Date

23 Mar 2017

Attendees

Raw notes

https://pad.disroot.org/p/foodsharing-dev-2017-03-23

Summary

Introductions

Two new people Lorenz (@zner0l) and Peter Tönnies.

Lorenz is studying physics in Göttingen, joined foodsharing a few months ago, not very active though. Found out developers were needed and is experienced in php, frameworks, etc and a web designer too. Looking for a project to get involved in. Does volunteering for kids/adult educational summer camps too. Interest to get into coding on project, specifically with symfony and doctrine.

Peter is from Köln, now living in Zurich. Works as a hydroelectric engineer, codes in C#. Been involved with foodsharing there for 1.5 years (irrc). Got involved quickly, a BOT after half a year. Evenings and weekends on foodsharing. Saw use of git at foodsharing, looked at issue tracked, interested in the tech (docker, etc), useful to learn for work too. Also hobby musician. Most interested in helping find more volunteer programmers, writing issues, bug reports, being pedantic

(see raw notes for introductions for existing people).

Mobile app development

Peter registered foodsharing at http://socialcoder.org (they say: "We connect volunteer programmers with charities for specific, goal-oriented projects"). Has contacted socialcoder and he/they seem nice and supportive, but Peter is concerned that the entire foodsharing.de website as it is would not be suitable for socialcoder as it's not open source, is not really a standalone project, and going through many other changes too. So the idea is to request a mobile app that would be a complete standalone project. They help to make estimate of time/cost and finding a project lead and collaborators.

Peter explained the issues that they are facing in Zurich - that is mostly about contacting members. Peter wants to delegate tasks and get support from other members, but it is too difficult to contact them and ends up doing things himself. Most members switch off email notifications and replies might take 2 weeks which is then too late. Various stores teams then switch to using individual whatsapp/telegram/etc groups. There is one big whatsapp group (~400 people) but this is not ideal as most messages are spam to most people so people might leave that too. People try and use the website with a mobile browser but with not such a great experience, Matthias had statistics showing 30% of page views were from mobile (although includes people visiting site, not actively using it).

So main feature set for a mobile app that would be useful he says is mobile push notifications for messages and maybe a map showing stores/fairteilers/food baskets/etc. Other features for later would be to make it easy to see who you are collecting food with, and have their contact information easily available. Kristijan points out you can click on the picture of people on the website and it will prompt you to call them - it's not very well known but it's not something people complain about in his region. Lorenz points out if the features aren't clear then that's part of the problem, it should be easy to understand. Kristijan wondered about adding forum features which are used in his region, but not so much in Peters.

Some discussion then followed about the features and protocols for chat and push notification. The existing app (website or mobile?) used to have push notification but Matthias disabled it as it was not working properly. Code was not very easy to follow. Peter wonders if it's possible to use an existing chat protocol so people can use their own chat clients - XMPP and Matrix were discussed. Matthias pointed out that the there are complicated permissions to decide who is part of each chat and that may be difficult to integrate with other chat protocols. Matrix seems the most likely to be able to be usable in this way. Lorenz pointed out that if the chat has a good API, it would be possible to build a bridge to other protocols. General agreement that a good API is important, Matthias points out the existing chat API is one of the better bits of foodsharing code at the moment, so might be sufficient.

The question was raised whether it could be worth developing a new mobile-friendly web app instead of a native app. It could be packaged to go in the mobile app stores, or used with a mobile browser - Lorenz described feature in Chrome for mobile notifications that can be packaged with app. Could start with a minimal feature set that would be useful for Peter. Generally it seemed a good option, but all agreed best left to whoever will do the development to input on this decision. Kristijan reminded us of what happened in Malo when we tried to reimplement foodsharing but got overwhelmed with the large featureset.

Nick mentioned looking at https://onesignal.com/ which has free features for mobile notifications (although sells user data).

Kristijan mentioned one example of extended featureset which is the quiz, Matthias wondered if the quiz can be retired now as the questions are not always kept up to date. Lorenz thinks we shouldn't be deciding things like this. Kristijan proposed to start a discussing in the ambassador forum about the topic to see the response.

Some debate between Peter and Kristijan whether the most important initial featureset for the app would be foodsavers or foodsharers and the possibility for separate apps focusing on different user types.

Outcomes:

Peter will continue progressing with a socialcoder.org application and ask for feedback/collaboration from others as required
Kristijan will consider opening ambassador forum discussion about whether the quiz is needed any more

Adding ORM in backend

Lorenz has been interested to implement Doctrine ORM for the backend. In the last meeting we decided it would also be a very useful step for generating fake/test data, and would in turn help us improve deployments.

Lorenz has worked with Symfony before and likes working with ORM. He would like to investigate the codebase a bit more first, then discuss proposals to integrate an ORM, Matthias is happy with approach. Lorenz is a little unsure of his time commitment though. Nick suggests to schedule a call between him, Lorenz, and Matthias to discuss how the database is accessed and code structure to help with motivation, Nick is happy to write notes/documentation.

Outcomes:

setup call for nick/Lorenz/Matthias to discuss foodsharing code
Lorenz to continue looking into foodsharing code

Websites for foodsharing regions

@dominik has started a discussion with <some foodsharing group> about creating simple brochure websites for each region where they can explain the project and edit the content in a way that will be found by google, probably with a common look and feel. Peter says it could also be a way to embed a calendar with related events (instead of trying to add features to foodsharing site itself).

Nick suggests wpmu (wordpress multi user - which is part of the core now https://codex.wordpress.org/Create_A_Network), Lorenz thinks this is too heavy for such simple sites and a static site generator or simpler CMS (e.g. picocms) might be a better approach, using markdown formatted content files, also that wpmu is simple to get started but gets slow quickly and is clumsy. Matthias thinks most foodsharing regions would not have people that are comfortable with markdown, and a wysiwyg editor is needed, finds a wysiwyg plugin for picocms, and intends to set up a demo installation to show other people.

Peter wants to know if subdomains for foodsharing.de (and other sites?) will be available, Matthias confirms they will be.

Outcomes:

Matthias to setup demo picocms (and/or other solutions)
preset dominik with our thoughts for consideration

A bit of chitchat at the end

Peter offered to do proof reading, translation, etc. Gladly accepted by all and pointed to #translation channel in slack.