To give the background for the (possible) development
of LM data connectivity, I will describe the
present situation and some possible upcoming improvements.


TODAY
-----

1) LM ------- Paradox Data Base System
|
--- XML files ------------> LM
|
--- Dump text files




This is the present situation (LM2000 and LM Pro alike).

The primary database storage is a conventional PC data base.
It is a competitor (or rather forerunner) to MS Access.

Being within the Borland environment, we use a system called "BDE"
instead of "ODBC" to connect between the Application (LM.exe) and
the database. The Paradox data base consists of separate files; a
little more oldfashioned than the Access ".mdb "(But do not bother;
Paradox and Access are old reliable working horses. Viewed
from a program like LM.exe, they just do what to be expected
from any of them.)

LM can print all item texts to a huge text file (Utilities menu).
This is a primitive way to tap the information or save it
some other way.

LM can also export single items or the items presently on the desktop
to a XML file, which in turn can be imported to (another)
LM data base. Words and concepts will be matched as exact
as possible.

This export-import system does not pass on the LM projects that
might be attached to the LM items.


A XML file is a text file, very similar to a HTML document.

XML files are very primitive and space-consuming
in a way, but since they are "flat texts", they can go
everywhere.



TOMORROW?
---------

2) LM ----- Paradox Data Base System
|
--- XML files ----(items)--------> LM
|
--- Dump text files
|
--- XML outline files ---(complete LM Outline)--> LM
|
--- Public format of other software --> Other program


LM outlines can be exported with all data required.
That is, an outline in Machine 1 will export the outline
to XML. When this XML is imported in Machine 2, the
outline will appear exactly as in Machine 1.

(Here we talk about LM Pro 1.1 outlines with a text pane.)

Depending on the matching process, the connections to existing
LM Dictionary and set of LM Concepts may be of varying quality.

(Other plans for footnotes/references and more, may make the
outline in a coming LM Pro 1.2 look even more like a readable
and workable ebook.)


In addition to this, it could be possible to export/import
LM Outlines to and from other established programs.

Since the Outline in program 1 contains other things than
the Outline in program 2, the exchange could not be complete.

However, requirements are that the proprietary formats of
such programs are public and that there are no legal
complications in connecting to them.

This is why we need more information than just a wishlist
containing popular outliners -- the data format must be
open and workable.

A first glimpse on "TreePad" gives the impression that this
is a candidate.


I have looked a little at works with general (XML) outline
standards. My impression is that this is of limited
interest -- the Outline idea is so variable that
the proposed standards appear as very specific.

I hope that this is of interest to some of you.

Hälsningar

Gunnar Sommestad