Re: Is Oracle is CASE SENSITIVE

Well, this thread is almost dead, but I'll jump in anyway, because I couldn't let this one ride for free. <g>

There is a well-defined methods for doing case-insensitive comparisons. Each character is assigned a collation number, so the comparison is done on the string of collation numbers. In caseinsensitive  comparisons in the Latin alphabet, 'a' and 'A' have the same collation number. Nothing hard about that. And don't worry about the performance. The translation is a very fast lookup, e.g., if the character has numeric value 55, then look in the 55th position of the collation table to get its collation value. Some (most?) processors can convert an entire string in a single instruction (a "translate" instruction).

I have to say, the case-sensitive thing seems awfully limiting and - dare I say it? - old fashioned to me. I don't see how anybody who has had to define a second column to hold the upper()ed or lower()ed data could argue that the limitation doesn't have its drawbacks. But I'm sure that we have here a good example of technology shaping solutions: if case-sensitivity is required, you learn to work with it until it seems natural and good. But choice is also good.

The next time you (the generic you) get angry that someone is asking about this, ask yourself why the question comes up so often. The answer is probably either:

1. People are stupid and lazy. If this is what you think, then you may as well get over it. People aren't going to change. The questions will keep coming.
2. The answer is not sufficiently publicized. Perhaps all the Oracle ads should state this in fine print. <g>

Another thing suggested by the many questions is that this is a feature that some people would appreciate. They shouldn't be put down for wanting a feature.

In article <8ucesp$ceqg$1_at_reader1.wxs.nl>,   "j.m.spit" <j.m.spit_at_uptime.nl> wrote:
> Hahahahahahhahahahahahahahahaha
>
> I will disclose one of the best known IT secrets (this is guru stuff):
> character glyphs are stored in the hardware as numbers. The ASCII
 table is
> one of the most widely used mappings between character glyphs and
 numbers,
> but certainly not the only one. If two character glyphs are not the
 same,
> the mapped numbers are not the same. This must obviously hold for all
> mappings. A computer cannot do magic. Its a computer, it operates on
> numbers. it stores numbers, not character glyphs.
>
> I just get sick and tired of people asking questions in this
 newsgroup that
> they can lookup easily themselves.
>
> I also get sick and tired of people that attempt fundamental
 discussions on
> subjects they know nothing about.
>
> Finally, if u do know something about Oracle (in this case), don't
 answer
> silly questions. It will save those who really use this newsgroup the
> FUCKING MESSAGE HEADERS.
>
> (Sorry but this one really pissed me off :)
>
>

Sent via Deja.com http://www.deja.com/
Before you buy. Received on Tue Nov 14 2000 - 17:47:10 CST