in reply to Re: PostgreSQL, Emacs, and other groupieware
in thread PostgreSQL, Emacs, and other groupieware
|
|---|
| Replies are listed 'Best First'. | |
|---|---|
|
Re^3: PostgreSQL, Emacs, and other groupieware
by BrowserUk (Patriarch) on Aug 01, 2004 at 21:11 UTC | |
Oh well. For almost anything that you can easily do with flat files, I'd strongly recommend them over an RDBMS anyway. Once pure flat files become limiting, using simple tied interfaces to files will buy you a lot of simplicity and give you access to structure. Especially, the most useful recommendation that I picked up from merlyn, which is DBM::Deep. By the time your application find the limitations of that (other than size), you'll probably be in need of a proper RDBMS anyway--but that's usually a way off at the beginning. | [reply] |
|
Re^3: PostgreSQL, Emacs, and other groupieware
by diotalevi (Canon) on Aug 02, 2004 at 14:17 UTC | |
Say, I never got the impression that cygin PostgreSQL was something to run in production anyway. You'd want a proper *NIX for that. When you've asked about getting cygwin PostgreSQL running I've gotten the impression that you take both the hard way (which often seemed to be getting native compiles of this stuff when Win32 native compiles aren't supported yet) and now that you might have never noticed that there is no production PostgreSQL on Win32. I can have a seemingly production quality client connection to PostgreSQL from Windows 2000 but that's over the ODBC drivers (http://gborg.postgresql.org/project/psqlodbc/genpage.php?downloads). You'd want to use the DBD::ODBC module then to connect since that'd be your local client interface. Anyhow, I use PostgreSQL and emacs because they let me do startlingly cool things that I can't get otherwise. That is, with PostgreSQL I use cascading updates/deletes to use natural primary keys instead of synthetic keys. I use triggers, rules, stored procedures, and views. Robert Treat did a nice demo of the sorts of things I'm talking about in his recent OSCON presentation Applying Advanced Database Features. I do these sorts of things all the time with my own install. I've merged R with PL/R to add support for statistical computing to PostgreSQL. The net effect is that I end up with a really, really capable platform. I need to have to a *nix machine to make this work well but that's just the nature of this stuff - *nix people write the code and that's what its targetted to. I use emacs because its something I expect to grow into over a few years. I haven't figured out how to write my own functions or rebind any keys sequences to functions or really any of the fancy stuff. I just know that when I need to, this thing will already be as powerful as I need it to be. I just haven't needed it yet. Appended: My example of PL/R is at Chatterbox conversational clusters. The full source code including R inside PostgreSL (called from a perl CGI) is available there as well. Appended: Also, I use GNU emacs on Win32. ftp.gnu.org was broken into in August of last year and all the precompiled packages were removed. The only version available there is for some older version. I lived with it for a while and recently got a copy of the latest from someplace. I forget the location. So emacs on Win32 is a matter of either picking the older package from ftp.gnu.org, compiling it yourself with MSVC, or finding a copy someplace. Its a world of hurt for the Win32 users. | [reply] |