The Parsing team at the Wikimedia Foundation that develops the Parsoid
service is deprecating support for node 0.1x. Parsoid is the service
that powers VisualEditor, Content Translation, and Flow. If you don't
run a MediaWiki install that uses VisualEditor, then this announcement
does not affect you.
Node 0.10 has reached end of life on October 31st, 2016  and node
0.12 is scheduled to reach end of life December 31st, 2016 .
Yesterday, we released a 0.6.1 debian package  and a 0.6.1 npm
version of Parsoid . This will be the last release that will have
node 0.1x support. We'll continue to provide any necessary critical bug
fixes and security fixes for the 0.6.1 release till March 31st 2017 and
will be completely dropping support for all node versions before node
v4.x starting April 2017.
If you are running a Parsoid service on your wiki and are still using
node 0.1x, please upgrade your node version by April 2017. The Wikimedia
cluster runs node v4.6 right now and will soon be upgraded to node v6.x
. Parsoid has been tested with node 0.1x, node v4.x and node v6.x and
works with all these versions. However, we are dropping support for node
0.1x right away from the master branch of Parsoid. Going forward, the
Parsoid codebase will adopt ES6 features available in node v4.x and
higher which aren't supported in node 0.1x and will constitute a
Subramanya Sastry (Subbu),
Technical Lead and Manager,
 Node.js Long Term Support schedule @ https://github.com/nodejs/LTS
This is a Final Call for Comments on the RFC on Content Model Storage . If
no new and serious objections are raised within a week, the Architecture
Committee will approve this RFC and drive its implementation.
The RFC on Content Model Storage was originally approved in 2015, but was then
postponed in favor of another RFC, which proposes to create a separate content
meta-data table  as part of the move towards multi-content revisions (MCR) .
However, MCR in turn got stuck on database performance concerns. So we now
propose to go ahead with implementing the original RFC. The idea is to assign a
number to every content model (and content format), and then use these numbers
to refer to the models and formats in the database, instead of repeating the
same string millions of times (which is my fault btw, sorry about that).
Since the original RFC was already approved, and the situation does not seem to
have changed since then, we see no need for another round of discussions. If
nobody raises any new and serious objections within a week, this should be good
Senior Software Developer
Gesellschaft zur Förderung Freien Wissens e.V.
On the second day of the Wikimedia Developer Summit (January 10) there will
be a Q&A session with Victoria Coleman (Wikimedia Foundation CTO) and Wes
Moran (VP of Product). It is a plenary session and it will be
The questions for this session are being crowdsourced at
can propose questions and vote, anonymously, as many times as you want. At
the moment, we have 25 questions and 451 votes.
An important technical detail: questions posted later have also good
chances to make it to the top of the list as long as new voters select
them. The ranking is made out of comparisons between questions, not
accumulation of votes. For instance, the current top question is in fact
one of the last that has been submitted so far.
Why posting or voting a good question? One obvious reason is to encourage
the Foundation's Technology and Product top managers to bring a good answer
in a public session with minutes taken and video recording. :) Beyond
that, if the ranking of questions makes sense and is backed by
participation numbers, it has a serious chance to influence plans and
discussions beyond the Summit.
The current ranking does make sense, but maybe you could help covering more
areas, other perspectives?
1. How do we deal with the lack of maintainers for all Wikimedia
2. Do we have a plan to bring our developer documentation to the level
of a top Internet website, a major free software project?
3. For WMF dev teams, what is the right balance between pushing own work
versus seeking and supporting volunteer contributors?
4. During the next year or so, what balance do you think we should
strike between new projects and technical debt?
5. When are we going to work on a modern talk pages system for good?
6. Whose responsibility is to assure that all MediaWiki core components
and the extensions deployed in Wikimedia have active maintainers?
7. How important is to have a well maintained and well promoted catalog
of tools, apps, gadgets, bots, templates, extensions...?
8. Will MediaWiki ever become easier to install and manage? (e.g. plugin
manager à la Wordpress). How much do we care about enterprise users?
9. What should be the role of the Architecture Committee in WMF planning
(priorities, goals, resources...) and are we there yet?
10. In addition to Community Tech, should the other WMF Product teams
prioritize their work taking into account the Community Wishlist results?
The full list:
Engineering Community Manager @ Wikimedia Foundation
I'm doing some renovations on recitation-bot and running into trouble when
the time comes for pywikibot to upload article data to wikisource and
commons. The thread doing so hangs without any sort of informative error. I
made sure that the unix user under which the web service that is using
pywikibot is running is logged into each wiki per Max's advice but I still
have the problem. I'm going to try to get more information about what's
going on but would also appreciate pointers about what might be going
wrong. Particularly, the web service is now running under Kubernetes rather
than sun grid engine, so I suspect that the login state might not be making
it into the container - can anyone advise on where the login state is
maintained and whether this will be transferred into the kubernetes
I want a small popup to be displayed when I hover over a link to a
Note that the popup has nothing to do with the page to which it links.
So for example, I want a link to A:X and when I hover the mouse over it
I want to display B:X in a small popup (where A and B are namespaces).
Note that B:X is a short page, it should be displayed in the popup
completely (not just its announce or description).
Note also that I need the popup to be "stable" enough, so that I could
move the mouse (without the tooltip disappearing) and click a link in
I am considering to write a new extension for this myself. But first I
want to consult its syntax and API.
For me personally the best mode of functionality is the following:
In LocalSettings.php write an associative array mapping of namespaces:
A=>B, etc. Then make popups for ALL links to objects in A (on every
page which links to namespace A).
However, other users may want different API and syntax, for example:
Please discuss with me which APIs and syntaxes should be available.
Victor Porton - http://portonvictor.org