[ 
https://issues.apache.org/jira/browse/SVN-4645?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Remi Andruccioli updated SVN-4645:
----------------------------------
    Description: 
Hello,

Using SVN in team development for more that 6 years I recently noticed a 
strange issue.
This issue implies two users A and B.

To reproduce:

* User A creates a new file (eg. "foo.c") on his working copy with some content 
inside.
* User B creates a new file *with the same name* ("foo.c") on his working copy 
with some *different* content inside.
* User A svn-adds his file and svn-commits it.
* User B svn-updates his working copy tree. His local "foo.c" file gets marked 
as "Versioned" and thus gets effectively versioned.
* User B svn-commits his file (no need to svn-add because the file has been 
versioned in the previous step).

Result:
The content of file "foo.c" (initially by User A) is now erased by the content 
of User B *without* reporting a conflict. So information is loosed.

Question(s):
Is it a normal behavior ?
What I think about it is that SVN should have reported a conflict (or tree 
conflict) when User B svn-updates his local tree.
Am I wrong ?


Thank you.

Remi Andruccioli

-----------------------------------------------------------------------------
Information that can be useful:
* Server:
** Windows Server 2008 r2
** Visual SVN Server Manager 3.3.1
** SVN 1.8.13
* Client:
** Windows 7
** Tortoise SVN 1.9.2
** SVN 1.9.2

  was:
Hello,

Using SVN in team development for more that 6 years I noticed a strange issue.
This issue implies two users A and B.

To reproduce:

* User A creates a new file (eg. "foo.c") on his working copy with some content 
inside.
* User B creates a new file *with the same name* ("foo.c") on his working copy 
with some *different* content inside.
* User A svn-adds his file and svn-commits it.
* User B svn-updates his working copy tree. His local "foo.c" file gets marked 
as "Versioned" and thus gets effectively versioned.
* User B svn-commits his file (no need to svn-add because the file has been 
versioned in the previous step).

Result:
The content of file "foo.c" (initially by User A) is now erased by the content 
of User B *without* reporting a conflict. So information is loosed.

Question(s):
Is it a normal behavior ?
What I think about it is that SVN should have reported a conflict (or tree 
conflict) when User B svn-updates his local tree.
Am I wrong ?


Thank you.

Remi Andruccioli

-----------------------------------------------------------------------------
Information that can be useful:
* Server:
** Windows Server 2008 r2
** Visual SVN Server Manager 3.3.1
** SVN 1.8.13
* Client:
** Windows 7
** Tortoise SVN 1.9.2
** SVN 1.9.2


> No conflict on new file crated by 2 users at the same time
> ----------------------------------------------------------
>
>                 Key: SVN-4645
>                 URL: https://issues.apache.org/jira/browse/SVN-4645
>             Project: Subversion
>          Issue Type: Bug
>    Affects Versions: 1.8.13
>         Environment: Windows
> Visual SVN
> Tortoise SVN
>            Reporter: Remi Andruccioli
>              Labels: conflict, tortoisesvn, windows
>
> Hello,
> Using SVN in team development for more that 6 years I recently noticed a 
> strange issue.
> This issue implies two users A and B.
> To reproduce:
> * User A creates a new file (eg. "foo.c") on his working copy with some 
> content inside.
> * User B creates a new file *with the same name* ("foo.c") on his working 
> copy with some *different* content inside.
> * User A svn-adds his file and svn-commits it.
> * User B svn-updates his working copy tree. His local "foo.c" file gets 
> marked as "Versioned" and thus gets effectively versioned.
> * User B svn-commits his file (no need to svn-add because the file has been 
> versioned in the previous step).
> Result:
> The content of file "foo.c" (initially by User A) is now erased by the 
> content of User B *without* reporting a conflict. So information is loosed.
> Question(s):
> Is it a normal behavior ?
> What I think about it is that SVN should have reported a conflict (or tree 
> conflict) when User B svn-updates his local tree.
> Am I wrong ?
> Thank you.
> Remi Andruccioli
> -----------------------------------------------------------------------------
> Information that can be useful:
> * Server:
> ** Windows Server 2008 r2
> ** Visual SVN Server Manager 3.3.1
> ** SVN 1.8.13
> * Client:
> ** Windows 7
> ** Tortoise SVN 1.9.2
> ** SVN 1.9.2



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to