An article by Edward Hurley called Is It Time to Get Lazy?
about Simon Williams' company, Lazy Software, published on the
TechTarget/SearchDatabase site, has recently come to my attention. I published
a couple of exchanges here (see links below) on Williams’ so-called
"Associative Model of Data" (AMD). Hurley had asked for my advice on
the subject, but none of it made the article. So I contacted the editor there
(who also happens to be the editor of my Against
the Grain series, which was recently terminated; see On the Trade
Media’s “Balance’ – Another One Bites the Dust and argued in quite
strong language that at the very least there should have been some
reference to criticism of Williams's “data model”, be it mine or others’ (since
the editor deems me “irascible”, I should give him reason for it). The editor
said that “that was a news piece, and wasn't meant as analysis” and that
Hurley, as “a journalist, and not a DBA or database guru, relies on contacts
and experts to give him their opinions”; because I refused to offer such, I
should not complain.
1. I will get shortly to the important issue of whether
journalists should know anything about the technology they cover and the
obvious implications for the readers if they don’t. Did I really refuse to
provide advice? You be the judge.
·
On July 7 last year I posted an exchange between myself
and Lee Fesperman regarding Williams' claims:
·
Lee drew my attention to some online exchange in which
Williams was making the very same claims which, as Lee pointed out to him, had
already been posted as one of my site’s Quotes-of-the-Week;
·
In his reply to Lee, Williams invoked positive
assessments of his AMD by two “respected technical analysts", Butler and
Bloor, and invited a debate on the subject;
·
I replied that the number of errors that Williams
managed to squeeze into one paragraph and the amount of nonsense in those
errors were impressive; I included Chris Date's comments on Bloor that throw
into question his respectability; and I issued the following challenge: “If
Williams knows what a data model is, let him provide the definition and then
explain how his so-called
"associative model" complies with that definition."
2. On August 25 I further debunked Williams' claims in a long
exchange with a reader who had read Williams' book and responded to the July
exchange.
3. It was much later that I heard
from Williams (see On What Is a Data
Model):
·
He did not initially provide a definition, but tried to
convince me that his AMD was superior to the relational model;
·
I countered that until he demonstrated he understood
what a data model is, there was no point to any debate;
·
Finally, he emailed me what he considered to be a definition;
·
My assessment of his "definition" was as
follows: "... it is too long. It
includes unnecessary historical details. Only the very last part contains what
can be deemed a definition, and my guess is that much of the text is lifted
from other sources. Be that as it may, even that lacks one critical
ingredient--the theoretical foundation." I stated the correct
definition and I issued a more specific challenge: "The question is does
Williams understand the "definition" that he cites? If so, and
if Williams’ AMD is a data model, then:
·
On what theory is it based?
·
What are – precisely, please--the structural,
integrity and manipulation equivalents of his "model"
and how are they different, let alone better, than the relational ones?"
I am still waiting for his response.
4. In December I received an email from Edward Hurley, a
reporter for TechTarget, requesting my input for an article he was writing on
Lazy Software: "I'm looking to do a story about the product and Lazy's
claims about its Associative Model of Data. Is the model a viable alternative
to relational databases? What makes the AMD model different than relational
databases? Would you be available this
week to discuss the AMD model with me? I can send you some detailed questions
if that would help. Thank you for any time you can spare."
Note that he was looking for an analysis/evaluation of
Williams’ claims, not just to report on them.
I responded as follows:
"The answer is very simple, but in order to really
understand it and write intelligently about it, you need to do some homework
before I can discuss it with you. Unfortunately, there are no shortcuts to
this.
AMD is not a data model at all and, therefore, it
cannot possibly be an alternative to the relational model. Like a vast majority
of practitioners, Simon Williams and those who refer to AMD as a data
model--including "respected analysts"--do not know what a data model
is. Hard to believe, but true. [Ed. Comment:
See also this week’s exchange More On Williams’s
“AMD”-–hinting that indeed, AMD is really not a data model after all).
I am attaching my exchange with Williams. I am still waiting
for his reply to my questions that I asked him both in a public article and in
the attached exchange. Don't hold your breath.
I suggest you read the following material--the first two
items are on AMD specifically:
On "Respected
Technical Analysts"
On the So-Called
"Associative Model of Data"
Something To Call
One's Own
Models, Models,
Everywhere, Nor Any Time To Think
This is important not just for this particular article, but
for anything you will write on database management. If you focus just on
products and vendors and don't know the fundamentals, you are likely to fall
into the same trap that most trade journalists fall into. Furthermore, almost
everything you hear or read on the subject of databases is simply wrong, but
you have no way of detecting this without adequate education on the subject.
I will be glad to answer any specific questions you might
have after you review the material, and even help you along in your
future endeavors. In particular, I will advise you what questions you should
ask to determine whether the claims have any legs to stand on.”
I never heard from Hurley again. So much for his reliance on
experts. I will leave it to the reader to decide whether I refused to offer
advice; furthermore, those who read DATABASE DEBUNKINGS know that they
answer precisely the questions that Hurley was asking.
So here’s what I told the TechTarget editor:
- Journalists should not be experts, but should have
an
idea about the subject matter and recognize when they're being sold a
bill of goods.
- Anybody can wake up in the morning and say
something--how do you establish that is news? If I were to declare
AMD is nonsense, would that be news worth publishing? What are the
criteria for what is news, other than that some vendor is the source?
- A journalist--particularly if he does not know
anything on the subject--should do some research to educate himself--what
research did Hurley do? How does he know Williams is not selling a bill of
goods (which he is)?
- The reason I am usually reluctant to offer advice
is
because each time I did, it was either ignored, or bastardized. I have published
on Williams and his AMD, did Hurley do any search?
The problem is not only that journalists don’t have minimal
knowledge, they don’t care to acquire it—they simply regurgitate what
vendors say, which is easier than spending time and effort assessing it.
As to Hurley’s article, stay tuned for my assessment.
(Continued in Part 2)
Posted:
08/04/02
[ABOUT]
[QUOTES]
[LINKS]
