Re: how to create/design your repository/project
[EMAIL PROTECTED] wrote: How should I set up my repository to best benefit the development/management effort? should we have multiple repositories? each repository represents one of our clients. or should we have just 1 repository, define multiple modules and each module represents one client/project combination? My company has played around with various approaches. We started out with the one-repository-many-modules model, which we didn't particularly like for some reason that now escapes me, and then moved to a one-repository-per-project model. I don't really like that either. Finally, it looks like we're settling on a one-repository-per-functional-group model, so the infrastructure group has their own repository with potentially several modules in it, the client-facing teams have their own repositories with potentially several modules, and so on. Hope that helps. Cheers, Laird
Re: how to create/design your repository/project
At 17:59 -0400 8/16/00, [EMAIL PROTECTED] wrote: Hi, I'm new in CVS. After reading lots of newsgroup and reading Karl Fogel's CVS book, I'm ready to import our corporate codes into cvs. We have multiple clients and each client has multiple projects. How should I set up my repository to best benefit the development/management effort? should we have multiple repositories? each repository represents one of our clients. or should we have just 1 repository, define multiple modules and each module represents one client/project combination? We have multiple clients, too. We have a contractural obligation to protect their Intellectual Property, and we want to protect ours, too. Therefore, we use different repositories for different clients. Our work processes, as well as CVS's behavior, inhibit checking out modules from multiple repositories under the same directory tree. This helps us avoid the accidental mixing of IP from different places. It also lets us segregate the development to a fixed set of machines, which we must do if they are to be delivered to the customer. When we have several projects from one client, we sometimes make separate repositories for different projects, and sometimes make different modules within the same repository. It all depends on the amount of sharing of the code base we expect, on the whim of the project leader, and on the phase of the moon :-). Fred == Fred Brehm, Sarnoff Corporation, [EMAIL PROTECTED]
WinCVS In the way Update Problem
Greetings, I'm using WinCVS 1.1b13 on NT in pserver mode against a Solaris 2.6 server running cvs 1.10.8. I have a checkout area in which I only perform "cvs update -Pd" (no editing is done in this area). Occasionally for some directories I get the following error for each file in the directory, such as: cvs server: Updatingics/infra/src/com/ec/infra/busproc/addressbook/exceptioncvs update: move awayics/infra/src/com/ec/infra/busproc/addressbook/exception/AddressbookBusinessException.java;it is in the wayCics/infra/src/com/ec/infra/busproc/addressbook/exception/AddressbookBusinessException.javacvs update: move awayics/infra/src/com/ec/infra/busproc/addressbook/exception/AddressbookException.java;it is in the wayCics/infra/src/com/ec/infra/busproc/addressbook/exception/AddressbookException.java I understand why this error is normally generated (when the user has manually created a directory which CVS does not recognize from its Entries file as having created, and CVS subsequently tries to create the directory as a result of a cvs update). For this case, however, no manual editing has been done in the working area. Looking at the CVS/Entries file under addressbook, there is no directory entry for "exception"; therein lies the problem. For some reason, during the checkout or update -d in which the exception directory was originally created, the Entries or Entries.log file was not properly modified to include "exception". The workaround is to delete the exception directory and update again, at which time the Entries.log is created and then on subsequent update, the Entries file is appropriately modified. Has anyone else seen this problem? Thanks, David
Re: error in cvstreedefault
[EMAIL PROTECTED] wrote: keep getting the following errors : Error: $cvstreedefault points to a repository not defined in %CVSROOT (edit your configuration file /usr/local/apache/conf/cvsweb.conf) in cvsweb.conf : %CVSROOT = ( 'Development' = '/cvs', 'Configuration' = '/cvs/CVSROOT/config', 'HTML-files' = '/usr/local/apache/htdocs' ); $cvstreedefault = 'Configuration'; what did I do wrong ? any idea ? Thank you. the cvsweb.conf file is read in by the .cgi and if you have a typo in there, the variable names get mixed up. personally, i made the error not to escape some chars (e.g. @) in the mailto: url in the address variable assignment and in the long instruction variable (why the hell doesn't disable end-of-text-marking the need for escaping chars???). for me perl seems like "there's more than one way to do the same error". eric.
merging main trunk to branch?
Hi, I am new to CVS and I use WinCVS as the front end on a local repository. I have a module with a main trunk. There is a branch off that trunk that I use as experiment code. Is there a way to merge the lastest changes on the main trunk to the branch code? I don't want the experiment code to affect the main trunk. I only found instruction for the other way around. Thanks.
How to get removed from this mailing list
hi, how to get removed from this mailing list? thanks -john -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] Sent: Thursday, August 17, 2000 10:31 AM To: [EMAIL PROTECTED] Subject: merging main trunk to branch? Hi, I am new to CVS and I use WinCVS as the front end on a local repository. I have a module with a main trunk. There is a branch off that trunk that I use as experiment code. Is there a way to merge the lastest changes on the main trunk to the branch code? I don't want the experiment code to affect the main trunk. I only found instruction for the other way around. Thanks.
Re: merging main trunk to branch?
On Thu, Aug 17, 2000 at 05:30:32PM -, [EMAIL PROTECTED] wrote: Hi, I am new to CVS and I use WinCVS as the front end on a local repository. I have a module with a main trunk. There is a branch off that trunk that I use as experiment code. Is there a way to merge the lastest changes on the main trunk to the branch code? I don't want the experiment code to affect the main trunk. I only found instruction for the other way around. Thanks. It's pretty much the same method, actually. Think of merging not so much as bringing two points together but, instead, taking the differences between any two arbitrary points in the system, and applying those same changes to your current code. If you've played in the unix world a lot, and used to people throwing around patches, it works just like that. Sometimes I find it conceptually easier to think of the process as 1) doing a diff between two versions, 2) applying that diff as a patch to the current version. Actually, there was one time where I actually had to do that because CVS couldn't handle the merge itself, so I had to use Larry Wall's actual patch program to apply it. Anyway, obviously it tends to only make sense to generate diff/patch from two version on the same branch. But you can, in general, apply that patch to any branch you want. So, for your situation, yuou have something like this: Branch point v 0+ Main | +--- Test Now, hopefully you labled the branch point with something when you created the branch. Otherwise, this is a pain. I will assume that the regular tag is called test_bp, and the branch tag is called test. First, get onto the main branch: cvs up -A Tag the current versions: cvs tag merge_to_test_1 Now, onto the test branch: cvs up -r test Now merge: cvs up -j test_bp -j merge_to_test_1 validate, then commit: cvs commit -m "merged in changes from test_bp to merge_to_test_1" Now, the reason for using test_1 is so, later on, when you want to do that again, you would do: cvs up -A cvs tag merge_to_test_2 cvs up -r test cvs up -j merge_to_test_1 -j merge_to_test_2 cvs commit -m "merged in changes from merge_to_test_1 to merge_to_test_2" Hope that helps! mrc -- Mike Castle Life is like a clock: You can work constantly [EMAIL PROTECTED] and be right all the time, or not work at all www.netcom.com/~dalgoda/ and be right at least twice a day. -- mrc We are all of us living in the shadow of Manhattan. -- Watchmen
Re: merging main trunk to branch?
Mike Castle wrote: So, for your situation, yuou have something like this: Branch point v 0+ Main | +--- Test Now, hopefully you labled the branch point with something when you created the branch. Otherwise, this is a pain. (Allow me to advertise my patch again) My patch will alleviate some of this pain, (though my patch may imply some changes that need to be made to wincvs or other cvs clients besides the command line cvs client, specifically in the area of allowing "cvs add" to work with the ".trunk" tag.) http://www.geocities.com/dotslashstar/branch_patch.html Then you could merge from trunk to branch "foo" like this, using the ".trunk" and "branchname.origin" pseudo tags that the patch implements. cvs rtag -r .trunk trunk_merge_point modulename cvs co -r foo modulename cd modulename cvs update -j foo.origin -j trunk_merge_point cvs commit cvs tag foo_post_trunk_merge -- steve __ Do You Yahoo!? Send instant messages get email alerts with Yahoo! Messenger. http://im.yahoo.com/
Pb with module deletion
Hello, I'm using a sourceforge account and I'm learning CVS at the same time. My problem stays in deletion of directories in the CVSROOT directory. In my case cvsroot is cvs.somewhere.else:/cvsroot/myrepository. I have now there the following directories : speedshooter (the good one) speeshooter (a bad one) ... (other bad ones) when I use the "cvs [EMAIL PROTECTED]:/cvsroot/myrepository co .", The server send me the emptied directories I just talk about previously. Sure I can make a "cvs update -P" to get rid off this dirs. By the way, although it is clear for me, It'll may be not for the other developpers I'm working with. Please, can you give a solution to destroy definitively this directories, or better reinitialize my repository (hoping my access rights are ok to do this). Thanks in advance, excuse me for my poor english (I'm french, that may be the reason my mistakes :-) ). Good bye.
Re: merging main trunk to branch?
One of the tricks I sometimes use, to avoid having to keep track of branch points, is to only merge from any given branch once. Thus, in this case, I would create branch2 from the main trunk and merge the changes from the first branch forward onto the new branch. cvs up -A# get the main trunk cvs tag -b test2 cvs up -rtest2 -jtest And use test2 from now on. By doing that, though, you will lose access to CVS's ability to organize log messages along a single development path by keeping them on a single branch. Derek -- Derek Price CVS Solutions Architect ( http://CVSHome.org ) mailto:[EMAIL PROTECTED] OpenAvenue ( http://OpenAvenue.com ) Mike Castle wrote: On Thu, Aug 17, 2000 at 05:30:32PM -, [EMAIL PROTECTED] wrote: Hi, I am new to CVS and I use WinCVS as the front end on a local repository. I have a module with a main trunk. There is a branch off that trunk that I use as experiment code. Is there a way to merge the lastest changes on the main trunk to the branch code? I don't want the experiment code to affect the main trunk. I only found instruction for the other way around. Thanks. It's pretty much the same method, actually. Think of merging not so much as bringing two points together but, instead, taking the differences between any two arbitrary points in the system, and applying those same changes to your current code. If you've played in the unix world a lot, and used to people throwing around patches, it works just like that. Sometimes I find it conceptually easier to think of the process as 1) doing a diff between two versions, 2) applying that diff as a patch to the current version. Actually, there was one time where I actually had to do that because CVS couldn't handle the merge itself, so I had to use Larry Wall's actual patch program to apply it. Anyway, obviously it tends to only make sense to generate diff/patch from two version on the same branch. But you can, in general, apply that patch to any branch you want. So, for your situation, yuou have something like this: Branch point v 0+ Main | +--- Test Now, hopefully you labled the branch point with something when you created the branch. Otherwise, this is a pain. I will assume that the regular tag is called test_bp, and the branch tag is called test. First, get onto the main branch: cvs up -A Tag the current versions: cvs tag merge_to_test_1 Now, onto the test branch: cvs up -r test Now merge: cvs up -j test_bp -j merge_to_test_1 validate, then commit: cvs commit -m "merged in changes from test_bp to merge_to_test_1" Now, the reason for using test_1 is so, later on, when you want to do that again, you would do: cvs up -A cvs tag merge_to_test_2 cvs up -r test cvs up -j merge_to_test_1 -j merge_to_test_2 cvs commit -m "merged in changes from merge_to_test_1 to merge_to_test_2" Hope that helps! mrc -- Mike Castle Life is like a clock: You can work constantly [EMAIL PROTECTED] and be right all the time, or not work at all www.netcom.com/~dalgoda/ and be right at least twice a day. -- mrc We are all of us living in the shadow of Manhattan. -- Watchmen
Re: Fw: Returned mail: User unknown
On Thu, Aug 17, 2000 at 04:08:23PM -0700, Sandra Wittenbrock wrote: Some .gif files havn't been being checked in as binary, even though they are specified in the cvswrappers file. (*.gif -k 'b') Where they done with some other capitalization, such as .GIF? I usually do: *.[gG][iI][fF] -k 'b' mrc -- Mike Castle Life is like a clock: You can work constantly [EMAIL PROTECTED] and be right all the time, or not work at all www.netcom.com/~dalgoda/ and be right at least twice a day. -- mrc We are all of us living in the shadow of Manhattan. -- Watchmen
Re: cvs history // cvs log
I'm not going to test this, but something like this: cvs co .# get a full repository cvs history -c -u$USER -Dyesterday tmp$$.txt outfile=lineitems-`date +%Y%m%d`.txt while read line; do echo $line $outfile revision=`echo $line |awk '{print $6}'` dir=`echo $line |awk '{print $8}'` file=`echo $line |awk '{print $7}'` cvs log -r$revision $dir/$file |sed -ne '/^revision/,$p'$outfile done tmp$$.txt rm tmp$$.txt Don't run it twice in a row without deleting lineitems-date.txt. No warranties and all that jazz. Derek -- Derek Price CVS Solutions Architect ( http://CVSHome.org ) mailto:[EMAIL PROTECTED] OpenAvenue ( http://OpenAvenue.com ) -- #! perl @a = ( 0x2E805,0x6B39,0x15B3,0x45993,0x153C,0x1D9F ); for ( @a ) { ( $s, $i )=( 'a', 0 ); $s++ while $i++ $_; print "$s" } TC wrote: I work as an indepdent contractor, most clients require detail line item billingwhich requires me to itemize the scr files i work on each day. I have been just updating my own work order system.I realized a cvs commit -m sdfsd filename at the end of each day, gives methe basic's for my line item billing info but do not seemd to be able to get a datesorted list with the comit messages using either cvs log or cvs historycvs history -c -uME would be perfect if I could just get the message for eachone of these commit messages. Anybody know a combination to get cvs history to spill messages for each commit ???
RE: Returned mail: User unknown
Some .gif files havn't been being checked in as binary, even though they are specified in the cvswrappers file. (*.gif -k 'b') I was looking at some of the old messages for this news group, and there is some discussion of the -k 'b' being disabled for wrappers in client/server CVS. I am running CVS 1.10.7 (client/server). Has this feature been disabled? If not, does anyone have any ideas about why the files aren't being committed properly? Just a guess, but maybe mumble.GiF is committed as text, while fumble.gif is committed as binary. If so, try *.[Gg][Ii][Ff] -k 'b'
RE: Returned mail: User unknown
I know how to fix the problem though... go to the repository and fix the ,v file in there. do a " cvs update -k 'b' " on the file. [EMAIL PROTECTED] on 18/08/2000 10:08:17 To: [EMAIL PROTECTED], [EMAIL PROTECTED] cc: (bcc: Aditya Sanghi/Net) Subject: RE: Returned mail: User unknown Some .gif files havn't been being checked in as binary, even though they are specified in the cvswrappers file. (*.gif -k 'b') I was looking at some of the old messages for this news group, and there is some discussion of the -k 'b' being disabled for wrappers in client/server CVS. I am running CVS 1.10.7 (client/server). Has this feature been disabled? If not, does anyone have any ideas about why the files aren't being committed properly? Just a guess, but maybe mumble.GiF is committed as text, while fumble.gif is committed as binary. If so, try *.[Gg][Ii][Ff] -k 'b'
Re: merging main trunk to branch?
Again, I think my "get date from file" patch would help here. Do -jbranch:date:.last right?