Re: [PATCH 0/8] Making reflog modifications part of the transactions API

2014-12-14 Thread Michael Haggerty
On 12/12/2014 10:16 PM, ronnie sahlberg wrote: > On Fri, Dec 12, 2014 at 11:17 AM, Michael Haggerty > wrote: >> [...] >> What am I missing? > > My original idea was to clean up a bit of the reflog handling API and > have one single transaction API for all ref and reflog operations. > > Think f

Re: [PATCH 0/8] Making reflog modifications part of the transactions API

2014-12-12 Thread ronnie sahlberg
On Fri, Dec 12, 2014 at 11:17 AM, Michael Haggerty wrote: > On 12/06/2014 03:46 AM, Stefan Beller wrote: >> This goes on top of Michaels series. The idea of this series is make the >> reflogs being part of the transaction API, so it will be part of the contract >> of transaction_commit to either c

Re: [PATCH 0/8] Making reflog modifications part of the transactions API

2014-12-12 Thread Stefan Beller
On Fri, Dec 12, 2014 at 8:17 AM, Michael Haggerty wrote: > On 12/06/2014 03:46 AM, Stefan Beller wrote: >> This goes on top of Michaels series. The idea of this series is make the >> reflogs being part of the transaction API, so it will be part of the contract >> of transaction_commit to either co

Re: [PATCH 0/8] Making reflog modifications part of the transactions API

2014-12-12 Thread Michael Haggerty
On 12/06/2014 03:46 AM, Stefan Beller wrote: > This goes on top of Michaels series. The idea of this series is make the > reflogs being part of the transaction API, so it will be part of the contract > of transaction_commit to either commit all the changes or none at all. > > Currently when using

Re: [PATCH 0/8] Making reflog modifications part of the transactions API

2014-12-08 Thread Jonathan Nieder
Hi, Stefan Beller wrote: > How do we view an empty reflog, i.e. an empty file at logs/refs/heads/master? That's a good question. I'm a little concerned about what 'git reflog expire --updateref' would do in this case. It looks like the longstanding behavior is for 'git reflog expire' to expire

Re: [PATCH 0/8] Making reflog modifications part of the transactions API

2014-12-08 Thread Stefan Beller
So I reviewed and examined this series and run into some problems. How do we view an empty reflog, i.e. an empty file at logs/refs/heads/master? I was told this is not a valid state for a reflog. However even the test suite sometimes produces an empty reflog files. Below there is a patch, which hi

[PATCH 0/8] Making reflog modifications part of the transactions API

2014-12-05 Thread Stefan Beller
This goes on top of Michaels series. The idea of this series is make the reflogs being part of the transaction API, so it will be part of the contract of transaction_commit to either commit all the changes or none at all. Currently when using the transaction API to change refs, also reflogs are c