# CC@Home: Mega CodeRefinery online workshop - How to scale up online workshop to reach 100 learners?
## Code of Conduct
All hosts, speakers, and attendees of the CarpentryCon@Home are expected to adhere to the Carpentries Code of Conduct.
* Use welcoming and inclusive language
* Be respectful of different viewpoints and experiences
* Gracefully accept constructive criticism
* Focus on what is best for the community
* Show courtesy and respect towards other community members
CoC facilitator: Anne Fouilloux, a.c.fouilloux@geo.uio.no
For more details, please refer to [Carpentries Code of Conduct](https://docs.carpentries.org/topic_folders/policies/code-of-conduct.html)
## Presentation slides
<https://cicero.xyz/v3/remark/0.14.0/github.com/coderefinery/carpentrycon-2020/master/slides.md/>
## How to edit in HackMD

If you see this page in view mode like the screen shot above, do either of the following to enable editing.
1. **Edit mode**: Ctrl+Alt+E, or click the button with a pencil on the left-top to swich to edit mode
2. **Both mode**: Ctrl+Alt+B, or click the button with a square with a vertical line in the middle
## Roll call + Check in
* name / email / pronoun / social media handle
* Naoe Tatara / naoe.tatara@ub.uio.no / she, her / Tw: @NaoeTatara, GH: @naoe-tatara
* Paul Harrison / paul.harrison@monash.edu / they|he / Tw: @paulfharrison GH: @pfh
* Radovan Bast / radovan.bast@uit.no / he,him / Tw: @__radovan, GH: @bast
* Anne Fouilloux / annefou@geo.uio.no / she, her, hers / Tw: @AnneFouilloux
* Matúš Kalaš / matus.kalas@uib.no / he, his / Tw, GH: @matuskalas
* Petri Jehkonen / petri.jehkonen@aalto.fi / he / Tw:@jehkonen GH:@pjehkonen
* Thor Wikfeldt / kthw@kth.se / he, his / @KTWikfeldt
* Thomas Arildsen / tari@its.aau.dk / he, his / Tw:@ThomasArildsen GH:@ThomasA
* Richard Darst / Aalto University
* Serah Rono / serah@carpentries.org / she, her, hers / Tw: @serahrono GH: @serahrono
* David Pérez-Suárez / dps.helio@gmail.com / he,him,his / 🐙 @dpshelio - 🐦 @DVDGC13 - 🐘 @dvdgc13@octodon.social
* Renato Alves / renato.alves@embl.de / he,him / tw: [@renato_alvs](https://twitter.com/renato_alvs) gh: [@unode](https://github.com/unode)
* Samantha Wittke / samantha.wittke@aalto.fi / she, her / @swittke1
* Rachel McKay / rmckay@usc.edu.au / she, her / University of Sunshine Coast, Australia
* Clark Gaylord / cgaylord@vt.edu / he / Virginia Tech and George Washington University
* Raphaela Heil / raphaela.heil@it.uu.se / she, her/ @RaphaelaHeil
* John Shepherdson/john.shepherdson@cessda.eu/he:him/jwshep.ukda (Skype)
* BBBBBBBBBBenson Muite
## Ice breaker question
### What’s the largest you have taught online?
Please write "+" on the option that is applicable below.
1. <10 learners: +++
2. <20 learners:
3. <30 learners: ++
4. <50 learners: +
5. <100 learners: +
6. '>= 100 learners: +
## Introduction
## About CodeRefinery
* Ask questions like this
* You'll get an answers and discussion like this
* What can we do to bridge the gap between beginner and intermediate lessons? how can Carpentries start developing intermediate and advanced resources? any advice...
* How to have bespoke learning path from beginner to expert?
*
### Questions to audience
*
#### Is there a need for more intermediate/advanced Carpentry-style material for learners who have completed a basic Carpentries training?
* Yes! :) (I've seen some lessons in the incubator that fills some gaps, but CodeRefinery is more structured to fill that gap)
* yes!community members that have been in The Carpentries for a long time (instructors and trainers) could benef
* Yes - there are many beginners who are only striving for "awareness" but learners who want to become producers need to go beyond the Carpentries/novice-level of expertise
#### What topics/courses are needed?
#### What keeps our material from being used already?
### A vision of reaching many people at once
How can large be more inclusive? How is it less?
* Room 1:
* key difference: with ~70 participants we need a lot more staff
* used many breakout rooms to prevent it from seeming crowded
* a lot more verbosity is needed to connect things that an instructor would otherwise say
* is it possible to design materials to satisfy both universes?
* Large groups, sending information works well and is preferable. breakout rooms work best...
* more interactivity by using breakout rooms often, but still some people may not be heard
* Technical obstacles: mostly social or how things are arranged.
* Room2:
* More inclusive more people to attend
* Less inclusive makes it more difficult to be heard. Easier to get behind
* When you have small group breakouts for help, you lose track of the main screen. Would be good to be able to multi-stream.
* It is a good idea to have a designated helper in the workshop who keeps an eye on the class chat and interrupts the instructor when necessary. This means that the instructor does not have to keep an eye on the chat while presenting.
* Room 3:
* Be able to see everybody on the screen
* Challenge is to make learners comfortable in a digital environment
* More difficult to get enough interaction going in online environment
* Online workshops made it possible to access material for participants from distant campus locations
* Barrier is possibly lower for those who are curious but unsure whether they can commit to a travel and full workshop
* Recordings are appreciated
* Less feedback for the instructor who "talks into the void"
* Room 4
* Inclusive
* room for more to attend
* tool you can use: person doing the presentation should appoint another to keep an eye on the chat and alert the instructor if there are questions or other input from the audience
* q: how do you solve the instructor and learner interaction without too much interruption? A: Have a dedicated helper who serves as a filter (or takes / answers questions)
* q: which channel is best to manage learner / instructor interaction in an online setting? A: the helper / goalie has power to interrupt instructor verbally, but other learners have to go through the goalie
* Less inclusive
* harder to be heard / get the attention of instructors and others in the room
* easier for people in a larger workshop to get lost / left behind in the lessons
* difficult to read the room / read expressions as people do not enable videos. Non-verbal cues are lost. From observation in a 75+ learner session, in larger rooms, people typically have their videos off
* consider letting people know there are no expectations ( it is ok to join in from wherever :) )
* consider using breakout rooms often
* breakout rooms are great for learning, but problematic for trouble shooting, as the learner in the breakout room misses the main session, and remains in a constant state of being left behind
* harder for helper to gauge urgency / need for help by learners in a large room i.e. is it needed immediately or can it wait?
* Room 5:
* Less inclusive:
* instructor will have more difficulties to interact with learners (and check what is done by learners)
* more technical problems (hackMD)
* More inclusive:
* if there are enough helpers/co-instructors
* broader range of people but can be divided as groups only
* can have people from multiple countries at once
What technical obstacles does it have?
* Room 1
* Zoom has proved reliable but has questionable data practices
* Timeouts keeps us always on time - regardless of consequences :D
* No need to keep class in sync if run in a self-learning format
* capacity to go up to high number of participants
* possible issues with privacy and data collection, fact that it's run by company bothers some people/organizations
* jitsi doesn't have breakout rooms (yet!, stay tuned!)
* BigBlueButton is an alternative
* HackMD or the open source alternative CodiMD. The latter can be self-hosted.
* ...
* Room 2:
* A tool that you can have helpers to keep an eye on the chat. So you can bring on the audience.
* A designated helper to filter the "noise" and interrupt the instructor if necessary.
* Don't be able to see the people, larger rooms tend to disconnect the cameras, so you don't get the visual feedback.
* Problem with breakouts for questions can be challenging as the class is moving on.
*
## How we did it
* Ask questions like this
* You'll get an answers and discussion like this
* How did you check if helpers were suitable?
* there wasn't any vetting of helpers, maybe it would be good to somehow check this in advance
* there was some negative feedback from people in some breakout rooms regarding the helper, so this is actually important
* another possiblity is to rotate helpers on breakout rooms so the same helper doesn't always get assigned the same room
* None of us know every single thing perfectly, either...
* absolutely, the "expert helpers" were actually people who had taught the material many times before, but there are always questions one can't answer
* The main difference of "expert helpers" was they could take the time to go deeper into problems and call in the right person to solve things.
* And it is actually good if a helper did not know the answer to a question/problem right away but learners could follow the way how the (expert) helper figured it out
* Has anybody used Didact (https://github.com/redhat-developer/vscode-didact) to create hands-ons tutorials for machine configuration/setup? I haven't used it yet, but am starting to evaluate it.
* Have not, looks interesting +1
* Have you used it?
* Any problems with zoom-bombing so far?
* No we had waiting room and used twitch for anonymous learners
* If zoom URLs are not publicly announced this is a non-issue
* Have you felt at any point that recording the sessions impaired participation?
* [name=CR] I'm biased because I was an organizer, but I don't think it was any much more than having 100 people in a room already...
* maybe participation as in asking questions in the main room, but the hackmd helped a lot there
* +1 on 2 screens (or two computers). Definitely recommended for instructors.
* I have used one screen and a tablet
* I used one screen most of the time, but during the workshop I learned how to efficiently use the different desktops and switch between them :)
* how much time did the different roles used on preparing for the workshop? Also how much extra time compared to in-person workshop?
* [name=CR] i think the CodeRefinery staff didn't spend much more time compared to earlier non-mega workshops. We had already tried giving online workshops on a smaller scale in 2-3 earlier occasions, so the CR instructors and helpers had practiced online teaching
* It did take more time with the helper preparation meeting though, and all the organizing work with group-registration etc took a lot of time
* would be interesting to hear from helpers, how much time they spent on preparing
* [name=CR] (helper) I went through the whole material once before the course, also to see if there was anything new to me. Then I attended the helper preparation workshop and helped with installation session. Could have maybe spent more time in preparing for the exercises and learn a little bit about how things work on Windows (Linux user) -> few hours
* How much time took to prepare it? In terms of setting the whole thing, to advertise the event, get helpers, select the material, etc.
* [name=CR] I coordinated registration, and it was a lot of work, but looking back not *that* much more than before. I had to deal with teams and pre-assign breakout rooms, but that wasn't too bad. It spent more time figuring out what questions to ask when registring, what to write in emails, to prevent complete chaos. Now, it would be much faster.
* [name=CR] since there wasn't yet a shared vision of the mega-workshop, there was a lot of time we spent building a shared understanding of the vision. That would be much easier now. +1
* In terms of days/hours? First one ~month, future ~week. Less than in person in proportion to the people reached.
* How many people were live streaming via Twitch, and how did people find out about the session?
* there were around 15-25 people on average on twitch
* wasn't advertised much
* Do you have any demographics on streaming participants? Wondering if it covers something the other approaches don't.
* good question, i don't think we knew who the streaming participants were
* We basically advertised to the same audience that registered, so
* This was for a 2 full days workshop, right? Anything regarding breaks?
* 2 weeks, 3 half days/week. 9-12:00 CEST (UTC+2)
* if there was problems with exercises in the breakoutroom, sometimes the breaks became a bit short, so its important to make sure yourself also that you get a break
* I would be interested in taking part via Twitch as a learner depending on how bad the time difference is.
* great, welcome! we'll have (at least) 2 mega online workshops this autumn
* Do you feel streaming has advantages over recording (offline - post-processed) + live communication channels (chat, hackmd)? - For instance, to overcome timezone challenges...
* what is the best channel/information location for people interested in the great work you are doing to keep up to date?
* we have a newsletter: https://coderefinery.org/#stay-informed-and-connected
* for more involvement, sign up for our zulipchat! https://coderefinery.org/get-involved/
* thank you!
* note that there are many streams under this zulipchat and you can decide which ones you subscribe to
* There is also Twitter (but we can be more active) and these days we write more blogs
* Would or forum be effective?
* you mean a chat forum? hackmd really served us well for questions and answers
* Ok. asHave you used Zulip much as welOk.
* HackMD has Git integration for history, I think... could help with loss of content issues.
* we did run into the hackmd suddenly vanishing! it was restored from the in-built version control
* `[name=]` tags, as well as code syntax and line numbering are some of the reasons we chose CodiMD for a local deployment over etherpad.
* Good point. We haven't used CodiMD yet
* CodiMD is basically HackMD, but the open-source version. A few less features but mostly same experience. See [CodiMD features](https://demo.codimd.org/features) vs [HackMD features](https://hackmd.io/features)
* +1 for asynchronous communication via shared doc / HackMD
* Definitely!
### Feedback survey
https://carpentries.typeform.com/to/kdJ99sD3
One good thing about the presentation:
- ...
- ...
- It was fantastic to hear the experience from all the different levels of helpers, learners, instructors etc.
One thing that should be improved:
- ...
- I do like hearing experience/questions via voice sometimes, but can also see the value in using HackMD (no not really improvement just a comment)
-
---
Always ask at the bottom
If you want to mention name but want it to be deleted afterward, please use
`- [name=XYZ]` so that it shows like;
- [name=XYZ].