Disney-comics Database

[Fredrik Ekman]

Harry wrote:
> Here's my proposal for a Disney-comics Database format.

And here are my comments. Let me just say first that I think you have
done a brilliant job on this so far. Most of my comments are just
minor remarks.

> Story code
> ~~~~~~~~~~
> [...]
> For the other publishers, we use mainly the codes they deliver themselves,
> maybe with some normalization:

Yes, some sort of normalization would be necessary. Perhaps also a
reasonably "smart" program to take care of such normalization on behalf
of the user (ie, you don't have to remember the exact format for a code
when you make a search).

> Database files
> ~~~~~~~~~~~~~~
> [...]
> (Our 'don-rosa' and 'van-horn' lists will not be generated from the database,
> because they have a lot of extra information that doesn't fit (?))

At least we'll keep them until we see that we CAN generate more or less
identical files from the database.

> Story files
> ~~~~~~~~~~~
> These files contain one line per story with information that can be used
> in the various lists, and zero or more lines with extra information that
> will not be accessible for other lists (e.g. exact dates for pencilers/
> inkers in the Mickey Mouse newspaper files).

Why not just state that every line beginning with a semi-colon (or
whatever) is treated as a comment and let this be true for the issue
files as well? That way you can squeeze in all the extra info you
want.

> Data fields:
> 
> [...]
> - number of entire pages (2 characters should be enough...)

Should is the keyword here. However, as you know, there are some Italian
comics with well over 100 pages. True, these are mostly split up in
several parts, each with their own unique code. BUT, when some of these
are published in thick paperbacks in Sweden and other countries, they
are in one part and often with all codes save the first removed. Thus,
if some lunatic were to index these in the future and did not know (or
did not care) about this, he/she would need three characters. So if we
can afford three characters for this field, I suggest we use that, just
to make sure.

> - indication for number of broken pages:
>   space = none, '+' = somewhere between 0 and 1 (need we be more exact?)

No, '+' should do.

> - (start) date, in the form yy/mm/dd

Isn't the international standard yy-mm-dd? If I didn't get some kind of
explanation, yy/mm/dd would confuse me at least.

> - title hero (2 [or 3?] characters. DD, US etc. We'll need a list of
>   Disney characters. I like Tryg's abbreviation for Gyro: GY is better than
>   the GG or GyG we used so far.

Two characters should be enough. Little Bad Wolf and a few others could
cause problems, but LW is good enough, I think. I, too, like GY.

> - title tag (2 characters):
>   T: original title
>   R: story has no original title, title of one of the reprints given
>   D: story never had a title, some "official" Gladstone description given
>   O: story never had a title, the indexer gives an own description
>   H: Dutch title or description (I plan to create H-lists and D-lists from
>      my Dutch index, but I don't plan to translate all the titles! If a
>      story is reprinted in the USA, that title can be listed. Or should
>      we list both?)
> 
> - title or description
>   I know Per prefers additions like 'in' and quotes, as in:
> 
>   T:DD in "No Such Varmint".
> 
>   I myself prefer a title without any additions, and all lower case, like:
> 
>   T:no such varmint
> 
>   It seems desirable to separate the title hero from the (rest of the) title.
>   That would mean titles like:
> 
>   T:in "No Such Varmint"
> 
>   which I think makes less sense.
>   How shall we do this?

This is where I have my only major disagreement with you. I think this
should be in the issue files. Unnecessary? Well, that field will be
filled with spaces in the issue files anyway, right? And that way, it
will be easier to fit in (say) the Danish titles of Ole's index or any
other future non-american indices. The story files would include a
description, normally the original English title in the same form that
you propose above. The issue file title would then be in Per's format
(which I too prefer, if we have to go with your proposal). The obvious
disadvantage with all this would obviously be that someone will have
to type in BOTH the story file description AND the issue file title
for any new entries. Well, you can't have everything. :-)

I'd like your comments on this one, Harry.

> - remarks
>   Is it OK to put Tryg's information about occurring characters here, or
>   should it be a separate entry? Remember: we only have that information
>   for a very limited set of stories. It seems not suitable to reserve
>   a special field for this in all story-files.

Why not just on comment lines like those I suggested above?

> Issue files
> ~~~~~~~~~~~

No comments here, except for those already given above.

> Current issue files (to be transformed):
> - gladstone-disney
> - swedish (from Crone's list, and maybe from the NAFS(k) list on ftp)
> - danish (only if Ole wants to do it...)
> - dutch

If noone beats me to it (I've quite a lot to do these days) I will also
make issue files from the Swedish info in the Rosa and Van Horn indices,
so that it will be possible to make database searches on those stories,
at least.

> That's it for now. Let the reviews begin...

The only thing that I really missed from your post was info about how
you plan to make the programs, especially the user interface. Will you
have command line parameters, or do you plan to invent your own
database language? The latter would certainly be preferrable if we
still plan to make an email service out of this (like someone proposed
when this was discussed in the past). Also; will the user be able to
do case-insensitive searches and the like?

Another feature I'd like to see in the program(s) is generation of
(optional) legends for each output produced. Such a legend would
consist of a short natural language description of what the output
consists of and explanations for used abbreviations. Here is an
nonsense example of how that could look:

"This file is created with the [name of program, et c]. It contains
information about:

Stories by Carl Barks published in WDC 100-200.

Used abbreviations for creators:

CB = Carl Barks

Used abbreviations for comics:

WDC = Walt Disney's Comics and Stories

Used abbreviations for characters:

DD = Donald Duck
[et c]"

Info of this sort is really quite easy to generate. It would of course
require that the program have access not only to the story/issue files
but also to the files explaining all those abbreviations.

  /F