[ZODB-Dev] BTrees and Mutables, was Re: [IndexedCatalog] bug in default indexing

[Christian Reis]

On Mon, Mar 03, 2003 at 03:30:17PM -0500, Nicholas Henke wrote:
> On 03 Mar 2003 15:23:01 -0500
> Jeremy Hylton <jeremy@zope.com> wrote:
> 
> > On Mon, 2003-03-03 at 15:14, Christian Reis wrote:
> > > I don't understand how this could make any difference. Which
> > > instruction might trigger an exception? What should it matter if
> > > abort() is issued-- the state will be reset to the original state
> > > anyway, won't it?
> > 
> > When the state is reset, the connection reloads any objects that were
> > marked as modified.  If you modify a non-persistent sub-object and you
> > haven't marked the container as modified, the container will not be
> > reset.
> > 
> 
> Isn't that nice and goofy? Is there anyway that the indexes could change
> to be OI/IO/OO Sets/trees ( one variety or the other ) ?

(Thanks for the prompt answer, Jeremy!)

That is better addressed to me or J, since Jeremy mightn't know about
IndexedCatalog indexes ;) 

We could, but the lists are so much faster and smaller, I hardly think
it's a win. Now that we know the tricky issues related to lists, I think
lavish commenting around this area might make things safe enough for the
future. 

Take care,
--
Christian Reis, Senior Engineer, Async Open Source, Brazil.
http://async.com.br/~kiko/ | [+55 16] 261 2331 | NMFL