Please do not take my continued questioning as a sign of not liking the vfs -> jdbc provider. I just like to question because I still don't completely understand.
> E.g. if it is not possible to setup a ftp/ssh server but already have a working > database connection (e.g. in webshop environments) it might be nice if > VFS could use this database as filestore e.g. for pictures too. If I have a webapp getting it's data from a database and that same database also has the pictures in it. Then how is it easier to get the picture data through vfs layer and the other data directly through jdbc? > Or maybe in an distributed environment. Multiple clients working with > the same filesystem - again without setup for a fileserver. So jdbc is seen here more like a transport protocol rather than vendor independent api for retrieving data from sql database. This is possible but the transport protocol is side effect of jdbc. But I accept it. Still very often if someone opens for you the not so secure jdbc port(s) (for DB2/AS400 jdbc driver one needs to open around 10 ports) he is quite willing to open ssh port. > Or what about fileupload forms? We could use the database as storage and > dont have the risk someone uploads and executes a trojan. Hmm... > We could leave the reorganisation to the database administrator. I'm sure he's going to be happy about it :) > It introduces no new dependency so why NOT do it and see if there are > others finding other usages for it. It just seems to expand vfs into the area of universal access interface. Next thing you know we access ldap servers, mail servers over pop3 etc. Now it has been accessing filesystems in heterogenous and distributed environment. Maybe I am a little afraid of bloatware. zip, jar and other kinds of providers have been still somewhat filesystem oriented. > I'm interested in using it e.g. to store the > thumbnails that are being used in one of my applications. Is it somehow easier to store data through vfs api than through jdbc api? > Also for some versioning systems it may prove > useful. This I understand. Because database is more flexible than more or less predefined filesystem structure, database enables someone to build a more versatile filesystem that can support a lot of metadata (attributes) for files. And we could think history info of a file as it's metadata. But... Ok, I leave this party spoiling now and hope success and fun for this provider. I just could not keep my mouth :) - rami --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
