[issue1881] strip of mq patches leaves bogus state

[Mads Kiilerich]

New submission from Mads Kiilerich <mads at kiilerich.com>:

"hg strip" with mq patches applied removes the mq changesets, but for
example "hg qap" still reports them as applied. 

What is worse is that after a pull other changesets will be tagged as being
the mq patches. 

In my case after other changes has been pulled too then qpop failed with
"abort: popping would remove a revision not managed by this patch queue".
That seems to be absolutely correct, but also a message I don't know what to
use for. The message should suggest how to resolve the situation. I guess
that it could qfinish it - or at least suggest to do it.

Trying to reproduce that I got something different but similar:

[mk at localhost tmp]$ hg init Z
[mk at localhost tmp]$ cd Z
[mk at localhost Z]$ echo foo > foo
[mk at localhost Z]$ hg add foo
[mk at localhost Z]$ hg ci -m foo
[mk at localhost Z]$ hg qnew bar
[mk at localhost Z]$ echo bar >> foo
[mk at localhost Z]$ hg qref
[mk at localhost Z]$ hg qap
bar
[mk at localhost Z]$ hg strip tip
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
saving bundle to /tmp/Z/.hg/strip-backup/a1cde8e42a91-backup
[mk at localhost Z]$ hg qap
bar
[mk at localhost Z]$ hg up -r qtip
mq status file refers to unknown node a1cde8e42a91
mq status file refers to unknown node a1cde8e42a91
abort: unknown revision 'qtip'!
[mk at localhost Z]$ hg pull -f ~/hg
...
[mk at localhost Z]$ hg qpop -a
abort: trying to pop unknown node a1cde8e42a91

----------
messages: 10818
nosy: kiilerix
priority: bug
status: unread
title: strip of mq patches leaves bogus state

____________________________________________________
Mercurial issue tracker <bugs at mercurial.selenic.com>
<http://mercurial.selenic.com/bts/issue1881>
____________________________________________________