Bug#863578: : Re: Bug#863578: Initial upload to experimental should probably still merge history

Russ Allbery writes ("Re: : Re: Bug#863578: Initial upload to experimental should probably still merge history"): > When uploading a package with dgit to unstable for the first time, the > recommendation is to use --overwrite, which will add a pseudo-merge of the > existing h

Bug#863578: : Re: Bug#863578: Initial upload to experimental should probably still merge history

Ian Jackson writes: > I have reread this bug report and I'm afraid I don't understand what > you are trying to do, what you tried, and why the most obvious thing > didn't work. > AIUI you are trying to upload a package to experimental, which has > * never been uploaded with dgit > * never

Bug#863578: : Re: Bug#863578: Initial upload to experimental should probably still merge history

I have reread this bug report and I'm afraid I don't understand what you are trying to do, what you tried, and why the most obvious thing didn't work. AIUI you are trying to upload a package to experimental, which has * never been uploaded with dgit * never been uploaded to experimental I

Bug#863578: Initial upload to experimental should probably still merge history

On Sun, May 28, 2017 at 03:26:09PM -0700, Russ Allbery wrote: > Oh, hm. I actually think the second is better, isn't it? Since it allows > anyone who had cloned dgit's representation of 3.9.8.0 to update cleanly > to the current dgit tree. H. Now I'm quite confused with the

Bug#863578: Initial upload to experimental should probably still merge history

Sean Whitton writes: > For the first upload of policy 4.0.0.x to unstable, I'm suggesting > --deliberately-not-fast-forward so that the history is a linear > descendent of the upload you made to experimental today. With > --overwrite, instead, the parent of HEAD would

Bug#863578: Initial upload to experimental should probably still merge history

On Sun, May 28, 2017 at 03:12:11PM -0700, Russ Allbery wrote: > Oh, okay, I think I understand. And then, when uploading to unstable, it > would then work? Or would I need --overwrite on the upload to unstable? In the context of uploads that are new to a suite, it is best to think of

Bug#863578: Initial upload to experimental should probably still merge history

Sean Whitton writes: > On Sun, May 28, 2017 at 02:59:16PM -0700, Russ Allbery wrote: >> Does that merge the dgit history of the package from the unstable >> branch into the history in the way that dgit-maint-native implies is >> supposed to happen? Or by "be a

Bug#863578: Initial upload to experimental should probably still merge history

On Sun, May 28, 2017 at 02:59:16PM -0700, Russ Allbery wrote: > > This is already possible: have your git history be a descendent of > > unstable, and use --deliberately-not-fast-forward instead of > > --overwrite. Documentation of this is waiting in #856402. > > Ah, yes, I presume that would

Bug#863578: Initial upload to experimental should probably still merge history

Sean Whitton writes: > This is already possible: have your git history be a descendent of > unstable, and use --deliberately-not-fast-forward instead of > --overwrite. Documentation of this is waiting in #856402. Ah, yes, I presume that would work. Does that merge

Bug#863578: Initial upload to experimental should probably still merge history

On Sun, May 28, 2017 at 01:34:04PM -0700, Russ Allbery wrote: > If a package that hasn't been using dgit is uploaded using dgit to > experimental, when it hasn't been uploaded to experimental before, > one cannot use the --overwrite option and merge previous dgit history. > Presumably (I haven't

Bug#863578: Initial upload to experimental should probably still merge history

Package: dgit Version: 3.10 Severity: wishlist If a package that hasn't been using dgit is uploaded using dgit to experimental, when it hasn't been uploaded to experimental before, one cannot use the --overwrite option and merge previous dgit history. Presumably (I haven't gotten that far yet)