Thanks for the pointer. I briefly looked at this Product and the
CVSMixin Product as well, but have one requirement that unfortunately I
_didn't_ explain clearly in my proposal.
I am not using Zope, only the ZODB so a Zope Product will not help me.
My first goal with the proposal was to get feedback on if my idea is
correct and what implementation problems could occur, i.e. transactions
and file locking.
My second goal is to inspire someone to implement it and merge it into
the official ZODB. I think that the idea has enough merit and could be
useful for a large number of applications.
We might have the time and need to implement PersistentBlob ourselves,
but I can't say for sure now.
Gregor Heine wrote:
> > Hi all,
> > Yesterday I put up a new proposal on the http://dev.zope.org site that
> > documents my vision of PersistentBlob. The main idea is the be able to
> > mark blocks of strings (text or binary) content for storage in a file
> > system as individual files. (Or blobs in a relation table.)
> > The two motivations for this are keeping a ZODB FileStorage small and
> > exposing blocks of string content to external processes (i.e. full-text
> > indexing).
> > Thanks for taking a look.
> > John
> Hi John!
> Have you had a look at the ExtFile Product
> It basically does all the things you want to achive with the PersistentBlob
> (i.e. storing files externally and making them accessible for external
> processes), except that it's not that deeply integrated into the Zope
> machinery and stores the files only in the file system and not in an RDBMS
> (which nevertheless could be implemented).
> Cheers, Gregor!
John D. Heintz
Zope-Dev maillist - [EMAIL PROTECTED]
** No cross posts or HTML encoding! **
(Related lists -