What happens if we run out of letters in the alphabet?
Nu and xi were skipped, the latter apparently for political reasons…
What happens if we run out of letters in the alphabet?
Nu and xi were skipped, the latter apparently for political reasons…
nu was too close to “new” and xi is too common a last name, and the WHO don’t want to put that on people
I would say that intentionally avoiding insulting a segment of the population is political; that is why we call it political correctness…
Since the initial political decision to use the greek alphabet instead of the name of the place of origin of a variant, did they not forsee xi or did they not think we would go through so many of them?
Why not pick some other famous ordered sequence such as the names of elements or the IUPAC alkane prefixes?
Call it SARS-CoV-2 variant: B.1.1.529 if you’d rather. It’s more developer friendly.
You’re wasting bytes. And that’s apparently really bad, at least judging by the table and field names chosen by virtually every DBA I’ve ever had the misfortune to work with. You get bonus points for obfuscation though.
Nah, it’s just the essentials. What project, what fork, what commit.
Sounds like you’ve dealt with AS400 databases before.
One of the major DBs at work stores phone numbers in 3 different fields - area code + exchange + last 4. The actual phone number is derived : [area code * 10,000,000] + [exchange * 10,000] + last 4. The table diskspace savings made a degree of sense … in the 1980s … nowadays it’s just dumb.
Yep. And nowadays treating number-like strings as numbers is idiotic. A zip code is not a number, nor is a credit card “number”.
My guess was that type of atomization of information/normalization had as much to do with query efficiency as well, e.g. extracting all phone numbers in an area code or local exchange (when those things were more closely tied to geographic location). Agree that nowadays it’s silly.
With the exception that, for security considerations, credit card numbers contain check digits which are numerically calculated on other elements of the card number. But still, with today’s cheap computing (CPU) and storage resources, the real TCO driver is analyst/developer/programmer time and effort, especially in the maintenance of the life cycle.
They do contain check digits, but they are calculated by considering each digit of the credit card number as a single byte/number. You can save a few bytes by storing a credit card number as a number but it buys you nothing toward calculating that check digit.
In the 1980s telephone number provisioning automation was both Hard To Do™ and a tremendously rigid process. There was one area code, the switch had only so many exchange codes - generally doled out in a largely deterministic fashion - and the last 4 were basically allocated on a next available basis. A business might be able to cough up happy happy money for a custom phone number - call Joe’s Carpet Cleaning at 214-JOE-CRPT - if it fit into the available ‘address space’ or could be welded into place somehow, but that’s about the limit.
The DB structure then could refer to the switch, pull the method to assign exchange, then the last 4 assigned to the subscriber.
@jast I love the titles you choose when splitting threads out. Cheeky and funny all at once. 10/10.
I swear this place is filled with a bunch of nerds for some reason
Because pentadecovid is much more clear to the average person than omicron is…
Heard on the radio today:
omicron is an anagram of moronic.
Heard on the TV today:
omnicron - President Joe Biden
He may have had better luck pronouncing pentadecovid.
Omnicron, that is the name of my next band - me
For some reason, omnicron bothers me much less than nucular.
With the exception that, for security considerations, credit card numbers contain check digits which are numerically calculated on other elements of the card number. But still, with today’s cheap computing (CPU) and storage resources, the real TCO driver is analyst/developer/programmer time and effort, especially in the maintenance of the life cycle.
Credit card processing is still largely done in COBOL, a programming language which distinguishes numeric (picture 9) content in a data field.
For some reason, omnicron bothers me much less than nucular.
It bothers me less than Hamberders.
It doesn’t matter what name or naming convention was chosen; people will see what they want to see in it.
Interesting space saving discussion. I flashed back to S/370 packed decimal format. Space was so very very tight back in the early days!