Talk:Relational database/Archive 2

From Wikipedia, the free encyclopedia

Archive This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.
Archive

Archives
Current Talk 1 2

Contents

New page

I've moved in the changes from my rewrite, for which there was no opposition, and at least one encouragement to be bold. Corresponding to this, I moved the talk page discussion to an archive. The entirety of the old discussion can be found there. I look forward to your constructive criticism so that we can make this page better. McKay 05:04, 7 August 2006 (UTC)

JA: Do not edit articles by blanking pages and substituting new content. Proceed by incremental improvement so that others can track the changes and decide whether they think it is an improvement. Your edits had the result of simply deleting the hard work of many others. Thanks, Jon Awbrey 05:10, 9 August 2006 (UTC)

Nothing was deleted, as was evidenced by your revert. The rewrite was announced and discussed here. No one commented. Now that you are here, I expect you to express your concerns with the rewrite. --Ideogram 12:37, 9 August 2006 (UTC)
A great deal of the deleted "hard work of many others" was mine. And you may have noticed that I encouraged and helped in the rewrite. WP:BOLD is a widely accepted guideline. Furthermore, drafts of rewrites to pages are not allowed in article space (see "Disallowed uses #1"). I think McKay might preferably have used a Talk subpage instead of a user subpage, but I think the route he took is better than the one you suggest for a major rewrite. I would strongly support going back to McKay's version. JA, you say nothing at all about what you feel is not in the new article which was "lost" from the old article. We're not going to stay on the current revision forever, so if there's some important point you'd like to keep you should say what it is.--Craig Stuntz 12:58, 9 August 2006 (UTC)
Concur with Craig My decision in using my userpage for the rewrite (over a talk page), was that I would be doing the bulk of the work in the rewrite. McKay 14:47, 9 August 2006 (UTC)

Main template

I changed the main template usage to use a single line instead of two lines just to save vertical space, and Mckaysalisbury (talk • contribs) reverted, commenting, "both sets of "main"s are correct. Both should have the same reference. Mathematical before "perversion"." I'm not sure I understand this comment at all, and I don't agree with the change, so, though it's a minor point, I'm asking here for clarification. For clarity of discussion, here's an example of the two options. First, the original, and the current version:

Main article: Relvar
Main article: View (database)

Second, my changed version:

Main articles: Relvar and View (database)

As far as I can see, the second is pretty much standard and avoids redundant wording. See, for example, this case in the MOS. --Craig Stuntz 12:36, 8 August 2006 (UTC)

Sorry, you're totally right on this, I thought that I saw that your change made it look like this
Main article: View (database)
and I thought I saw this in the article. I must have been mistaken. I'll put it back to your change. McKay 17:11, 8 August 2006 (UTC)

Boolean

Finally, we get a reasoning attached to the changes. I've reverted again based on the following reasons:

  1. the wikipedia article "boolean" doesn't capitalize uniformly
  2. google "define:boolean" top five (that loaded) didn't conclusively upper-case. i.e. in each of the the top 5 hits (that didn't 404), they either used "boolean" in lower-case form, or all items in lists that "Boolean" was in, there were other capitalized words that aren't normally (e.g. "Integer" or "Bookshelf" ).

McKay 00:12, 9 August 2006 (UTC)

In this context, I agree with you. I would use 'boolean' as an adjective, and 'Boolean' as the name of a datatype. RayGates 02:10, 9 August 2006 (UTC)

JA: It is fairly common practice in mathematics to quit capitalizing adjectives derived from proper names after a decent interval of a century or so, hence, abelian, archimedean, boolean, euclidean, fuchsian, and so on. This is partly a mark of additional respect for truly venerable terms, but also because that much time usually means that the honorific eponym no longer amounts to a literal attribution in any exact sense. Jon Awbrey 02:20, 9 August 2006 (UTC)

Normalization

The last line of this section read:

All reduce functional dependencies, and therefore, various kinds of data duplication, and therefore ease of correctly modifying the data.

I.e. All reduce ... ease of correctly modifying the data.
I don't think that is what is intended. I have amended it accordingly. RayGates 02:17, 9 August 2006 (UTC)

Blankety Blank

JA: The page blanking occurred here. That is just not a legitimate edit and should have been reverted immediately. Major rewrites demand major discussion. These did not occur. Jon Awbrey 13:08, 9 August 2006 (UTC)

What do you mean no discussion occurred? Did you read any of this? We discussed this rewrite for weeks, rejected one draft almost entirely, and tweaked the second fairly heavily. Eventually it got to the point where I suggested it was good enough to be visible to the public and that any further revisions should happen in article space. --Craig Stuntz 13:20, 9 August 2006 (UTC)
Re: the blank. Yes, there was a blank, but if you looked at the next edit (where I incorporated the changes from the rewrite) this happened in the same minute! Prolly within 10 seconds. Also, I made mention of this in the edit history ("momstly Blanking (briefly), emphasising that my major change (next) is all new.", "Pulling major changes in from a rewrite page.")
Discussion occured, [/Archive 1] covers most of this. That which wasn't there was mentioned on my talk page, and Ideogram's. This went on for several weeks. McKay 14:47, 9 August 2006 (UTC)

JA: There are more than 5000 pages on my Watchlist. I'm not about to start adding User Pages just to keep track of article changes. The only practical way to keep track of things is by using the Diff tool, and that is messed up when you do page blanking. We all use sandboxes, but it's just plain good sense to import things in chunks. Do not blank pages. Jon Awbrey 16:00, 9 August 2006 (UTC)

That's nonsense. All of the discussion occurred right here. If you had been following this page you would have seen it as it went on for weeks. Here's Ideogram's post of 8 June. Your insistance that no "major discussion" occurred is at odds with two months of history which anyone can examine. McKay and Ideogram discussed every change they made to their drafts here. You didn't need to watch their user space to see it, you just needed to pay attention to the conversation here.
Blanking before totally changing the article does not affect the diff at all when you compare the appropriate versions. Here is the diff between McKay's original rewrite and the last non-blank version posted prior to the rewrite publication. Having the blanked page in between these two versions doesn't affect the diff at all, because that revision isn't examined at all by the diff tool in the link I gave. If you'd like help using the diff, just ask.
There is no policy or guideline that bold changes be done in small chunks. --Craig Stuntz 16:57, 9 August 2006 (UTC)

JA: WP has many heuristic slogans. Like most heuristics and deep truths they come in complementary pairs. One says Be Bold. Another say Dont Be A Dick. Being bold is no excuse for destroying the integrity of an article. In articles that have been under long development, common sense dictates moving incrementally. Anybody can go out of their way to compute arbitrary Diffs, but routine watching means clicking on the adjacent transitions. Do not blank pages. Jon Awbrey 17:22, 9 August 2006 (UTC)

That would make sense if the article was, in fact, "destroyed." I personally think it was improved, not destroyed (despite the fact that I personally wrote a significant chunk of what was replaced), and you have only provided the vaguest hint of what a problem with it might be, namely that it doesn't seem to conform to some plan for a group of articles that you have but somehow didn't mention in the two months we discussed the rewrite here on this page. You have not made a case that the article was destroyed, only an assertion. Moreover, you say a number of things which are provably incorrect. I personally agree with both "be bold" and "don't be a dick," but I think McKay has been very well behaved in spite of the fact that you reverted his work. You don't own this article (nor do I, or McKay, or...). If you want to contribute, how about proposing a version which incorporates your concerns as well as the accessibility to non-expert readers concerns expressed by Ideogram in the first place? --Craig Stuntz 17:42, 9 August 2006 (UTC)

JA: Blanking pages is not being well-behaved, period. Ideogram was not even here way back in Jan-Feb-Mar when I worked this page up as a tutorial. When I left off active work on the page, it was mostly being fine-tuned by folks who were experts and who recalled the planning discussions that we had had, so it seemed like it was in good hands. That does not appear to be the case at present, and a quick scan reveals many howlers that are just plain wrong, plus all the stuff that belongs somewhere else. Jon Awbrey 18:20, 9 August 2006 (UTC)

I can't read the above as anything other than, "It's different than my version, so it must be wrong." Do you have any substantive criticism of McKay's version? Vague remarks about "howlers" and even vaguer appeals to authority don't help improve the page. If you want to keep the article the way it is now forever, then give up; wikis will change. If you're interested in improving the article, on the other hand, why not respond to the repeated requests for constructive feedback rather than just generally trashing others work? I'm going to ask you one last time to explain why you don't like McKay's rewrite, and then I give up asking. You've given us no reason to believe anything you say at all. --Craig Stuntz 18:43, 9 August 2006 (UTC)

JA: When I get ready to edit the main page then I will do so. Right now I'm just trying to ensure that a certain message is received and understood. Do not blank pages. Jon Awbrey 18:54, 9 August 2006 (UTC)

You already edited the main page by reverting other editors' work; reverts are edits. Don't expect it to stay that way until you feel like editing again if you steadfastly ignore all requests to explain your actions. WRT blanking, you are making an enormously big deal over an edit which I personally find almost entirely inconsequential (I would have felt differently if he had left it blanked, but he didn't), perhaps because it is one of the only verifiable thing you've said so far which isn't provably incorrect — namely, the fact that McKay blanked the page for under a minute. From what you write it's difficult to believe that if a Steward removed that edit from article history it would make you satisfied with the substantive concern about what should be in the article. Hence, I see no point in discussing it; it was immediately "unblanked" (by the same editor), I gave you a link so that you don't have to compose a diff URL yourself, it's in the past now, and nobody is proposing blanking the article again. I'm much more interested in the future of the article than its past. --Craig Stuntz 19:27, 9 August 2006 (UTC)

JA: Reverting vandalism is not an edit. Blanking pages is normally considered vandalism. I made an extra effort to assume good faith, but you are proving to be resistant to common sense advice. Do not blank pages. Jon Awbrey 19:38, 9 August 2006 (UTC)

from WP:Vandalism:
  • "Vandalism is any addition, deletion, or change to content, made in a deliberate attempt to compromise the integrity of the encyclopedia."
  • "Any good-faith effort to improve the encyclopedia, even if misguided or ill-considered, is not vandalism."
Same article under "What vandalism is not":
  • "Bold Edits
    • Wikipedians often make sweeping changes to articles in order to improve them — most of us aim to be bold when updating articles. While having large chunks of text you've written deleted, moved to the talk page, or substantially rewritten can sometimes feel like vandalism, it should not be confused with vandalism."
What I did was not vandalism. Any questions?
I don't question my judgement of the moment, but I wish that I wouldn't have blanked the page. Only because of this inane argument of yours. Sure, I blanked the page, for less than a minute. I was worried that someone would have seen the change during the 10 seconds it was down, but I never possibly occured that someone would have complained that it messed up the diff, that it destroyed content. That I am a dick. I take offense to that last one, as it is essentially a personal attack. have you even read meta:Don't be a dick? DBID specifically referrs to ignoring all rules. Sure, I blanked the page. Sure, blanking is generally considered vandalism, but I figured that I would ignore that simple rule to make the wikipedia history slightly better. By blanking the rule it allows the possiblity (without removing other possiblities) too emphasise that my changes are substantial enough to warrant the notice that, in effect, I removed all of the old changes, and added a bunch of new content of my own. Technically, you continuing to say that I shouldn't blank pages (a fact that I know to be true already), means that you need to better understand the policy of not being a dick. meta:Don't be a dick. (yeah, read the whole thing). McKay 01:01, 10 August 2006 (UTC)

JA: Good. I am gratified to find that you've been thinking about the issue. Jon Awbrey 04:00, 10 August 2006 (UTC)

That's it? After 6 comments of you complaining that I blanked pages (some directly calling my work vandalism, or indirectly calling me a "dick" also). I respond with
  • directly quoting WP policy
  • calling your comments inane
  • indirectly calling you a dick
  • saying I wish I wouldn't have pushed your button
  • saying that I would have made the same decision of blanking the page again
all I get is a I'm glad you're thinking about the issue? I was expecting one of the following:
  1. further complaints (nothing else seemed to have stopped your rambling)
  2. rebuttal (a possibly scientific response to my claims)
  3. silence (from a defeated spirit)
  4. an apology (admission of guilt)

I've made some bold claims (mostly in response to your bold claims). Such a "middle of the road" comment seems out of place. McKay 06:02, 10 August 2006 (UTC)

Organization of Articles Related to Relational Databases

JA: There was a lot of thought and discussion early this year on the coordination of the articles related to relational databases, especially these three, in order of decreasing abstraction:

  1. Relational algebra
  2. Relational model
  3. Relational database

JA: The relational model article was intended to form a middle ground between the abstractions of relational algebra and the concrete specifics of relational databases, but still keep to the general principles of the underlying logical model. The relational database article RDBMS was intended to deal with specializations of and the departures from the relational model that arise in practical implementations. Somewhere along the line that organization has gotten totally messed over. There is now a lot of stuff in the relational model article that just plain does not belong there, and should be taken up in the relational database RDBMS article. Jon Awbrey 13:46, 9 August 2006 (UTC)

Did you read this mention on the talk page? I wouldn't mind if my changes there were reverted. I somewhat agree with this coordination you mention. Relational Algebra has nothing to do with the perversions of it in SQL. Even in pseudo-relation land, people refer to the "relational model" as a comparison for their perversions. Relational Database is the first article that definitely has some discrepancies among pseudo-relational and truly-relational adherents. I think that my new rewrite covers those differences, makes the article more readable, has a brief mathematical explanation that isn't overly complicated by dense mathematical treatment (which is, or should be, available elsewhere).

JA: The important thing is the current state of the coordinated articles — and that is now a mess, no matter how you look at it. I went back and reviewed the discussions from early in the year, and it appears that the situation was more like this:

  1. Relational algebra
  2. Relational model
  3. Relational database
  4. Relational database management system

JA: In other words, it was the RDBMS article was supposed to collect all the concrete implementational stuff. Several readers had expressed a need for a simple introduction to the mathematical concepts of the basic relational model, using lots of concrete examples. It was decided to use the relational database article for that purpose, based on the idea that a relational database is an instantiation implemented according the relational model. I personally put a lot of work into doing that. This has all been trucked over with stuff that should be moved to the RDBMS article. Jon Awbrey 15:24, 9 August 2006 (UTC)

Here, I would have to disagree with you. Very vehemently. Relational Model is where I think more intense mathematical treatments should go. When someone is trying to explain the difference between SQL-Relational systems, and Truly-Relational systems, the comparison always revolves around the relational model (i.e. "SQL fails to conform to the relational model because of X"). Relational Algebra is merely an algebra for manipulation of the relational model. When people started making pseudo-relational DBMSs, particularly SQL DBMSs, they called them "relational" (I will note that I disagree with their usage, but they did it nonetheless). All of the databases made in RDBMSs are Relational Databases. No one goes around saying "I made a pseudo-relational database in my Relational DBMS". There is definitely some confusion in the world as to what a relational database is, just like RDBMSs.
Also, is the problem of the layman. All too constantly people (even those with some familiarity with the subject) confuse the terms "Relational Database" and "Relational Database Management System". People say "relational database" when they mean "RDBMS". I think it is critical for this article to be simple, and to explain that difference. McKay 01:01, 10 August 2006 (UTC)

JA: If what you say were so, then the Relational model article is even more messed up than I said it was. But that is not the way that these words are commonly understood. Relational algebra is very abstract. The laws of this algebra constitute a type of formal theory and the array of relational models, as there are many of them[citation needed], are analogous to models in the sense of model theory. The formal theory of relational algebra does not specify many concrete details that have to be fleshed out in the particular relational model of choice. Usually there is what we think of as the canonical model, or the only sensible model, and so we tend to focus on what we call "the" relational model. It is the RDBMS that amounts to a software system for managing approximations to relational models, and they will have many more arbitrary features that are not dictated by the theory or the canonical model, but by the pragmatic demands of actual computation[citation needed].

JA: The Relational database article was specifically intended to meet the expressed needs of readers for a presentation that explained the underlying mathematical theory in a "Keep It Concrete And Simple" (KICAS) way. But that aim has now been waylaid in an article that propagates all of the common misconceptions that come from trying to reverse engineer the theory from wrangling with database packages. Jon Awbrey 04:22, 10 August 2006 (UTC)

Okay, while you may have "designed" the articles to be this way. I can't see how this is based on reality, or at least verifiability. Google "define:relational model".
I might be out of the loop here, but I don't recall ever hearing about several relational models. In fact, I've heard SQL proponents (those who call their SQL-based DBMSs "relational") use the relational model as a baseline for their references. None of them claim that theirs adheres to "a relational model". I realize that Date is a truly-relationist, but SQL doesn't implement relational model. The relational model is the de facto basis for comparison of pseudo-relational and truly-relational systems. I'm not saying that relational model isn't messed up as an article, I'm here defending the article I wrote on Relational databases. Why do you propose that the math belongs here? Google "define:relational database" None of these articles have a mathematical treatment.
Google hits:
  1. relational model 23.2 million
  2. relational database 41.4 million
  3. "relational model" 0.843 million
  4. "relational database" 16.9 million
  5. relational model tuple 468 thousand (2.0%)
  6. relational database tuple 488 thousand (1.1%)
  7. "relational model" tuple 121 thousand (14.3%)
  8. "relational database" tuple 327 thousand (1.9%)
In other words, only 2% (1-2%) of articles about relational databases reference tuples (the basic mathematical construct of a relation), but 14% (sure, fight the 2% all you want, but it's not in quotes, and "model" is a rather common word, it occurs in about 1/2 of all articles containing "relational") of all articles about the relational model reference tuples. So I think I stand by the fact that the math belongs in the relational model article.
Also, you mentioned "But that aim has now been waylaid in an article that propagates all of the common misconceptions that come from trying to reverse engineer the theory from wrangling with database packages." I'm not entirely sure what you mean by this. But I think I might. This article explains what a relational database should be (as proposed by Codd), mixed in with what we can deduce is in a "relational" database because of how much (we've reversed engineered) SQL databases actually meet that expectation; they fall short in several areas. If I'm understanding you correctly, you're complaining that I'm analyzing the conformity of "relational" database packages to what I call the relational model. The reason that I do this, is because SQL DBMSs have slapped the "relational" appelation on themselves, and I don't think they deserve it. Sure, it's a de facto standard now, so we include that definition. As a result, we mathematicians, have to determine what "relational" actually means because of this new de facto definition. The only way I see of doing this, is "reverse engineering" "relational" DBMSs. Am I understanding you right?
And to go back to what you said slightly before that, yes, it is the DBMS that enforces these restrictions, and there will be other potential features (like "do they allow us access to the file system? an imperative language? First class functions? Strongly-typed or weakly-typed? Pass by value or pass by reference?), but we can still gague it's adherence to the relational model by several means. Codd's 12 rules, an actual comparison to the relational model, what Fabian Pascal has to say, or something else. In any case, a relational database is usually intrinsically tied to the RDBMS that it is created in. McKay 05:44, 10 August 2006 (UTC)

JA: Short comment before long nap. I don't believe in trying to reverse engineer adequate articles from superficial web statistics, though that is a good way to identify the most common misconceptions. But the term Relational database was singled out for the gateway article simply because that would be the most common search term for beginners. That was also the reason for the extra guidance paragraph that once led off the article, which may seem superfluous to non-beginners, but was put there in answer to real bewilderment on the part of some readers. ZZZZZZ Jon Awbrey 06:04, 10 August 2006 (UTC)

Sorry, I can't buy that either. Compare the first paragraph of my article, and yours
  1. A relational database is a database structured in accordance with the relational model. Strictly speaking the term refers to a specific collection of data but it is invariably employed together with the software used to manage that collection of data. That software is more correctly called a relational database management system (RDBMS).
  2. A relational database is simply a database that conforms to the relational model. The term is used to refer to the data, and the structure of that data. The software used to create a relational database is called the Relational Database Management System (RDBMS), but sometimes that software is mistakenly called the relational database.
These seem substantially identical. Sure, there are parts of mine I like better, and a part of yours that I like better (I'm changing the live article to reflect this). If you actually mean the paragraph removed by this edit, I think that it's edit summary speaks for itself. McKay 07:13, 10 August 2006 (UTC)

Revert

All in all, I think that Jon's revert was not mutually agreed upon? McKay 14:47, 9 August 2006 (UTC)

Well, he didn't ask, so it's hard to say what people "agree" on. Only three others (you, me, Ideogram) have commented so far. For me the biggest consideration is not what various editors agree on, but rather which version of the article (or combination) best describes the subject. I've asked Jon to explain which bits of the original he felt were valuable and appropriate here but not adequately covered in the rewrite, but he hasn't answered yet. --Craig Stuntz 15:13, 9 August 2006 (UTC)

JAs bold claims refuted.

In JA's ramblings, he made some bold claims. Here I refute them. don't get me wrong. He did put a lot of work into the page, which brought it out of stub status. I'm just refuting his claims made here on the talk page.

One example

"JA: Blanking pages is not being well-behaved, period. Ideogram was not even here way back in Jan-Feb-Mar when I worked this page up as a tutorial. When I left off active work on the page, it was mostly being fine-tuned by folks who were experts and who recalled the planning discussions that we had had, so it seemed like it was in good hands. That does not appear to be the case at present, and a quick scan reveals many howlers that are just plain wrong, plus all the stuff that belongs somewhere else. Jon Awbrey 18:20, 9 August 2006 (UTC)"

Extensive Discussion in Jan-Mar

So, I can't find any discussion anywhere about what this article should be. Where this is proposed? anyone? McKay 14:23, 10 August 2006 (UTC)

Other nearly-related talk articles JA touched Jan-Mar:
Jan
(Enter his work on Relational database)
ooh, Relational database is mentioned here. JA mentions that maybe there should be a hub page, like the one Relational Database was at the time of his posting his comment. The one without any real substance. No decision was made here about this article.
Feb coming soon
McKay 21:15, 10 August 2006 (UTC)
admitedly, I'm only looking at articles he's made edits to the talk page. He could concievably have just seen the discussion, and not taken part in it.

Gah! Blank Sections

Examples of "experts" improving the site after Jan-Mar (all non JA edits Feb-May)

February

March

April

Note: JA stops editing

May

June

Note, Craig is now on the scene, Ideogram enters in the second edit in June

.....

Complaints of JAs version

April

June

Comments

This page started tongue in cheek. I wasn't expecting to actually submit this page. The first few things I wrote here were mostly worthless ("discussion", "blanking", and "complaints"), but when I really started looking at it. I couldn't help but refute all of these claims. I'll mention that I think that he means well. He wants this article to be good. He's just made some claims (against me) that had no substance. I wouldn't be suprised if he was thinking of another article. I'd look at the edit histories of Relational algebra, Relational model and Relational Database Management System if I had the time (I do not). I await his valuable contributions. McKay 07:02, 10 August 2006 (UTC)

What the real problem is that someone seems to have a big ego. Signs of people with big egos include user pages full of poetry, prefixing their initials to everything they write, ending every speech they make in a call to destroy Carthage, and thinking they own the article. People with big egos are never wrong. They are, were, and will always be right, and they never made a mistake. - ∅ (∅), 12:45, 10 August 2006 (UTC)

JA: The ego fits, ergo wear it out. Jon Awbrey 14:44, 10 August 2006 (UTC)

I think everyone here wants the article to be good, but that can't happen unless we actually discuss what needs to be improved or added. Vague insults to other people's work will not help the article. (And that goes for ∅, too.) --Craig Stuntz 13:55, 10 August 2006 (UTC)

JA: I see no insults so far, only criticisms. Deleting a mass of another's work is the severest form of criticism. Jon Awbrey 14:48, 10 August 2006 (UTC)

Yes, I removed a lot of work from the page. The article said that it was in need of an expert. The most active participant was Ideogram, he and I had a nice discussion, and it was recommended that I do a rewrite, so I did. I'm not saying that the old article had no merit. If there's something in the old article that merits inclusion, let's put it in!
Sorry, but most of the work I "deleted" wasn't yours Diff: JA's last version, version McKay deleted Similarities:
  • Some of the opening paragraph (previously established that this is mostly intact)
  • Schematic Example of a relational database
  • 1/2 of the see also section
So, like we said earlier, what needs to be improved? Craig, Ideogram, others, and I have made (non-trivial) contributions. We're hoping JA has something to add, besides non-substantiative complaints about the article. Please, help the article get better. McKay 18:04, 10 August 2006 (UTC)

SiteGist

JA: The clearest indication of what I think should be in the article is given by what I already put in the article, now mostly gone. I am sure that I said at least some of this already, but the angelic doc-workers who were discussing this congeries of related articles — including the coordinate articles from the principalities of logic and math — were hardly dancing on this stubby pinhead of an article at the time, but decided to develop it as a tutelary gateway to the kingdom beyond. Now, there wasn't a whole lot of dicussion — not like this — because, after all, a word to the wise is sufficient. Jon Awbrey 20:00, 10 August 2006 (UTC)

I don't think anyone agrees with you with the table that was in there earlier. The bit on k-adic tuples, should probably go in relation. Most everything else is actually in the article. If you disagree with any of these points in specific, let's analyze that, and put it in the new article. Please, give specific examples of what should go in the article! Feel free to actually make those changes. The only change you've made to the extensive rewrite was to revert my changes. I don't believe that you seriously think the whole article I wrote is trash. Let's see your actual substantive edits.
I believe that I've done a decent job in wikiresearch suggesting that this discussion you mentioned didn't actually take place. I assumed good faith for a while, but I've since delved in all of the articles you mentioned were part of the discussion. I would love to see such a discussion and add it's contributions to the current articles, even if it is one line from Larry Ellison, but I'm going to need to actually read that discussion. McKay 20:26, 10 August 2006 (UTC)

JA: I do have a natural curiosity about the integrity of my own memory, and I don't usually expect too much perfection of it, but this topic is kind of back-burner for me. If it rains this weekend I may get time to finish compiling the data — but a quick scan doesn't turn up anything that would help me to understand your apparent loss of faith.

JA: Begin compiling data: Jon Awbrey 21:56, 10 August 2006 (UTC)

I've looked through each of the edits for this article a few times now. Feel free to keep posting data here, but I'm not sure why you're doing it. You mentioned my "lack of faith". Don't get me wrong. I think that you made some somewhat useful changes to the page. I don't doubt your edits. I still will assume good faith in the changes you make to the article. What I have lost faith in, is what you say about the article.
Faith-in-JA things that McKay still has:
  1. JA's useful contributions to this article in the past.
  2. JA's ability to usefully contribute to the article still to come
Faith-in-JA things that McKay has lost:
  1. A discussion in the past about this article, regarding what should be in it.[#JAs_bold_claims_refuted.]
  2. That when JA left this article in March 06, there were experts working on the page.[#JAs_bold_claims_refuted.]
  3. that JA has read most of this new talk page (questions for him abound, e.g. anything about the new article that he doesn't like).
Now JA has said that he wants Relational database to be a "gateway article", and I think that the {{main}} template stuff does a decent job of that, just like you had done before. McKay 04:54, 11 August 2006 (UTC)

JA: Brief remark. It's likely that your target audience and my target audience are somewhat different populations. Aside from the usual suspects that we're both familiar with, the types of interdisciplinary discussions that I have been engaged in for the last couple of decades, whether in vocation or avocation, frequently require me to explain things like the theory of relations and relational databases to folks who have backgrounds mostly in humanistic and philosophical areas. How this comes about is too long a story for the moment. But when I came to WP and started writing articles that required the same sort of stretch across disciplines, the right sorts of links at the right sorts of levels were just nowhere to be found. Now, maybe you have seen enough flat database tables to imagine one on your own recognizance, but that is just not the case for everybody. And that is just one aspect of the need that I found myself needing to address in this or whatever article. Way 2 late. Jon Awbrey 05:16, 11 August 2006 (UTC)

So, what do you think needs to be added / changed to make the page more "cross-discipline friendly"? I don't think that a description of flat fliles should be included in this article, but a link to flat-file databases should be present, and made available on first use of the term. McKay 14:10, 11 August 2006 (UTC)