- From: Jeremy Orlow <jorlow@chromium.org>
- Date: Thu, 30 Sep 2010 10:16:44 +0100
- To: Jonas Sicking <jonas@sicking.cc>
- Cc: public-webapps WG <public-webapps@w3.org>
- Message-ID: <AANLkTikPZWW0N-TJNHnc4ejdfWAR2uDrYVVCg9S4w5OU@mail.gmail.com>
Hm. Actually, I think I like Jonas' proposal better than 1 or 2 (and Shawn, I see your point about why 2 is better than 1). I also think that Shawn's example of doing multiple schema upgrades should still work 3 since we fire onsuccesses in the order that items were queued up. Will file a bug in a day or two if no other comments. J On Wed, Sep 29, 2010 at 10:59 PM, Jonas Sicking <jonas@sicking.cc> wrote: > On Wed, Sep 29, 2010 at 6:12 AM, Jeremy Orlow <jorlow@chromium.org> wrote: > > What should we do when setVersion is called while a setVersion > transaction > > is currently active? > > Off the top of my head, I can think of two behaviors we might want to > spec: > > 1) Have the subsequent setVersion simply throw an error. 2) Have the > > subsequent setVersion adopt the existing setVersion transaction and > change > > the version. (i.e. whatever the last setVersion call sets as the version > > string will win.) Any others? What do you guys think is the most sane > > behavior? > > My initial reaction was > > 3) Schedule another version transaction which is started after the > currently running version transaction (and any other already scheduled > transactions) are done running. > > That was actually my initial reaction, though that is biased by what > our implementation naturally would do unless special care is taken to > do something else. > > In general I don't feel strongly either way and am fine with all > currently proposed solutions. > > / Jonas >
Received on Thursday, 30 September 2010 09:18:02 UTC