On 2007-04-26 10:30, Mike Kupfer <[EMAIL PROTECTED]> wrote:
> I spent a little time review Mercurial's open issues list
> (http://www.selenic.com/mercurial/bts/), to see if there is anything
> that we need to have fixed before we move ON to Mercurial.

Nice list, thanks for taking the time to make it :)

> After some discussion, Steve, Rich, and I have a draft list of
> showstopper bugs, and another list of non-stoppers that seemed
> important enough to at least mention.

>                   Draft Migration Stoppers (Mercurial Issues)
> 
>   Link to as http://www.selenic.com/mercurial/bts/issueNNN.
> 
>   +--------------------------------------------------------------------------+
>   | issue# |            synopsis             |             notes             |
>   |--------+---------------------------------+-------------------------------|
...
>   |--------+---------------------------------+-------------------------------|
>   |        | Mismerge involving two          |                               |
>   | 455    | concurrently renamed            |                               |
>   |        | directories                     |                               |
>   |--------+---------------------------------+-------------------------------|
...
>   |--------+---------------------------------+-------------------------------|
>   | 226    | Strange hgmerge invocation      | could lead to mismerges.      |
>   |        | behaviour                       |                               |
>   |--------+---------------------------------+-------------------------------|
>   | 122    | commit succeeds if merge fails  | makes mismerges easier than   |
>   |        |                                 | with Teamware.                |
>   |--------+---------------------------------+-------------------------------|
...
>   |--------+-----------------------------+-----------------------------------|
>   |        |                             | makes merging significantly more  |
>   |        | hg merge -f doesn't work    | brittle than with Teamware, but   |
>   | 507    | after a failed merge        | probably not an issue as long as  |
>   |        |                             | most people use a merge tool that |
>   |        |                             | produces a sane exit code.        |
>   +--------------------------------------------------------------------------+

There were some merge fixes recently, which were pushed to the
Mercurial CrewRepository as the following changesets:

  [ "hg log --style compact" output ]

  4401   84cd52b48f94   2007-05-03 17:24 -0500   mpm
    merge: reorganize some hunks in findcopies
  
  4400   93652499bed3   2007-05-03 17:24 -0500   mpm
    merge: simplify a check in checkcopies
  
  4399   3b7e284b8f28   2007-05-03 17:24 -0500   mpm
    merge: expand and simplify the invalid handling for directory moves
  
  4398   9fe267f77f56   2007-05-03 17:24 -0500   mpm
    merge: fix a bug detecting directory moves
  
  4397:4387   c04c96504a12   2007-05-03 17:24 -0500   mpm
    merge: clarify the findcopies code

Now that these are part of the official tree, I'm going through the
merge-related issues in BTS again -- especially those related to renames
-- to see if any of them are now fixed as a side-effect of the
changesets listed above.

In particular, what these changes fix in Mercurial (and any release
which will happen after the current 0.9.3 builds), is the handling of
merges for cases where:

  - the child workspace clones the parent
  - the child renames all the files in a directory (or the directory
    itself)
  - the child adds a new directory with the same name as the origin of a
    previous rename
  - the parent modifies one or more files in the origin directory of the
    rename
  - the child pulls
  - the child runs "hg merge"

A test case for Mercurial, which better illustrates what was fixed, and
can be used with a recent 'crew' installation of Hg to check that indeed
the merge is done correctly is available online too, at my outgoing Hg
changes tree at: http://hg.hellug.gr/mercurial/gker/rev/622410e4dc62

If anyone feels like helping me with a check for the merge-related
issues in Mercurial's BTS which may affect OpenSolaris, make sure you
have a 'mercurial-crew' installation and feel free to drop me an email ;)

Regards,
Giorgos

_______________________________________________
tools-discuss mailing list
[email protected]

Reply via email to