Hi all,

ok, here's my first attempt for the wiki basics. I don't know if I'm on
the right track, so please anybody with experiences in wiki usage give
your comments.

The E/R sources can be seen at "http://picolisp.com/5000/@src?er.l";.

Currently, there are just two subsystems: Role/Users and Content. The
User will probably need to be extended in the future with further
attributes. Currently we have name, email and picture.

Content consists of Categories, Threads, Documents, and Markup Text.
Categories will be things like "Forum", "Articles/Essays", "News", etc.
Each category may have many threads, and a thread consists of a number
of documents.

Initially, I was tempted to put the markup text of a document simply
into a blob. But I think the current version, where we have a separate
"markup" object holding a list of strings, has several advantages:

- It is faster, because a blob is a separate file that needs to be
  opened and closed.

- It saves space, as it allocates (and grows by) the DB block size of
  that class (currently 1024 bytes), while a blob file needs an inode
  entry and uses larger block sizes (disk blocks, 8k?).

- It is not a flat text, but a (+List +String) relation. So it has a
  certain structure which may be handy in further processing. Each
  string could be a line, or a paragraph, for example.

Each markup object has a joint to the previous version. This is to
maintain an edit history. Each version has user, date and time, and is
also searchable by these attributes.

Do these structures make sense? And if so, what else is needed?


As before, this system is running at "http://wiki.picolisp.com";. You can
log in as "admin" / "admin".

The Content stuff still has no GUI (perhaps after the weekend). The
Roles and Users are functional so far. Everybody is invited to create
his own login account, add a photo, etc.!

Cheers,
- Alex
-- 
UNSUBSCRIBE: mailto:picol...@software-lab.de?subject=unsubscribe

Reply via email to