On Wed, Sep 11, 2002 at 12:37:58AM +0100, Matthew Toseland wrote: > It may make sense to implement several different types of store. > a) Native filesystem, with one file per key, and the filename derived > from the keyname (may need to escape some chars). Partially implemented > already, needs more implementation as well as debugging > b) Native filesystem, with a mapping from filename to keyname, and a > pool of truncated files (hack to support broken windoze JVMs). Would be > a descendant of the class implementing a). > c) Everything in a single file, the current datastore that we all know > and hate. Implemented fully but has the DSB. > > What are the relative urgencies of these options? Should I tackle a) > and b) before debugging c) ? If a) and b) are implemented, is c) even > necessary?
I don't think it makes sense to implement 2 or 3 different types of datastore - that would be a wasteful duplication of effort, we only need one - but obviously it must be one that works. To be honest, I don't know enough about how the current datastore is implemented to make a judgement call as to whether it would be easier to start from scratch than to debug it, but we need to settle on one course of action and focus on that, rather than diluting our efforts over several redundant paths. Ian. -- Ian Clarke ian at freenetproject.org Founder & Coordinator, The Freenet Project http://freenetproject.org/ Chief Technology Officer, Uprizer Inc. http://www.uprizer.com/ Personal Homepage http://locut.us/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 232 bytes Desc: not available URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20020910/f89b35f1/attachment.pgp>
