# _You are one of the hardest communities to find!_ (mcorr)

Developing a user-focused Pulp community strategy

## What Pulp users in 2020 are saying about Pulp

_"Pulp is the only open source binary repo I see has a potential to beat every other paid binary repo out there. You have a long way to go but I am very happy to use Pulp."_

_"So far the team has been doing a good job with the application I feel the workflow could be more intuitive for users"_

_"When I told DevOps folks that I was working on installing Pulp, they reacted very negatively. It seems that Pulp 2 has a really bad reputation with those that were required to use Katello. They said that all issues with Katello were traced back to Pulp 2. Some kind of campaign might be required to change the perception of Pulp 3."_

## What are the overall goals of a user-focused community strategy?

_<Please add/modify any that you can think of>_

* Growth in user base
* Improved user experience
* Quicker onboarding: Can a user go from learning about to installing Pulp within the same hour/afternoon/day? What would be our ideal and demonstrable goal?

As an aside for another day: How will we measure our growth?

## Areas of interest for a user-focused community strategy

_<Please add any others that you can think of>_

* Communication
* Usability

## Communication

### Challenges

<Please add/modify as you see fit>

* Our project name makes SEO hard
* Our mailing-list-focused method for info sharing doesn't help with SEO
* Our mailing list is reportedly hard to discover
* We're on Freenode IRC; potential user base may be elsewhere (Matrix/Telegram)
* Our changelogs can be cryptic
* We don't have opportunities to talk to people at conference booths
* Low levels of social media engagement

### Communications - What @mcorr has been trying

* Create a consistent cadence of content
* Make pulpproject.org a reliable and up-to-date source of information about Pulp
* Develop a network of FOSS blogs & contacts in which to distribute Pulp news and content
* Started a digestable monthly summary of events
* A pipeline of content to publish on opensource.com

#### Channels

|Plugin release & content |Avenue|
| --- | --- |
| Ansible | The Bullhorn - Ansible Community News
| RPM | CentOS Community Blog & Newsletter
| Python | Python Weekly News

I have reached out to the Debian community.
I'm trying to do this for all open source communities whose content we interact with.

### Channels of Communication

#### IRC & Mailing Lists

Should the Pulp team sign up to Matrix and open a range of channels and see where we can court new users?

Should we set up a Telegram channel?

What do we do about our mailing list?

### Improving Perceptions

Pulp has been around a while and has gone through much change.

How do we shift perception of Pulp away from Pulp 2 to Pulp 3?

Do we know which Red Hat FOSS communities have tried earlier versions of Pulp?

Should we think about some Red Hat focused information sharing?

* Community Central
* Open By Default
* OpenShift AMA

This week, it was mentioned that some internal folks who tried Pulp 1 or Pulp 2 still perceive Pulp as Pulp 1/2.

Should we think about some Red Hat focused information sharing?

What can we say or do to help shift this perception?

### What the Heck?

Would running something like this in the community be an idea?
could we use this as an opportunity to gather more high level user-scenarios we can use to shape further community docs/content?

https://weblog.rubyonrails.org/2020/5/7/A-May-of-WTFs/
https://www.home-assistant.io/blog/2020/08/18/the-month-of-what-the-heck/

## Usability

### Challenges

<Please add/modify as you see fit>

* Installation
* With flexibility comes complexity - how can we facilitate new users and allow them to grow into their Pulp usage?

### Documentation

A lot of effort has been put into docs.
58.8% of users reported finding it difficult to find information about Pulp.

From the survey:

_"Pulp 1.x docs still available but totally different from 2.x (and totally different again from 3.x). Non-response from mailing list. Lack of real world examples. Unhelpful error messages."_

* Error messages are important - in IBM actionable error messages significantly reduced the cost of support
* how do we gather and integrate more real-world examples?

_"Documentation sometimes is difficult to find for the required plugin"_

_"for me it was very hard to get started with the shell script and httpie stuff. could be easier for newbies"_

_"The API docs are too vague in places. "String" does not tell me if there are character restrictions."_

How to we measure whether docs serves users well? Direct documentation feedback.

Topic-based writing - task, concept, reference (as a follow-up to Mike D's comments on man pages )

How can we test the documentation?

How do we capture more meaningful workflows?

### CLI

From the survey:

* All Pulp 2 users expressed the desire to have a Pulp 3 CLI
* 50% of Pulp 2 users cited no CLI as a blocker to migrating to Pulp 3
* 44% of Pulp 3 users cited no CLI adds to the difficulty of using Pulp 3

### Web UI

70% of Pulp users reported that they do not need a web UI