The Wikimedia Legal team is interested in revising, updating, and
projects on labs.
We have opened up Round 1 of our community consultation
can, but the main purpose of this round is to hear all your thoughts. After
the feedback round, we will prepare a draft revision of the Terms based on
that feedback and other minor revisions to clarify statements in existing
the Terms. We will then engage in a community discussion about the revised
We plan to leave the discussion open until June 9, 2016. Thank you for all
your help and feedback.
149 New Montgomery Street, 6th Floor
San Francisco, CA 94105
NOTICE: This message might have confidential or legally privileged
information in it. If you have received this message by accident, please
delete it and let us know about the mistake. As an attorney for the
Wikimedia Foundation, for legal/ethical reasons I cannot give legal advice
to, or serve as a lawyer for, community members, volunteers, or staff
members in their personal capacity. For more on what this means, please see
our legal disclaimer
I had originally planned to release 1.28.0-rc.0 today. Instead we'll be
getting a security release!
This release will cover all active branches. The new versions will be
1.23.14, 1.25.6 and 1.26.3. Master and the REL1_27 branches will also get
There's quite a few fixes going into this release so all installs are
strongly advised to upgrade.
The RC for 1.28 will happen on Monday. The reasoning is half for the
security release and the other is because I wanted a few more days for the
AuthManager code to stew on beta before releasing it.
TLDR: If you care about Selenium tests that run daily, please take
ownership of repository (or repositories) that you care about.
Our Selenium tests provide useful feedback, finding problems in several
places, when there is something wrong with:
#1 the repository they are testing (broken production or test code),
#2 beta and/or production clusters,
#3 continuous integration.
It is really important that all tests that run daily are green all the
time. I have started cleaning up tests[2-3].
If there is a failure, we have to investigate and fix it as soon as
possible, ideally in one business day. To be able to do that, we need
contact information for each repository. I have started collecting contact
details  from various sources.
What does taking ownership mean? Your e-mail address will be added to
Jenkins job repository configuration and you will receive one e-mail a
day per repository and job, but only if there are any failed jobs. In
short, if you are an owner of one repository with one job, you will receive
0-2 e-mails a day, depending on the stability of the job. It will be your
responsibility to fix the job. If you need help, I will be glad to help.
The contact e-mail address can be one or more people, or a team mailing
list. All I need is a reply if I send an e-mail message to that address,
without having to be a member of a lot of mailing lists.
Jobs for repositories without contact person will be running while they are
passing, and will be deleted when they start failing.
To become a contact person, submit a patch or let me know off-list.
Questions? Comments? Please do let me know.
Gerrit is commonly used as a place to share works in progress early.
This is great, but it has an unfortunate side effect of making it
harder for would-be reviewers to find patches that need reviewing
using this query:
Could I ask that as a norm, if you post a WIP patch that you also self -2 it?
In addition to this, if a patch is open for longer than several months
you may want to abandon it - it's much more useful to link to an
abandoned patchset in a phabricator task which has all the context for
someone who might be able to solve the problem it tries to solve.
I'd love to get us to a place where
is manageable that code gets merged left right and center :)
I'm a self employed programmer who's taking a bit of time off from his
current project as a way to gain some fresh perspective and I thought I'd
use that free time to help out a community I really appreciate. Wikipedia is
one of the best things the internet has to offer and it has certainly helped
me out a bunch of times, so it seemed like the place to volunteer.
My problem is that is that I haven't done any sort of collaborative
opensource project before, (the closest I've come to opensource is writing
the compare plugin for Notepad++), and I'm not entirely sure what the best
way to go about it is.
I was wondering if someone there is willing to play the role of onboarding
manager. You know, show me the ropes, assign some good relevant tasks to get
a feel for things, the standard stuff. I'd really appreciate it, and I know
I can contribute a lot to the community once I get on my feet.
As for my skill set, while I don't have a lot of experience with PHP, I do
have a lot of experience with just about everything else. I've done game
development, mobile development, web development and even some good old
fashioned desktop development, (not that anyone still uses desktop apps). I
have no doubt that I can pick up the language quickly enough and I'm sure
Google (the great teacher of all things) will help me figure out the rest.
I'm attaching my resume, so that you can get a feel for my experience.
(Plus it just seems like the thing to do in this situation.)
I'm looking forward to getting started!
I'm one of those programmers who got into it for the love of programming
and it shows. I've programmed on everything from the TI-83 to the Xbox 360,
and if there's a skill I don't have than all I need is an excuse to learn it
and I will. I turned a hobby project into a business that supported me for
several years so I've learned a lot about how to make an amazing product
that succeeds even in the ruthless market that is the app store, as well as
the softer skills that you need to run a business.
I'm looking for a good home. A start up or a startup like environment where
there's more work than time and everyone has to wear multiple hats, and
figure out how to get things done and even what can be done, because those
are the environments that I thrive in.
Visual Studio, C#,C++, Java, C, Visual Basic
4+ years iPhone, Android, Blackberry, Windows Mobile development
Unity, Flash, Actionscript 2.0, ActionScript 3.0, Flex framework, Silverlight
Thirteen years experience with .NET
Owner Pixelality, Seattle, WA (7/14 to present)
Created The Virtual Window a device that keeps track of your head and turns
your TV into a realistic window!
Used Kinect and Wii-mote based tracking
Modified Unreal Engine to support Off-axis projection
Used computer vision code to anaylize a video stream and find the head.
Owner SupportStream, Seattle, WA (4/15 to present)
Funding model that allows people to donate a dollar a month to their
favorite charities, artists or other things they care about.
Created Marketing campaign that will eventually be used to drive donators
to help their favorite causes : https://supportstream.solutions/ASP.NET backend.
Paypal integration for payments. Went through the application process for
Cross-Site scripting with iframes
Used JQuery for animations and popups.
Custom designed logo and various CSS changes
Contractor Big Finish Games (not Big Fish), Salt Lake City, UT (12/13 to 3/14)
Programmed puzzles for the continuation of my all time favorite video game!
Scripted both 2D and 3D puzzles using Unity and C#
Owner Comic Reader Mobi, Seattle, WA (4/09 to present)
Software allows fullsized comics to be read on small screens
Several unique and intuitive UI features to make using the program easier.
(Clicking on text to have it pop out. Realistic page turn effect when
dragging from sides. Drag page down from top to access controls)
Automatically detects text in image and pops out when touched (patent pending)
Out for the iPhone, Windows Mobile, Blackberry, Android, Windows Phone 7,
Linux, PC and Mac OS X (Even have a prototype for the Xbox 360!)
Original version built from scratch for each platform.
Latest version uses .NET/Mono technology to share code base among all the
devices (Except blackberry)
Was one of the top 100 grossing ipad apps
Created Silverlight version which allows you to demo the app without buying it
Reviewed on Gizmodo!
Implemented effective piracy prevention for iPhone version.
Resolution independent code scales to any screen size
Modified jpeg reader to handle large progressive jpegs on low memory devices
Built an entire object oriented UI library from scratch to ease
Mobile Developer Inrix, Seattle, WA (7/13 to 1/14)
Developed parts of the Inrix Traffic app for both iPhone and Android platforms.
Built a prototype turn by turn navigation system for the app.
Mobile Developer Pulsar Informatics, Seattle, WA (2/12 to 5/13)
Lead developer for app for iphone and Android that helped keep track of
sleep and brain performance.
Software was used by International Space Station, Fedex, Navy, and
Department of Transportation for sleep and alertness studies.
Worked with Bluetooth LE connectivity.
Flash Developer Microsoft, Seattle, WA (10/08 to 4/09)
Wrote commercial websites in flash to advertise various products.
Lead Developer for site: http://theguysmanual.msn.com/
Converted artistic vision into functional website
Software Engineer Interactive Alchemy, Phoenix, AZ (1/06 to 6/08)
Won “The Viagra Award for Performance Enhancement Under Pressure”
Full software release cycle. From prototype to final build.
Improved “Story Map” layout performance from several seconds to instant by
rewriting the layout algorithm.
Cut page load times in half by creating a custom C#/Flash socket connection
Wrote “Single Swf” implementation, which dynamically injects files into a
Wrote pure AJAX pager for website, which allows near instant page changes
Used XAML to allow the program to render content off the website as if it
were a winforms application
ActionScript) on a daily basis, depending on the required task
Modified network site to use NTLM/forms authentication, so network users
didn’t have to log in, but allowed outside users to continue using the Forms
Used double buffering, caching and other GDI tricks to significantly
improve the speed and appearance of the application
Extended many winforms controls to have new look and feel
Wrote memory file management system, preventing the need to write files to
a temporary location
Wrote custom logging for Flash application, dramatically simplifying
debugging problems on clients machines
Used ASP.NET’s URL rewriting to share common files among several different
Worked with clients on several projects which often required the
modification of unfamiliar and undocumented code
Interviewed several applicants for both the product and web teams
Programmer FLEXfone – Brussels, Belgium (6/03 to 12/05)
Client side is written in C#, server is written in ASP.NET, MySQL database
Makes use of alpha blending in order to allow the customer to change colors
of almost any item.
Automatically updates itself on startup
Client side is written in C#, server is written in ASP.NET using C#
Open source Projects:
Notepad++: Compare Plugin
Shows the differences between two documents
Written in C++
Highlights and aligns added, deleted, modified, and moved lines
Included in version 4.8 and higher
The University of Arizona – Tucson, Arizona
Bachelor of Science in Computer Science, December 2003
Minor in Mathematics and Japanese
Computer Science GPA: 3.424 Overall GPA: 3.035
Team leader for project to program the game Qubic
Graduated in 3.5 years
Awarded the Charles Blixt Scholarship
Uses unsafe code to directly access memory in order to speed up rendering
Peer to peer mode allows players to play each other over a lan
Written as a Java applet
Pseudo 3d rotating grid
Server side software allows online multiplayer matches
Graphics Renderer: (http://www.ibudesigns.com/Graphics.png)
Basic graphics engine
Renders parallelepipeds (blocks and cubes) and spheres.
Implements shading, basic, diffuse, and specular lighting and raytracing
= 2016-05-18 =
== Product ==
=== Reading ===
==== Web ====
Blocked on Ops - need ability to do a 50% roll out of lazy loaded images.
BBlack is looking into this, waiting for his feedback.
Popups (hovercards) A/B test to be run on Hungarian wiki - finalising pre
UX changes to search on mobile web
==== Android ====
==== iOS ====
* Crash fixes
==== Mobile Content Service ====
* Started implementing first "Explore feed" endpoints: featured article
* Naming of feed endpoints discussion in progress at
https://phabricator.wikimedia.org/T132597. (Make your voice heard by
Thursday, May 19)
==== Reading Infrastructure ===
* AuthManager core is merged! Also fixes for bundled extensions! And
backported to 1.27!
* Work continues on other WMF-deployed extensions, see T110282. Every
extension that we know needs updates has a subtask there.
** If you see in that list an extension you maintain, and have some time
this week, please help out! Many of those tasks are very simple and setting
up the environment / figuring out how to test would take up most of our
time. You can find an update guide at
** JsonConfig could use help from someone familiar with that extension
(yurik?), see T135074. If nothing else, confirm it won't affect WMF and we
can remove it from blocking T110282.
** TranslationNotifications could use help from someone familiar with that
extension (Language team?), see T110766.
=== Community Tech ===
* Working on CopyPatrol tool (currently at
* Collaborating with TCB team on RevisionSlider tool (
** Needs Security Review - probably beginning of July
* Still pushing on category sorting (
** jcrespo adding new indexes for us (
=== Editing ===
==== Parsing ====
(Subbu not going to be around .. just an update)
* Moves afoot to migrate Parsoid cluster to Jessie / node v4 -- discussion
on https://phabricator.wikimedia.org/T135176 about options since Parsoid
puppet code has hardcoded refs to upstart. This probably affects the beta
cluster config as well. Working with Services and Ops.
* Kunal has been working on Linker::link() rewrite -- see email discussion
on wikitech-l (Subject: Linker::link() rewrite)
* Option to split data-mw out of the HTML is now ready in Parsoid.
HTML2HTML to convert between Parsoid HTML versions close to being done.
Appreciate clients (CX, VE) thinking about how they want to work with the
new version where data-mw is no longer inline. This eliminates the
additional html2html conversion once this new version becomes the default
* Tidy replacement work ongoing.
https://www.mediawiki.org/wiki/Parsing/Replacing_Tidy will continue to have
==== Language ====
(Apologies, Kartik not available)
* Blockers: none
* Blocking: none
** Work on Compact Language Links out-of-beta continue.
** apertium->jessie in progress, Kartik should update first set of packages
==== Collaboration ====
** Working on External Store on Beta.
** Continuing notification work on:
*** Cross-wiki notifications successfully released. A couple post-release
bugs, but all dealt with
*** Echo HTML email formatter done
==== Multimedia ====
(MarkTraceur missed the meeting, sorry!)
* '''Blocking''': None
* '''Blocked''': None
** UploadWizard stabilising efforts
** File annotations in a proper extension (cf. hacky wikitext and gadgets)
** Gallery slideshow mode
=== Technology ===
=== Technical operations ===
Apologies, not gonna be around, updates:
** deploying ORES
** working on kubernetes on toollabs
** varnish4 migration ongoing after some setbacks
=== Fundraising Tech ===
* No blockers
* Testing Paypal Express Checkout
* PCI work on payments cluster (modsecurity)
* Civi de-dupe work continues
=== Security ===
* 2FA rollout today
* Security realease (no really, for realz this time) this week
=== Services ===
* Cassandra 2.2.6 roll-out
** RESTBase staging today
* Change propagation back in prod
** RFC meeting today about it - https://phabricator.wikimedia.org/T102476
** started rate-limiting in prod, logging only and locally
** will need help from Ops to set up firewall rules for global rate limiting
* Parsoid to move to Jessie and Node 4.x likely tomorrow
* MathML rendering default on test wiki, mw.org and dewikibooks
=== Release Engineering ===
* Blockers: none
* Blocking: ???
** CI outage in progress, working in #wikimedia-labs for fix
** Scap3 breaking change email just went out
** wmf.2 rolling to group1 (hopefully)
=== Analytics ===
* Blockers: none
* Blocking: nobody we know of
** kafka upgraded to 0.9
** jenkins now knows how to build java packages and add them to archiva (a
few manual steps remain, working on it)
** Cleaning up beta feature enable graphs so limn-flow-data, limn-ee-data,
and limn-language-data will all get a little cleaner in the next day or so,
ping me if you're concerned
** Druid analytics cluster is almost operational, we'll be using it for
wikistats 2.0 metric computation
== Wikidata ==
* Blockers: none.
* Wikidata Query Service was going down several times (
Will need updated resources soon.
* Deployed ArticlePlaceholder extension on 4 small Wikipedias (e.g.
* Experimenting with Timelines on Wikidata Query Service (talk to Lydia).
== Discovery ==
* No blockers
* Updating maps servers
* Blocking: JsonConfig authentication update (Zero), backend script auth
(Zero), SCAP3 for all services
This is a change that affects services that have moved to deployment
via Scap3 (not MediaWiki deployments).
The 3.2.0-1 release that is currently live makes an important change
to the stages in which custom checks may be run. There is now a new
stage called `restart_service` that occurs after the `promote` stage.
The `promote` stage no longer does a service restart. This change is
outlined in the Scap3 docs.
This change likely means that you need to move any custom checks (in
scap/checks.yaml) that were intended to run post-service restart to
use the stage `restart_service` rather than `promote`.
For example this check, which depends on a service restart to work correctly:
command: curl -Ss localhost:1234
Should now be written as:
command: curl -Ss localhost:1234
Sorry for any inconvenience. For future releases, changelog highlights
will be sent to the list prior to release.
There's a thing I've been doing for exactly one year now, and some people
on this list may find it interesting: I've been counting how many article
space edits in the Hebrew Wikipedia added a <nowiki> tag.
These tags are very rarely needed in articles, but they are often added in
edits that go through Parsoid (VisualEditor and ContentTranslation).
Experienced editors complained that they are added too frequently and they
have to fix them manually, so I started meticulously counting _how_
frequently, and also _why_ are they added, so I'd be able to report Parsoid
/ VE / ContentTranslation bugs with the hope of reducing it.
I did the counting by checking Recent Changes every day for edits tagged "
nowiki" (added by a locally-defined AbuseFilter if a main space edit has a <
nowiki> tag in the new text), and checking every diff.
The full analyzed and sorted results are at https://he.wikipedia.org/wiki
/WP:VE/nowiki . I did my best to translate the most essential parts to
English, but please ask me if you have any more questions.
A summary of findings:
* There are on average about 3000 article-space edits in the Hebrew
Wikipedia per day.
* There are on average about 450 edits with the VisualEditor tag in the
Hebrew Wikipedia per day.
* There are rarely more than 20 edits per day that have <nowiki>, and
usually much less than that.
* The most common reason for the appearance of <nowiki> is writing two
apostrophes ('') instead of a double quotation mark ("). It's remarkable
how many people make this mistake, although it's possible that it's more
common in the Hebrew language because of the peculiar ways in which quote
characters are used in it and how they appear on common keyboards.
* The other most common reason is what I call "bad links" and "wrong
links". Both involve letters added after internal links, with a <nowiki/>
added immediately after the closing ']]'; for an explanation about the
difference "bad" and "wrong", see the linked page. Counted together, these
two categories of errors is the most common cause for the appearance of <
* After the above reasons, the most common are vandalism (and I don't
consider it an issue on VisualEditor or Parsoid) and making mistakes in the
wiki syntax of template parameters.
As a result of this work I reported many Parsoid and VisualEditor bugs, and
their excellent developers fixed a bunch: Wiki syntax pasted in
VisualEditor is now correctly auto-converted in a DWIM way; empty runs of <
nowiki>'''</nowiki> are not created any longer if somebody makes text bold
but doesn't write anything; _some_ bugs related to ISBN and external links
handling were fixed (though a few remain); and more.
Something similar was also being done in the French Wikipedia for some
time, but not updated since August 2015 :(
I wish I could do it for other languages, but there's no chance that I'll
find time for that. However, if anybody volunteers to do it for the
Wikipedia in their language, I'll be very happy to help you get started.
I'd be super-interested to know how it is in English, Spanish, Dutch,
Polish, Czech, Russian, Hungarian, and any other language. Takes no more
than 5 minutes per day with the volume of edits in Hebrew, but the time for
other languages will probably be different.
P.S. I'm stupid, please correct my queries if they are wrong.
 select substring(rev_timestamp, 1, 8) rev_date, count(rev_id) from
revision, page where page_id = rev_page and page_namespace = 0 and
rev_timestamp > 20160100000000 group by rev_date order by rev_date;
 select substring(rev_timestamp, 1, 8) rev_date, count(rev_id) from
revision, page, change_tag where page_id = rev_page and page_namespace = 0
and rev_timestamp > 20160100000000 and ct_tag = "visualeditor" and
ct_rev_id = rev_id group by rev_date order by rev_date;
Amir Elisha Aharoni · אָמִיר אֱלִישָׁע אַהֲרוֹנִי
“We're living in pieces,
I want to live in peace.” – T. Moore