i050809a: Navigational Data Model

Status	Date	Description

	2005-09-04	Version 0.75 is cited in the current thesis draft. If I make any changes I must updated the thesis draft to use the new version number unless I freeze i050809c and move on to i050809d. There are also improvements in the flow and simplicity in Appendix C of the dissertation that it would be useful to fold back into the on-line version. And, this could all wait until later after other under-nourished areas have been addressed.
	2005-09-04	0.75 Errata: The (Fowler & Scott) bibliographic entry is out of sequence.
	2005-09-03	Get permission to use the Barker quotation.
	2005-09-02	The diagrams should have their own linked pages where larger versions, perhaps with accessibility explanations too. The availability of the masters for the drawings could be indicated there.
	2005-09-01	Move the completed work items as of inclusion in the dissertation into an archive backup of this page, so that only open and newer work items remain.
	2005-09-01	Create a material package for all of the material as used in the dissertation thesis
	2005-09-01	Verify that everything works in taking the diagrams and text into the MS Word thesis document and from their into PDF.
	2005-09-01	Create a version in the dissertation appendix on the navigational data models.
	2005-09-01	In section 5.2.3, the identification of attributes or constituents should make no difference with regard to functional dependencies. It doesn't seem necessary to say anything about that, but keep an eye on it anyhow.
	2005-09-01	And for that matter, I might actually be able to make sense of MOF this way. Hmm. I wonder why I didn't think of that before.
	2005-09-01	Hmm, I just realized that the WebDAV data model might be expressible in this way too.
	2005-09-01	The section 1.1 notion about encompassing and the integrated notion of data types also applies to patters and their fusions too. This is an useful example to cite in the pattern-language description.
	2005-09-01	Make a link to TROST materials and bootstrapping in the i050809c synopsis.
	2005-08-31	Consider expanding section 1 into subsections with finer granularity of paragraph level, more like section 3.
	2005-08-27	Creating the Visio stencil for navigational data models is a project that can be done under TROSTing /dev and moved to SourceForge. That's the first development case that has come up.
	2005-08-27	Incorporate a version in the appendix of the thesis draft and confirm that it and the graphics render properly in MS Word and in PDF.
	2005-08-27	Publish on the site and let people know it is available for review. Hay and Shneiderman. Also Vinoski if this is one of the ways he runs into impedance mismatches. [dh:2005-09-03 and Bill Kent.]
	2005-08-27	Summarize, at the beginning, the intended use in patterns for TROSTing. Later on, connect to worked examples in patterns.
	2005-08-27	Illustrate the "overlay" case with permalinks and version progression, both of which will be implemented in the i050809c document [heh!].
	2005-08-27	Beside the few examples included here, there should be links to worked cases from different sections of the description.
	2005-08-27	The Information Architecture of ODMref 1.0 can be expressed this way, including as a mapping onto a conventional hierarchical file system (NTFS). Link to that when we get to it.
	2005-08-27	It would also be interesting to see whether this helps with "~~Metro~~XPS." It might be too hard. I'm not sure. There might be something like an ~~Metro~~XPS InfoSet. It certainly might help to depict Reach since the navigational map it works with ODA and some of the disk-dataset cases that Bachman looked at.
	2005-08-27	It would be great to build a navigational data model of the XML Information Set. It would also be interesting to create an RDF model in this manner that shows how RDF involves distributed knowledge of a sort.
	2005-08-24	Show off-page incoming connector in a non-data-navigational sense (it's a meta-diagram navigation). [dh:2005-08-27 I'm not sure this is appropriate. The from-page i case is different than whether there is navigation and which way. We want to keep this simpler.]
	2005-08-24	Create a Visio stencil for the diagrams I want.
	2005-08-21	Having a navigational structure coming down versus a relational association going up is a big deal.
done 0.75 2005-09-03	2005-09-02	0.60 Errata: The figures all need alternate texts to go with their captions.
done 0.75 2005-09-03	2005-09-02	0.60 Errata: 6.7 last sentence needs something more.
done 0.75 2005-09-03	2005-09-01	At some point, the diagrams should be redone with different typography for their use of data-entity names too. It should certainly be done by the time a Visio stencil is introduced.
done 0.75	2005-09-03	Along with the typographical clean-up and nuances around that, introduce "focal attention" and refer to Polanyi. Refer to Bill Kent's Data and Reality as pertinent here too.
done 0.75	2005-09-03	0.60 Errata: "This version: 0.50" should have been "... 0.60". This will be caught in the move to 0.75
done 0.75	2005-09-03	I wonder if the schematic use of variables should be in fixed-pitch just because the specific names are. I've done that in other schematic forms (function prototypes for ODMA, etc), and it makes the path-name example work better, I think. I will use that as the determining case. [dh:2005-09-03 I settled on fixed-pitch in the models, ordinary in the schematic, and italic or not in models depending on the nature of the element. It is a bit confused to speak about, but I trust it is only disruptive in the intended way in practice.]
done 0.75 2005-09-03	2005-09-02	I'm thinking that using a Hungarian prefix, like d for data would also be useful in differentiating data entities from the entities they are intended to carry data about as part of the application of the data model. This is worth doing as I update the Visio versions of the diagrams in preparation for making 0.65 and using a derivative of the material in my thesis draft. [dh:2005-09-03 I did that, although "e" would have been good too. I also used lower-case first letter consistently to see what that gets us. There are some cross-over cases I am not clear about and I don't know that it will ever be clean.]]
done 0.60 2005-09-02	2005-08-31	Update the calalog for i050809 when the cover page is updated.
done 0.60 2005-09-02	2005-09-01	Incorporate the overview as created on the blog as an overview for the i050809 folio cover page.
done 2005-09-02	2005-09-01	Blog about the Navigational Data Model
done 0.60	2005-09-01	Complete updating to 0.60 and publish to the hosted site.
done 0.60 2005-09-01	2005-08-27	When Martin's "Standard ~~symbols~~ Diagramming Techniques" arrives, double check for crows-feet there and see if that is also where I noticed them first (if not in Martin & McClure). [dh:2005-09-01 I can't believe this is where I saw them first, and maybe they were in Martin & McClure. But (Martin 1987) clears up a few things. It is odd, now, to see how much the ability to produce diagrams on the output devices of the time was influential in choice of approaches.
done 0.60	2005-09-01	Apply all changes noted on #47.15 and also make additional changes as noticed.
done 0.60	2005-09-01	Apply all changes noted on #47.14 and also make additional changes as noticed.
done 0.60	2005-09-01	0.50 Errata: 4. and elsewhere, switch to fixed-pitch normal font in text so that these namings are clearly about data artifacts, even though they suggest an external interpretation of the data as being about something in the world (conceptual or concrete).
done 0.60	2005-09-01	Add section 4.3 on typography to make clear why we use a form suggestive of formal, digital use.
done 0.60	2005-09-01	Convert all uses of data-entity type names in the text to fixed pitch fonts, further emphasizing that this is about digital data structures and schemes for them (mainly), and that the data-entity names are not references to entities in the world, even though the model may be so-interpreted.
done 0.60	2005-09-01	Apply all changes noted on #47.13 and also make additional changes as noticed.
done 0.60	2005-09-01	Apply corrections from #47.12 and clean up additional items noticed while making repairs.
done 0.60	2005-09-01	0.50 Errata: 1.2 "one-of" no hyphen
done 0.60	2005-09-01	0.50 Errata: 1.2 "so long as an no opposing arrow is ~~not~~ traversed."
done 0.60	2005-09-01	0.50 Errata: 1.2 "afforded directly by the structure under some tacitly-understood operation (section 5.1)."
done 0.60	2005-09-01	0.50 Errata: 1.1, first sentence "immediately-~~superior~~encompassing" and elsewhere.
done 0.60	2005-09-01	0.50 Errata: Synopsis "Because the TROST bootstrapping cases ..."
done	2005-09-01	Proofread i050809c 0.50 as it appears in my browser, placing notes in notebook #47.12-15.
done	2005-09-01	Review all citation links and all URLs to material on the internet in i050809c 0.50
done	2005-09-01	Check all permalinks and cross-references on i050809c 0.50.
done 0.50	2005-08-31	Update on the site so that 0.50 material can be incorporated in a blog announcement and then on the cover page of this InfoNote.
done 0.50	2005-08-31	Put the permalink in section 6.7 of 0.50.
done 0.50	2005-08-31	Correct the version number in the i050809c 0.50 current, latest, linkage block.
done 0.50	2005-08-31	Switch Diagramming 0.50 in as the lastest in the b-page.
done 2005-08-31	2005-08-21	These are similar to ERD and UML diagrams for data models. Point out the key difference in regard to navigation and relationships via foreign keys, and the different namings of links based on viewpoint (direction of navigation). (Internal relationships correspond to use of foreign keys as part of primary keys for pointing up-structure). [dh:2005-08-24 I may as well use the example that Hay provides in his comparison, as well as my favorite illustration of the so-called recursive parts explosion.]
done 2005-08-31	2005-08-22	Include a reference and resources and include the work that I did in the 70's through 90's. [dh:2005-08-26 The references are being completed and I scanned in the handwritten pages from 1993 where I still have the originals. These will be stored in the archive for i050809c.]
done 2005-08-31	2005-08-27	Fill in the basic organization so that I can start numbering sections and checking off completion of materials.
done 2005-08-31	2005-08-27	Introduce a qualified-name notation for constituent data-entity types. Add a section that explains this approach. ~~Refer to it from section 1.~~
done	2005-08-31	Cite (Whitehead & Russell 1962) for the use of converses of relations in conjunction with associations.
done	2005-08-31	Adapt the figure A-8c from (Hamilton 1993) as an illustration of associations.
done 2005-08-31	2005-08-27	Remember to feature that any of the block forms can have constituents although some of those are relatively uncommon.
done 2005-08-31	2005-08-28	Build a figure of block types, followed by definition of how relations are denoted by "tunnel" entrances.
done	2005-08-31	Add the Tsichritzis reference that I found in an older bibliography of mine. I have ordered the book.
done	2005-08-28	Use workgroup shared-file nodes as a way to illustrate the relative uniqueness of keys and the use of path/compound keys.
done 2005-08-28	2005-08-27	Build the table of connectors ~~from individual connector symbols to be kept on a stencil.~~
done	2005-08-28	Quote Barker on effective entity modeling.
done	2005-08-28	Add presence of (Hay 1996) Foreword to the citation of the book on i050809c. Make citation for (Barker 1996).
done	2005-08-28	Change "Barker" approach to "Barker-Ellis" approach, based on comments in the Foreword to (Hay 1996).
done 2005-08-27	2005-08-24	Make the hierarchic organization diagram from my (1993) fig. A3-1~~, and see how to make the elements part of a stencil (or at least anticipate that). I need them as separate Visio drawings, and then as components of stencils.~~
done	2005-08-27	Introduce Figure 1 and section 1 on the basic hierarchical structure of navigational data-entity types.
done	2005-08-27	Reflect the title change on the i050809b placeholder
done	2005-08-27	Retitle i050809c to Diagramming [placeholder] (with version to be added when we have one)
done	2005-08-27	Modify the Fig. A3-1 conversion so that the descending line to constituents extends into the open side of the encompassing data-entity box.
abandoned 2005-08-27	2005-08-22	Find a reference on navigational models; Probably in something of Gupta. [dh:2005-08-27 I could find nothing that gripping. I'll need to distinguish the navigational case and leave it at that for now. If a decent resource crops up, I will add it.]
done	2005-08-27	Find the ANSI/SPARC definition of levels so that I can cite it, or simply use the 4-level degrees of abstraction from Rob & Coronel. The HDM is slightly on the conceptual side of intermediate but not so conceptual as ERD or the Barker model. [dh:2005-08-27 Rob & Coronel is good enough.]
done 2005-08-26	2005-08-24	Provide citations for the James Martin database design book and the Martin & McClure book. [dh:2005-08-26 I found (Martin 1982) in the library, and that is good enough for initial calibration. I have another book on its way.]
done 2005-08-26	2005-08-21	Find an old punched-card or similar illustration (maybe from some RPG -- Report Program Generator -- materials) that indicates the physical analog of these, ignoring the foreign key usage. [dh:2005-08-26 I just drew a "stack" in a comparison diagram of the two connector styles.]
done 2005-08-26	2005-08-21	Find a reference to James Martin's Database book that had me introduce these so I could tell what I was looking at. [dh:2005-08-23 There two database books and a book on diagramming techniques that it may have been. I don't recall the Information Engineering view, but I did have the diagramming book and it would have been like me to look at several sources and then pull out a composite that did what I wanted. I keep thinking that the earlier James Martin Database books didn't use the crows-feet and that was what had me do something else, though it may also be that I adapted my span model after seeing crows-feet and imagining them as suggestive of spanning a set of things. I will probably never know, at this point.] [dh:2005-08-26 I found (Martin 1982) in the Seattle Central Library yesterday. That's good enough. There is a little polemic against crows feet, so Martin wasn't about to use them, it seems.]
done 2005-08-26	2005-08-24	Do the A3-1 diagram as a Visio diagram. Improve the connectors and geometry, alter the shapes somewhat.
done	2005-08-24	Use a subroutine-like structure for the off-page connector. This might better be a different shape rather than a decorated rectangle. Check the flow-chart symbols and see if that works. [dh:2005-08-24 Yes it still does.]
done	2005-08-24	Show one optional constituent, one none-or-more constituent
done	2005-08-24	Use metric dimensions and make sure they copy into Word in US units appropriately. (boxes are 30mm wide, 15mm high, with 10mm spacing vertically, 5mm space separation horizontally)
done	2005-08-24	Check Bachman & Bouvard (1972) for anything that may apply here or elsewhere from their architectural definition approach. [dh:2005-08-24 The ACM Digital Library doesn't have a scan of the pages, but the abstract suggests that it is not going to do more on data models, although (Bachman 1969, Bachman 1972) are both cited and a file processing example is used. We'll pass here.
done 2005-08-24	2005-08-23	There's an on-line article about making Crows-Feet diagrams with Visio. I want to look at that and see what I can do for my navigational data model drawing templates. [dh:2005-08-24 This is just using the Database Diagram template as furnished. We don't want to use that.]
done 2005-08-24	2005-08-23	Include a reference to the Hay book and also to his comparison of data models.
done	2005-08-22	Include what I can of Bachman's approach to data models (data-structure diagrams).
done	2005-08-22	Add references to Codd definition of data models.
done 2005-08-22	2005-08-21	Customize b- and c- pages as placeholders
done	2005-08-21	Create this page ready for recording Diary & Job Jar items of this InfoNote.