Re: What do you need to excel with databases?

"it" <it_at_it.com> writes:

> Hi,
> thanks to a downturn I am considering a career in DB programming so I wonder
> do you need any special education or type of mind/talent to excel with it?
> I suppose you dont have to be very much creative but you need a strong
> logic. No numerical skills but strong concentration?
> Another question is are realtional databases her to stay or object oriented
> are the future? But what about Lyee programming methodology ?
>
> thank you
>
> IT
>

Well, it all depends where you want to work and what sort of problem domain area you are going to work in - I don't believe database programming is any different from any other sort of programming in that you need

1. A good understanding of the theoretical basis of the technology (e.g. database theory - all elvels, network, hierarchical, relational OO etc).
2. A good theoretical understanding of the problem domain you are working in. There is a big difference between the database requirements for a geographical information system, a financial information system, an on-line transaction processing system, a data warehouse application etc.
3. Prepared to get to know the the tools you are working with. don't treat all DBMS as the same. If your programming appications for an Oracle DB, then do it the Oracle way, if its for MS SQL Server, do it their way, if its DB2 then do it their way etc. This means having to read manuals, learn new techniques or ways of solving problems suitable for the tools you are using etc.
4. Creative and imaginative - I can't imagine anyone being any good as a programmer, no matter what the domain and tools, without this!

Wtih respect to methodologies - well, they come and they go - some are good, some are not. Some are useful in some problem domains and not in others and vice-versa. I have to say I'm a little skeptical of lyee's claim that it can do a lot of the programming itself without human intervention - I've seen this promise too many times before. I don't know much about it as a methodology yet, but it seems to me its really just an attempt to automate design pattern application. I'm probably just a bit skeptical, but I've yet to see a good CASE tool that was able to handle all problem domains equally well - to me the best tool for development is still that lump of grey matter between your ears and it will be a long time yet before a computer can do as well. this is not to say such tools and methodologies are not useful, they can be very useful and can help productivity, but not when used by someone who doesn't understand what they are doing and the problem domain they are working in.

Tim

-- 
Tim X.
tcross (at) northnet com au