[ABE.pm] recap

Ruth A. Kramer rhkramer at fast.net
Wed Apr 14 01:13:02 CDT 2004


Ricardo SIGNES wrote:
> > > rjbs has a killer name for a pet project of his:  Acropolis.  Acropolis
> > > is a (proposed) system of books, articles, and comments that are
> > > revision controlled and all linked together.  I'm sure rjbs could give a
> > > better synopsis.

> > I'd like to find a tool that is a sort of combination of AskSam (free
> > form database, Windows only (but for several years now available as a
> > "web server", but, IIRC, no "explicit" revision control), a wiki
> > (TWiki is my current favorite, it can be used as a free form database,
> > and does have revision control), and a large text file with "custom"
> > delimiters (actually, TWiki ML, but much faster for data entry than
> > calling up a web page over a dialup line).
> 
> Explain how you'd use it? :)

1. I already use it (WikiLearn; personal TWiki (currently dead, hope to
revive after recovering data from crashed (?) disk); AskSam (used while
working at BSC for all sorts of work related notes (along with ZyIndex);
large text file (used now because updating WikiLearn over a dialup is so
slow).

2. It may take a while to explain, so I'll give a generality now and
maybe more after my taxes are done. ;-)  I'd like to use it to store
every piece of information I come across that I don't think will be
retained in my wetware storage.  This includes:
   * tidbits of facts that I haven't yet integrated into "understanding"
(snippets or entire emails, web pages, "links" to hard copy information,
..., ..., ...) (i.e., a learning notebook, and collaborative, I'd like
to let others read it, make use of it, and add to / correct it).
   * Possibly things like lists of books, CDs, that I either own, have
read / heard, like, dislike, whatever.  (I.e., both lists of, e.g.,
books, and short "book reviews" that might only help me remember that I
did read such and such a book (probably along with "spoilers" to help me
remember what it was about).)
   * Something like blogging of in-progress projects.
   * A code repository (for things I write or might help with), or
pointers to things like CVS trees for that code.
   * etc., etc., etc.

Changing from Windows to Linux made me abandom AskSam and ZyIndex for
the most part.  After a quite unrewarding search, I found wikis and
figured they were the closest thing I'd find in Linux (at least at the
time).  

Aside: AskSam is a free-format database, with good searching tools, and
available as a group tool (as early as 1985, IIRC, AskSam had a (local)
network version which I and some of colleagues used.  Somewhere along
the line it also became an Internet product, but fairly expensive, IIRC
($1500???).  ZyIndex is an indexing tool that will (would, don't know if
it is still sold / supported) index a variety of file types (mostly word
processors, plain text, and I have a vague recollection maybe it started
supporting spread sheets (or I kludged it somehow) for easy search. 
Made a point of putting all project documents I generated (specs,
requirements, design documents, etc.) on a PC and ZyIndexing them, along
with any documents I could beg, borrow, steal, or require as new
contracts were placed.

I find TWiki (both on the Internet and locally a little too cumbersome
for various reasons):

   * Slowness over dialup (being hosted at SourceForge doesn't help).
   * Find page, click edit, edit, preview, save cycle is too
cumbersome.  (In AskSam you simply found or started a record, edited
that record without a mode change (IIRC), and that record was
automatically saved when you "surfed" to a new record (or on demand,
IIRC).
   * Wikis require that you label the data before entering it (find an
existing page with a title or create a new page with a title).  AskSam
(and the plain text file) allow me to enter data without naming the
record immediately, I can do so after I enter the data.  This is very
helpful, as quite often I start recording some data (copy and paste an
email or snippet thereof) before I've really figured out what "topic" or
"category" I want it filed under.  
   * This also means that I have less need to find an existing record
(in either AskSam or the gigantic text file) to add data to -- as long
as I eventually name the record (or included explicit or implicit
keywords) similarly, I can access all data on a particular topic by
virtue of the sorting and searching functions.

Other things on the wishlist for such a tool would include lots of
special import tools to facilitate copy and paste of specific types of
data while maintaining proper attribution, and allowing editing and
commenting of that record before storing.  

For example, it would be nice to copy an email, and when I pasted it
into the "tool" it would parse and format it in a prespecified manner (I
might have more than one specification for various email lists or
categories) -- in general it would parse all the headings and save and
format only those specified, in a specified format, perhaps something
like this:

<quote>

---++ : : 

   * [[][]]; ; ; &mdash;
<blockquote><blockquote><pre>
Subject: Re: [ABE.pm] recap
Date: Wed, 14 Apr 2004 09:08:14 -0400
From:  Ricardo SIGNES <rjbs-perl-abe at lists.manxome.org>
</pre>
> > rjbs has a killer name for a pet project of his:  Acropolis.  Acropolis
> > is a (proposed) system of books, articles, and comments that are
> > revision controlled and all linked together.  I'm sure rjbs could give a
> > better synopsis.

> I'd like to find a tool that is a sort of combination of AskSam (free
> form database, Windows only (but for several years now available as a
> "web server", but, IIRC, no "explicit" revision control), a wiki
> (TWiki is my current favorite, it can be used as a free form database,
> and does have revision control), and a large text file with "custom"
> delimiters (actually, TWiki ML, but much faster for data entry than
> calling up a web page over a dialup line).

Explain how you'd use it? :)
</blockquote></blockquote>

</quote>

This is actually the format I use in my plain text "database".  Some of
the markup is TWiki ("---++") is a second level heading, other is HTML. 
Keeping records formatted like this is for a number of reasons, I needed
record delimiters of some sort, and in this format, a record (or
several) can easily be copy and pasted to a TWiki page.  Some day it
would be nice to have an automated import tool to move selected records
to a TWiki.  

Notice that I edited out some of the message headers and most of the
text.

The "---++" serves as a record delimiter, a reminder to enter a topic,
and an easy way to keep the database sorted alphabetically (using some
outline / fold macros I created for Nedit), maybe (in this case)
something like:

---++ Databases: Free Format: Acropolis

After that line I allow myself to add other keywords that I might want
to seach by, and anywhere within the record I can include my own free
format comments of any sort, so maybe the final "record" (in the text
"database") looks like this:

<quote>

---++ Database: Free Format: Acropolis??

This topic came up on the ABE.pm mail list, I've written to Ricardo to
find out more about what he had in mind:

<blockquote><blockquote><pre>
Subject: Re: [ABE.pm] recap
Date: Wed, 14 Apr 2004 09:08:14 -0400
From:  Ricardo SIGNES <rjbs-perl-abe at lists.manxome.org>
</pre>
> > rjbs has a killer name for a pet project of his:  Acropolis.  Acropolis
> > is a (proposed) system of books, articles, and comments that are
> > revision controlled and all linked together.  I'm sure rjbs could give a
> > better synopsis.

> I'd like to find a tool that is a sort of combination of AskSam (free
> form database, Windows only (but for several years now available as a
> "web server", but, IIRC, no "explicit" revision control), a wiki
> (TWiki is my current favorite, it can be used as a free form database,
> and does have revision control), and a large text file with "custom"
> delimiters (actually, TWiki ML, but much faster for data entry than
> calling up a web page over a dialup line).

Explain how you'd use it? :)
</blockquote></blockquote>

</quote>

Sorry for the length of this.

Randy Kramer



More information about the ABE-pm mailing list