jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Brian Thompson <elephant...@gmail.com>
Subject Re: How to start thinking in JCR
Date Sat, 24 Mar 2007 03:35:19 GMT
Nice relational modeling - your data below would fit quite neatly into a 
database.  Keeping in mind that JCR nodes fall naturally into a tree 
structure, how are you thinking of organizing the nodes that will store 
your Users, BlogEntries, etc?

Also, do you have ideas on how to guarantee unique IDs for your data and 
maintain referential integrity?  In my mind, there are some obvious 
answers, but it may be more valuable for you as a learning exercise to 
ponder these questions a bit :)

It might also be valuable to consider doing a different sort of example 
application.  When I started investigating Jackrabbit, I saw one or two 
articles (with code snippets) that used a simple blog application as the 
example to demonstrate basic Jackrabbit operations.

Good luck on your GSOC project!

-Brian



Nandana Mihindukulasooriya wrote:
> Hi Xin,
>             Thanks a lot for giving me the starting point. Those
> discussions you started on your application are really valuable for me and
> I'm learning a lot from them.
>              As a start, If we consider only a simple blog first with least
> features we have users, blog entries and images attached to blog entries,
> comments and rating.
> 
> Users has
> 1.    User ID
> 2.    Nick name
> 3.    Email
> 4.    Password
> 
> Blog entry has
> 1.    BlogEntryID
> 2.    Owner  :  An existing user
> 3.    Text Content
> 4.    Image Attachments
> 5.    Rate
> 6.    Date created
> 
> Image Attachment has
> 1.    ImageID
> 2.    BlogEntryID : Image should be related with exiting blog entry
> 3.    Image : Binary Image
> 
> Comment
> 1.    CommentID
> 2.    BlogEntryID : Comment should be related with existing blog entry
> 3.    CommenterID : Comment should be made by existing user
> 4.    Comment : A text comment
> 5.    Date created
> 
>              All the ids used here will have to be unique.  And the
> specially mentioned fields must have referential integrity.
>              I would like to have discussion on what are options I have
> when I do object content mapping and what are the things I should consider
> in more general context . If there are more options what are pros and cons.
> I can spend some time on design phase as GSoc officially starts on April 9
> and the date scheduled to start coding is May 28. Even though I would start
> much before that, I would like to have a in-depth discussion as it would be
> good for the new comers to the community.
> 
> Thanks a lot,
> Nandana
> 

Mime
View raw message