a bug with "hg update"

[Patrick Mézard]

Hello,

Zhikai Cheng a écrit :
> the Mercurial is a good tool I really like it and enjoyed to use it.
> I wish it get better.
> 
> I did the following with Mercurial:

Let's try with a toy example:

# Create a base repository t
$ hg init t
$ cd t
$ echo a > a
$ hg ci -Am adda
adding a
$ cd ..

# Clone t to t2, then change t/a
$ hg clone t t2
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ cd t
$ echo b > a
$ hg ci -m changea
$ cd ..

# Move a to b and update
$ cd t2
$ hg mv a b
$ hg st
A b
R a
$ hg pull
pulling from /Users/pmezard/dev/mercurial/hg/tests/t
searching for changes
adding changesets
adding manifests
adding file changes
added 1 changesets with 1 changes to 1 files
(run 'hg update' to get a working copy)
$ hg up
remote changed a which local deleted
use (c)hanged version or leave (d)eleted? 


The last line asks you whether you want to retrieve the updated "a" to replace "b" or do nothing. I would have expected it to fire a merge on "b" but it does not happen.

When you say "(file.1 is an existing file in the Mercurial repository.)", was "file.1" tracked by mercurial ?

--
Patrick Mézard