3 ms·

I've always been a fan of PostgreSQL- We've seen the pain that mysql rebuilds can cause, with companies being down for hours. I've seen this in companies I've

by e1ven 19y ago

I've always been a fan of PostgreSQL- We've seen the pain that mysql rebuilds can cause, with companies being down for hours. I've seen this in companies I've worked at, and on public companies such as wikipedia. We've had good luck with psql internally. The speed isn't as much of an issue as some people had predicted, once we eliminated row counts. We're doing most of our processing on the server side, rather than using stored procedures, so we may have a different use case than most. Slony should be fine for replication for most people. There's also the heftier Mammoth replication, but that's for-pay. We've talked with Command Prompt (commandprompt.com) a few times, and they've been great at supporting pgsql- They have top notch pepole on staff, who know the details of slony setup, or tweaking db files. I believe they employ two slony devs on staff. The only real annoyance I've found with postgresql is capitalization. they're being faithful to the SQL standard, but they require that all column names are lower-case, unless you put each part in quotes. select * from "Product"."Table", where "Field", etc. It'd be great if there were a way to allow more permissive names throughout. In all though, I'd highly recommend it. -Colin

neilc 19y ago
WRT capitalization, lower-case identifiers are not actually required in PostgreSQL. The case-folding rules are: 1. unquoted identifiers are folded to lower-case 2. double-quoted identifiers have their case preserved This applies to all SQL queries, whether DDL or otherwise. Therefore, you can freely use mixed-case or upper-case identifiers, as long as you consistently either don't double-quote them, or always use the same case when you double-quote them. So it is perfectly feasible to use non-lower-case identifier names, you just need to be aware of the rules. In practice, most Postgres users just use lower case identifiers all the time, but that's mostly a matter of style I think. (FWIW, Postgres' case-folding behavior is actually one of the few areas where it is not conformant with the SQL standard.)
- e1ven 19y ago
  Wow, thanks for the info. I'm glad to learn when I've got the wrong idea- I must have misread their forum posts on why they did it that way. When we tried using mixed case without quotes it always converted it to lowercase internally- It reminds me a bit of OS X's file system handling, in how it worked- It accepted Mixed case, it just didn't care ;) I'll look more into it, thanks. -Colin

by e1ven 19y ago

neilc 19y ago

WRT capitalization, lower-case identifiers are not actually required in PostgreSQL. The case-folding rules are: 1. unquoted identifiers are folded to lower-case 2. double-quoted identifiers have their case preserved This applies to all SQL queries, whether DDL or otherwise. Therefore, you can freely use mixed-case or upper-case identifiers, as long as you consistently either don't double-quote them, or always use the same case when you double-quote them. So it is perfectly feasible to use non-lower-case identifier names, you just need to be aware of the rules. In practice, most Postgres users just use lower case identifiers all the time, but that's mostly a matter of style I think. (FWIW, Postgres' case-folding behavior is actually one of the few areas where it is not conformant with the SQL standard.)

e1ven 19y ago
Wow, thanks for the info. I'm glad to learn when I've got the wrong idea- I must have misread their forum posts on why they did it that way. When we tried using mixed case without quotes it always converted it to lowercase internally- It reminds me a bit of OS X's file system handling, in how it worked- It accepted Mixed case, it just didn't care ;) I'll look more into it, thanks. -Colin