but we’ve recently surpassed an important milestone: as of March 2013, the combined sites hosted by the Wikimedia Foundation reached more than 500 million monthly unique visitors, according to the latest comScore Media Metrix data. Our traffic increased to 517 million in March, five percent higher than our previous record: 492 million in May 2012.

While more people are coming to Wikipedia and the other Wikimedia sites, they are also staying longer and reading more. Over the past 12 months, Wikipedia monthly page requests increased from 17.1 billion to 21.3 billion, with the mobile share increasing to roughly 15 percent of the total, or more than 3 billion monthly views (data). We’re also gratified to see growth in significant target areas: in India, traffic as a percentage of our worldwide total increased from 4.0 percent to 4.8 percent; in Brazil it increased from 3.6 percent to 5.9 percent.

To reach the entire planet, we will need to not only continue to expand our mobile offerings, but also eliminate barriers to access. With Wikipedia Zero, we’re partnering with mobile providers in the developing world to reduce or eliminate data fees for accessing Wikipedia on a mobile phone. In March, we announced the fifth major Wikipedia Zero partnership, which means that the program will be available to 410 million mobile users around the world.

For those who don’t have an Internet connection at all, Wikimedia movement contributors are enabling offline access to Wikipedia, such as the work by Kenyan volunteers who travel to rural schools and install copies of the encyclopedia on computers there. And now, there’s also an open source application for Android phones and tablets that makes it easy to download and read offline copies of Wikimedia content.

The idea of enabling every single human being to freely share in the sum of all knowledge is still as audacious as ever — but it’s also starting to look like an achievable goal, if we come together to make it happen.

Sue Gardner, Executive Director, Wikimedia Foundation

The Kiwix app empowers Android users to download and view large sets of Wikipedia content on their devices while offline.

Kiwix isn’t just Yet Another Wikipedia app. Once you install it on your Android device, you can actually use it to download the whole website, and access its millions of articles even when you’re offline (or when you don’t want to use your data plan). You can also download smaller (and custom) selections of pages, if you’re short on storage space.

Providing access to Wikipedia and other Wikimedia sites to as many people as possible is one of the Wikimedia movement’s core goals. While their regular desktop, online version is enough for most users, it is inadequate for many others.

This is the reason why, throughout the movement, Wikimedians are working to diversify and facilitate access to Wikipedia and its sister sites:

• Kiwix is a software that allows users to browse full snapshots of Wikipedia (and numerous other resources) from a personal computer that isn’t connected to the internet.
• The Wikipedia Mobile App allows smartphone users to browse Wikipedia on mobile devices, save articles and much more.
• The Wikipedia Zero initiative allows cellphone users in Africa and Asia to access parts of Wikipedia without incurring data charges.
• The Kiwix-plug empowers African students to access full snapshots of Wikipedia and Wikisource right from their campus, on their own laptop or phone.

Kiwix for Android is the latest innovation in this series: This app, available from the Google Play Store or from the Kiwix website, allows users of Android-powered devices to browse offline content from Wikipedia and its sister sites.

The market of Android-powered devices is exploding. It is not limited to mobile phones and tablets, but it also powers TV, appliances, “USB computers” and an increasing range of cheap computers.

In places where connectivity is a difficult (at least 30 countries on the sole African continent), the only way to access Wikipedia content is Kiwix Desktop, but it still requires a computer and electricity. Cheap Android devices might spread more quickly, and we really want to facilitate access to free content everywhere.

That said, Kiwix for Android can also be just as useful to Westerners who want to walk around with the entirety of the world’s largest encyclopedia in their pocket (if they have a big enough SD card), accessible at no cost or data fees.

Where do I get it?

It’s very easy to use:

1. Download the app from the Google Play Store, and launch it;
2. Click on the “Open” button, select a ZIM file from the list (from your device or SD card);
3. That’s it! You’re already browsing offline content.

If you don’t already have a ZIM file, Kiwix leads you to its online repository so you can download one. You can also use any of the Books created on Wikipedia.

Being a mobile version of a feature-rich desktop software, the Android version is lightweight. It provides access to the most essential features: opening and reading a ZIM file, search with auto-completion on article titles, in-page search, random navigation, zoom in & out, and access to the mobile-friendly ZIM file catalog.

How does it work?

Kiwix for Android uses a native Android user interface in Java (which offers a nice look & feel) and is built on the Kiwix and libzim C++ code. We combine these tools using the Java Native Interface.

It’s very easy to improve the app. Just download the code from our code repository and read the COMPILE file. You don’t even need to compile the C++ code: it’s available as a pre-compiled file. You can directly start hacking the UI and easily rebuild the APK package file.

What’s next?

This is the beginning of Kiwix for mobile. This Android version was developed during a one-week hackathon sponsored by Wikimedia CH (Switzerland) and our focus was on providing a stable and comfortable experience.

Our strategy for mobile development is not set in stone. We have some ideas but are also expecting feedback from users. Give it a try, and tell us what you want: new features, new paradigm, new platform target, etc.: We’ll try to make it happen.

Beside the Kiwix app itself, we are also working to provide more content in the ZIM format, so that every Kiwix-enabled device can become a large library of free content.

About Kiwix

Kiwix is an offline content reader working with ZIM (OpenZIM) files. It works on Windows, Linux, Mac and now Android. Additionally, it is available as a Web Server solution and powers the Afripedia Plugs.

Renaud Gaudin, Kiwix Developer

The Wikimedia Language Engineering team completed its recent development sprints with a focus on feature completeness of ‘Translate User eXperience’ or ‘TUX’ for deployment and also made preparations for its new community outreach project. Below are some of the highlights from the sprint.

Work is underway on the design of the new main page for translatewiki.net

Translate Editor – TUX, has been deployed: The Language Engineering team has been blogging on improvements to the Translate Extension which has now been deployed to WMF sites and been enabled as the default editor on translatewiki.net. Issues encountered while using the new editor can be reported via Bugzilla. More details about TUX’s design features can be found in our previous blog post.

New Language Outreach Program: A new outreach and support program to facilitate feedback from our language wiki projects is also currently in development. The program would facilitate focused feedback about the tools being developed and reaching out with solutions to the larger communities working on Wikimedia projects in various languages. The pilot phase is scheduled for launch in the coming weeks and an announcement with the details of the program will be shortly made.

Updates to the MediaWiki Language Extension Bundle (MLEB): In other news, Amir Aharoni announced the release of the new version of the MediaWiki Language Extension Bundle (MLEB). Besides development updates to TUX, version 2013.3 of MLEB includes bug fixes to the Universal Language Selector (ULS) as well as new fonts for Hebrew, Javanese and Arabic. Last but not least, a Marshallese input method has been contributed to jQuery.ime by Nick Doiron.

Testing plan for language tools: A preliminary outline of a testing plan for all internationalization tools currently maintained by the team was completed. Guided by earlier discussions with the Fedora Localization Testing Group (FLTG), the initial draft includes setup of test environment, preparing test scenarios, collecting feedback and analyzing results.

Preparations for visualization of language coverage information: Work also continued on the Language Coverage Matrix, a collection of data about the availability of language tools for different languages in Wikimedia projects. Currently, the focus is to prepare a technical specification for automated presentation and access to this information.

Event participation and other news: Siebrand Mazeland represented the Wikimedia Language Engineering team at the Internationalization and Localization Conference organized by Lingoport in Santa Clara. (More details: presentation slides, and talk recording)

The ongoing development sprint will focus on creating a new design for the translatewiki.net homepage, launch of the pilot phase for Language support outreach program, and publication of the Language Engineering roadmap for the next fiscal year 2013-2014. The Language Engineering team is also looking to hire JavaScript and PHP engineers with deep experience in i18n and l10n technologies. Additionally, a few of the projects that are open for participation have been listed here.

The Language Engineering team is available to answer any technical questions you may have about the tools it develops. You can join us at our monthly office hour scheduled this month on April 10, 1700 UTC and 1000 PDT or find us on irc.freenode.net at #mediawiki-i18n. Logs from the last office hour held on March 13, 2013 can be found here.

Runa Bhattacharjee, Outreach and QA coordinator, Language Engineering

Wikimedia consists of many projects, Wikipedia most notable among them. However, the name “Wikimedia” suggests a world beyond text. Indeed, Wikimedia Commons, our repository of freely-licensed media files, already contains more than 16 million images, sound files, and videos.

Well, mostly images. Right now, there are fewer than 30,000 video files, and fewer than 170,000 audio files. And while Wikipedia articles are often richly illustrated, they still share the old-school feel of a print-based experience. Projects like Snow Fall by the New York Times show what an immersive reader experience can look like, with video elements prominently featured and blended into the core of the content. In contrast, Wikipedia articles rarely have videos, and if they do, those videos are usually very short and included at the bottom of the article.

Of course, well-written text forms the foundation of most high quality educational content.  Text is versatile, adaptable, accessible, efficient, and relatively easy to collaborate on.  It will form the core of the Wikimedia experience for a long time to come. Still, we can greatly improve the educational value of our sites by empowering everyone to share media, collaborate on improving that media, and using that media well throughout our sites.

In the last three years, Wikimedia has seen some very significant multimedia developments:

• The Wikimedia movement has launched successful photo contests and competitions, notably the “Wiki Loves Monuments” competition, which was recognized as the world’s largest photo competition by the Guinness Book of Records. In the 2012 competition, more than 350,000 photos were taken by volunteers. It was organized by Wikimedia chapters and volunteers in 33 countries (see jury report).
• Wikimedia chapters and volunteers have also formed partnerships into the cultural sector (e.g. museums, galleries, archives), resulting in hundreds of thousands of photographs, reproductions of paintings, and other media being made available on Wikimedia Commons.
• Wikimedia Foundation has developed a number of enhancements and features focused on multimedia:
  • the Upload Wizard, an easy-to-use tool for uploading media files that’s been used to upload more than 2.2 million files to Wikimedia Commons;
  • upload features for the mobile web that make it easy to enrich any article requiring a photograph using a smartphone;
  • a new HTML5 video player with support for the open WebM video format and encoding of videos in multiple resolutions;
  • dedicated upload apps for iOS and Android are in development;
  • a feature to import photographs from Flickr (started as a Google Summer of Code project)
  • an experimental feature to upload files up to 500MB in size.

In combination, these efforts have already borne fruit. The number of contributors to Wikimedia Commons has increased significantly in the last 3 years.  In January 2010, only 13219 users had contributed at least one upload.  That number increased to 20161 users by January 2013.

At the same time, we haven’t invested enough. With the exception of the work of our mobile team, much of the above work has been done by one or two developers at a time, often in between other priorities or by engineers working as volunteers. There has never been a well-resourced team fully dedicated to multimedia engineering work at the Wikimedia Foundation. This is about to change.

The Wikimedia Foundation is hiring at least three engineers and additional product/design support to fully focus on improving the user experience for contributing, curating and reviewing multimedia. Right now, you can apply for the following positions:

• Software Engineer – Multimedia Systems
• Software Engineer – Multimedia User Interfaces

Here are some of the key challenges for the new team:

• further improvements to the upload experience. Contributing an image or video to an article while you’re editing should not require leaving the “edit mode” — it should be integrated with the editing process.
• solidifying experimental features such as large file uploads;
• improving transcoding features for video files to reduce the learning curve for video uploaders;
• improving media search and discovery;
• improving display of images, videos and sound files in Wikipedia articles, including a standard lightbox viewer for media embedded in an article and related media from Wikimedia Commons (building on some of the excellent submissions in our October 2011 Coding Challenge).

As we continue to provide new means for uploading media, we need to ensure that the Wikimedia community is empowered to curate and categorize the images. Curation includes removal of content that is out of scope or incorrectly licensed. To more effectively patrol content, the development of curation tools similar to the Page Curation feature developed for Wikipedia may become necessary.

Beyond Wikimedia’s category system, we will likely want to explore implementation of lightweight tagging systems, possibly in partnership with the Wikidata team.

As if this weren’t enough, the long term frontiers for multimedia include web-based editing of images, video and sounds, improvement for subtitle editing, browser-based audio recording features, and more.

In short, breaking through walls of text and creating a richer media experience for all our projects will keep the Wikimedia Foundation and the Wikimedia movement busy for many years to come. Please help us expand our library of freely-licensed educational media, and help us ensure it gets used effectively on the world’s fifth-most popular website.  Apply today.

Rob Lanphier, Director of Platform Engineering
Erik Möller, Deputy Director; Vice President of Engineering and Product Development

The Nardis Waterfall (Cascate Nardis) in Trentino, Italy, uploaded via mobile.

Wikipedia isn’t just the encyclopedia anyone can edit—it’s also the encyclopedia anyone can illustrate. Starting this week, logged in users browsing the mobile web on smartphones with upload capability will see a new feature: the ability to add images to articles that lack them.

In one easy step, you can upload an image from your phone’s camera or image library and add it directly to an article that has no images. You can also donate images for use on articles that already have images but may need more.

Images enhance the visual appeal of Wikipedia and its sister projects and help bring our content to life, but they’re also a powerful educational tool. There’s no better way to describe a notable building or landmark than with a current photo. Not only will you be illustrating knowledge, you’ll also be sharing your photographs with billions of people around the globe.

An example of an article lacking images on the mobile English Wikipedia.

For example, a quick snapshot from your smart phone’s camera can showcase the beauty of the Nardis Waterfall in Trentino to Wikipedia readers who have never set foot in Northern Italy. With millions of articles on Wikipedia, no matter where you are, it’s likely that you have an important piece of knowledge to illustrate right in your backyard.

Smartphones with cameras are becoming increasingly prevalent, and more mobile sites and apps are focused on getting people to explore and photograph the world around them, so it made sense for our mobile web team to bring this feature to Wikipedia.

Unlike many other image sharing sites on the web, images donated via Wikimedia mobile are released under a free license and can be shared and reused by anyone, anywhere, for free. When you donate images to Wikimedia projects, you’re not just sharing photos with your friends, you’re sharing them with everyone in the world.

Help make Wikipedia more beautiful, vibrant and educational for all our readers! Log in or create an account on any one of the over 280 language Wikipedias or sister projects to try out this feature. And stay tuned for more opportunities to contribute via the mobile web, coming soon.

Maryana Pinchuk, Associate Product Manager
Mobile Web

This post was originally published in German on Wikimedia Deutschland’s blog by Nicole Ebber; it was translated by Denise Jansen.

Wikimedia Nederland is going to be host to Wikimedia Hackathon Amsterdam, the international Wikimedia developers conference, on May 24–26, 2013. The Netherlands Chapter invites MediaWiki developers, coders, hackers and other technically-inclined Wikimedians to spend a week-end in Amsterdam.* The event is open to everyone who is involved in areas such as tools, gadgets, bots, bugs, extensions or templates — regardless of how long they have been active.

Proposals for workshops, presentations and sessions are currently being gathered on the event page.

Focal points will be, among others:

• VisualEditor and Parsoid
• Lua
• Design
• Wikidata
• RENDER and Cosyne
• GLAM and Wiki Loves Monuments
• OpenStreetMap
• Outreach and Promotion
• Migration from Toolserver to Labs/Tool Labs
• Cooperation between skilled developers and the project communities

With more than 40 staff members of the Wikimedia Foundation taking part, as well Wikimedia Deutschland staff involved in Wikidata, RENDER and the Toolserver, the Amsterdam Hackathon will provide a great opportunity for exchange and cooperation among organisations and communities.

If you are interested in the Toolserver, or its future alternative Tool Labs, you will get the chance in Amsterdam to meet the entire team that is currently involved in the development of Tool Labs and the imminent migration.

This team will offer a Tool Labs introduction workshop and will be ‘approachable’ in the Hacking area. If you want to try out Tool Labs in Amsterdam, it’ll help if you set up an account beforehand.

Registration for the Amsterdam Hackathon is open until April 20; participation is free. As in previous years, there is a scholarship programme for participants who need support to cover the costs of travel and stay. This scholarship programme is supported by the Wikimedia Foundation and Wikimedia Deutschland.

* In previous years the Hackathon was hosted by Wikimedia Deutschland. Wikimedia Nederland is thrilled to be able to host the event this year and is cooperating with Wikimedia Deutschland in the preparations.
 

Wikimedia Hackathon in Amsterdam 24-26 mei 2013

Wikimedia Nederland is in 2013 gastheer van ‘Wikimedia Hackathon Amsterdam’. Deze internationale conferentie voor ontwikkelaars vindt plaats van 24 t/m 26 mei 2013. Wikimedia Nederland nodigt MediaWiki ontwikkelaars, codeurs, hackers en overige technische Wikimedianen uit om samen een weekend in Amsterdam door te brengen.* Het evenement is toegankelijk voor iedereen die zich binnen MediaWiki bezig houdt met tools, gadgets, bots, bugs, extensions of templates. De uitnodiging geldt voor ervaren gebruikers én starters.

Op de eventpagina worden op dit moment voorstellen voor Workshops, presentaties en sessies verzameld. Hoofdonderwerpen zijn onder andere:

• Visual Editor en Parsoid
• Lua
• Design
• Wikidata
• RENDER en Cosyne
• GLAM en Wiki Loves Monuments
• Open Street Map
• Outreach en Promotion
• Migratie van Toolserver naar Labs/Tool Labs
• Samenwerking tussen WMF ontwikkelaars en de projectgroepen

Door deelname van ruim 40 medewerkers vanuit de Wikimedia Foundation en de aanwezigheid van medewerkers van Wikimedia Deutschland die zich bezig houden met Wikidata, Render en Toolserver biedt de Hackathon Amsterdam uitstekende mogelijkheden voor uitwisseling van kennis en voor samenwerking tussen organisaties en de gemeenschap.

Geïnteresseerden in Toolserver of het toekomstige alternatief Tool Labs hebben de kans om het team te ontmoeten dat zich momenteel bezig houdt met de bouw van Tool Labs en de op handen zijnde migratie. Het team biedt een Tool Labs introductie workshop aan en is benaderbaar in de Hacking ruimte. Het is ook mogelijk om Tool Labs uit te proberen, daarvoor is het wel noodzakelijk van te voren een account aan te maken.

Registreren voor deelname aan de Hackathon is mogelijk tot en met 20 april 2013, toegang is gratis. Zoals in voorgaande jaren is het mogelijk om een sponsoring aan te vragen voor deelnemers die financiële ondersteuning nodig hebben voor reiskosten en/of verblijf. De Wikimedia Foundation en Wikimedia Deutschland maken deze financiering mogelijk.

* In voorgaande jaren is de Hackathon gehost door Wikimedia Deutschland. Wikimedia Nederland heeft de eer om dit jaar de organisatie te mogen verzorgen. Wikimedia Deutschland ondersteunt Wikimedia Nederland bij de organisatie van deze Hackathon.

Wikimedia Hackathon in Amsterdam am 24.-26. Mai 2013

Wikimedia Nederland ist Gastgeber der internationalen Entwicklerkonferenz Wikimedia Hackathon Amsterdam, die vom 24. bis 26. Mai 2013 stattfindet. Das niederländische Chapter lädt MediaWiki-Entwickler, Coder, Hacker und andere technisch versierte Wikimedia-Interessierte für ein Wochenende nach Amsterdam ein.* Die Veranstaltung ist offen für alle, die sich — egal ob schon lange oder erst seit kurzem — mit Themen wie Tools, Gadgets, Bots, Bugs, Extensions oder Templates im MediaWikiversum beschäftigen.

Auf der Eventseite werden zur Zeit Vorschläge für Workshop, Vorträge und Sessions gesammelt; Schwerpunkte sind unter anderem:

• Visual Editor und Parsoid
• Lua
• Design
• Wikidata
• RENDER und Cosyne
• GLAM und Wiki Loves Monuments
• Open Street Map
• Outreach und Promotion
• Migration vom Toolserver zu Labs/Tool Labs
• Zusammenarbeit der WMF-Entwickler mit den Projektcommunitys.

Aus der WMDE-Geschäftsstelle werden unter anderem Mitarbeiterinnen und Mitarbeiter aus den Projekten Wikidata (Lydia Pintscher, Katie Filbert, Daniel Kinzler, Jeroen De Dauw), Render (Johannes Kroll) und Toolserver/Tool Labs (Silke Meyer) sowie aus dem Team Communitys und außerdem ich selber vor Ort sein. Sehr erfreulich ist auch, dass mehr als 40 angestellte Entwickler der Wikimedia Foundation teilnehmen. Es bietet sich also eine tolle Gelegenheit für Austausch und Zusammenarbeit mit den Organisationen und den Communitys.

Wer sich für den Toolserver und die zukünftige Alternative Tool Labs interessiert (Kurierbeitrag), hat in Amsterdam die Gelegenheit, das ganze Team kennen zu lernen, das sich jetzt um den Aufbau von Tool Labs und den mittelfristig anstehenden Umzug kümmert.

Dieses Team bietet Tool Labs-Einführungsworkshop an und ist im Hacking-Bereich ansprechbar. Wer Tool Labs dort ausprobieren möchte, sollte sich vorher einen Account besorgen.

Die Registrierung ist bis zum 20. April 2013 geöffnet, der Eintritt ist frei. Wie in den Vorjahren gibt es für Teilnehmende, die ihre Reise nicht aus eigener Tasche oder mit Unterstützung ihres Chapters tragen können, Stipendien für Anreise und Unterkunft. Wikimedia Deutschland unterstützt dieses offizielle Stipendienprogramm, so dass Communitymitglieder aus Deutschland sich direkt dort bewerben können. Wir freuen uns über zahlreiche Teilnahme!

* Und um Spekulationen vorzubeugen (in den Vorjahren war Wikimedia Deutschland Gastgeber des Events): Wir sind hocherfreut, dass Wikimedia NL in diesem Jahr die Gastgeberrolle übernimmt und unterstützen sie bei den Vorbereitungen mit den Erfahrungen und den Materialien der Vorjahre. Für uns ist der Hackathon ein schönes Beispiel für gute Chapterzusammenarbeit und Wissenstransfer.

The Wikimedia Language Engineering team has been regularly reporting updates about improvements to the Translate editor, as part of the “Translate User eXperience” project, or “TUX”. Pau Giner, the team’s UX expert, has also conducted online sessions to talk about these features. If you have missed these updates, here is a summary of what we are changing about the way the Translate editor is used.

The main editor screen of Translate’s new version, with Spanish translations in List view.

Translate is a MediaWiki extension that is used for translating software and wiki pages. Besides providing translations through the web-based editor and proofreading features, it also supports export and import of gettext files for offline translation. The editor provides various features to assist in translation, such as:

• Message documentation, also known as “context”;
• Suggestions from translation memory and machine translation;
• Checking translations for common syntax mistakes;
• Translation status of messages.

Originally created by Niklas Laxström, this extension has grown in features through contributions made by other contributors, as well as by the Wikimedia Language Engineering team. The extension uses a continuous development model and, if you use the extension on a wiki you administer, you are encouraged to update it periodically using the MediaWiki Language Extension Bundle (MLEB).

The workflow and features for Translate were recently redesigned to provide users with an improved experience. The development was done based upon the designs in the workflow specification document. This included changes in navigation, editor look and feel, translation area, filters, search, and color & style. Here are some of the notable new features and changes:

Editing Modes: The translation editor will now provide two translation modes and one proofreading mode. For translation, the user will be able to choose between the ‘List’ view, more suitable for smaller messages, or the ‘Page’ view, designed for longer pieces of text like paragraphs of a wiki page. The proofreading mode will allow users to view translations by other users and mark their accuracy. Although users can view the messages translated by themselves in this mode, they cannot mark them as accepted.

Message status-based filtering: Users will have the option to select and only view messages that match a filter, depending on their status. In the editor, users can choose an appropriate filter to quickly access ‘Translated’, ‘Outdated’ or ‘Untranslated’ messages in translation mode, and ‘Translated’, ‘Outdated’ and ‘Unproofread’ in proofreading mode. Translations marked as ‘Outdated’ (equal to the jargon term “fuzzy”) need attention, for example because the source message has changed.

Message editor and translation aids: The messages in focus are shown within an editing area that is divided into two separate sections: one for translation, and the other for translation helpers, like context documentation, suggestions from previous translation and external translation services. The layout aims to make optimal use of available space and also provides users with the additional option to focus better on a message by expanding the size of the editing area to the entire width of the editor. The navigation to the next message, the ability to save drafts and the display of warnings make the translation process more fluent. Development of some exciting features for improving context-related translation aids is also on the cards.

Search and edits: Users can search translatable strings using the search field at the top of the edit section. The search results are displayed within various categories like ‘source’ or ‘translated’ messages. An additional overview displays the languages and message groups where they occur and users can further filter them based on the sub-groups. Users will be able to directly go into ‘Translation mode’ to make changes to the messages in the search results. A navigation arrow can bring them back to the list of results.

Not all of these features are available on Wikimedia wikis yet, but they will be soon. The current development version is available on translatewiki.net. If the new editor is not visible, appending “&tux=1” to the URL will enable the new features. Appending “&tux=0” will disable them.

While redesigning Translate’s User experience has been a significant project, development is continuously carried out to make the extension even better to use. And for this, we are always looking for valuable feedback from our users. Bugs and features requests can be filed through bugzilla; additionally, one can write to me at runa at wikimedia dot org with their feedback and suggestions.

Runa Bhattacharjee, Outreach and QA coordinator, Language Engineering