Re: Resiliency To New Data Requirements
Date: Mon, 14 Aug 2006 15:58:44 GMT
Message-ID: <U41Eg.44891$pu3.554383_at_ursa-nb00s0.nbnet.nb.ca>
erk wrote:
>>erk wrote: >> >> >>>dawn wrote: >>> >>> >>>>Perhaps it would be more precise to say that they "threw the baby out >>>>with the bathwater." I found arguments, but no emperical data, that >>>>showed the relational model to be beneficial to those using it. If you >>>>are aware of any emperical data, particularly any that shows that the >>>>RM is more flexible over time than either MUMPS or PICK that predate >>>>it, I would really like to see that. >>> >>>It doesn't exist. >> >>Bullshit! Look, if you are going to engage the ignorant cranks, at least >>call them on their bullshit. Don't swallow it!
>
> I'm not swallowing any arguments about flaws in the RM, but I don't
> know of evidence for the benefits of RM that would fit the definition
> of "empirical data" she seems to be using. Then again, that sort of
> data is also lacking for the benefits of any programming language,
> operating system, etc., making every decision on every technology
> similarly "arbitrary," and reducing the entire thing to nonsense. That
> line of argument is a rabbit's hole, so I ceded the argument - "it"
> (meaning what she's looking for) doesn't exist, though it has no
> bearing on the arguments.
>
> None of this has any bearing on my knowledge that the RM is superior,
> both objectively (via logic, thought experiments, reading, etc.) and in
> my personal experience. But I can't call that "empirical data" in the
> sense in which she means it.
Do you suppose she means it in some sense other than "relying on or derived from observation or experiment" or "guided by practical experience" ?
Empirical evidence abounds.
One could argue that lines of code and number of bugs are not valid measures for comparing two solutions, but the argument would be very weak. One could argue that man-years of effort is a loose measure and one would be correct. If one were comparing two things where the measures were close, one could not conclude much. However, when one compares two things where the effort for the equivalent product is orders of magnitude different, the measure suffices for coarse conclusions.
Network model and NFNF models require orders of magnitude more effort for the same outcome. Proponents of these logical data models invariably deliver buggy code with an order of magnitude or more complexity while ignoring concurrency, integrity and a whole host of other concerns dealt with by relational solutions. The red and blue car cross-post thread offers empirical evidence that Pick proponents lack the cognition to recognize the faults in their beliefs.
Others have recently mentioned the examples from the Great Debate comparing a CODASYL solution with a much simpler relational solution.
Anyone who claims that no empirical evidence exists to shred network model and NFNF models is either ignorant or in denial. Received on Mon Aug 14 2006 - 17:58:44 CEST