DataTalksClub / DataTalksClub/website
No way to generate article, podcast and event assets in the current style
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 0
- Forks
- 0
- PR merge metrics
- No merged PRs in 30d
Description
Articles, podcast episodes and events all carry imagery, and there is no repeatable way to produce it in the site's current visual style. Courses have one — _docs/design/illustration-assets.md describes generating light and dark variants through Codex's imagegen — but nothing covers the three editorial content types.
This blocks production readiness: content published after launch will either ship without imagery or with imagery that does not match the site.
Why now
- Podcast covers are being replaced wholesale. 213 covers in
content/public_projection/media/podcast/are deliberately excluded from the CDN migration because new ones are coming. Something has to make them. - Media keys are becoming ours. The CDN key is assigned at ingest rather than inherited from the upstream filename, so generated assets need a defined place in that pipeline rather than being dropped into a directory.
- The three content types are moving to database-backed ingest, synced from a push. A generator that writes files into a repository will not fit that shape without thought.
Current media, for scale
| prefix | files | |
|---|---|---|
authors/ |
438 | people faces — carried across |
posts/ |
407 | blog illustrations — carried across |
books/ |
196 | book covers — carried across |
podcast/ |
213 | being regenerated |
What "done" looks like
- A written procedure, in the same spirit as
_docs/design/illustration-assets.md, covering articles, podcast episodes and events - Output matches the current design system (
_docs/design/design-system.md) — not the pre-redesign style - Light and dark variants where the surface needs them, as courses already do
- Dimensions, aspect ratios and formats defined per content type, so a generated asset drops into the existing layouts without per-item adjustment
- No ImageMagick for dark-mode derivation — dark variants are generated, not filtered. Cropping is the only sanctioned image-tool use
- A stated path from "asset generated" to "asset in the CDN under its assigned key", so this fits the ingest pipeline rather than sitting beside it
- Reproducible by someone who did not write it — a prompt or procedure recorded, not a one-off
Related
- #309 — production hosting and DNS migration checklist (this is listed there as a blocker)
- #310 — data ingest consolidation
- #301 — serve public projection media from object storage
_docs/runbooks/production-data-migration.md— step 8 covers media into the CDN_docs/design/illustration-assets.md— the existing course-illustration procedure this should extend
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with _docs/design/illustration-assets.md and _docs/design/design-system.md, then read _docs/runbooks/production-data-migration.md step 8 and the related ingest issues #310 and #301. Define a reproducible procedure for article, podcast and event assets, including variants, dimensions, formats, and the path from generation to the assigned CDN key.
Written by the indexing model from the issue text.
Assessment
- Domain
- cloud, devops, documentation
- Issue type
- Documentation
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 48/100