Notes from the workshop.
How the catalog is sourced, what gets checked before a quote ships, and what we’re building next.
Juzʾ ʿAmma is not 9% of the Qur’an
Almost every memorisation tracker counts verses. Juzʾ 30 has 564 of them and juzʾ 6 has 110, so counting verses tells someone who has finished Juzʾ ʿAmma that they know 9% of the Qur’an. They know 3.3%.
The shortest sūra to memorise is not the one with the fewest verses
Al-ʿAṣr has three āyāt. Al-Ikhlāṣ has four. Al-Ikhlāṣ is a third less text. If you are picking what to memorise next by verse count, you are picking with the wrong ruler.
Al-Mukhtaṣar or Ibn Kathīr: which tafsir to read first
One gives you a few sentences on every āyah in three languages. The other gives you several thousand words on a passage, in two. They answer different questions, and the numbers make the difference concrete.
- Español
Solo hay un tafsir en español
Revisamos las cuatro bibliotecas públicas de tafsir que existen. Entre las 108 obras que reúne QUL hay 55 en árabe, 5 en inglés y exactamente una en español. Esto es lo que significa para quien lee el Corán en castellano.
Arabic in next/og: which fonts work, and why yours probably doesn’t
Satori renders Arabic with unjoined letters, backwards word order, or throws outright — depending on the font. Here is what fails, what passes, and the two rules that fix it, measured over 90 real Arabic strings.
The slug rule that deleted a letter from the name of Allah
A URL slugifier that kept Arabic letters and separated on everything else looked correct and was not. It silently dropped alif wasla — turning ٱللَّهِ into لله on 102 of 250 rows.
Splitting a Next.js sitemap 404s the URL you submitted to Google
generateSitemaps() emits the parts and no index. Every part serves correctly, /sitemap.xml stops existing, and nothing in the build tells you — the only thing that breaks is the address Google is actually polling.