Re: Semi-structured data

From: Leandro Guimaraens Faria Corsetti Dutra <leandro_at_dutra.fastmail.fm>
Date: Fri, 25 Jun 2004 09:58:45 -0300
Message-ID: <pan.2004.06.25.12.58.42.198858_at_dutra.fastmail.fm>

Em Thu, 24 Jun 2004 16:56:05 -0700, Costin Cozianu escreveu:

>> 	I am asking you to provide what you consider to be the precise
>> meaning of semistructured data.  Meaning I want to do some reading.

>
> I'm sure you can find easily a CS bibliography on semi-structured data.

        Probably. Is it any good? That’s the point. Check the other reaction to your post.

>> 	The reason I ask you instead of Google is that Google returns
>> nothing useful.

>
> Or maybe you are utterly unskilled at researching a subject for
> bibliography.

        Maybe this is a newsgroup, not a paper.

>>>Semi-structured means whose structure is only partially exposed to the 
>>>processing application (where application can be database systems, 
>>>client app, middleware, etc), i.e. it has hidden structure as far as the 
>>>application is concerned.
>> 
>> 	So it is not semistructured at all.  It simply means that not
>> all structure is known or relevant to the system.

>
> It means that the system can work with partial knowledge of the whole
> structure. You don't get to decree what is "semistructured" and what is
> not. The label associated with definition that I gave has already
> imposed itself in common usage both in the CS community, and the
> software engineering practice.

        It still fails Occam’s Razor, that is, it introduces a new concept that serves nothing over what there was before.

        Your definition, while useful as being precise, really adds nothing to encapsulation, and is not different from structured data at all given we don’t have exhaustive knowlege and (or) are often not interested in all details.

> Unless you can claim that you have the perfect sense of English
> language, and the rest of the CS community and the industry simply
> doesn't, then arguing over the chosen name is trolling.

        Well, they keep calling SQL relational…

> Even in that case all you can do is whine about the label, but not about
> the content under the label, and you can do that all day, but, of
> course, nobody will notice you, and it ain't gonna change just because
> you think it is a poorly chosen name.

        Words are important. They are used by people with hidden (or not-so-hidden) agendas to confound meaning. OO, XML, all of them have their Newspeak: classes, objects, semistructured and the like, galore.

>>>>>which is accepted by the overwhelming majority of CS community.
>>>>
>>>>	That says nothing about its preciseness or usefulness.  
>>>
>>>It says something very precise about the legitimacy of wannabe trolls on 
>>>c.d.t who want to wave their magic hands and pretend there's some kind 
>>>of truth or useful knowledge in their baseless claims.
>> 
>> 	I don’t follow your reasoning here.  I suspect because there’s
>> more anger than reasoning in this your paragraph.

>
> Or maybe because you don't want to read, and it is convenient to ignore
> it. Fine.

        It is really anger. You are loosing your reasoning faculties, intelligent and informed as you are.

        Just reread the above posts and see that you fell into reasoning from authority, and furthermore you made a personal attack on whomever doesn’t agree with your reasoning from authority.

        Oi Vei.

>> 	If you were right, there wouldn’t be so much SQL, OO, XML crap
>> all around.

>
> Non-sequitur.

        Not at all. If CS had high standards, it would influence more the state of affairs. It is when people compromise that they get accepted by the world at large, but then their ideas loose effectiveness.

>>>Who decreed that there's any good in the majority of CS community
>>>using FP languages ?
>> 
>> 	OT… and irrelevant.  I only painted a broad picture of the
>> field, I’m not wanting to argue every single point of it here.

>
> You're broad picture is patently false in both the essentials and the
> details, and is not done in good will, plus you have absolutely no
> standing to "paint a broad picture of the CS community".

        If you really want to argue that, fine, but I won’t bother until you actually make your point instead of jumping to conclusions. Just do it in personal email, this is not an adequate forum for personal feuds.

> >>Just grow up, will you ? Comp.database.theory has been in useless
> >>troll mode for months now, and you can't in all honesty blame it all
> >>on Pick fans.
> >
> > Don’t feed the trolls.
> >
> > Unless you yourself is the said one.
> >
> > Does not accepting mumbo-jumbo qualifies as trolling?
>
> I don't feed the troll

        Because there is none here.

> I'm just correcting a matter of fact issue.

        Trying to. AFAIU, you failed at that.

        Or better yet, the matter of fact you corrected indeed, but that did not make the concept you champion any more useful or less harmful.

> You can troll all you like regardless, and I ain't gonna do a thing
> about it.

        I fear I’m not the troll here.

-- 
Leandro Guimarães Faria Corsetti Dutra           +55 (11) 5685 2219
Av Sgto Geraldo Santana, 1100 6/71        leandro_at_dutra.fastmail.fm
04.674-000  São Paulo, SP                                    BRASIL
http://br.geocities.com./lgcdutra/

Received on Fri Jun 25 2004 - 14:58:45 CEST