The Make automation that was slower than manual
A Make scenario that cross-posted Ghost articles to Dev.to looked efficient. API churn, broken canonicals, bad timing, and key exposure made manual faster.
My writings
Dispatches from a life spent online.
This week on richardlemon.com — July 24, 2026
Hi there,
Here's what went up on richardlemon.com this week:
A publishing system is not a writing
A tiny script that adds my internal links
A 40-line Python script adds internal links to my Ghost drafts via a shell alias. It is crude, fast, and only saves minutes, which is exactly why it works.
Content breakpoints for a single hero component
A category hero only started working once its breakpoints came from the heading, cards, and nav, not from tablet and phone widths.
Three things I delete from almost every AI draft
When I use AI for first drafts, I spend most of my editing time deleting: fake origin stories, forced AI angles, and tidy morals the model invented for me.
A publishing system is not a writing system
Automation can move your words into Ghost with perfect formatting. It cannot notice what mattered this week or decide whether it is true.
The practice that fell apart after ten minutes
A beautifully structured 90-minute U12 baseball practice fell apart in ten minutes. The fix was not a better spreadsheet, but a one-sentence pocket plan.
When I Stop Correcting a Young Player
How I decide when another coaching cue will help a young player and when it will only create noise, freeze them up, and block real learning.
Why my posts now start with five bullets
I stopped asking my automation to invent posts from topics and started feeding it five ugly bullets from real life instead. The system did not change much, but the writing did.
Why I Prototype Motion Before Polishing Layout
Static layouts can look perfect and still refuse to move. By prototyping motion early, spacing, hierarchy and components evolve around how the interface actually behaves.
What I Write Down During a Youth Baseball Game
During youth baseball games, I let the app track the score and use my notebook for the human data: faces, failed drills, invisible players, and one actionable line per kid.
A Month of Ignoring Daily Body-Fat Numbers
I hid the body-fat tile on my smart scale for 30 days and watched only a 7-day rolling weight average. The results were calmer mornings and steadier habits, not worse progress.