Archives of the TeradataForum

Message Posted: Wed, 20 Oct 2004 @ 08:11:26 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Research in the field of commercial software is not always easy as most vendors do not publish detailed technical specs of their products. They remain black boxes. One of the way to get reliable results is to take a few boxes and compare them in similar situations. Such a comparison might have a scientific value, if done properly. As you don't have hands-on experience with TD, why not take another database system and compare them side by side. It will mean you'll have to learn the basics of the two systems but SQL will only differ in small details if you keep it on a reasonable level of simplicity, and that should help.

In this context, if you like the idea, I'd like to suggest to take Netezza as the second database. It might be a bit provocative suggestion as Netezza is a newcomer on the market. However, it seems to be TD influenced in some ways, it is young and hip, and it seems like fun to do the research on. Having some fun while doing research is important to keep you going.

Well, if that sounds good so far, then the number of practical topics you can take is virtually unlimited. You can describe the similarities and differences in the technical architecture. Technical test cases and benchmarks. You can make a survey of applications where only one of the systems shines and describe the areas where they are likely to compete in the near future. Some predictions for the future should be a part of any research, in my opinion. You could also consider the business side and compare costs of ownership for both systems and perhaps sketch typical customer / application profiles for them. Hopefully, all this still can be a subject for a research paper, and it won't lead you to troubles with the vendors, - check with your supervisor on that.

As Netezza is hardware based it might be difficult to use it. You'll either have to travel to one of their sites or use it via the web.

This is just one of the possible suggestions. For what it's worth, I for one would be very interested to read such an independent research paper and perhaps use it in my consulting practice. I think you'll find other interested readers, too. Having a wide readers' circle for your first paper is not a bad thing either. Over to you.

Regards,

Victor

	<Prev	Next>		<<First	<Prev	Next>	Last>>