b4 tui: Old serieses not removed

b4 tui: Old serieses not removed

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 506 bytes --]

Hi,

As part of an effort to sync review state between machines I have
arranged for the set of b4/review branches on the destination machine to
be the same as it was on the source.  When I open b4 tui and run 'U' to
update all branches it reports the existance of several hundred more
branches, looking at them going past and what's in the UI I think most
but not all of them are thanked branches which I later archived on the
source machine, the branches are gone but they remain in the UI.

Thanks,
Mark

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: b4 tui: Old serieses not removed

[Konstantin Ryabitsev]

On Mon, Apr 27, 2026 at 04:56:52AM +0900, Mark Brown wrote:
> As part of an effort to sync review state between machines I have
> arranged for the set of b4/review branches on the destination machine to
> be the same as it was on the source.

So, let's call these "machineA" and "machineB". You do most work on machineA,
but machineB runs a periodic git fetch to get all the review branches from
machineA, correct?

> When I open b4 tui and run 'U' to
> update all branches it reports the existance of several hundred more
> branches, looking at them going past and what's in the UI I think most
> but not all of them are thanked branches which I later archived on the
> source machine, the branches are gone but they remain in the UI.

This is the part that's confusing to me, so I need clarification. I assume the
following happened:

- These series were reviewed on machineA (creating b4/review/* branches for
  them).
- MachineB runs periodic fetches from machineA, which creates these
  b4/review/* branches on machineB.
- The seriews are applied, thanked, and archived on machineA
- When you open "b4 review tui" on machineB, these branches come back to life
  as still being reviewed

Is this the correct description of your workflow? If so, are you running the
sync script with --prune, so that archived branches are properly removed on
machineB to match machineA?

Just trying to fully understand the sequence of events.

-- 
KR

Re: b4 tui: Old serieses not removed

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 2300 bytes --]

On Mon, Apr 27, 2026 at 11:33:48AM -0400, Konstantin Ryabitsev wrote:
> On Mon, Apr 27, 2026 at 04:56:52AM +0900, Mark Brown wrote:

> > As part of an effort to sync review state between machines I have
> > arranged for the set of b4/review branches on the destination machine to
> > be the same as it was on the source.

> So, let's call these "machineA" and "machineB". You do most work on machineA,
> but machineB runs a periodic git fetch to get all the review branches from
> machineA, correct?

The use case is more that I have machines that I switch between - mostly
I'll just work on one machine, but sometimes I'll move that to a
different machine and use that for a period before switching in the
opposite direction.  The sync is done by mirroring the current state of
the active machine to a git server, then syncing the set of b4/review
branches with that on the machine that's picking things up.  It's more
than just fetches, I've also started clearing out old branches that were
archived.

> > When I open b4 tui and run 'U' to
> > update all branches it reports the existance of several hundred more
> > branches, looking at them going past and what's in the UI I think most
> > but not all of them are thanked branches which I later archived on the
> > source machine, the branches are gone but they remain in the UI.

> This is the part that's confusing to me, so I need clarification. I assume the
> following happened:

> - These series were reviewed on machineA (creating b4/review/* branches for
>   them).
> - MachineB runs periodic fetches from machineA, which creates these
>   b4/review/* branches on machineB.
> - The seriews are applied, thanked, and archived on machineA
> - When you open "b4 review tui" on machineB, these branches come back to life
>   as still being reviewed

IIRC they were actually listed as thanked, some of them would've been
synced prior to actually being archived (since that's a fiddly manual
process).

> Is this the correct description of your workflow? If so, are you running the
> sync script with --prune, so that archived branches are properly removed on
> machineB to match machineA?

Yes, the branches are being pruned (not with git prune).  I can't see
any local record of them and when I deleted the b4 review database they
were gone.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]