enforce-single-head = yes prevents commit, although there are no multiple heads

[Georges Racinet]

On 12/17/20 8:46 AM, Uwe Brauer wrote:
> Hi
>
>
> I am not sure whether I will be able to produce a script that can
> reproduce the error described below because it occurs on a repository 
> with some 800 changesets and various branches
>
> When I tried 
>
> [experimental]
> enforce-single-head = yes
For the record, it is activated by default server-side in Heptapod [1].
>
> On a small repository it did what I exspected, namely it prevented me by
> using say topics to create multiple heads on a branch.

This is already odd : with evolve and topic active, the "single head" is
assessed per branch/topic combination.

> However when I used this setting on  this huge repository and tried to
> commit I obtained 
>
> Waiting for Emacs...
> transaction abort!
> rollback completed
> note: commit message saved in .hg/last-message.txt
> abort: 3 heads on "default"
> (cbf9827aaa0b, c5126e027526, aa50036267da)
>
> So I deleted the option and commited again, however when running 
> hg heads
>
> It returned one head not three 
>
> Can anybody please enlighten me?
Do you have bookmarks perhaps?

[1] or at least was until Heptapod 0.18.0 (currently release candidate)
and actually reimplemented directly since 0.17.2 for performance reasons.

-- 
Georges Racinet
https://octobus.net, https://about.heptapod.host, https://heptapod.net
GPG: BF5456F4DC625443849B6E58EE20CA44EF691D39, sur serveurs publics