This thread is to brainstorm talk topics for the 2024 Open edX conference. Please share all of the ideas you can think of - even if you won’t be the one who presents it. The idea is to build a big pool of topics, so our team can pick the very best ideas to submit as talk proposals.
I will be creating time-boxed tasks for each of us to start brainstorming talk topics and posting them here. You can browse similar threads from previous years at the following links: 2023, 2022, 2020. Feel free to start suggesting topics anytime from now, if your schedule allows it.
Once we have a list of ideas, we can start working on scheduling tasks for submitting proposals. The deadline for submitting talk proposals is January 17th, so the sooner we can start collecting ideas, the better!
P.s. Other community members who read this forum are also welcome to share their ideas!
One idea I have been thinking about is either creating a Teacher Assitant Xblock leveraging the OpenAI GPT API or enhancing existing Xblocks such as Problem Builder.
In the following image, we can see that it is straightforward for ChatGPT to learn the multiple-choice problem format and assist with writing new questions while respecting that format.
You may have already seen the email from Eden about early submission of talk proposals:
The Call for Speakers is open now until January 17th, 2024. However, if you submit your presentation proposal by December 13th, 2023, you will seek the benefit of having early feedback related to your presentation content, in order to optimize your presentation submission.
This is a great opportunity! Getting feedback on our ideas, and the chance to revise and resubmit our proposals increases the chance of our talks being accepted . With that in mind, please submit your ideas on this thread before December 1st. Once your idea has been given the go-ahead by our team, please submit the first draft of your proposal before December 13th.
You can see more detail in the tasks I have scheduled for each cell (with sub-tasks for each team member) in the upcoming sprint:
Handling OpenEDX instances at scale – what it means for an operator to host and manage 10s of instances – opportunity to introduce Harmony (and Grove)
User Experience Design and Interface Optimization in OpenEDX – I’m probably not the right person to do this talk, but I would be interested for sure! (maybe @Ali or @cassie? finds inspiration in this)
Customizing MFEs using Tutor – again, I don’t have the context for this, but I’ve seen many MFE related questions on the forum lately
GitLab’s Minimal Viable Change approach to product design has a lot of benefits - not only for designers and developers, but for users as well. I would like to investigate how we could borrow some of their processes to improve the current approach to product design in the Open edX community (I think an MVC approach would have saved a lot of grey hairs during the Content Tagging MVP!)
IDEA 2
Cognitive Load and UX: Simplifying Complexity
Examining how UX design can mitigate cognitive load for a more effective learning process. We could perhaps contrast a well-designed course and a poorly-designed course (removing any identifiable logos etc)
@antoviaque I was wondering how “approval” works for the talk ideas posted here. Will you be commenting on everyone’s ideas, letting them know which idea they should submit a proposal for?
FYI: We’ve asked everyone to submit their ideas by Friday, 1 December (that’s in 2 days’ time). I will send out a reminder tomorrow.
IDEA 1 Building a Collaborative Learner Culture with Better UX
Explore how UX design can foster better collaboration among online learners
Showcase examples of features that promote interaction and discussion within the online learning platform. Interweave the work we’re doing on Graded Discussions
IDEA 2 HMX Custom Pathways: Navigating the Journey from Product Spec to Functional Product
A deep dive into how we managed the HMX Custom Pathways Project. While we’re only in the beginning stages of the process, this could be a good case study showcasing how OpenCraft managed and shipped a new feature
From crafting the product specification, to UX and UI design, to seamless development, we can explain how our team brought the vision to life. This shows how OpenCraft engages in a thoughtful product process
I think a comparison of Open Source LMSes would be a good idea. Most of us spend all day in Open edX and aren’t familiar with the likes of Moodle and Canvas. It would be good to have a presentation on some of these differences, like:
Features
Community composition/focus
Technical Requirements
Ease of Customization
Ease of Deployment and Maintenance
Understanding the differences-- including where we shine, and where our competitors have a better niche, can be helpful in understanding where we might improve the platform, and can help in determining cases where maybe Open edX isn’t the best option for a team-- which we should be honest about when we encounter them, lest we create grumbling from teams who are convinced to use Open edX and sink time and effort into it when it’s not the tool best suited for their needs/budget for their needs (yet). That can sour public relations.
@Cef That could be a good idea - but for talk proposals it’s better to talk about things we have done, seen, experienced, etc. rather than describe feature requests. It’s more likely to be accepted, and probably more interesting to the audience than something which could exist - but doesn’t yet. We also wouldn’t have time or budget to develop specific new things for the conference, while we do have a lot of things we work on already as part of our Open edX project, which many might not know about, or not much.
Since we have to condense a bit the approval this year to be able to submit something by the end of the sprint, I’ll try to review all the proposals early next week - likely Tuesday on the first day of my upcoming sprint. I’ll comment in the meantime to try to guide a bit the proposals I see going off; and to keep things simple, I can simply put a “like” on the posts where all the proposals could work?
Btw - for core contributors, don’t forget to count the time spent for conference stuff/prep as “core contributor”.
@jill@gabor Your proposals look good! Just a comment about some of them:
They are great topic, but I would try to avoid presenting them as a workshop (or being requalified as one) - they are a lot of work, with a very small audience.
@Cef Just want to add that instead of depending on chatgpt or some external provider, one can make use of Ollama or similar tool to run a model like mistral or similar general llm and provide assistance.
Harmony project: running Open edX instances at scale – an overview of what production-ready deployments look like, what problems you might run into, features that are missing in Tutor, introduction to the Harmony project…
Dealing with large django projects and breaking down them down to be modular multi-app and even multi-project components that are imported in. Talk about the benefits of it, separation of concerns, development can progress in parallel, well tested plug-n-play modules. Probably mention the current state/progress on the migration path towards that. An example of this is the openedx-learning repo and how it is handling the tagging functionality
How investing time upfront in developing sample data generation that can be imported into local dev and sandbox environment, pays dividends in speeding up developing/testing new features and collaboration across the team. An example of this is how we developed the sample-taxonomy-data repo and how its being used by the whole team for continuous development and testing of new content taxonomy/tagging features.
These ideas are mostly based on our experience in the last year, and problems that we faced and solved. Some of them are the topics I would love to know more about, but not much is out there, or not easy to digest from the perspective of developer/client infra maintainer.
Grove - anything, since we are heavily relying on it for our clients, we have a lot of experience.
Could be a workshop where we go through setting up a grove cluster (how do we make sure it fits into 90 minutes?).
Day to day management; tips and tricks.
How Grove works, breakdown of components and their relationships to each other, future changes (e.g. Harmony)?
tutor’s single file plugins - lightning talk about this little know but sometimes very useful and convenient feature; can provide examples of how we use it.
Multi-site setup on tutor (and grove), i.e. eox-tenant - basic setup, as well as more advanced features that are not well known, like filtering the courses by org, language, theming etc. Heavily inspired by one of the client’s setup. Would require help from someone who knows a lot about it, like @navin or @kaustav. Could be a 45 min talk. Could also be a workshop, but hard to set up so that everyone can participate, and relies a lot on external factors - grove cluster and instances, GitLab pipelines, etc.
Existing and future options for theming MFEs - @kshitij is probably the most competent to talk about it.
MFE Config API
Roles in Open edX; adding new roles - many clients want more granular control of the roles, but afaik, it’s not currently possible; however, we (mainly @demid and @piotr) were able to introduce new roles for clients.
h5pxblock - we went through a lot of troubles getting it to work for our clients, so might share what we’ve learned and changed to make it work. Would be a very short talk though.
Certificates in Open edX - not much documentation and it is very vague. You learn the most by trial and error and by reading the source code.
OARS - one of the new clients wants specific analytics data. Insights is no longer maintained (for some time now), but as a developer/maintainer, I don’t know enough about OARS (even after watching the @jill’s great talk) to know what it is, how to work with it, how to set it up for the client, and what data I can pull and display for them. In other words, more practical and detailed presentation.
Modular Learning - catch term for more flexible course paths and composable units. A few clients in this and last years wanted very custom experience, which was not possible back then (without huge investments into development). I wonder what’s new and what is possible now or will be in the near future.
Importing courses from old instances - we’ve done it quite a lot recently. We encountered and solved different issues, so could share valuable experience.
XBlocks for coding problems (or lack of them :)) - there is currently no easy answer for that.
The only thing I could think of that hasn’t been mentioned yet is the plugin for cloning courses that @pooja and @paulo built, which might be relevant for people interested in content reuse. It’s a small tool, though, and probably wouldn’t provide enough material for more than a lightning talk.
. With that in mind, please submit your ideas on this thread before December 1st. Once your idea has been given the go-ahead by our team, please submit the first draft of your proposal before December 13th.