Marketing site for an Open edX instance

[Pierre Mailhot]

Hello everyone,

Here at EDUlib, we want to build a minimal open source "marketing site" on top of the Open edX catalog which is available with the LMS. Unfortunately, not everybody has the ressources (staff, money and time) to build a new front-end.

For example, what we need is something similar to a minimal clone of the https://www.edx.org site (see also https://github.com/edx/marketing-site or edx-mktg) but built on Django and based on the architectural vision of Clinton Blackburn, i.e. with all the information coming from the course discovery service (see https://github.com/edx/course-discovery).

Has anyone in the open source community started a project like that?

If yes, can you share what you are doing?

If no, who would be interested in sharing the efforts in making such a "marketing site" available to the open source community?

What would be a good starting point?

Is the work Clinton has done last year would be a good starting points? See https://github.com/edx/marketing-site.

Any ideas?

Any suggestions?

Anyone interested?

Thanks in advance for any ideas, suggestions or comments.

[Clinton Blackburn]

I'm excited folks are interested in my prototypes!

The work done in https://github.com/edx/marketing-site is a Django/Wagtail-based CMS. I initially went down that path because it was Python/Django and offered the ability to export data to S3 for a static site.

Since that time I have also done some work with Node.js and Lambda. This initially began as an effort to do server-side rendering of React (hence the use of Node.js), but that made less sense to me as I continued my investigation. I blogged about this work at https://engineering.edx.org/serverless-984cee7797e1. There still seems to be a push toward server-side rendered React at edX, so we will most likely go down this path for pages that can be powered by the Discovery API—course, instructor, subject, school, search—and continue using Drupal for fully-customizable pages such as the homepage and landing pages.

I'm happy to consult on either approach if someone is interested in leading the efforts. In the interest of full transparency, this is an area where edX.org probably won't be leading the efforts. All marketing site changes on our end will be a gradual evolution toward a hybrid between Drupal and some other static marketing site (most likely powered by Node.js to support React). We simply rely on Drupal too much to make a large shift. I would not be surprised if Drupal is still in use a year from now.

Clinton

[Pierre Mailhot]

Clinton,

Any reason why you chose Wagtail instead of Django-CMS or Mezzanine for example? Is it just because of the ability to export data to S3 for a static site.

[Clinton Blackburn]

The static site, facilitated by wagtail-bakery, was a factor. Also, I wanted the ability to create completely custom pages outside of the CMS to avoid having to sync information between the CMS and Discovery (which is a pain point for our Drupal implementation). It's been a while, but I vaguely remember Django CMS and Mezzanine not really allowing that behavior.

One item I don't think I addressed with Wagtail is i18n. I don't like the idea of duplicating fields for each language, but didn't explore using wagtail-modeltranslation. You can learn more at http://docs.wagtail.io/en/v1.13/advanced_topics/i18n/index.html.

I'm happy to discuss the advantages/disadvantages of all the CMSs. It would also be worthwhile if we gathered a list of features and categorized them as must-have, nice-to-have, and wont-do (if any).

Clinton Blackburn

--
You received this message because you are subscribed to a topic in the Google Groups "General Open edX discussion" group.
To view this discussion on the web visit https://groups.google.com/d/msgid/edx-code/096dfd39-d2e0-4ff3-b9a6-e7b2aa7a83a4%40googlegroups.com.