Send Info-cvs mailing list submissions to
address@hidden
To subscribe or unsubscribe via the World Wide Web, visit
http://mail.gnu.org/mailman/listinfo/info-cvs
or, via email, send a message with subject or body 'help' to
address@hidden
You can reach the person managing the list at
address@hidden
When replying, please edit your Subject line so it is more
specific
than "Re: Contents of Info-cvs digest..."
Today's Topics:
1. RE: "cvs commit: nothing known about 'filename'" (Teala
Spitzbarth)
2. RE: Is this taginfo problem a permissions thing? (Teala
Spitzbarth)
3. Re: CVS 1.11.2 Released! (Matthew M. Ogilvie)
4. Re: another CVS problem (address@hidden)
5. Need help (Rohan Bairat)
6. Re: Standard Styntax for TAG (Muhammad Shakeel)
7. Re: FW: Help w/ CVS (Matt Riechers)
8. Re: cvs on MacOSX (Lee Fellows)
9. Re: CVS Question ? (Oliver Fischer)
10. Re: problem when using shell script to checkout huge
binary files
(>70MB) (Lee Fellows)
11. Re: problem when using shell script to checkout huge
binary files
(>70MB) (Frederic Brehm)
12. Re: Need help (Lee Fellows)
--__--__--
Message: 1
Subject: RE: "cvs commit: nothing known about 'filename'"
Date: Mon, 20 May 2002 17:26:45 -0700
From: "Teala Spitzbarth" <address@hidden>
To: "Sven Sandberg" <address@hidden>,
<address@hidden>
Have you done a "cvs add" of the file first?=20
Looks like you are trying to commit a new file without running
the add=20
first.
Cheers,
Teala
-----Original Message-----
From: Sven Sandberg [mailto:address@hidden
Sent: Saturday, May 18, 2002 4:41 PM
To: address@hidden
Subject: "cvs commit: nothing known about 'filename'"
Hello,
When I run
cvs -d:ext:address@hidden:blah commit -m 'message' filename
I get the message:
cvs commit: nothing known about 'filename'
cvs [commit aborted]: correct above errors first!
It only happens with some files, not all. Does anyone know what
it means? Thanks in advance,
Sven Sandberg
_______________________________________________
Info-cvs mailing list
address@hidden
http://mail.gnu.org/mailman/listinfo/info-cvs
--__--__--
Message: 2
Subject: RE: Is this taginfo problem a permissions thing?
Date: Mon, 20 May 2002 17:49:37 -0700
From: "Teala Spitzbarth" <address@hidden>
To: "Duane Morin" <address@hidden>,
<address@hidden>
Are you sure your CVSROOT variable was pointing to the test
repository when you checked out CVSROOT & commited the taginfo
file?
If taginfo did not change in your test repository, then it sure
sounds=20
like it got commited to some other repository...
Cheers,
Teala
-----Original Message-----
From: Duane Morin [mailto:address@hidden
Sent: Friday, May 17, 2002 10:30 AM
To: address@hidden
Subject: Is this taginfo problem a permissions thing?
For the life of me I can't get taginfo to work although I think
I'm
doing everything right. I put together a quick repository in my
home
directory,
/home/dmorin/repository. I have
/home/dmorin/repository/CVSROOT/taginfo and taginfo,v.
I create a /home/dmorin/localcheckout and do "cvs co CVSROOT". I
add
this
line:
ALL '/home/dmorin/tagger.sh $1 $2 $3 $4'
(tagger.sh is just a quick shell script:
#!/bin/sh
echo $1 $2 $3 $4 >> /tmp/taglog)
I checkin taginfo. But no matter how many tagging operations I
do,
this file is not getting executed -- I would think that I would
at
least see a /tmp/taglog file of length 0 if the program was
being
executed but had nothing to echo.
The weird part is that when I look at
/home/dmorin/repository/CVSROOT
I see that taginfo has *not* changed at all, bit taginfo,v seems
to
indeed know about my change! What's the scoop with that? That's
the
part I can't figure out. I don't get any errors from cvs when
I
commit the file. I thought it might be a permissions thing (I
had
r--r--r-- when I first looked at the repository, but then changed
it
to rw-rw-rw- and it's still a problem), but no luck.
Anybody point me in the right direction? The online tutorials
I'm
finding aren't helping me with this -- they all just say
"Checkout
CVSROOT, edit the file you want, then check it back in." Well,
I'm
doing that, and it's not working.
Thanks!
Duane
_______________________________________________
Info-cvs mailing list
address@hidden
http://mail.gnu.org/mailman/listinfo/info-cvs
--__--__--
Message: 3
Date: Mon, 20 May 2002 20:13:59 -0600
From: "Matthew M. Ogilvie" <address@hidden>
To: address@hidden
Cc: address@hidden, address@hidden, address@hidden,
address@hidden
Subject: Re: CVS 1.11.2 Released!
Reply-To: address@hidden
>
>Message: 3
>Date: Mon, 20 May 2002 14:55:29 -0400
>From: Derek Robert Price <address@hidden>
>Subject: Re: CVS 1.11.2 Released!
>To: Noel Yap <address@hidden>
>Cc: address@hidden, address@hidden
>
>Noel Yap wrote:
>
>>--- Derek Robert Price <address@hidden> wrote:
>>
>>>Noel Yap wrote:
>>>
>>>>Grrr! It looks like the advisory locks patch still
>>>>didn't make it in. I think the patch have all the
>>>>necessary requisites (doc and test patch). Is it
>>>>still missing something?
>>>>
>>>>Noel
>>>>
>>>I've fallen a little bit behind on the email lists,
>>>I won't get into
>>>why, but I just found your patch on SourceForge.
>>>I'll try and find time
>>>to take a look at it so that you can at least get
>>>feedback and maybe get
>>>the patch into the next release.
>>>
>>>Thanks for your patience.
>>>
>>
>>Hey, so long as it's not being ignored outright ...
>>:-)
>>
>>The LAG (latest and greatest) of this patch was
>>submitted by Matt Ogilvie (who added docs and tests).
>>Just to be sure you get the right one, here is the
>>URL:
>>https://sourceforge.net/tracker/index.php?func=detail&aid=471046&group_id=4680atid=304680
>>
>>Thanks,
>>Noel
>>
>
>The addition of documentation and test cases is great, but this
patch is
>still way too big. I'll keep trying to deal with it if
somebody will
>chop it into little bite size chunks, one bug fix or feature,
its doc,
>and its tests per chunk.
>
>For instance, the bug fixes Mathew Ogilvie mentions he made as
he was
>working on this should have their own patches. Making edit
report
>editors should be in its own patch. Advisory locks should have
its own
>patch, possibly several when it makes sense, but I didn't read
enough to
>give further advice. The sockPlayer separation is nice too -
keep it
>separate.
>
>Sorry about that, but I really don't have the time for so much
unpaid
>work at the moment.
>
>Derek
>
>--
>
Derek,
It is hard to be certain, but your message leaves me with the
impression
that you were looking at an older version of the patch. There
are several versions at the web page, you want the latest
version,
which is currently editCheck-25feb2002.diff:
https://sourceforge.net/tracker/index.php?func=detail&aid=471046&group_id=4680&atid=304680
The 25feb2002 version includes much of what you ask for:
sockPlayer
is gone completely, (you can always get it out the older version
if
you want [might be useful when overhauling the test
suite??]), and the patch itself is split into logically
seperate
subpatches within the file.
Take your time - as Noel said "Hey, so long as it's not being
ignored outright ...". Thanks for looking at it, though...
--
Matthew Ogilvie address@hidden
--__--__--
Message: 4
Subject: Re: another CVS problem
To: Sudhaker P <address@hidden>
Date: Mon, 20 May 2002 22:39:16 -0500 (CDT)
Cc: address@hidden
From: address@hidden
Organization: Are you kidding?
> Hi,
>
> Larry's solution helped me out of the problems .Thanks a lot
folks, that
> really helped.I was able to go to the directories but I'm
facing a new
> problem when I'm checking out files from the repository:
> "cvs server: Updating /DesignDocuments
> cvs server: failed to create lock directory for
> `/usr/local/cvsrepos/DesignDocuments'
> (/usr/local/cvsrepos/DesignDocuments/#cvs.lock): Permission
denied
This is because you do not have write permissions for that
directory. I don't know why you wouldn't, and there are
several
possibilities. Are you supposed to have full or read-only
access
there? What access method are you using? Do you have a real
account on the server machine? What OS is the server running?
Is the administrator using Unix groups to control access, or
ACLs?
David H. Thornley | If you want my
opinion, ask.
address@hidden | If you don't, flee.
http://www.thornley.net/~thornley/david/ | O-
--__--__--
Message: 5
Date: 21 May 2002 04:39:19 -0000
From: "Rohan Bairat" <address@hidden>
Reply-To: "Rohan Bairat" <address@hidden>
To: address@hidden
Subject: Need help
Hi
I am trying to configure CVS...i am going through the manual
....In the manual in exercise 3 i guess it talks abt rsh
thing..
i am not able to find .rhosts file.... i dont understand what
to
do.. I created one and wrote the IP add of the other machine
and
usrname in that basically i am operating on user root.. so i
created that file ..
and tried to run rsh from remote machine ie
rsh -l username ipof the server 'echo $PATH'
still it is not workin plz help
rohan
_________________________________________________________
Click below to visit monsterindia.com and review jobs in India
or
Abroad
http://monsterindia.rediff.com/jobs
--__--__--
Message: 6
Date: Tue, 21 May 2002 11:21:53 +0500
From: Muhammad Shakeel
<address@hidden>
To: "Glew, Andy" <address@hidden>
CC: address@hidden
Subject: Re: Standard Styntax for TAG
--------------020008050606020406010303
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Thanks for the information, but there should be a version number
of the
module in case of release level tag. I designed the following
syntax.
Any comments/suggestions ?
Release Tag Syntax:
Rel_ProjectName_x_y_State
e.g.
Rel_Mpeg4_1_0_Final
Branch TAG Syntax:
B_ BranchName_ProjectName_ TAGName_State
e.g.
B_ Group2_ H263_TestOK
Devoper TAG Syntaz:
ProjectName_TAGName_State
e.g.
Mpeg4_Test1OK_Alpha
Regards,
shakeel
Glew, Andy wrote:
>>Is there is any standard syntax or pattern for Tags , for
>>following type of tags or some other importnat one:
>>
>>Developement level Tags
>>Release levelTags
>>Branch level Tags
>>
>
>I don't know of any universal standards, but I try to enforce
>the following on projects I am involved with.
>Also, I have several scripts, like cvs-make-branch,
>that automatically mandate some of these
>"standards".
>
>All branches should have "branch" in them,
>e.g. the branch tag should be suffixed
> *--branch
>
>All branches should have a tag "branch-base"
>associated with the contour they branched from
>e.g. the branch tag, suffixed
> *--branch-base
>
>Less often, but more and more recently,
>I use the words "contour" to distinguish
>tags that identify a self consistent set of versions
>e.g. some meaningful name suffixed
> *--contour
>
>I nearly always want to have the date and time
>in my tags. I indicate this as
> *--YYYY-MM-DD-HHhMM--*
>I only use local time at the moment, although
>have been forced to use GMT in the past.
>
>I only do not put the year and date
>a) for floating tags
> e.g. regression-test-latest
>although for such floating tags I usually have
>a timestamped version
> e.g. regression-test-2002-05-17-16h10
>or b) for long-lived tags, typically branches,
>e.g.
> windows-branch
>vs
> linux-branch
>However, overall I find long-lived branches a bad idea,
>and prefer to unify them via the build system.
>I more often regret not having the timestamp in a branch tag
>than I regret having it in a tag.
>
>At the front of a tag, of course, I try to put something
meaningful.
>I have learned that it is best to put the most distinguishing
part
>of the tag first rather than last, so that name completion
tools
>work well --- e.g.
meaningful-text--2002-05-17-16h20--branch
>rather than tag-2002-05-17-16h20--branch--meaningful-text.
>
>
>I have tried to enforce naming conventions wrt merge points,
>where other branches are merged in, but have never found a
good
>one.
>
--------------020008050606020406010303
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit
<html>
<head>
</head>
<body>
<br>
Thanks for the information, but there should be a version number
of the module
in case of release level tag. I designed the following syntax.
Any comments/suggestions
?<br>
<p class="MsoNormal"><b>Release Tag Syntax:</b> </p>
<p class="MsoNormal">Rel_ProjectName_x_y_State</p>
<p class="MsoNormal">e.g.</p>
<p class="MsoNormal">Rel_Mpeg4_1_0_Final</p>
<!--[if !supportEmptyParas]--> <b>Branch TAG Syntax:</b><br>
<p class="MsoNormal">B_ BranchName_ProjectName_
TAGName_State</p>
<p class="MsoNormal">e.g.</p>
<p class="MsoNormal">B_ Group2_ H263_TestOK</p>
<p class="MsoNormal"><b>Devoper TAG Syntaz:</b><br>
ProjectName_TAGName_State</p>
<p class="MsoNormal">e.g.</p>
<p class="MsoNormal">Mpeg4_Test1OK_Alpha</p>
<br>
<br>
Regards,<br>
shakeel<br>
<br>
<br>
Glew, Andy wrote:<br>
<blockquote type="cite"
cite="mid:address@hidden">
<blockquote type="cite">
<pre wrap="">Is there is any standard syntax or pattern for
Tags , for <br>following type of tags or some other importnat
one:<br><br>Developement level Tags <br>Release levelTags
<br>Branch level Tags <br></pre>
</blockquote>
<pre wrap=""><!----><br>I don't know of any universal
standards, but I try to enforce<br>the following on projects I am
involved with.<br>Also, I have several scripts, like
cvs-make-branch,<br>that automatically mandate some of
these<br>"standards".<br><br>All branches should have "branch" in
them,<br>e.g. the branch tag should be suffixed<br>
*--branch<br><br>All branches should have a tag "branch-base"
<br>associated with the contour they branched from<br>e.g. the
branch tag, suffixed<br> *--branch-base<br><br>Less often,
but more and more recently,<br>I use the words "contour" to
distinguish <br>tags that identify a self consistent set of
versions<br>e.g. some meaningful name suffixed<br>
*--contour<br><br>I nearly always want to have the date and
time<br>in my tags. I indicate this as<br>
*--YYYY-MM-DD-HHhMM--*<br>I only use local time at the moment,
although<br>have been forced to use GMT in the past.<br><br>I
only do not put the year and date <br>a)
for floating tags<br> e.g. regression-test-latest<br>although for
such floating tags I usually have<br>a timestamped version<br>
e.g. regression-test-2002-05-17-16h10<br>or b) for long-lived
tags, typically branches,<br>e.g.<br> windows-branch<br>vs<br>
linux-branch<br>However, overall I find long-lived branches a bad
idea,<br>and prefer to unify them via the build system.<br>I more
often regret not having the timestamp in a branch tag<br>than I
regret having it in a tag.<br><br>At the front of a tag, of
course, I try to put something meaningful.<br>I have learned that
it is best to put the most distinguishing part<br>of the tag
first rather than last, so that name completion tools<br>work
well --- e.g.
meaningful-text--2002-05-17-16h20--branch<br>rather than
tag-2002-05-17-16h20--branch--meaningful-text.<br><br><br>I have
tried to enforce naming conventions wrt merge points,<br>where
other branches are merged in, but have never found a
good<br>one.<br><br></pre>
</blockquote>
<br>
<pre class="moz-signature" cols="$mailwrapcol"></pre>
<br>
</body>
</html>
--------------020008050606020406010303--
--__--__--
Message: 7
Date: Tue, 21 May 2002 07:39:01 -0400
From: Matt Riechers <address@hidden>
To: Caroline Neal <address@hidden>
CC: "'address@hidden'" <address@hidden>
Subject: Re: FW: Help w/ CVS
Caroline Neal wrote:
>
> I am working on a project that requires customer headers and
footers and
> cannot find where to edit them in the CVS program. Is it
possible and if so
> how?
What headers and footers would those be? Are you trying to use
keyword
expansion? Please be more specific.
-Matt
--__--__--
Message: 8
Subject: Re: cvs on MacOSX
From: Lee Fellows <address@hidden>
To: address@hidden
Date: 21 May 2002 08:23:22 -0400
Hi Valerio,
I do not have MacOSX experience, but the problem your seeing
is not
OS specific. Inetd or xinetd are not listening for the
cvspserver, or
you have a firewall or other software that is preventing
outside
connections to port 2401. If you changed inetd.conf or
xinetd's
configuration (could be a .conf file or a directory containing
service
definitions), then you need to signal the superserver to re-read
its
configuration. Check your man pages for the correct procedure.
If
inetd/xinetd has re-read its configuration and you still cannot
connect,
look for a firewall script on the host to see if that is
causing
the problem.
HTH
On Thu, 2002-05-16 at 07:25, Valerio Ferrucci wrote:
> Does someone has experience on installing cvs server on
MacOSX?
> I followed instructions at
> http://www.geektimes.com/macintosh/os/x/and/cvs_install.html
> but when I login the server responds:
>
> cvs [login aborted]: connect to localhost:2401 failed:
Connection refused
>
> Any idea or helpful links?
>
> Thanks
>
>
==================================================================
> Valerio Ferrucci Tabasoft Sas
> address@hidden
http://www.tabasoft.it
--__--__--
Message: 9
From: "Oliver Fischer" <address@hidden>
To: <address@hidden>
Subject: Re: CVS Question ?
Date: Tue, 21 May 2002 14:28:27 +0200
Hi,
you have to lock the branch - if you have one. A tag is not what
you
are looking for.
Bye
Oliver
----- Original Message -----
From: address@hidden
To: address@hidden
Sent: Thursday, May 16, 2002 4:38 PM
Subject: CVS Question ?
Is there a way to lock a tag so that no one can make
any commits into that tag?
After the release, we need to kind of freez that tag.
Thanx, Sean
--__--__--
Message: 10
Subject: Re: problem when using shell script to checkout huge
binary files
(>70MB)
From: Lee Fellows <address@hidden>
To: address@hidden
Date: 21 May 2002 08:39:20 -0400
Hi Veronica,
Have you tried the shell script from the command line, not
calling
it from inside Java? What were the results? In calling this
shell
script from java, is this an automated process (i.e., a cron
job)?
If so, have you tried running the java process from the command
line?
What were the results?
More questions then helpful suggestions, I know, but it would
help
to know what your environment is telling you. I am suspicious of
a
timeout, possibly being imposed by the java process, but your
observations can help confirm or eliminate that possibility.
On Mon, 2002-05-20 at 13:54, Veronica Zhu wrote:
> Hi,
>
> I use Java Runtime.getRuntime().exec() to execute a shell
script.
> This shell script does cvs checkout of project repositories
> into user directories for web based file browsing.
>
> It works great for all files EXCEPT huge binary files whose
size is
> about 70 Mega Bytes. The directory, which should contain
those
> large files, was checkout as an empty directory. (No errors
observed.)
> However, at the same time, other directories with smaller
files
> were checkout normally with expected files.
>
> It doesn't look like the Java problem, because I can use the
> same Java function to execute a shell script that copy the
huge
> file from one folder to another one. Well, cvs checkout
> works OK for the large files when doing so in win-cvs
windows.
> The server is Red Hat Linux 7.1 and uses
> Concurrent Versions System (CVS) 1.11.1p1 (client/server).
>
> I have searched mailing-list of cvs with no luck.
> Any one has any ideas ?
> Any input will be greatly appreciated!!
>
> Veronica
>
>
> _______________________________________________
> Info-cvs mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/info-cvs
>
--__--__--
Message: 11
Date: Tue, 21 May 2002 08:58:15 -0400
To: address@hidden
From: "Frederic Brehm" <address@hidden>
Subject: Re: problem when using shell script to checkout huge
binary files
(>70MB)
At 12:54 -0500 5/20/02, Veronica Zhu wrote:
>I use Java Runtime.getRuntime().exec() to execute a shell
script.
...
>It works great for all files EXCEPT huge binary files whose
size is
>about 70 Mega Bytes. The directory, which should contain
those
>large files, was checkout as an empty directory. (No errors
observed.)
>However, at the same time, other directories with smaller
files
>were checkout normally with expected files.
Here's a few things to try...
Make sure all files are on the same branch.
Try running the shell script from the command line of the client
and
watch for error messages. Try the same command line on the
server
(you might have to adjust CVSROOT).
Use the -t switch of CVS and watch what happens when traversing
the
directories with the large files.
Very large binary files can be a problem in CVS. Updates to
binary
files are not stored very efficiently in the repository (unlike
text
files where diff works well). CVS reads the entire repository
file
into virtual memory in order to process it. That means your 70M
binary file might be much larger in the repository--too large for
VM
on your server. Check the file size of the repository file (if
the
file is named foo.gif, then the repository file will be named
foo.giv,v).
Good Luck,
Fred
--
Fred Brehm, Sarnoff Corporation, address@hidden
http://www.sarnoff.com/digital_video_informatics/vision_technology/index.asp
--__--__--
Message: 12
Subject: Re: Need help
From: Lee Fellows <address@hidden>
To: address@hidden
Date: 21 May 2002 09:16:42 -0400
Hi Rohan,
I would not recommend using rsh. Having said that, what error
do you
get? Does the command 'rsh -l username ipof the server 'echo
$PATH''
work?
On Tue, 2002-05-21 at 00:39, Rohan Bairat wrote:
> Hi
> I am trying to configure CVS...i am going through the manual
> ....In the manual in exercise 3 i guess it talks abt rsh
thing..
>
> i am not able to find .rhosts file.... i dont understand what
to
> do.. I created one and wrote the IP add of the other machine
and
> usrname in that basically i am operating on user root.. so
i
> created that file ..
>
> and tried to run rsh from remote machine ie
> rsh -l username ipof the server 'echo $PATH'
>
> still it is not workin plz help
>
> rohan
> _________________________________________________________
> Click below to visit monsterindia.com and review jobs in India
or
> Abroad
> http://monsterindia.rediff.com/jobs
>
>
> _______________________________________________
> Info-cvs mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/info-cvs
>
--__--__--
_______________________________________________
Info-cvs mailing list
address@hidden
http://mail.gnu.org/mailman/listinfo/info-cvs
End of Info-cvs Digest