Comments on: Waiting for 8.4 – case insensitive text ( citext ) http://www.depesz.com/index.php/2008/08/10/waiting-for-84-case-insensitive-text-citext/ Thu, 29 Jul 2010 21:40:44 +0000 hourly 1 http://wordpress.org/?v=3.0 By: depesz http://www.depesz.com/index.php/2008/08/10/waiting-for-84-case-insensitive-text-citext/comment-page-1/#comment-26505 depesz Mon, 11 Aug 2008 09:23:19 +0000 http://www.depesz.com/?p=1242#comment-26505 @Lukas: without a doubt it would be good. On the other hand - I just recently spent 3 hours trying to load a dump from mysql (to another mysql), and finally failed. Because of problems with setting collation/charset. It is most probably due to lack of knowledge of MySQL, but it looks very complicated. @Lukas:
without a doubt it would be good.

On the other hand – I just recently spent 3 hours trying to load a dump from mysql (to another mysql), and finally failed. Because of problems with setting collation/charset.

It is most probably due to lack of knowledge of MySQL, but it looks very complicated.

]]> By: Lukas http://www.depesz.com/index.php/2008/08/10/waiting-for-84-case-insensitive-text-citext/comment-page-1/#comment-26504 Lukas Mon, 11 Aug 2008 09:16:51 +0000 http://www.depesz.com/?p=1242#comment-26504 Nice addition. Would be nice if PostgreSQL could one day resolve this in a more "global" manner by allowing users to choose the charset/collation similar to MySQL, where one has the choice between binary, case sensitive and case insensitive. This works all the way down to the column level. One of the few areas where MySQL is actually ahead of PostgreSQL in features. Nice addition. Would be nice if PostgreSQL could one day resolve this in a more “global” manner by allowing users to choose the charset/collation similar to MySQL, where one has the choice between binary, case sensitive and case insensitive. This works all the way down to the column level. One of the few areas where MySQL is actually ahead of PostgreSQL in features.

]]> By: depesz http://www.depesz.com/index.php/2008/08/10/waiting-for-84-case-insensitive-text-citext/comment-page-1/#comment-26503 depesz Mon, 11 Aug 2008 04:37:14 +0000 http://www.depesz.com/?p=1242#comment-26503 @Theory: Sorry for keeping you waiting. I had a backtrack "thanks" to my vacation, but now I think it's all done (unless I missed something). As for compilation requirement. Well, when you're installing pg84 from binary packages (when it will be release), you'll just do apt-get install postgreqsl-contrib-8.4. If you're compiling from source - then you usually already know how to get contrib modules :) But the point is still valid - I should have mention compilation requirement. @Theory:
Sorry for keeping you waiting. I had a backtrack “thanks” to my vacation, but now I think it’s all done (unless I missed something).

As for compilation requirement.

Well, when you’re installing pg84 from binary packages (when it will be release), you’ll just do apt-get install postgreqsl-contrib-8.4.

If you’re compiling from source – then you usually already know how to get contrib modules :)

But the point is still valid – I should have mention compilation requirement.

]]> By: Theory http://www.depesz.com/index.php/2008/08/10/waiting-for-84-case-insensitive-text-citext/comment-page-1/#comment-26501 Theory Mon, 11 Aug 2008 01:46:13 +0000 http://www.depesz.com/?p=1242#comment-26501 Hey depesz, I've been waiting for you to blob citext. Thanks! I have just a couple of things to add: First, it needs to be compiled. That usually means `cd contrib/citext && gmake && gmake install`. Then you can install it in whatever database you like. I always put it into template1, myself, since I will be using it in damn near every app I write from here on in. The second thing to note is a caveat: citext is not a true cast-insensitive data type. Rather, it does exactly the same thing you do when you use `LOWER()` to get the same result. The downside is that it's slower than a true case-insensitive data type would be. The upside, however, is that you can now create case-insensitive primary keys and constraints without having two indexes! And it's a wee bit faster than doing `LOWER()` manually, as we've been having to do for lo these many years. Thanks for the coverage, I appreciate it! —Theory Hey depesz, I’ve been waiting for you to blob citext. Thanks! I have just a couple of things to add:

First, it needs to be compiled. That usually means `cd contrib/citext && gmake && gmake install`. Then you can install it in whatever database you like. I always put it into template1, myself, since I will be using it in damn near every app I write from here on in.

The second thing to note is a caveat: citext is not a true cast-insensitive data type. Rather, it does exactly the same thing you do when you use `LOWER()` to get the same result. The downside is that it’s slower than a true case-insensitive data type would be. The upside, however, is that you can now create case-insensitive primary keys and constraints without having two indexes! And it’s a wee bit faster than doing `LOWER()` manually, as we’ve been having to do for lo these many years.

Thanks for the coverage, I appreciate it!

—Theory

]]>