hi guys,

i'm doing an assignment which is relatively simple and would like to get some help with some of the concepts in database design. There are a few questions i am reseraching and any answers would be much appreciated. I'm guessing that these are simple questions for peoeple who work with databases all the time?

1. What are 3 main advantages of an object oriented database over a relational database?
2. What are 3 different situations where relational databases don't perform well and what other approaches can be used to address these issues?
3. How are data warehouses managed differently to transaction processing system databases?
4. Explain concurency management and why it is a compromise between maximising concurreny and minimising delays.

1A) OODB provides that DBA with travel opportunities on an expense account as they try to find ways to get the OODB to work
1b) OODB provides the DBA and DBM with opportunities to meet many new consultants brought onsite trying to get the OODB working.
1c) OODB provides opportunities to test disaster recovery procedures.
2a) Improperly configured hardware, Operating Systems, and Database engines
2b) Unrealistic expectations
2c) Naive developers using techniques that "Worked fine on product X" on product Y.
3) DW managers tend to be slower to react, but more drastic. There is frequently more perceived value in the DW than the OLTP database, and until upper management is re-educated they frequently make poor choices.
4) Concurrency management is the process of managing concurrents. These evil fruit are know to cause felonious delays of data, often when that data is needed most.

-PatP

thanks for your insightful answers Pat Phelan,

i look forward to some more opinions on these questions,

1A) Front end developers intuitively understand them with no need to learn all that pesky relational theory
1b) Managers no longer need to employ specialist front and back end developers since it is all "one technology"
1c) You can pronounce it "oh ooohhh" database.
2a) Server is turned off
2b) New developer writes a query that creates a cartesian product of your 20 largest tables.
2c) You use your RDBMS to add gradient effects & remove red eye from your family pictures.
3) You need more mouse traps in the warehouse.
4) I don't know what a "compormise " is. Presumably then you didn't cut & paste so credit where it is due - you must have typed out the questions all by yourself.

pootle flump ur codings are working excelent.

yep, you can tell because he calls it "orientated"

Well, give the guy a break fellas. He's not asking you to do his English 101 assignment for him. He just wants you to do his database assignment for him.

okay, since he's new, we'll cut him some slack

john_6767, we will not do your database homework for you

we will, however, help you with your english and/or your spelling, for free!!

I don't mind helping with spelling gratis...but I'll be damned if I'll help with English for free! For crissake y'all, I had to PAY for my dang-nab ainglish edukashun, I ain't gonna give it away free to all y'all.

Thanks!

-PatP

lol, thanks guys

its just all the database links i have found are in IT jargon and are hard to understand, but i'll keep going with it. Please let me know of any urls that have easily understood coverage of these topics.

cheers

Point understood, John...unfortunately, the answers to the questions you have will likely be expected in "IT Jargon" as well. It's really hard to answer a question like "What are 3 main advantages of an object oriented database over a relational database?" without some discussion of databases (probably using some IT Jargon).

Some of your answers are readily available, and probably in your class notes or textbook. Part of the learning battle is getting to the answer - seriously...and that search is pretty much the only way to gain the understanding. Like those spankings your Mom and Dad used to divvey out (oh, you're probably of the "spanking is a no-no" generation, so nevermind that, or replace the "spanings" reference with "time-outs" ) it's for your own good (well, and the self preservation of we who will be your co-workers at some point).

At least you have taken it all in good humour and this is the redeeming post.

Rule #1 Google is your friend.

Q1 & Q2:
http://www.google.co.uk/search?hl=en...ted%22&spell=1
Q3 (at least one good link here):
http://www.google.co.uk/search?hl=en...s+oltp%22&meta=
Q4:
You get the idea

pootle flump ur codings are working excelent.

Now that several people have weighed in against OORDBMSs, here's at least a hint as to why they're not always popular.

Look at a typical imperative programming language like C. Now look at what C++ or Objective C added to it.

If you're knowledgeable in this stuff, look at vanilla LISP or something even more functional. Then look at what an object system (CLOS) adds to those languages.

OOP's main achievements in these languages:

* you can write bigger programs with OOP because data and code are neatly organized

* you can write more "dynamic" software thanks to polymorphism and encapsulation

* you can employ richer data structures with OOP

Now, the RM isn't perfect. However, those are largely areas in which it is *not* lacking. That's why OODBMS's are not a big win.

Actually, I'm not against the idea of OODB at all. I had great hopes for a group of geeks that were working on one a couple of years ago, and thought it might be the "SQL Killer" that the relational purists have been waiting for for years. While I don't have a crystal ball, I'd be willing to bet that the next major breakthrough in database engine design will be object oriented, but I don't expect it in the next 3 years.

The problem is that all of the existing OODB engines that I know of are development liabilities instead of assets. While some of the OOP wrappers that interface to RDBMS technology work well, the pure OOP engines either never get off the ground, or they self-destruct soon after they get going.

The academics (including the academic in me) love OODB engines, but the real world is focused on getting work done, and I haven't seen an OODB engine that is really ready for that yet.

-PatP