On Wed, Jun 10, 2009 at 1:20 PM, Ville M. Vainio <vivai...@gmail.com> wrote:
> > On Wed, Jun 10, 2009 at 8:37 PM, Edward K. Ream<edream...@gmail.com> > wrote: > > >> In summary, the "safe" way to recreate vnode uAs is storing both gnx > >> and archivedposition. If archivedposition fails, it can still look up > >> the node by gnx, provided that it has only one node with that gnx. If > >> it has several, then it's ok to fail. > >> > >> But, the only safe uA is the tnode uA... > > > > Can we do this for b2? (Might as well make b2 truly memorable :-) Iirc, > the > > code is already somewhere in Leo. Do have a link to the discussion? > > There is a proto for one version: > > Scripts-->@thin leoScripts.txt-->Prototypes-->pos_to_archive, > archive_to_pos > > But, this causes more lengthy archived positions than what would be > usable in xml files (gnx1.gnx2.gnx3...) Why do we care about the length? > currently p.archivedPosition gives 1.2.4. Perhaps it should be changed > to give 1.2.4:gnx. That way, resolveArchivedPosition could remain > compatible with old archived positions. Couldn't resolveArchivedPosition be made compatible in either case? > Shouldn't be too risky - and improving data integrity is worth it. I agree. > We > can still fail if the gnx can't be found in the subtree being scanned > (like we fail now when the position can't be found). Sure. Representation can't make up for truly missing data. Edward --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "leo-editor" group. To post to this group, send email to leo-editor@googlegroups.com To unsubscribe from this group, send email to leo-editor+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/leo-editor?hl=en -~----------~----~----~----~------~----~------~--~---