Re: Backward compatibility (was Re: Announcing Newt DB, the amphibious database)
Jim Fulton <[email protected]>
| Newsgroups | gmane.comp.web.zope.zodb |
|---|---|
| Message-ID | <CAPDm-FhZ8MQ_u4wt+uxeu3y0o_DRn7T8YP6frun9ELE-DVdWpg@mail.gmail.com> |
On Thu, Feb 23, 2017 at 11:47 PM, Jim Fulton <[email protected]> wrote: > > > On Thu, Feb 23, 2017 at 2:51 PM, Hanno Schlichting <[email protected]> > wrote: > >> On Thu, Feb 23, 2017, at 20:22, David Glick (Glick Software) wrote: >> >> If you make a branch of transaction with your proposed fix, I can run the >> Plone tests with that (unreleased) plus current releases of >> ZODB/ZEO/BTrees. >> >> >> I just tried updating Zope 2.13 to the latest >> ZODB/ZEO/BTrees/persistent/transaction releases. By default it still >> used ZODB 3.10, so didn't even have the split into separate projects yet. >> >> I hacked transaction locally to accept both None and forcefully decode >> bytes from utf-8 for the transaction metadata. That made most tests work. >> >> But it does still fail with import errors for tempstorage. I tried to >> update tempstorage master at some point but had to give up, as I don't know >> enough about the ZODB protocol. This project implements a full custom >> storage, main code is at https://github.com/zopefoundat >> ion/tempstorage/blob/master/src/tempstorage/TemporaryStorage.py >> >> Unfortunately in Zope 2.13 one of those temp storages is always present >> and added to the ZODB, so even though nobody really uses it anymore, it's >> still there. >> > By "nobody really uses it anymore", I assume you mean, "some people use it". :) Is that right? If not, why do we care if there are failing tests? We could just remove them. Jim -- Jim Fulton http://jimfulton.info -- You received this message because you are subscribed to the Google Groups "zodb" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.