Re: [O] Best practice for providing an Org-based application?

[Jean Louis]

* John Kitchin <address@hidden> [2019-09-09 18:48]:
> It would be helpful to get a better sense of what is the private
> data, e.g. is it something like org-contacts, or some structured
> data in a heading? and what are the workflows that might be generic.

My personal problem is that Org files are scattered around, in the
file system. Some people choose to have one or few Org files for many
activities. I have it different, I have one Org file per person, per
company and per more complex project, such project alone produced PDF
of 116 pages or more. I cannot mix various activities in one or few
Org files.

All my contacts are in the database, thus opening of Org file does not
depends on my decision, rather on a simple algorithm, so I am not
"finding a file", rather "finding a person" and then deciding to open
the Org file belonging to the person, or to open the directory
belonging to person, or to account (group, company).

It is based on ID numbers, thus /some/directory/People/1/1.org becomes
the Org file for the person. I am not browsing directories to find it,
computer does that and just opens the file for me.

Org-contacts would never work for me, I have 192,000 contacts to
manage. It works only with the SQL (PostgreSQL) database in the
background. 

> I have used org for lots of organizational things in the past,
> ranging from conference organization, teaching classes, running job
> searches, organizing special issues in scientific publications and
> lately to organize some things for a cub scout troop. They have all
> been pretty different, but I am sure I have reinvented pieces of it
> each time. I am interested in learning more about what you are
> doing.

As you mentioned structured data, I know that I work better with
structured data, for me is that very necessary. For purposes of finely
grained hyperlinking, which can be done with the Org mode, I am
developing system HyperScope, when it becomes ready for public I will
publish it.

It is based on principles as envisioned by Doug Engelbart:
http://dougengelbart.org/content/view/140/000/ -- but the system he
had for finely grained hyperlinking was proprietary software, and
later sold to somebody.

This system could be considered finely grained data tree based driven
menu of hyperlinks. It is based on easier, structured data in the
database, opposite to files.

One video is demonstrating how it is using author names to find other
entries of the same author:
https://open.tube/videos/watch/ed9c47e9-29b7-4624-8d71-3bca68078a0e
and same is valid for tags.

Hyperlinks can be to anything possible, it could be notes, web links,
YouTube links, PDF links, especially PDF by query links or PDF by page
number links.

It is a thought processor that helps in finely grained navigation,
viewing and linking of documents or collections of documents.

If I write a link in the Org file, that link is just written in that
one Org file. There are searches that can be performed, but the link
is usually nowhere else indexed or classified. Same link I need to
repeat for many other people in their Org files, so that means I would
need to keep them somehow in Org files indexed, or in any other
system, so that I can find the link and place it to their Org file
when necessary.

Instead, I keep links in the structured SQL database, PostgreSQL and
collections of documents are organized in the data tree. When
necessary, I can quickly choose structured groups of links by their
tags, categories, groups, and create an Org file, which then can be
inserted or moved to better location.

Here is video demonstrating quick creation of Org files:
https://open.tube/videos/watch/d6a03e79-8b33-4d1f-b840-6a5df0f5f802

Then of course I can have database backed Org files as well.

Or I can link from one list of hyperlinks to various other Org files
by their section heading or by query search or similar.

More important than software is the concept of finely grained
hyperlinking. You read a PDF file, but there is just few pages that
you wish to bookmark. Rarely some PDF reader allows some bookmarking,
but even then, such is not centrally managed, bookmarks are rather
somehow related to files. They cannot be easily shared or published,
or converted into something usable.

Then there are hundreds and thousands of PDF files with their
corresponding thousands of pages, but not all of the knowledge in one
book is useful, maybe just few pieces could be useful in certain
application.

It would be nice to write it all in the Org file, yet for me such
data shall better be in the database. And links are then copied from
central place to Org files or any other files finally, or to email, or
shared by SMS, or published as web pages when necessary.

Jean