Global Gathering 2019 - Day 2 - Pipe session

What is the purpose of this session?

Do we want to review / establish our 12-month priorities? 6-month priorities?
Actual pipe? … noting that we can’t really do this until after the technical sessions on performance, security, etc.

What outcomes/deliverables do we want?
12-month priorities, maybe also 6-months

Who is facilitating? Kirsten

Who is scribing? Gen


  • We discussed the 4 pillars and suggested that these might be reviewed, but no decisions made here
  • Reviewed the Milanote together, knocked off some tasks that are done/advanced elsewhere, added some new tasks
  • Turned each of these tasks into post-it notes and plotted on a value matrix
  • Decided on next steps for the most important of these tasks according to the value matrix
  • Agreed that this will be developed and updated over the course of the global gathering as we learn more


  • Schedule separate session on establishing a dedicated second place where non-dev tasks can be managed. Links to the issue raised re: lacking comms and marketing team.
  • Schedule separate session on network prototype - breakdown sub-tasks, potential quick-wins
  • Schedule daily check-in with the value matrix
  • Matt to check ordinary cost of security check so we can decide whether it’s really urgent to pounce on the free trial now
  • Spree 2.0 - just do it! Review/write up previously agreed plan
  • Sys admin standardization - determine next steps in tech decision meeting, and record in dedicated Discourse post
  • Review roll-out plan - determine next steps in tech decision meeting, and record in dedicated Discourse post
  • Invoices legally compliant - Rachel to create a github issue
  • SKU - Lynne to create a github issue
  • Moblie - Luis to break down the Mobile UX improvements task into different PRs
  • Mobile - schedule working session on mobile browser device compliance discussion (Rachel, Lynne)
  • Mobile - Matt and Maikel to get started on the front-end work
  • Customer name on the report - Lynne to boundary the scope of the existing wishlist
  • API to do list - Luis will update us on this tomorrow (Wednesday 14th)
  • S3 top 5 (easy) - Rachel will create an epic
  • Rachel will close existing epics
  • Rachel will deal with product import, subscriptions, users’ access to enterprise fees
  • Spree 2.1+Rails 4 - schedule an action session

Detailed notes:

Lynne - time manager
9:00-11:00 break 11:15-12:30

4 pillars:

  • Be a Single Cohesive OFN Unit
  • Make What We Have Great
  • Future-Proof the Tech Platform
  • Build the Network in the OFN Platform
    5th pillar on interoperability? Or should this fit under Make What We Have Great?

Maikel asks what are the deliverables we want on the tech side. Kirsten responds that we won’t know until we know our priorities. Lots of resources have gone into the Spree upgrade, so we need to pause and think about how to best use dev resources going for the next period.

Lynne asks whether the pillars are necessary for planning work. It seems that the work often comes first, and subsequently it can be fit under one of the 4 pillars. Some dispute on this - e.g. necessary to prioritize performance over some other objectives. Rachel suggests that right now we’re mixing building a roadmap and building an action plan. Luis argues that the roadmap should be prioritized by return-on-investment.
Rather than starting from the existing pillars, maybe we should start organizing our priorities from scratch? But most of the room is confident we’ll land on the same pillars.

We all hop in a Google Hangout to look at the Milanote together

Allow enterprises to bulk upload product information

  • We discuss whether we should avoid opening dev tasks in relation to this until there’s real value to allowing users to do this directly.
  • Kirsten - How do we prioritise tasks not for devs? Second pipe? At least having a dedicated second place where these things can go. Links to the issue raised re: lacking comms and marketing team - spin out to separate session.
  • Focus on creating guidelines for users so users/others troubleshoot, rather than having product import tasks clogging up the dev pipe
  • 1 issue remaining on GitHub - known fragility around special characters
  • How do we support customer service people to help users troubleshoot around this
  • Mostly done for v1, but outstanding v1 issues re: e.g. subscriptions

Sys admin standardization

  • Whole session after lunch on this
  • Maikel - not done, but could be done in 1 or 2 months
  • Luis - we should have 1 pipe and ensure that we prioritise within that pipe
  • Kirsten proposes a new decision-making framework - Sys admin tasks should go into Dev Ready
  • Luis - we need to establish epics for each of these task groups so we can be clear on overall goals
  • Kirsten - any Dev or Tech Admin should be able to jump into any instance to fix things
  • Francois - we’ve actually made massive progress on sys admin - it’s much more integrated now than it used to be

Report that isolates various admin fees

  • Kirsten closes this task, with the understanding that there will be bugs in the upgrade

Enable better navigation and task completion when shopping on mobile devices - Mobile UX improvements

  • Luis can do the part of breaking the task into different PRs
  • Is there anyone who’d like to work on mobile redesign? No volunteers, so we need to find someone
  • Lynne points out that someone at the UK Dev Hack was excited to get involved in this - CSS so easier for a newcomer. Kirsten highlights that it’s complicated by integration with the full system though, so better for one of the existing devs to do it… It’ll be in the queue and someone will take it

Make OFN invoices legally compliant + Make information on quantity ordered / delivered / invoiced / voided / paid accurate to improve packing operation efficiency

  • Split into two separate tasks:
  • “Make OFN invoices legally compliant” is the outstanding priority
  • “Enable quicker packing and invoicing” marked as complete

Enable users on multilingual instances to see all info in their selected language

  • Luis highlights that there are three levels to address: static translation (missing translations), dynamic translations (categories and tax categories), multilingual catalogue
  • Lynne - should this really be an “up next” task? Rachel - let’s just have all the tasks laid out first, and then deal with priorities

Subscriptions v1.1 and v2

  • new task emerging out of “Allow enterprises to bulk upload product information”

6-12 month outlook for Global Priority List: In Play

Mining things already on the list, brainstorming what else needs to be here
Added: Network prototype, customer name on report, minimum spend, internal credit…
Luis: Performance is not just one thing - bulk order management, shop products list, reports, front-office, order cycle, bulk product edit - to be discussed further in another session
Rachel: Not only performance, also UX
Matt points out that the free two-week trial of the security check might flag issues to be addressed urgently. Francois suggests this should happen after the upgrade. Kirsten suggests that even if the free trial is only available right now, it might be better to wait until the correct moment to run it, and just pay later (if it’s only a few hundred dollars - to be checked)

We then use the wall sheet for value matrix [see above image]
Green post-its - must be discussed in tech session
This will be developed and updated over the course of the global gathering as we learn more

Single instance

  • Wasting lots of money on 10 servers
  • Fewer instances, fewer servers?
  • Currency is the key block here (so maybe just merge shared currency regions)
  • Language is a second key block (customers/ordinary users (S3), super admin, multiple names to products)
  • Also would need to deal with tax and email differences…
  • Connects to interoperability issue

Admin UX and Onboarding

  • How important is admin UX?
  • Losing potential users due to these issues?
  • But also think about how clunky competitors are, what experience users are coming from
  • Nick suggests that clunky admin UX serves to prove that OFN isn’t venture capital-funded like competitors (e.g. Food Assembly)
  • Redesign of onboarding is maybe more important here to get new users used to system. 1-on-1s with new users is a workaround, but this is too time-intensive for the already overworked…
  • Direct personal contact with OFN team perhaps compensates for lack of slick interface, and adds its own value separately. Pau is skeptical of this, 1-on-1 time not sustainable…

Divergences between instance-specific and global community importance:

  • “Make OFN invoices legally compliant” - v important for France
  • “add SKU to order confirmation” - important to UK

Front-end vs back-end development:

  • Front-end important, but more messy, subjective
  • 2-3 months for a front-end dev to get up and running and become useful to OFN, according to Kirsten
  • More efficient to do that, or have one of our existing back-end devs do it instead?
  • Luis suggests he could do Spree upgrade on voluntary basis so as to free up budget for front-end dev
  • Kirsten - this would be great, but not if it risks losing Luis in the long term
  • Nick - too risky to hire a font-end dev if we don’t get the money in July
  • Alternatively Matt, Maikel can implement existing front-end design
  • Maikel - but if there’s any indication that money will come through, we should bring a new dev on
  • Angular 1.5 and CSS separate issues to address - parallel to the angular question