Scrum of scrums/2014-12-03

Facilitating: Andrew Otto


  • Deployed fix for template expansion with older MW versions.
  • Deployment today: Fixes to v2 API endpoint, fix to API v2 ids leaking into v1 API.
  • Need CSS help on Cite CSS patch, more info at


  • Working with product and analytics on defining metrics and targets for CX beta feature deployment
  • Work on adding cx_ tables to beta wikishared completed; patch for review
  • Finalizing features and stabilizing release for beta feature deployment in January



  • ContentHandler stlll blocked on us?
  • T1223 Unified diff on mobile - Is this still valid


spage Can't attend, first standup on Phabricator

  • still working on TOC
  • need wgContentHandlerUseDB in late December, haven't followed up with csteipp


  • David Chan visiting SF until Dec 14, working on IME support with Roan
  • No particularly exciting things happened in wmf11
  • Exciting things will happen in wmf12: IME support improvements, find and replace
  • Roan working with Ori on JS gateway for statsd


  • Helping people with dashboarding (you're next!) - some basic documentation coming soon
  • Event logging outages and matching reports available for those interested. All sent via email already, but just in case you didn't hear
  • Data Warehouse progress is good, will not be making it available in labs right away though
  • We are working around labsdb problems - so it's no longer a problem for us in the short term


  • HHVM 100% of non-API requests; API underway blocked on T758
  • LabsDB issues (T76075) - in progress
  • Phabricator
  • BZ leftovers (mainly handling of security tickets)
  • service level discussion, T76446
  • RT migration to come next
  • T1228 Restbase - in progress(?)
  • T1254 CX Databases- in progress(?) -- Is this still an ops task?
  • T67683 - deployment cache purge
    • Need to ping bblack
  • T76086 - labs deployment-prep ldap user for NFS4
    • In progress, Andrew Bogott did somethign, need to check if it works.

Fundraising TechEdit

  • Whole team in SF this week
  • Souping up internal dashboard
    • Watching the money roll in, scrutinizing load and logs
    • Started full-strength campaigns Tuesday. $3.5M raised in December so far.
  • Watching CentralNotice closely
    • So far, any odd allocation behavior has been due to banner-specific javascript
  • Minor updates to mailing list data, payment forms, and reporting
  • GeoIP issues Monday? saw incorrect lookups from HHVM and from Varnish, but Varnish seems correct now
    • Updated FundraiserLandingPage and LandingCheck to rely on cookie from Varnish


* Otherwise: 
  • refactoring MobileFrontend tests
  • fixing linting errors everywhere
  • training people to write browser test

Mobile WebEdit

  • Experimenting with pulling data and images from Wikidata
  • Concluded WikiGrok A/B test in Beta mode
  • Planning to conduct some limited testing in stable mode next week
  • Nik's Wikidata Query team is working on testing graph databases
  • Talking with Analytics about replacing limn for dashboards
  • Still having problems with Beta Labs outages/instability


  • Done: Deployed Accept-Language aware redirects to mdot for Wikipedia Zero users
  • Yuri to add X-Analytics logging across sister projects and desktop
  • T76626: HTTPS-to-HTTP downgrade option interstitial for external Referers via Varnish upcoming



  • SHA-1 patch is ACTUALLY HAPPENING THIS WEEK. Or Gilles will have to answer to ME.
  • UploadWizard was throwing fake stash errors for about two weeks, apparently nobody noticed, it's fixed now.
  • We're shifting away from other projects to seriously focus on UploadWizard now, so...yeah.
  • Also, Commons is using chained thumbnails now, and apparently nothing is broken *knocks on wood* so that will continue being awesome and hopefully faster!
  • We're thinking about bringing OOUI into UW a little bit, but it's not quite ready yet...VE team, watch out, I'll be coming for you soonish.