Semi-OT: Database for "average" users
Kevin C. Krinke
kckrinke-eqjHHVKjh9GttCpgsWEBFlaTQe2KTcn/ at public.gmane.org
Mon May 2 18:16:48 UTC 2005
On Mon, 2005-05-02 at 13:04 -0400, billt-lxSQFCZeNF4 at public.gmane.org wrote:
> Neither side is right or wrong its a matter of preference.
> I prefer to be able too control what I put into my systems.
That is precisely the tipping point; control.
Some are more apt to assume that their vendor knows best and is looking
out to provide them with the right amount of control over their data.
(Read: I use spreadsheets for everything and it's brother.)
Others prefer to stick to their vendor not because they assume anything
but instead because that is all they know and are not of the mentality
to endeavour to learn something new when what they already know does the
job just fine for them. These people don't want "more" control as they
perceive that they've got enough control already. (Read: I use the
entire office suite as it was "meant" to be used. Spreadsheets for data,
slides for presentations, documents for reports, etc.)
Control and trust go hand in hand. The above mentioned persons accept
their available levels of control and work within those boundaries while
trusting that what they're doing with the software is "The Right Thing"
for them.
However there are the paranoid; they do not trust what they explicitly
do not have control over. These people like the ability to "look under
the hood" in order to gain that trust within their software. Whether or
not they actually read the code isn't important per-se but the mere fact
that they can is more often enough to support their confidence. (Read: I
use OpenOffice.org on Mac/*nix/Windows.)
And then there are those that are paranoid but are also capable of
learning new things and are not afraid of doing so. (Read: I use
OpenOffice.org whenever someone else sends me an office-type document,
otherwise I do my own thing that suites my needs perfectly and if it
happens to help you to that's a bonus and I use *nix whenever possible.)
I myself am of that last mentality and here's how I use the tools
available for the job at hand:
Spreadsheets:
Frequency: Rarely, if ever.
Examples : Excel, Calc, Gnumeric, etc.
Reasoning: Only when someone else requests it or I need to
"tweak" a .csv file before inserting the data into a
database of some sort (flat or server).
Flat Files:
Frequency: Sometimes.
Examples : Perl using ASCI/UTF-8 text files.
Reasoning: Conceptually easy to understand. Directories and
sub-directories are used for categorization and each
file contains one or two (three at the most) bits
of data. Only basic "querying" is performed. Only
simple updating/inserting is performed. Only suited
for local hosting of data.
UI Driven Database:
Frequency: Extremely rare.
Examples : MS Access, OO.o Base
Reasoning: Relies on version dependant software. Encourages bad
UI design elements. Not really network friendly.
Database Server:
Frequency: Common.
Examples : MySQL, PostgreSQL, Oracle, etc.
Reasoning: Performance, scalability and general lack of system
dependant bottlenecks. Group data by tables and have
as many columns as required to represent that data.
Advanced querying and updating is permitted. Equally
suited for local and remote hosting of data.
Web based UI:
Frequency: Common.
Examples : CGI, HTML, XML, RSS, etc.
Reasoning: Easily scaled for massive audience. Can be
integrated with an existing web solution. Relatively
intuitive interface due to the ambiguity of the web.
Easily networked with a database server.
Native UI:
Frequency: Common.
Examples : Gtk2, wx, Tk, etc.
Reasoning: Easily customised interface that is specific to the
solution required. Rich interface elements. High
level of application interactivity. Easily
networked with a database server.
Anyways, those are my preferences for control.
--
Kevin C. Krinke <kckrinke-eqjHHVKjh9GttCpgsWEBFlaTQe2KTcn/@public.gmane.org>
Open Door Software Inc.
--
The Toronto Linux Users Group. Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml
More information about the Legacy
mailing list