Uhmmm....
What a nice dream.
After I awoke from my Scyther-induced slumber I decided to work a little more on the database model. I was working on some details like what information does a story convey, how the chapters are separated and how a private message system can be assembled. So I think I can safely say the model is quite advanced already and is at least
safe enough to be loaded onto the computer systems. Then I will at least try to create some basic test pages (login, search and the like).
In the meanwhile I'll leave you all with a sample of the model. Maybe you won't understand a thing, but that's why I'm attaching a picture and some explanations about how the model works.
So, here it is:
«The Quality Fanfiction Archive Database Model, Mark 1»:
[ASSISTANCE="Database Sample (nicely explained)"]
The model consists of three
sections: the site features, the site members, and the site contents (the actual fanfics).
The basic unit of participation is a "
Member" or user. Basic information about the user is registered, and an optional "Profile" with contact information or preferences can be attached.
A Member can opt in to become an "
Author", that is, a user who publishes stories. In that case the Author can have (at least) a
penname or seudonym that can differ from their actual username (for artistic/anonimity/marketing purposes).
A Member can also opt-in to become a "
Juror", that is, a user who reviews and rates stories. I' still working on how will Jurors be selected, although the base premise is that of previous works and peer acknowledgement.
The
central unit of content is that of a "
Story", which is essentially a published (or in-publishing) work by a specific Author. An Author can publish one or more Stories.
A Story has (contains) "
Chapters", which are the actual content of the story, fragmented in units of publishing and development. Each chapter can be given an unique name for the system to automatically create a Table of Contents.
Two items I'm still working on are "Genre" and "Tag". A "Genre" is essentially a unique element of qualification for a work; that is, a Story can be linked to only one Genre. For example, "romance", "action", etc. Whereas a "Tag" is a
user-defined label attached to stories by their readers, to ease clasification and searching. For example, readers may attach the tag "Shaymin" to all stories where a Shaymin is prominently portrayed, or may attach the tag "deathpool" to stories where a lot of characters die.
I'm still trying to find the adequate method to link tags to stories.
A
Judge (or, specifically, a group of them) will review a story and rate it for the site, emitting a "
Evaluation Report". Each one of the Juror's reviews will be graded acording to a score (probably jeffback's grading system), and will feature comments about the development, flaws or interesting points. Review scores will be averaged and the resulting number will be stored as the published work's "site score". A Juror can evalute many stories (although with the limitations of the model, they won't be currently able to review the same story twice), and a Story can and will be evaluated by at least one Juror.
Other minor elements I am working for the site are:
- Private Messaging
- Author Profiles (as in self-published information)
- Guilds (Author groups/clans based on common publishing styles such as genre)
- Site Wall (think Facebook's wall, but for full-open conversations)
- Support for chapters as attachments, supporting the HTML and ODT formats as FF.net does because those formats are free, well documented and easy to parse.
- Semiautomatic announcement system: a feature which will allow the site to "publish to itself" regarding site status, downtime, publishing issues and the like.
Timeframe for development of the current model: 14 to 30 days.
ETA for the next model: at least 2 months ( = Jan 2009)
[/ASSISTANCE]
I hope this computer-sciences mumbo jumbo won't get you dizzied. I just wanted to report, since I see all the other people hands on (
sp?) the job. And also so that you can have a less abstract model of operations and features to work on with new ideas.
Well... no more mumbo-jumbo. Oh wait... the model screenshot is attached.
Now I'll go away for a while... keep rocking on, people!