Discussion:
CVS Vitrual Project
Rez P
2009-06-08 18:01:54 UTC
Permalink
We're running CVS on a Redhat linux server and our users all use Windows machines either using CVS command line or WinCVS or TortoiseCVS. I'm not clear on the CVS manual about this. Is it possible to create or have a project in CVS that consists of only files(NO folders or subfolders) from other existing projects?

NewProject
File1 (a symlink to $cvsroot/projectXYZ/file1)
File2 (a symlink to $cvsroot/projectFooBar/folder/subfolder/xfile)
.
.
.
File N (a symlink to $cvsroot/project123/folder/subfolder/file123)

And users can check out this new project and once they modify the files in it, the updates would actually occur to the existing files in other projects where they point to?

Should such a project be created directly on the server with symlinks pointing to the actual versioned files with correct permissions configured. Or should it be configured on the client machine and imported to CVS as a project or module? Could someone please give me detailed instructions or point me to a link.

Thanks


_________________________________________________________________
Hotmail® has ever-growing storage! Don’t worry about storage limits.
http://windowslive.com/Tutorial/Hotmail/Storage?ocid=TXT_TAGLM_WL_HM_Tutorial_Storage_062009
Risman, Mark
2009-06-08 18:47:06 UTC
Permalink
CVS looks at the "CVS" subdirectory of the working directory for its
metadata, and among other things, that metadata tells CVS where in the
repository to find the files you're checking out or in. This means that
even with the symlink, the CVS file operations will be looking at what's
in your "NewProject" directory for this metadata, not at projectXYZ or
projectFooBar.

You could still go to the other project checkout directories and check
in from there, but it sounds like that's not what you want.

________________________________

From: info-cvs-bounces+mark.risman=***@nongnu.org
[mailto:info-cvs-bounces+mark.risman=***@nongnu.org] On Behalf Of
Rez P
Sent: Monday, June 08, 2009 2:02 PM
To: CVS
Subject: CVS Vitrual Project



We're running CVS on a Redhat linux server and our users all use Windows
machines either using CVS command line or WinCVS or TortoiseCVS. I'm
not clear on the CVS manual about this. Is it possible to create or
have a project in CVS that consists of only files(NO folders or
subfolders) from other existing projects?



NewProject

File1 (a symlink to $cvsroot/projectXYZ/file1)

File2 (a symlink to
$cvsroot/projectFooBar/folder/subfolder/xfile)

.

.

.

File N (a symlink to
$cvsroot/project123/folder/subfolder/file123)



And users can check out this new project and once they modify the files
in it, the updates would actually occur to the existing files in other
projects where they point to?



Should such a project be created directly on the server with symlinks
pointing to the actual versioned files with correct permissions
configured. Or should it be configured on the client machine and
imported to CVS as a project or module? Could someone please give me
detailed instructions or point me to a link.



Thanks






________________________________

Hotmail(r) has ever-growing storage! Don't worry about storage limits.
Check it out.
<http://windowslive.com/Tutorial/Hotmail/Storage?ocid=TXT_TAGLM_WL_HM_Tu
torial_Storage_062009>





**********************************************************

MLB.com: Where Baseball is Always On
Risman, Mark
2009-06-08 19:45:18 UTC
Permalink
I haven't tried it, but it sounds like that should work.

________________________________

From: Rez P [mailto:***@hotmail.com]
Sent: Monday, June 08, 2009 3:31 PM
To: Risman, Mark
Subject: RE: CVS Vitrual Project


Bummer :(

So could I stitch together such a new project comprised of subfolders
from other existing projects with their cvs subdirectories intact? And
when the files in the new project are modified the changes will really
happen to their place in the existing projects?


________________________________

Date: Mon, 8 Jun 2009 14:47:06 -0400
From: ***@mlb.com
To: ***@hotmail.com
CC: info-***@nongnu.org
Subject: RE: CVS Vitrual Project


CVS looks at the "CVS" subdirectory of the working directory for its
metadata, and among other things, that metadata tells CVS where in the
repository to find the files you're checking out or in. This means that
even with the symlink, the CVS file operations will be looking at what's
in your "NewProject" directory for this metadata, not at projectXYZ or
projectFooBar.

You could still go to the other project checkout directories and check
in from there, but it sounds like that's not what you want.

________________________________

From: info-cvs-bounces+mark.risman=***@nongnu.org
[mailto:info-cvs-bounces+mark.risman=***@nongnu.org] On Behalf Of
Rez P
Sent: Monday, June 08, 2009 2:02 PM
To: CVS
Subject: CVS Vitrual Project



We're running CVS on a Redhat linux server and our users all use Windows
machines either using CVS command line or WinCVS or TortoiseCVS. I'm
not clear on the CVS manual about this. Is it possible to create or
have a project in CVS that consists of only files(NO folders or
subfolders) from other existing projects?



NewProject

File1 (a symlink to $cvsroot/projectXYZ/file1)

File2 (a symlink to
$cvsroot/projectFooBar/folder/subfolder/xfile)

.

.

.

File N (a symlink to
$cvsroot/project123/folder/subfolder/file123)



And users can check out this new project and once they modify the files
in it, the updates would actually occur to the existing files in other
projects where they point to?



Should such a project be created directly on the server with symlinks
pointing to the actual versioned files with correct permissions
configured. Or should it be configured on the client machine and
imported to CVS as a project or module? Could someone please give me
detailed instructions or point me to a link.



Thanks






________________________________

Hotmail(r) has ever-growing storage! Don't worry about storage limits.
Check it out.
<http://windowslive.com/Tutorial/Hotmail/Storage?ocid=TXT_TAGLM_WL_HM_Tu
torial_Storage_062009>



**********************************************************

MLB.com: Where Baseball is Always On

________________________________

Windows Live(tm): Keep your life in sync. Check it out.
<http://windowslive.com/explore?ocid=TXT_TAGLM_WL_BR_life_in_synch_06200
9>





**********************************************************

MLB.com: Where Baseball is Always On
Rez P
2009-06-08 22:18:18 UTC
Permalink
Thanks for your feedback.



Subject: RE: CVS Vitrual Project
Date: Mon, 8 Jun 2009 15:45:18 -0400
From: ***@mlb.com
To: ***@hotmail.com
CC: info-***@nongnu.org




I haven't tried it, but it sounds like that should work.



From: Rez P [mailto:***@hotmail.com]
Sent: Monday, June 08, 2009 3:31 PM
To: Risman, Mark
Subject: RE: CVS Vitrual Project


Bummer :(

So could I stitch together such a new project comprised of subfolders from other existing projects with their cvs subdirectories intact? And when the files in the new project are modified the changes will really happen to their place in the existing projects?



Date: Mon, 8 Jun 2009 14:47:06 -0400
From: ***@mlb.com
To: ***@hotmail.com
CC: info-***@nongnu.org
Subject: RE: CVS Vitrual Project




CVS looks at the "CVS" subdirectory of the working directory for its metadata, and among other things, that metadata tells CVS where in the repository to find the files you're checking out or in. This means that even with the symlink, the CVS file operations will be looking at what's in your "NewProject" directory for this metadata, not at projectXYZ or projectFooBar.

You could still go to the other project checkout directories and check in from there, but it sounds like that's not what you want.



From: info-cvs-bounces+mark.risman=***@nongnu.org [mailto:info-cvs-bounces+mark.risman=***@nongnu.org] On Behalf Of Rez P
Sent: Monday, June 08, 2009 2:02 PM
To: CVS
Subject: CVS Vitrual Project



We're running CVS on a Redhat linux server and our users all use Windows machines either using CVS command line or WinCVS or TortoiseCVS. I'm not clear on the CVS manual about this. Is it possible to create or have a project in CVS that consists of only files(NO folders or subfolders) from other existing projects?

NewProject
File1 (a symlink to $cvsroot/projectXYZ/file1)
File2 (a symlink to $cvsroot/projectFooBar/folder/subfolder/xfile)
.
.
.
File N (a symlink to $cvsroot/project123/folder/subfolder/file123)

And users can check out this new project and once they modify the files in it, the updates would actually occur to the existing files in other projects where they point to?

Should such a project be created directly on the server with symlinks pointing to the actual versioned files with correct permissions configured. Or should it be configured on the client machine and imported to CVS as a project or module? Could someone please give me detailed instructions or point me to a link.

Thanks




Hotmail® has ever-growing storage! Don’t worry about storage limits. Check it out.

**********************************************************

MLB.com: Where Baseball is Always On



Windows Live™: Keep your life in sync. Check it out.



**********************************************************

MLB.com: Where Baseball is Always On

_________________________________________________________________
Windows Live™ SkyDrive™: Get 25 GB of free online storage.
http://windowslive.com/online/skydrive?ocid=TXT_TAGLM_WL_SD_25GB_062009
Larry Jones
2009-06-08 20:04:49 UTC
Permalink
Rez P writes:
>
> We're running CVS on a Redhat linux server and our users all use
> Windows machines either using CVS command line or WinCVS or TortoiseCVS.
> I'm not clear on the CVS manual about this. Is it possible to create
> or have a project in CVS that consists of only files(NO folders or
> subfolders) from other existing projects?

Sort of -- you can do that with the modules file:

<http://ximbiot.com/cvs/manual/cvs-1.11.23/cvs_18.html#SEC170>

Not everything works as well as you might hope, however. In particular,
if you do ``update -d'' to pick up new files and directories, you'll
likely get *all* the files in each directory rather than just the ones
selected by the module definition.

> NewProject
> File1 (a symlink to $cvsroot/projectXYZ/file1)
> File2 (a symlink to $cvsroot/projectFooBar/folder/subfolder/xfile)
> .
> .
> .
> File N (a symlink to $cvsroot/project123/folder/subfolder/file123)

This looks more like a mishmash of unrelated stuff than any kind of
coherent project. Could you be more explicit about exactly what it is
you're trying to do?
--
Larry Jones

Moms and reason are like oil and water. -- Calvin
Rez P
2009-06-08 23:32:00 UTC
Permalink
Thanks for your response. See below please.


> Subject: Re: CVS Vitrual Project
> To: ***@hotmail.com
> Date: Mon, 8 Jun 2009 16:04:49 -0400
> CC: info-***@nongnu.org
> From: ***@siemens.com
>
> Rez P writes:
> >
> > We're running CVS on a Redhat linux server and our users all use
> > Windows machines either using CVS command line or WinCVS or TortoiseCVS.
> > I'm not clear on the CVS manual about this. Is it possible to create
> > or have a project in CVS that consists of only files(NO folders or
> > subfolders) from other existing projects?
>
> Sort of -- you can do that with the modules file:
>
> <http://ximbiot.com/cvs/manual/cvs-1.11.23/cvs_18.html#SEC170>
>
> Not everything works as well as you might hope, however. In particular,
> if you do ``update -d'' to pick up new files and directories, you'll
> likely get *all* the files in each directory rather than just the ones
> selected by the module definition.



Is it then possible to create my own CVS/Entries folder cosisting of
>
> > NewProject
> > File1 (a symlink to $cvsroot/projectXYZ/file1)
> > File2 (a symlink to $cvsroot/projectFooBar/folder/subfolder/xfile)
> > .
> > .
> > .
> > File N (a symlink to $cvsroot/project123/folder/subfolder/file123)
>
> This looks more like a mishmash of unrelated stuff than any kind of
> coherent project. Could you be more explicit about exactly what it is
> you're trying to do?



Certain segments of our source code or perl scripts pair up with their own database packages. Currently, the source code and the db modules are entirely separated into their projects and I was going to make it easier for our developers by using this newProject approach as a one stop shop.


> --
> Larry Jones
>
> Moms and reason are like oil and water. -- Calvin


_________________________________________________________________
Windows Live™ SkyDrive™: Get 25 GB of free online storage.
http://windowslive.com/online/skydrive?ocid=TXT_TAGLM_WL_SD_25GB_062009
Loading...