Aubin Paul wrote:
On Fri, Jan 23, 2004 at 07:48:32PM -0500, Michael Ruelle wrote:
  
Having gone through in my professional service the transition from one
database to another I would advise against this. I was going to bring
this up with dischi, but will state my piece here.
    

Considering most of the db stuff has been mine so far, that's probably
best :)

  
I would suggest the plugin stategy, we have used successfully
elsewhere.  Failing that I would rather see a rpcxml (or some other web
service API) or at the very least the judicious use of the Business
Delegate design pattern to keep the SQL as far away as possible from
real code and just use abstract methods and objects from our software
rather than some class from the database interaction layer.
    

That's probably fine, but I'm still not sold on a network-aware DB
like MySQL.... the benefits are questionable when the webserver,
recordserver and GUI core are all happy to talk over HTTP/XML...
  
I think that it would be pretty nice to have the option to have an external database accesable from other systems. Just such a simple thing as my media library is remote means I would like to have my database remote. I think of freevo as thin frontend to my media library. Something I want to access from various places. I would like to be able to integrate freevo into my webpage (to complex to fit inside freevo) to be able to get things like "top 10 songs listend to last week". I would also like to have my music store index when I add new songs to improve speed. This might be posible with freevo, but again my media store is not located on the same computer that runs freevo. There are several other reasons as well so some form of pluginable persistance I think would be nice.

BTW are there any where I might find info in what the current DB persistance does/how to extend it ?
I have quite alot of features I would like to see added to it and am currently builiding my own, and it seems pretty silly do the work twice :)
What I would like to store inside the DB is such things like all ID3v2 tags, playlists, etc, etc...

// Michael Medin
  
This we can define the interface and not have to conform to some
individual piece of SQL softwares view of what the standard is. And
believe me they widely vary. And can use the best thing available. Also
someone may even what to use an object database too (right now it maybe
a stretch but one rarely gets a straight answer from the magic
eightball).
    

This is true, but also consider the fact that what is happening right
now is that SQLite is used to log music ratings and play times. None
of this is particularly complicated, but I'm using a database because
performing queries against a pantheon of mmpython cache files is time
consuming.

I don't like the idea of remote DB calls, and I think using the
existing network-aware functionality to handle network transparancy is
more reasonable. If we put the TV guide into a database, SQLite is a
good option because it's a single file with minimal requirements on
the user end. 

It's faster than a pickle and it's easier than MySQL.


-------------------------------------------------------
The SF.Net email is sponsored by EclipseCon 2004
Premiere Conference on Open Tools Development and Integration
See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
http://www.eclipsecon.org/osdn
_______________________________________________
Freevo-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/freevo-devel
  

Reply via email to