Shawn & Guys to fill in the blanks...

Having had some good inroads with Allaire / Macromedia I managed to to get
sneak a sneak insight of whats going on (not that anyone with their head to
the ground wont already know :-)

I have had a thread with Jeremy regarding the future of the Spectra COAPI
(Greylock) which will be built into the next CF server (Neo) ... here is the
plan...


Jeremy Allaire
----------------------------------------------
Hi Neil-

Great to hear from you.

Fundamentally, the COAPI was about structured programming and the
abstraction of data persistence/management from the developer.  Application,
data and component portability were also side-effects, if not goals of the
model.

These are all things we're carrying forward with Greylock, but we're trying
to fix known problems with the existing model, such as:

- WDDX storage made representing relationships difficult and often slow
- WDDX storage made using standard SQL for custom searching cumbersome and
also inflexible
- WDDX storage did not fully take advantage of the databases indexing engine

With Greylock, we do all the hard work of generating standard relational
schema, still preserving content ID uniqueness with GUIDs (therefore
allowing transparent component packaging and re-deployment), and managing a
lot of the inheritence and relationship data within the runtime (in EJB)
instead of in SQL against the WDDX object store.

This allows users to directly SQL against the database, and allows
third-party applications to more easily interoperate with the data (e.g.
reporting tools, etc.).

Addition
---------------------------------------
<jeremy>
We may ship with a database you can use (e.g. SQL Anywhere), but the APIs
will work with SQL Server and Oracle, at least.  The APIs will support type
packaging and deployment to/from any server that has Greylock.
</jeremy>
----------------------------------

So there you go, from the horses mouth...what can we extrapolate from it:
well thus: (snippets of text from a post from a MM engineer working on Neo)

The new API will form abstraction from the underlying queries and will
protect the user between versions of the schema. Also, the API will be much
more portable between projects.

The schema will be relational; a major addition will be to ensure that users
wont have to write their own SQL against a datasource. The API will do it
all - and yep, the API has taken into consideration more than just
"performance concerns" for why people wrote their own queries.

Users also wanted to do conditionals and their own sorting etc. But users
should forget about directly querying the relational model. Trust me,
without giving too much away, once you see greylock and all it can do for
you above the database layer, you won't care about the SQL!

Greylock will support many databases, not just the few it could with Spectra
1.x.x

These are extracts from mails from MM and they are bound by NDA, I have
included them without malice, only to tell the new list that MM have not
abandoned the CMS model they had with Spectra, merely they are enhancing it.

Hope this all helps, lets get rocking and start building :-)

Neil


<!------------------------------
Neil Clark
Senior Web Applications Engineer
Spectra / CF / XML / Cerebro
Macromedia Alliance Partner
MCB Digital Limited
[T]: 020 8941 3232
[F]: 020 8941 4333
[M]: 07719 448 227
[E]: [EMAIL PROTECTED]
[www] www.fluidik.com
-------------------------------->

'A goal is not always meant to be reached, it often serves simply as
something to aim at.'


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Structure your ColdFusion code with Fusebox. Get the official book at 
http://www.fusionauthority.com/bkinfo.cfm
Archives: http://www.mail-archive.com/[email protected]/
Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists

Reply via email to