Talk:Comparison of relational database management systems/Archive 2006

A function by any other name

Barring any objection, I'm going to merge the Function and Procedure into a single column, as there seems to be no clear distinction made between the two for the purposes of this article, and it just makes the table that much wider without really adding more information. Turnstep 15:45, 7 April 2006 (UTC)[reply]

Add section on XML indexing and other native XML capabilities?

I'm only familiar with SQL Server in this area; can others fill in data for other DBMSs?

GiN

I removed the GiN column, on the grounds that no released DBMS actually implements this feature, even PostgreSQL. Even so, it is just an idea that a few Postgres hackers came up with, it's hardly a proper index type that other databases could reasonably be expected to implement. It might make sense to have a column for full-text indexing, however. Neilc 05:12, 28 August 2006 (UTC)[reply]

Hmmmm ... maybe one column on full-text indexes, and another on on "other" indexes? The Other column would have to be a list or a number rather than yes/no. In addition to GIN, there's GiST (and both of those are algorithms, so other databases could implement them) Spatial, R-Tree, XQuery, etc. Jberkus 19:47, 23 September 2006 (UTC)[reply]

Nobody else has done anything about this, so I'm going to clean it up per my suggestion above. It's silly to have an index type column supported by only one database. Of course, I'm going to need other people to fill in the exotic index types supported by DBs other than PostgreSQL, since I don't know them. Jberkus 07:29, 16 December 2006 (UTC)[reply]

Domain

Not a specialist myself, I'd like to know what the “domain” object is (in paragraph Other objects). Is it data domain? If yes, can anybody add a link to this article, or I can do it myself, as soon as I'm sure it's the right “domain”. --Olivier Debre 19:34, 24 September 2006 (UTC)[reply]

A domain is basically a simple user-defined type: it includes a base type and some optional associated constraints (e.g. NOT NULL, CHECK, foreign key). It's not the same as a "data domain", AFAICS. See CREATE DOMAIN for PG's syntax and some examples. Neilc 20:14, 24 September 2006 (UTC)[reply]

Thanks! --Olivier Debre 10:29, 25 September 2006 (UTC)[reply]

Strange Choice of Databases?

Currently, this page contains the following database systems which I've never heard of before (given 12 years as a DBA) and suspect of having miniscule-to-nonexistant user bases: Pyrrho DBMS, H2, OpenLink Virtuoso, SmallSQL, WX2, SUPRA SQL. On the other hand, it omits the following DBMSes which have had wide usage for years: PervasiveSQL, Progress, MS Access, Adabas and FrontBase (there's probably others, too). Are we aiming to include everything, or are there criteria for inclusion/exclusion? Jberkus 07:46, 16 December 2006 (UTC)[reply]

I think you are right. As well as I wish to ask for somebody to ask RDBMS Linter (Linter Offical Website (Japaneese), Relex Inc. Offical Site, RelexUS site). Information on Relex can be found on those sites. I would do it, but to find all of information would be a bit difficult, even if I do use it. 71.229.243.190 17:52, 22 December 2006 (UTC) (Vmatikov)[reply]