How to version artifacts like object storage buckets for backups?
"-a" "-b" "-c" is really short. It would benefit to be expanded to a full letter name...
...but every alphabet is charged with luggage and a specific culture.
I'd suggest to craft a new naming system rooted in Phoenician origins, yet distinct and neutral.
If we're drawing from the foundational imagery of ancient Phoenician signs—simple, universal symbols like the ox or house—we're reimagining them through the lens of phonetic innovation and cultural detachment.
Why the Phoenician? The Phoenician alphabet (c. 1050 BCE) is the common ancestor of Greek, Latin, Hebrew, Arabic, Cyrillic, and most alphabets in use today. It's a shared human heritage that belongs to no living nation-state or political movement.
By starting from that root and deviating into new coinages, we get names that feel ancient and grounded without inheriting any modern cultural baggage.
Names of the letter should be easy to memorize, distinct so we can use them as spelling alphabet too (Think "Amaris Tavolan Celor" instead of "alpha tango charlie")
A would be Amaris for example.
GPT-5.6 Sol and Qwen3.8-Max have been involved in that construction.
Full table:
| Letter | Name | Pronunciation | Evokes | Status |
|---|---|---|---|---|
| A | Amaris | ah-MAH-ris | the open sky | frozen |
| B | Bolaran | boh-LAH-ran | the flat stone | keep |
| C | Celor | CHEH-lor | the hollow reed | keep |
| D | Darak | DAH-rak | the worn path | frozen |
| E | Evar | EH-var | the high peak | keep — accepted weak token |
| F | Faloren | FAH-loh-ren | the drifting leaf | keep |
| G | Gorivan | GOH-rih-van | the rolling wave | keep |
| H | Hakat | HAH-kat | the open wind | keep — revised |
| I | Inar | EE-nar | the drawn thread | keep |
| J | Javik | JAH-vik | the young shoot | keep |
| K | Kerin | KEH-rin | the deep core | keep |
| L | Lumar | LOO-mar | the deep water | keep |
| M | Mavon | MAH-von | the standing pillar | keep |
| N | Noru | NOH-roo | the running stream | frozen |
| O | Orek | OH-rek | the seeing eye | frozen |
| P | Porel | POH-rel | the open palm | keep |
| Q | Quolar | KWOH-lar | the bending hook | frozen |
| R | Rovas | ROH-vas | the strong current | keep |
| S | Siron | SEE-ron | the scattered seed | keep |
| T | Tavolan | TAH-voh-lan | the carved mark | frozen |
| U | Utan | OO-tan | the hollow cave | keep |
| V | Vakor | VAH-kor | the climbing vine | frozen |
| W | Wolun | WOH-lun | the deep well | keep — revised |
| X | Xekal | KSEH-kal | the river ford | frozen |
| Y | Yarin | YAH-rin | the spring source | keep |
| Z | Zorev | ZOH-rev | the sharp thorn | keep |
Number or order
For the real table, we have the choice between that Latin order, or to go back to the Phoenician one:
' / a, b, g, d, h, w / v, z, ḥ, ṭ, y, k, l, m, n, s, ' (o), p, ṣ, q, r, š / sh, t
If we take that 22 letters, we can extend to f, j / dʒ, x / ks, c / tʃ.
How to count to 4?
If we need 4 instances, we would get:
Or we could get, like in latin/greek/phoenician (think alpha beta gamma delta epsilon):
I was wondering how to handle collation if we chose the second system.
Seems it's not a problem for PostgreSQL:
CREATE COLLATION phoenician_order ( provider = icu, locale = 'und', rules = $$ &A < B < G < D < H < W < Z < C ... $$ );