Skip to content

Home

Jump to bottom
Chris Millar edited this page Jan 21, 2024 · 5 revisions

One Page

Abstract

Project Dark Alley is a research project. Its goal is to be a first-class content provider to AEM Edge Delivery Services. It aims to provide a fully integrated and authenticated experience for common tasks that are hyper-specific to WCMS needs.

While Microsoft SharePoint and Google Drive solve many common repository needs (compliance, authentication, authorization, versioning, collaboration, etc.) they are also not tailor made for the needs of AEM Edge Delivery Customers. When pushed to their limits, these existing content repositories provide a feeling of, "square peg in a round hole." Companies of a certain size with certain acceptance criteria may find that bolting on WCMS-friendly tools can feel cumbersome.

Screenshot 2023-11-28 at 10 13 48 PM

Dark Alley is

  1. An API to store docs, sheets, and media at scale.
  2. A UI for humans to create content.
  3. A content provider for AEM Edge Delivery Services.

Dark Alley is not

  1. A WYYSIWYG editor
  2. A markdown editor
  3. An SPA editor
  4. A SharePoint / Google Drive replacement.
  5. An AEM Assets replacement.
  6. Meant to host direct web traffic.

Dark Alley is inspired by

  1. AEM Edge Delivery Services - Serverless, Simple, Fast.
  2. Apache Sling - Where ever you POST, that's where your content will be. Wherever you GET, that's what you'll get.
  3. SharePoint & Word - The muscle memory of using Word is critical.

Dark Alley goals

  1. Provide a rich, performant, scalable, and stable API that eclipses what Microsoft Graph + SharePoint provides.
  2. Get feature parity with Word editing. Including muscle memory behaviors like keyboard shortcuts.
  3. Provide real-time content previews.
  4. Provide real-time collaboration.

Dark Alley principles

  1. No markdown. - Markdown is great for viewing and writing, but provides a poor developer experience when considering different flavors and the ubiquity of DOM APIs.
  2. Un-apologetically a document-based editor.
  3. The server should be as dumb as possible. The goal is to serve static (html) files with no manipulation.
  4. Build like it will last 20 years. Prototype like it won't last more than 20 days.
  5. If it doesn't work on mobile, it doesn't exist.

Dark Alley status

After much research, the Dark Alley POC is done. The next phase is to provide enough MVP features to entice a pilot customer there are enough benefits to make the switch.

  1. POC - Done
  2. Pilot - Not done
  3. Collect pilot feedback & iterate - Not done
  4. Security, privacy, etc. review. - Not done
  5. Customer #2. - Not done

Getting to the pilot

  1. Editor - still needs a tremendous amount of work
  2. IMS Authentication - DONE
  3. Browser - pivot to Web Components, support all common ops: move, delete, replace, etc.
  4. DAC / DAS - Better separation between the editor and the Worker - DONE

Research

  1. What type of content repository? Filesystem, Sling / AEM, MongoDB, Azure Blob Storage, S3-compatible storage.
  2. What editor? CKEditor, WYSIWYGHTML, TinyMCE, ProseMirror. - A relatively sane API with a lot of prior art, examples, plugins, collaborative features, comments, etc. You just have to build it all. Maybe most importantly is the UX is 100% customizable to provide an experience Adobe customers expect.
Clone this wiki locally