December 4th, 2018 @ Spiegel
Attendees:
- Brian Hall
- Jason Howard
- Marie Lauder
- Alessandro
- Logan Best
- Michael Donatz
- Jillian Murphy
- Willem Boning
I've loosly gathered points of conversation into topics, with some light annotation. Topics
- Volunteer Compensation
- Installation Procedures
- Onboarding Members
- Important Documents
- Equipmment Storage
- Core Tenets
- Meetup Structure
- Hardware
- Deployed Inventory Management
Contact Brian, or Willem, or perhaps post in #install-team for access to these docs:
- Install post-mortems
- Safety
- Ettiquette
Should volunteers be compensated? Yes; however, NYC Mesh cannot directly pay volunteers. The legal, insurance, and tax implications are prohibitive for us now. Instead we would like to:
- Give the installation fee to the Install Lead.
- Offer 2-way metro cards, Amazon gift cards, or something similar to the install participants.
- Swag (when we get around to making some)
What makes an install lead? Three installs, or when they feel comfortable enough with the install process to do it solo.
A suggested format for the training:
- First install: Shadow the install leader. Ask questions, help out, follow instructions.
- Second install: Take an active role in each of the steps.
- Third install: The install lead will shadow you. You take the lead on contacting the new Mesh member, sourcing equipment, and executing the install.
Differentiate between Mesh "Tourists" who we don't see again, from those who are invested. Can we prioritize install lead trainees in an equitable way?
- Have a pre-first install questionaire. With what exactly?
- Scale installers to size of install, but it's worth keeping the group smaller rather than invite everyone who needs training.
While not strictly related to installation, we talked about documenting/formalizing other ways people can contribute in maintaining and growing the mesh. We can lower the barrier to participation by describing how many of us already contribute to the mesh. This gives a role model for others to start from, and expand on.
This would include things like:
- Installation
- Software tools
- Firmware/hardware
- Marketing, outreach
- Mentorship
- etc...
We covered quite a bit of ground here... Ideally, these will be gathered into best practices documents as we go on more installs and refine these things.
How can we discourage late installers, or no shows?
Can we model our practices after other projects that go to people's homes and provide a service?
- Handy.com : They went from 10% to 2% and now with a goal of 0.2% no-shows
- Meals on wheels
Have a few days lead time, so we can arrange equipment.
Communicate the number of installers, and their names. Security and peace of mind is paramount. Try to arrive as a single group.
Main bottleneck in scheduling: surveys, lines of sight, and finding free install leads. We collected availability of the install leads present here.
Keep active communication in the days leading up to the install; this can promote attendance, or discourage last minute change in plans.
Design and share an installation plan: procedure, equipment list, primary/backup node connections
Set expectations for both the new member and the other installers. Checklists are great, so are expected outcomes. Take notes.
The signup form can be improved in a couple of ways. The first is reducing the friction for new users, the second is making sure we have the information we need.
A combination of redesign and tooling should work.
The form potentially requires more research/knowledge/context than new users might have. For instance, knowledge about rooftop access. Would it be better to have a simpler sign up form, just to get new members enganged? We can then get more specific with node details along the way. Shift the burden of research and knowledge from the initial contact, to when we've got them hooked.
We're running into some problems with Google's geo-coding. Parsing a street address including apartment number is difficult for Google. We also have some trouble with zip codes. We can either separate the input fields a bit, or tweak our geo-coding service.
How can we make equipment available to installers?
Given enough lead time: deliver equipment to the lead installer before install day.
We can also operate some caches around the city. Some requirements:
- Available on-demand: 24/7 and keypad (easier than making a bunch of keys)
- power outlets for charging equipment
- Located convenient to active install leads, or to hot neighborhoods
Marie offered space at her apartment. Additional ideas:
- Community centers (like Mayday)
- bars/restaurants
- Churches
- Coworking spaces
Better if any space we do use is also a Mesh member: an investment in the mission will help keep our equipment ready, and safe.
Replace NYCs ISPs with a non-profit, ethical, and engaged alternative.
Do we have a manifesto?
Beginning workshops are great. Let's have them more frequently. With that said, let's make sure we keep leveling up our game: let's get some more advanced training or collaboration going.
Make webinars to increase participation. Allow people to engage with the Mesh organization remotely and on their own schedule.
Meetups are a great space to trade equipment. It is also a great time to hunker down and get stuff done. Knock out some neighborhood planning with Google Earth, clean up and expand documentation. Do some member or community outreach.
Netbox is a tool that can help us answer the following questions: What equipment do we have deployed at a given node? What is its software configuration? Physical orientation?
This is a separate question from inventory management: What equipment and tools do we have?
A separate (but somewhat related) question still from network management.
Can every node be a hub? Theoretically, yes. However, it's the difference of ~$100. Additionally, some hub antennas (the omnis) have limited range; sector antennas are better suited for covering large areas.
Some talk of new point-to-point antennas. Didn't quite catch the details. Brian Hall can provide more details.
Who does hardware research? Everyone really. However, this is another example of silo'd, undocumented information. Currently, the best way to get involved in the hardware conversation is to find those who are having the conversation.
Have we considered hardware partnerships? Perhaps we can get volume discounts.