Re: how do I undo local changes
Hello, adding to what Jason and Rahul said: * On Fri, Jul 08, 2005 at 03:27:13PM -0700 MEENA SELVAM wrote: but in CVS I only find the cvs admin -o command to revert to older versions. But how can i simply undo the changes locally, without concerning about version? From your statement, I am not sure if you correctly understand the impliciations of the admin -o command. Thus, the following applies: WARNING: DON'T DO THIS! ;) There is almost never a reason to do a cvs admin -o, thus, be very careful unless you fully understand what that means. This command has NOTHING to do with what you want to achieve. Your task can be done by removing the file before doing an update, or using up -C (as Jason und Rahul said). Best regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Cannot remove directories from the CVS repository
Hello Peter, * On Wed, Jul 06, 2005 at 10:08:22AM -0400 Peter Desjardins wrote: Right now, if I checkout the module, CVS will create the old name directory and some of its subdirectories in my working directory. Then if I use cvs update -P immediately after the checkout, CVS will include the following message for the old name directory and each of those subdirectories: cvs update: cannot remove directoryX/subdirectoryY: Directory not empty Have you tried checking out with -P, that is, cvs co -P? Does it generate the same directories? The empty directory does not happen to be only different in case to another directory (or file) in your repository? Recent cvs versions have problems if this is the case. This might be your problem, too Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS Server Change
Hello again, * On Fri, Jun 24, 2005 at 02:51:08PM +0900 Soo-Hyun Choi wrote: What would be an appropriate way to merge the two CVS servers? I am just trying to simply copy the files (,v files) in one CVS server to the other CVS server. Would that cause any harm for the existing CVS files in the target CVS server? Ok, in some more details: You have a cvsroot directory, in your example, the local one is ~/cvsroot. In that directory, you have a CVSROOT directory. In your example, this is the ~/cvsroot/CVSROOT/ directory. This CVSROOT directory contains some information about your repository (for example, your modules file). If you happen to overwrite the modules file in the server CVS, then the modules there would be lost (or, to say it better, you would have to restore the file to be able to access them like before). Of course, I do not know if you ever changed anything in your ~/cvsroot/CVSROOT/ directory. If you did not, just move every ,v file to the server, leaving out all files in ~/cvsroot/CVSROOT/. Make sure you do not overwrite other files if they happen to already exist on the server! (directories with the same name, for example). If you made any changes to your ~/cvsroot/CVSROOT/ directory, I would suggest you should try to find out what you changed exactly, and put these changes in the server's CVSROOT by hand after thinking about what you need and what you do not need (or, what might break things on the server). HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS Server Change
Hello, * On Fri, Jun 24, 2005 at 01:13:28AM + Pierre Asselin wrote: Soo-Hyun Choi [EMAIL PROTECTED] wrote: I am a beginner in using CVS, and have a question on to change the CVS server. [ ... ] Would this work if I simply copy the CVS's ,v files with the directory to the machine S's cvsroot directory? Yes. This also makes your existing sandboxes nonfunctional, so be sure to delete them after committing any pending changes. Check out new sandboxes from the new server after the move. To the OP: If your server already has a CVS server (and, thus, a CVSROOT directory) and you want to merge both repositories into one, make sure you do NOT move that directory over to the server, but make sure you MERGE them in an appropriate way. If you just overwrite the old CVSROOT, you might get trouble with the data which is already stored on the server! Apart from this, it works exactly like you thought (and Pierre confirmed). Obviously, its best if you do not try to merge two cvs servers into one. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Referring the base of a branch
Hello Derrick, * On Thu, Jun 16, 2005 at 06:39:35PM -0700 Derrick Bass wrote: I did something dumb and created a branch without tagging the base of the branch first. Now I want to merge the changes on the branch into the mainline, but I can't figure out a way to refer to the base of the branch in a cvs update statement. (Using -j Branch-name gets me the tip, of course.) Is there a way to that? If not, what's the best way to merge the changes? I'm no expert on the CVS file format, so be warned to test any manipulation I suggest on a copy of your repository, not on the repository itself! You can get to the revision of each file where it was branched rather quickly. Do the following: - FOR each file DO: - cvs log file - grep the branch name out of the output of the above (for example, you get something like branch: 1.164.0.14 - Get the number from above and remove the 0 and the number after it. This is the base where the branch was created. For example, here, we get 1.164 - Now, tag the file on this revision: cvs tag TAGNAME -r 1.164 file As I told you, this SHOULD work, but I cannot be held reliable if it does no!. Thus, again, I advice you to test it with a copy of you repository, and wait if someone here corrects my posting. :-) HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: cvs 1.12.9 and new loginfo format
Hello Mark, first of all, thanks for your answer. * On Sun, Jun 05, 2005 at 09:39:07PM -0700 Mark D. Baushke wrote: Read https://www.cvshome.org/docs/manual/cvs-1.12.12/cvs_18.html#SEC188 Yes, I had already done this. Add the line 'UseNewInfoFmtStrings=yes' to CVSROOT/config I had already done this, too. Use: DEFAULT Mail -s %p %s [EMAIL PROTECTED] and you should get something that approximates your old behavior. Thanks, that did the trick. Re-reading the relevant section, I understand it now. I'm not sure why I was confused yesterday; perhaps, I should have wait until today before I wrote to this mailing list. Thank you very much! Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
cvs 1.12.9 and new loginfo format
Hello, I just upgraded to 1.12.9 (from 1.11.x) and I am a little bit confused about the new loginfo format. My current loginfo looks like: DEFAULT Mail -s %1s [EMAIL PROTECTED] Now, I've read https://www.cvshome.org/docs/manual/cvs-1.12.12/cvs_18.html#SEC188 and added the 1 as suggested there. Anyway, I I do not quite understand how to go from here. :-( What must I add to my loginfo line to make it behave like it used to behave with 1.11.x? I don't get it right from https://www.cvshome.org/docs/manual/cvs-1.12.12/cvs_18.html#SEC192. From my understanding, changing the line back to DEFAULT Mail -s %s [EMAIL PROTECTED] (that is, removing the 1 again) should do the trick for me. Anyway, I am confused if this is the right step, or if there are any side-effects I do not foresee yet? Thanks for any help, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: $id automatically inserted in file
Hello Klaus, * On Wed, May 25, 2005 at 12:50:29PM +0200 Klaus Hebsgaard wrote: I am using cvs for keeping some html files, in these html files i have strings like $id$. [...] However I don't want this to happen. https://www.cvshome.org/docs/manual/cvs-1.11.20/cvs_12.html#SEC100 HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: List of files changed between 2 tags
Hello, * On Wed, May 25, 2005 at 06:53:49AM -0700 [EMAIL PROTECTED] wrote: I am trying to get a list of files changed between two tags. I am aware of 'cvs diff' command but not interested to see in the output the lines that have changed, contextually or not, I just need to see _what_ files have changed, period. I have checked all options of 'cvs diff' including --brief, however without any luck. I'm not sure if there is a shorter option, but you could use the following: 1. checkout the first tag 2. perform a cvs up to the 2nd tag This way, you can check the output of the up command to see which files changed. HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Ignoring keywords during update
Hello, * On Wed, May 25, 2005 at 09:52:59AM -0400 [EMAIL PROTECTED] wrote: What is the option/syntax to ignore keywords during an update? cvs up -k??? cvs up -kk (see the double k) Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: $id automatically inserted in file
Hello, * On Wed, May 25, 2005 at 10:21:02AM -0400 Russ Sherk wrote: cvs admin is intended to be used on the server. No, it is not. It can be issued from a client, too. Furthermore, I can't believe TurtoiseCVS not being able to do this itself. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Can't checkout to my C: root directly *PANIC* Admin Files Missing
Hello, * On Tue, May 10, 2005 at 08:57:48AM -0700 S I wrote: OK, I wrote this email and looks like it was bounced back or didn't get through to the mailing list, however, if it did, please pardon the redundancy. This mail went through. Anyway, the mail server often needs some time to deliver messages, thus, I am Cc:ing to you directly. If I check out any of our CVS modules or projects from the main respository to any subfolder of my drive C: everything goes smoothly. However none of the versions of CVS I tried allow me to directly check out a project to the root. Well, I never use cvs directly from the CMD.EXE prompt, but always from inside of Cygwin (and it would need some modifications here to be able to access it directly). Anyway, just a wild guess because of your message: I get these bizzare errors: cvs checkout: in directory .: cvs [checkout aborted]: *PANIC* administration files missing Do you have the right to create directories on C:\? This might be a permission problem. HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Can't checkout to my C: root directly *PANIC* Admin Files Missing
Hello, * On Tue, May 10, 2005 at 09:51:50AM -0700 S I wrote: I don't have this problem on WinNT, however, only on WinXP or Win2003 servers. Well, ok. I use my XP (SP1) box (with a self-compiled cvs 1.11.17 on cygwin), set the paths correctly for accessing CVS and SSH, and tried it myself: C:\cvs co PROJECT (The server is a Linux box, cvs 1.11.20:) The checkout worked flawlessly. Thus, I still vote for the permissions problem. ;-) At least, I do not have an idea what else might be the problem. Remember: On Windows, the administrator does not have unlimited rights like on unix-style machines. You might not have the right to create directories in C:\ itself? I'm leaving the rest for the mailing list; perhaps there will be someone who has an idea and can jump in? I do have full admin access and what I discovered after I posted to the mailing list was that On XP/2003: If my project does not exist at the root, cvs co -P proj succeeds, however, if the proj already exists, cvs checkout does not word and produces the errors I mentioned. On NT: If cvs checkout detects the project already exists, it does an UPDATE instead. In both cases, I'm connecting to the main repository on the linux server. This is a weird behavior I've not seen before. But thanks for your help. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: removing a revision
Hello, * On Thu, Apr 28, 2005 at 10:36:02PM -0400 Igor Schein wrote: let's say I have 4 revisions of a file, 1.1, 1.2, 1.3 and 1.4, and I want to wipe out rev 1.2 completely, shiting the later 2 revisions: 1.1-1.1 1.2-gone 1.3-1.2 1.4-1.3 I can do cvs admin -o 1.2 filename, but I can't figure out an easy way to shift the other 2 revisions. Why do you want to do that? Think of these numbers as magic numbers and do not try to give them any meaning at all! That said, IF YOU REALLY KNOW WHAT YOU ARE DOING, you can do the following: Remove the revision 1.2 with cvs admin -o 1.2 filename, and edit the filename,v file directly in the repository. Anyway, I do not recommend this as this is very error prone. For example, you have to think about all the branches. Thus, my recommendation: Remove 1.2 and leave everything as it is now. HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Annotate binary files...
Hello Andreas, * On Sun, Apr 24, 2005 at 12:39:53AM +0200 Andreas Volz wrote: Am Sat, 23 Apr 2005 17:27:04 -0400 (EDT) schrieb Larry Jones: It should. What does ``cvs status'' say about the problem files? cvs status works without problems. What does cvs status tell about the status of the binary files? Are these really tagged as -kb? Best regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: 'cvs status' returns error if I delete a directory in my local working copy
Hello, * On Sat, Apr 23, 2005 at 02:38:08AM + Todd Foster wrote: If for some reason, Dir3 is deleted, then cvs status returns an error message: How do you delete it? I hope, you are using cvs release -d? Is this expected behavior? I'm asking because there are some inconsistencies in the 'cvs status' behavior. I don't think that this is expected behaviour, either, as it is inconsistent. Anyway, it is always better to remove a directory by running cvs release -d instead of deleting it directory. I believe your odd behaviour will be gone with this. HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: File size different with Wincvs
* On Fri, Apr 22, 2005 at 02:52:25PM +0530 Balaji D wrote: Whenever I import an executable file or a java archive file to cvs area and check out in Wincvs, there is a mismatch in the file size drastically, but the same exe happens to be perfectly checked out in Unix environment... A binary file has to be marked as such. For cvs (from cvshome.org), you have to specify cvs add -kb filename. I don't know for WinCVS (and many others here do not know, too, because this mailing list is meant mainly for the cvshome.org version). Quoting Arthur Barrett: ;-) In future if you direct questions regarding CVSNT to the CVSNT newsgroup you will probably get a more direct response: http://www.cvsnt.org/cgi-bin/mailman/listinfo/cvsnt or news://news.cvsnt.org/support.cvsnt HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: checking out multiple branches of a project
Hello Doug, * On Thu, Apr 21, 2005 at 09:59:30AM -0400 Doug Lee wrote: I have often wished for an automatic way to check out all existing branches of a given module with one command. Example: If I have module mymod with HEAD and branches named rel1 and rel2: SVN has such an option (if you follow the recommended style of creating branches). Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: updating Entries files only
Hello Jim, * On Tue, Apr 19, 2005 at 04:33:48PM -0400 Jim.Hyslop wrote: I see. That's a tricky one to overcome. Sometimes, you just gotta bend the rules a little: how important is it, really, for everything to be lower-case? Of course, this is all perfect hind-sight, and doesn't help you in your day-to-day work. The problem here is the following: I am managing some upstream sources in my own, local repository. I cannot change anything about what I get to be imported. As I must compile this project on Windows as well as on Linux, I cannot locally changes the names locally (or I will have to adjust many paths in the files). IMHO, this is a CVS bug, isn't it? Not really. CVS is at the mercy of the underlying operating system. It tries its best when you have two conflicting types of O/S, but sometimes it just cannot resolve the problems. This is one of those times. You have to step in and fix things up for it. Well, IMHO, CVS should be able to resolve this. As I always checkout with -P and one of the two directories is always empty, the first directory could have been already pruned when the second directory is checked out, or, if the directory order is reverses, the empty directory would not even need to be checked out (as -P is given). Thus, there would be a solution. Of course, without looking into the CVS sources, I do not know how much work this is. Actually, though, I should think a 'cvs update -P ABC' followed by a 'cvs update -ld abc' should do the trick for you. Thanks, Jim! This is the solution to my problem. At least on a first test, it worked exactly as expected. I always tried to solve this issue with cvs release -d ABC and cvs update -d abc, but this did not work. I thought your and my way would be the same, but it seems cvs release removes ABC from the CVS/Entries file, while cvs update -P does not. Thank you, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: updating Entries files only
Hello Frederic, * On Tue, Apr 19, 2005 at 04:31:39PM -0400 Frederic Brehm wrote: How about renaming the *repository* directory ABC to abc? Try it (in a copy of the repository) and see if it breaks anything that you care about. Thank you for this suggestion. I already tried it, but without success. On Windows, it worked, but it broke the build on Linux (obviously). Anyway, Jim's solution worked as expected. Thank you, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: updating Entries files only
Hello, * On Wed, Apr 20, 2005 at 10:15:42AM +0200 I wrote: Actually, though, I should think a 'cvs update -P ABC' followed by a 'cvs update -ld abc' should do the trick for you. Thanks, Jim! This is the solution to my problem. Doing some more tests, this is not a solution to the problem. It seems to depend upon the order in which cvs sees the directories ABC/ and abc/ if it works or not. Thus, modifying CVS/Entries seems to be the only reliable way to solve this. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: updating Entries files only
* On Tue, Apr 19, 2005 at 10:53:28AM -0400 Jim.Hyslop wrote: As a side note, though: don't be poking around the Entries file. There is rarely, if ever, any need to examine the contents of the file. You mean, do not poke around with them unless you are working on a Windows system, the server is Linux, and in the repository, there are two directories SOMEPATH\ABC and SOMEPATH\abc where ABC/ is empty (everything in the Attic), and abc/ contains some data? In this case, cvs cannot checkout abc/ because ABC/ is in the way. If you want to use cvs, you HAVE to fiddle around with the Entries file. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: updating Entries files only
Hello Jim, * On Tue, Apr 19, 2005 at 03:31:33PM -0400 Jim.Hyslop wrote: If you get into that situation, then the solution is to (a) go rap somebody's knuckles, then (b) fix the collision in the repository. Well, renaming ABC/ to abc/ was a deliberate decision as all directories should be lowercase on our side. We did not know that CVS has such a problem with this. Removing ABC/ from the repository would break history and the ability to build old versions, thus, it is out of question. Yes, we use co -P, thus, I would have assumed that ABC/ is already gone before abc/ is tried to be checked out. Anyway, this is not the case. IMHO, this is a CVS bug, isn't it? If you want to use cvs, you HAVE to fiddle around with the Entries file. I would disagree completely with that statement. I'm sorry, this statement was not meant as it seems to be. It's my fault, I should have formulated it more precisely. Thus, I reformulate: IN THIS SITUATION [given above], if you want to use cvs, you HAVE to fiddle around with the Entries file [because CVS will not work otherwise]. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Share A cvs repository between two OSes
Hello, * On Tue, Apr 19, 2005 at 12:56:50PM -0700 Mr. Question wrote: If there was some way to have windows map the ext3 drive so that it could be accessed from a drive letter, all would be well. While not CVS related, EXT2IFS (http://www.fs-driver.org/) might help you. It can mount ext3 (as ext2, of course: http://www.fs-driver.org/faq.html#acc_ext3). Anyway, I do not believe I would use this one on a production server. Of course, I would ensure that all CVS versions (on Windows and Linux) are *exactly* the same, to prevent different interpretations and possible repository corruption. To make a long story short: Personally, I would *NOT* use your setup, but use an external server. YMMV, and if you do try this, then I would like to see an experience report on this mailing list. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS import for executables and binary files
Hello, * On Mon, Apr 18, 2005 at 10:26:58AM +0530 Balaji D wrote: When I import few object files and exeuctables, through cvs import command through command line, I dont get these files reflected in the cvs area. Can some one help me to find out the option for importing those files. Most probably, they are ignored because of the build-in ignore list if CVS, cf. https://www.cvshome.org/docs/manual/cvs-1.11.19/cvs_18.html#SEC178 If you set -I ! in the command-line of cvs import, the binary files should be handled, too. Also, I import .jar files into the cvs area, they donot reflect the original size in the cvs area and even when I do a check out, their sizes seem to be less...(For example a .jar of 393 KB is reflecting as 7KB when I import to cvs and check it out). .jar files are compiled java classes, aren't they? In this case, they are binary files, and they have to be handled like this. See https://www.cvshome.org/docs/manual/cvs-1.11.19/cvs_9.html#SEC79 and https://www.cvshome.org/docs/manual/cvs-1.11.19/cvs_18.html#SEC166 HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Versioning modules instead of files?
Hello, * On Thu, Apr 14, 2005 at 11:45:30AM +0200 Matthias Kaeppler wrote: However, especially in early phases of development, revising single files is sort of too granular, because the changes are so huge. It would be a lot better, if the versioning would happen on the module base, i.e. you can commit the content of your sandbox as a new version of your program rather than new versions of the individual files. AFAIK, this is not possible with CVS. Anyway, Subversion, which many people regard as a replacement for CVS, has exactly the feature you are requesting. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Removing nonexistant mail account (was: Benachrichtung zum +ANw-bermittlungsstatus (Fehlgeschlagen))
Hello Jim, * On Fri, Apr 08, 2005 at 11:02:41AM -0400 Jim.Hyslop wrote: I have sent numerous requests to [EMAIL PROTECTED] asking to have him removed (that is the address listed at http://lists.gnu.org/mailman/listinfo/info-cvs as being the owner). I have neither received a response from info-cvs-owner, nor has the offender been removed from the list. I have not yet tried the alternate address, [EMAIL PROTECTED] A good hint, I totally forgot about this. BTW, at least the first mail address you mentioned is not the one listed on the given page. ;-) It's gnulist-ownrr (yes, there is a typo in it. This might be intentionally, or it might not.) I'll try it myself. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Fwd: Benachrichtung zum +ANw-bermittlungsstatus (Fehlgeschlagen)
Hello, * On Fri, Apr 08, 2005 at 12:16:29PM +0530 Satya Prasad D.V wrote:a Does this mean, no one receiving mails OR this guy's id no longer exists?? It's only that this guy does not exist anymore. As he does not exist for very long know, this might raise the question if anyone can remove him from the list? It is really annyoing to get these does not exist message (not to count the I am on vacation messages - it seems some people cannot handle their software correctly). Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Win32 cvs client - Unix cvs server
Hello, * On Sat, Apr 09, 2005 at 12:29:57AM +1000 Sisyphus wrote: Which brings me back to finding out how to get the SSH stuff working with the cvs-1.11.19 binaries. Any advice/links on that ? Yes, the link I already gave you: http://sourceforge.net/cvs/?group_id=122047 You might even want to follow the links given there. ;-) They handle very explicitly how to set up ssh, cvs, and the like, especially in the context of sf. Good luck, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Win32 cvs client - Unix cvs server
Hello, * On Fri, Apr 08, 2005 at 12:36:39PM +1000 Sisyphus wrote: The FAQ says that Win32 cvs clients don't get along with Unix cvs servers, and the mix should be avoided. Is that advice still correct ? I use this constellation myself very often, and I do not have this problem. Thus, from my point of view, this seems not to be correct anymore. But I'm sure others will have to say more on this topic. Perhaps, this was meant for cvsnt clients? I have cvs-1.11.19 (binaries obtained from cvshome website) on my Windows 2000 box, and I have cvs-write permission to a sourceforge project. With 'pserver' I can 'login', 'co' and 'update' - but I can't 'commit' any changes I make. All I ever get is: cvs [server aborted]: commit requires write access to the repository Well I do have write access - I assume the problem is that fact doesn't get established. No. The problem is that you did not read the sourceforge documentation correctly. ;-) Quoting http://sourceforge.net/cvs/?group_id=122047: Anonymous CVS Access [... telling about pserver ...] Developer CVS Access via SSH [... telling about writing to cvs with ssh] You might want to check out http://sourceforge.net/docman/display_doc.php?docid=14033group_id=1#howtoaccess: If you are a project developer and will be writing to the CVS repository, you will require an additional piece of software, an SSH client. I highly recommand you read the above mentioned web pages, they tell you all you need to set up cvs with sourceforge. Good luck, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: no such repository
Hello, * On Mon, Apr 04, 2005 at 06:00:26PM +0200 Vincent YSMAL wrote: ~]# cvs -d :pserver:[EMAIL PROTECTED]:/cvsroot login Logging in to :pserver:[EMAIL PROTECTED]:2401/cvsroot CVS password: /cvsroot: no such repository If you've got an idea, thanks No one have an idea ... sicc :(( I never used pserver, so this is just a wild guess: Does /cvsroot (at the root!) even exist on your server? Did you ever do a cvs init on it? Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Is there a safe way to do this kind of offline CVS management?
Hello Doug, * On Sun, Apr 03, 2005 at 04:13:24AM -0400 Doug Lee wrote: I very much prefer having the off-site history preserved in the main repo, for various reasons. I'll have to think through how important that really is though. Another option I use many times: Before going away from my CVS server, I checkout a sandbox (on an own branch) on which I want to work, in a directory dir/. After this, I copy dir1/ to another directory, for example, dir.1/. Now, whenever you want to check in something, copy dir/ to dir.2/, dir.3/, dir.4/, and so on. Furthermore, write the changelog you would want to apply into a file with a speaking name, for example, changelog.dir.2, changelog.dir.3, and-so-on. When you come back home, do the following: - go to dir.2, commit and use the changelog file changelog.dir.2 - go to dir.3, use cvs up, commit and use the changelog file changelog.dir.3 - go to dir.X, use cvs up, commit and use the changelog file changelog.dir.X This way, you preserve all the history (in a branch, but at least, you have it). This way, there is no need to fiddle with the CVS repository itself. Another option, which involves some manual steps but allows you to even ommit the branch: Do as before, that is, generate dir.1, dir.2, dir.3, and so on. After coming home, diff against the directories (diff -urN dir.1 dir.2 1-2.diff, diff -urN dir.2 dir.3 2-3.diff) and put the diffs into files. Now, go to dir.1 (the original copy), cvs up, and apply the diffs to the sandbox (patch 1-2.diff), commit and use changelog.dir.2. Now, apply the next diff, commit and use changelog.dir.3. As long as there are no conflicts applying the patches, this works as expected I hope I did not forget any step. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Idea for reducing disk IO on tagging operations
Hello, sorry for the late reply to this, but I was on vacation. Anything, I believe I might be able to contribute something to this discussion, which even resulted in some code. * On Sun, Mar 20, 2005 at 11:54:32PM + Dr. David Alan Gilbert wrote: OK, my conscience will let me carefully ignore NFS issues given the pain it causes me elsewhere (and I make my mechanism switchable). What happens if I only used the overwrite mechanism if none of the characters being modified crossed a 512 (e.g.) byte boundary offset in the file? Since the spaces were actually written in a previous operation we can assume that the space is allocated and no allocation operation is going to happen at this point (mumble filesystem journalling mumble!). IMHO, here, you are not correct. If I write X times a char Y into a file, I cannot assume that memory for X characters has been allocated. The file system can do some optimizations, compress the file (for example, run-length encoding RLE: First character tells that X times the same character will be written, and the character itself is written afterwards), or anything else. Furthermore, think of so-called sparse-files, which can be rather big - much bigger than your actual medium is itself. Because of this, even a block boundary in the file does not make much sense, IMHO, for the general case, that is, arbitrary file systems. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: recursive add
Hello, an answer to an older post: * On Tue, Mar 01, 2005 at 10:59:32AM -0800 Wei Shi wrote: Maybe this question has been brought up before. Is there a recursive add command in CVS? I'm not sure what you mean with recursive add. Possibly, the following commands might accomplish what you want to do: $ cvs add `find . -type d` $ cvs add `find . -type f` (I expect you are running on a Unix box, or with Cygwin on Windows) If there are too much files/directories, you might encounter problems with the maximum line length of the shell interpreter. In this case, replacing the second command with $ find . -type f -exec cvs add \{\} \; is a good replacement, neglecting the fact this this command will be much slower for a remote repository. HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: How to change the 'x' bit of a file in CVS?
Hello Jean-François, * On Mon, Feb 28, 2005 at 06:35:30PM +0100 JF Larvoire wrote: We added a file in our CVS system by mistake with the wrong 'x' flag value. I tried checking in an update with the correct 'x' bit, but this does not work. I tried deleting and adding back in the file, but this does not work either. If you have access to the repository, change the file attribute there to the ,v file. Notice that an cvs up in your sandbox will not set the attribute correctly; for this, you have to remove the file and do a cvs up HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Info-cvs Digest, Vol 27, Issue 40
Hello, * On Thu, Feb 17, 2005 at 09:45:40AM -0700 Mark E. Hamilton wrote: Using 'update -A' will reset any sticky tags. What if his file is checked out from a branch? Wouldn't this get him the wrong working copy? This is correct. here, you need to copy the original jar file back over the sandbox version, as the sandbox version is corrupted cvs commit Why would you need to do this? I thought that keywords were only expanded when the file was updated/checked-out/exported, so that the file in the repository should be okay; just the working version is corrupted. Not necessarily. If you are operating from a machine that uses CR/LF (for example, Windows), than the CR/LF was converted on check in, thus, the repository version is corrupted, too. The instructions I've given our users is to do this: cvs admin -kb file_name cvs update file_name which has always seemed to work in the past. Are there situations where the repository file could be corrupted, so that this wouldn't work? If you are working from Unix boxes solely, this works as expected. If not, chances are you have to check in a new binary file. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Error when import data..
Hello, * On Tue, Feb 15, 2005 at 01:00:19AM +0100 Cvs by Gsoft wrote: this is the error that appears when found to import a great project : cvsgui [import aborted]: reading CVS/Tag: Not a directory I do not know about cvsgui, but I believe the project you want to import already contains some CVS/ directories. Thus, remove every directory with CVS/ - for example, by typing $ find pathtoproject -name CVS -type d -exec rm -r -i \{\} \; - and try again to import HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Change a keyword for another keyword
Hello, * On Wed, Feb 09, 2005 at 09:36:52AM -0500 Jim.Hyslop wrote: You'll also probably want a way to filter out files in the CVS subdirectories - I haven't figured out how to do that with 'find' yet (if indeed it's possible). from the man page of find, it seems this is possible with -prune. Check out your man page. If this does not work as expected: Let find output the file names, grep -v through the list, and use xargs. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: RV: Newbie question. How to manage replaced files
Hello, * On Sun, Jan 30, 2005 at 09:47:30PM + Pierre Asselin wrote: If you can't determine the revision or tag from which to branch, you probably need to tighten your controls a little... Another possiblity is to tell all your external contributors to send in only patches, never the complete source file. I would suggest the unified diff format (-u). You would have to patch these diff files into your working copies. Furthermore, it is not you who must determine the version the external contributor used, but the contributor himself has to find this out. Regards, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Is it secure to use CVS over the Internet?
Hello, * On Mon, Jan 31, 2005 at 10:21:07AM -0400 Néstor Boscán wrote: Is it secure to use CVS over the Internet?. if you mean pserver access: There have been more than one warning not to use that protocol over untrusted networks. That protocol was not designed with security in mind. Anyway, if you use anything below cvs (for example, ssh), you get the same security the underlying protocol itself offers you. Thus, if you use cvs over ssh, your question reduces to the question: Is ssh over the internet secure enough [for me]? HTH, Spiro. -- Spiro R. Trikaliotis http://cbm4win.sf.net/ http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVSROOT: No such file or directory?
Hello, * On Wed, Jan 19, 2005 at 10:35:08AM -0800 Mark D. Baushke wrote: Second, I have seen something similar related to NFS operations on some hosts where the directory being released was on an NFS mounted directory and there were kernel messages like 'got bad cookie vp hexstring bp hexstring' with the user seeing something like: % cvs release -d foo You have [0] altered files in this repository. Are you sure you want to release (and delete) directory `foo': yes cvs release: deletion of directory src failed: Directory not empty % but I have not actually seen the error you are reporting. Just for the records: I have seen exactly the behaviour you are reporting on a Windows machine. Some poeple might remember my problems with getting CVS to run on DOS I posted some time ago. My setup is now as follows: I checkout everything on a Windows box, and completely work from the Windows share on the DOS box. That is, nothing is done locally in DOS, but all in the (shared) directory. Now, if I just worked in the directory in the DOS box and I am ready to stop working, I am not allowed to cvs release -d the directory. If I do this, I get exactly the message you mentioned above. Interestingly, I have now the following situation: 1. I cannot delete the directory from the Windows share until I have waited some time ( 15 minutes, it seems to me) 2. I cannot issue a cvs release -d afterwards, because CVS does not recognize the directory as a CVS repository anymore. It seems to me that some administrative files are already removed. (and: Yes, I have made sure on the DOS box that I am not in the directory on the network drive anymore: The sandbox can be found in r:\dos\SANDBOXNAME, and I issue a cd \ in r: before trying to release the sandbox.) Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Can't download CVS!
Hello, * On Thu, Jan 06, 2005 at 12:07:27AM -0600 jsWalter wrote: CVS (as in from cvshome.org) is as close to CVS on Linux you can get without running Linux. This is only client software, not the server. This is not true. If you install it on cygwin on Windows, you can use it as a server, too. Anyway, I would not recommend to to use a Windows server if any of the clients uses a case-sensitive filesystem. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Is import really necessary?
Hello, * On Tue, Dec 21, 2004 at 02:54:23AM + Mike wrote: cvs co -ldtop . cd top mv /home/user/newtree . find newtree -exec cvs add {} \; cvs ci Hmmm. I like this! Is it possible to use an import-less approach like this one if top is one's home directory (meaning that there will be a lot of other stuff there besides newtree)? Yes, almost... You can do a cvs co -ldtop . cd top mkdir newtree/ cvs add newtree cd newtree mv /home/user/* . find newtree -exec cvs add {} \; cvs ci BTW: Personally, I would prefer to not move (mv) everything, but to copy (cp -pr) it. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Changing CVS Notify from Mail to use ICQ?
Hello, * On Wed, Dec 22, 2004 at 12:22:56PM -0500 Rich Snowdon-Smith wrote: Has anyone been able to change the cvs notify from using email to another method of message delivery? I'm looking for delivering those notifications via icq. Any hacks out there? Attempted to email to icqnumber@pager.icq.com but got nothing. Any other suggestions or directions? There are tools for ICQ which can be used from the command-line. One of these tools is centericq. If you have a running instance of it, you can use the --send option to send a message. Another option would be to try to set a RSS/RDF feed. Again, centericq can treat these like a contact from some other person. Just some ideas, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list Info-cvs@gnu.org http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Is import really necessary?
Hello Jill, * On Mon, Dec 20, 2004 at 01:10:30PM + J Krugman wrote: But now I can't find this post via Google Groups. Maybe I dreamt it? If not, or if anyone knows what this import-less alternative may be, please refresh my memory. You might refer to that one: http://lists.gnu.org/archive/html/info-cvs/2004-03/msg00246.html (the link I posted in a follow-up is not working, please contact me off-list if you want to get that script). HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Too small permanent file space to co a project
Hello, * On Sat, Dec 11, 2004 at 05:18:12PM +0100 Gerard Cats wrote: Will this work? Yes. Are there things I have to be concerned about, e.g. the administration files in the CVS directories? As long as no CVS watches are set, you do not have any problems. With watches, you will have problems if you overwrite the CVS/ directories with each other, as you will have checkouts then releases. The CVS/ directory is very simple. Just look at the files in there (for example, with less) to see. HTH Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Too small permanent file space to co a project - again
Hello, * On Sun, Dec 12, 2004 at 04:53:28PM +0100 Gerard Cats wrote: In one way or another, the second 'cvs co' will need to know what branches/tags the first co referred to. Is there a standard method to accomplish this, or do I have to maintain my own administration file(s) in $HOME for this? The file CVS/Tag contains exactly the information you need for this. Just have a look into it. If it does not exist, you have checked out your file without using a tag. Else, it contains Ntag, where tag is the tag used for checking out. HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: cvs update: move away file; it is in the way
Hello, * On Thu, Dec 09, 2004 at 08:27:40AM -0600 Rachel Suddeth wrote: I got this in output window from WinCVS. I'm not sure how to say if a message came from CVSNT client, or the server running under Linux but I'm guessing server since path has /, not \ in it... cvs server: Updating BhCtrl/BHTypes cvs update: move away BhCtrl/BHTypes/BHTypes.csproj; it is in the way. You want to checkout (or, in this case, update) the file BhCtrl/BHTypes/BHTypes.csproj from the repository, but this file is already in your sandbox. Unfortunately, the file is not under control of CVS. Thus, CVS does not want to overwrite it, as you might loose the file if it is totally different to what is checked out here. Have you copied the file by hand into the sandbox? You can rename the file BHTypes.csproj to something different and look what CVS checks out for you. Furthermore, I have seen this problem on the CVSHOME version of CVS on Windows systems if there are files in the repository which differ only in the case. Well, IIRC, it was more if there are directories which only differ by case. For more details, see the thread I started here Date: Sun, 18 Apr 2004 18:37:32 +0200 Subject: Windows with Dir/ and dir/ From: Spiro Trikaliotis [EMAIL PROTECTED] Message-ID: [EMAIL PROTECTED] (sorry, the archive does not work now, so I cannot give you an URL) HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: getting files related to two tags
Hello, * On Thu, Dec 09, 2004 at 11:28:07AM -0500 Larry Jones wrote: Sandhya writes: Suppose I want to get files related to two tags. I can check out files based on one tag. And if I try to check out files based on the second tag, all the files related with the 1st tag is deleted. Is there any scenario where I can have files related to both the tags. No. That's not how tags are intended to work. There is a non-CVS way: At first, decide which of the tags A and B is the one from which you want the files if both exist. Assume you want B to take precedence. Then, export the proj with both tags in different directories: cvs export -d proj-A/ -r A proj cvs export -d proj-B/ -r B proj Now, copy everything from proj-B into proj-A. Voila, you have both projects mixed. Of course, this does *not* work if you want to have a real sandbox! Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Two different platforms sharing common code base: How to do so?
Hello, I have a question regarding two projects, which do the same for Windows and/or Linux, and share many files of the common code base. Unfortunately, the build systems or not compatible: GNU Make with gcc on the one hand, Microsoft's BUILD-Tool on the other. MS's solution needs the files to be co-located with the .C-files. To make things worse, GNU uses a file it calls Makefile, as well as MS's BUILD-Tools. So, a typical directory structure (should) look like this: dir1/ dir1/DIRS [Win] dir1/Makefile [Linux] dir1/dir2/SOURCES [Win] dir1/dir2/Makefile [Win] dir1/dir2/Makefile [Linux] dir1/dir2/abc.c[both] The problem is obvious: dir1/dir2/Makefile has to be two different files, which is not possible at all. I see the following solutions: 1. Let Linux use a makefile called GNUMakefile instead of Makefile. 2. Move the Makefile somewhere else, in other directories. No. 1 has been rejected, as the Linux guys thing: GNUMakefile is a name which is reserved for makefiles which use GNU-specific extensions. This is not the case here. No. 2 has been rejected, as MS's build tool does not allow having the SOURCES and Makefile files somewhere else than the .C files. They think it would make the Linux project look like a 2nd class project if the makefiles of Linux or somewhere else, but the makefiles of Windows are in the directory itself. Now, we ask ourselves how we can integrate both projects into one? Currently, we are using two CVS modules now. I do not really see a solution to this. It would be nice if we could have 3 modules: - common files (.C) - Linux build system (Makefile) - Windows build system (Makefile, dirs, sources) Now, if we checkout one module, for example the Windows version, we could checkout the common files, too. Anyway, as far as I understand it, this is not completely possible with CVS, is it? Can we mix files in one directory this way? From my understanding of the CVSROOT/modules file, this is not possible, is it? For example, an ampersand-module puts another module into a sub-directory, they do not overlap. Any advice is highly appreciated. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Two different platforms sharing common code base: How to do so?
Hello John (and the others who answered), * On Mon, Nov 29, 2004 at 09:55:32AM -0600 McNamee, John wrote: If you're not building device drivers, consider using a tool other than build.exe. There are plenty of make utilities for Windows, including a port of GNU make. Yes, I know. This project contains a driver, as well as user mode programs. Although the drivers differ significantly between Linux and Windows (obviously), I like to have one build environment for both types. Because of this, I find build.exe to be the tool of choice. (1) Use a different name for the Linux makefile. makefile.linux, for example. If the Linux developers alias build to make -f makefile.linux, then both Windows and Linux will use the same command to build the code. How much more equal can they be? :-) A good point, which was suggested by another one, too. I did not yet think about such a solution. I will suggest that one internally. (2) The Microsoft build.exe tool can compile code from the parent directory of the sources file. It can? Ok, I did not yet test for the parent directory, but it cannot compile code from another directory. A solution I used before was to #include the .c file, but IMHO, this is very error-prone. I will test this one, too. Thank you all for your suggestions, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Two different platforms sharing common code base: How to do so?
Hello, * On Mon, Nov 29, 2004 at 07:49:06AM -0500 Todd Denniston wrote: I thought it was possible with MS to have them not collocated, but you have to spend time figuring out how to convince the MS tool of that as it is not the DEFAULT mode of operation. Well, it depends upon the tool you're using. We are building drivers, and they need the build.exe tool, which is part of the DDK. That one can only compile files in the current directory. As John has pointed out, it might even be able to compile files in the parent directory, which would help here. 1) install cygwin or another unix like environment on windows, or possibly even just perl. We already use cygwin. 3) create either a makefile (if using cygwin above) or perl script (if using perl above), which: [...] While this would be a work-around, I consider it very error-prone. Thanks, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: new user - do people store compiled code under cvs
Hello, * On Mon, Nov 29, 2004 at 04:53:51PM -0500 Jim.Hyslop wrote: Your instincts are correct. Use CVS to manage files that cannot be recreated by an automated process. Well, yes and no. The problem with recreating some files is, that this recreation might not work at any future point of time (you've get a new compiler, and the old one is not available anymore), or you need exactly what you compiled some time before (for example, for crash dump analyses, having the exact same binary helps much). On the other hand, of course, you are right. CVS is not good in managing binaries. So, I would store them off-line, as Frederic suggested in the other mail. Either make a tarball (.tar.gz or .tgz) if you are on a unix system, or generate a zip if you're running on Windows. On other OS, use something similar. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: not all files get imported
Hello, * On Tue, Nov 30, 2004 at 06:56:04AM - manisha r sankpal wrote: [EMAIL PROTECTED] test_prj import -m log msg test_prj username start but this command fails to check in one directory into the cvs. this dir contains only one *.o file. Have a look here: https://www.cvshome.org/docs/manual/cvs-1.11.18/cvs_18.html#SEC178 By default, CVS ignores some files, even on import. *.o is one of the ignored file patterns. If you want to circumvent this, use -I ! on your cvs import. HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: cvs readonly checkouts
Hello, * On Thu, Nov 25, 2004 at 05:48:33AM -0800 Shri wrote: I want to checkout files such that later commit won't be possible. Is there any option (while checkout) which can set some sticky tag which would not allow me to commit anything, even new files? I do not know what you want to achieve, but exporting file files (instead of checking out) might be what you want. Have a look at cvs export. HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: -- How to delete a Branch ???
Hello, * On Thu, Nov 25, 2004 at 06:52:01PM - Max Bowsher wrote: You can force delete the branch tag *BUT DON'T DO THAT*. If you do, any revisions on the branch still persist, but only accessible by revision numbers - and there is no way to refer to the branch as a whole across all files - the remaining numbers may be unique to every file. Just a thought that came up just now: Could anyone use cvs admin -o to delete the branches revisions, and delete the branches afterwards? For example: cvs admin -o BRANCHTAG:: Not that I would recommend using CVS this way! Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: log
Hello, * On Sun, Nov 21, 2004 at 10:16:24PM -0300 Gleidson Sá Barreto wrote: What means state: Exp; lines: +0 -0 ? RCS knows some states of a revision, but I think this feature is hardly ever used with CVS. In your case, your revision is in state Exp (which should be Experimental), which is the default for RCS/CVS. the +0 means this revision adds exactly zero new lines to the previous revision. The -0 means that this revision deletes exactly zero lines from the previous revision. HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS for DOS?
Hello, * On Fri, Nov 19, 2004 at 10:04:48AM -0600 jsWalter wrote: Now! I have no idea if this will work on a DOS [only] box. Of course, it will not. The windows binaries are Win32 binaries, that is, they are PE executables. OTOH, DOS can only execute the older MZ format. Furthermore, they are using the Windows API, not the DOS API. Best regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS for DOS?
Hello, On Wed, Nov 17, 2004 at 04:32:45PM -0600, jsWalter wrote: But, I figure you can try and see if it works. Any more specific details are welcome. I could test it even myself. Best regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS for DOS?
Hello Arno, On Wed, Nov 17, 2004 at 11:26:37PM +0100, Arno Schuring wrote: Regarding ssh for dos: there is no widely-used tcp/ip stack available for DOS, so such network applications you are very unlikely to find. I know of at least two native DOS TCP/IP stacks: - The Microsoft TCP/IP stack, as delivered with MS LAN Manager - THe IBM TCP/IP stack, as delivered with OS/2 LAN Server I believe both stacks are almost the same. Furthermore, using Windows for Workgroups, you have a virtual stack for DOS (which uses the Windows part as back-end). All three are compatible with one another. Furthermore, I believe that the DOS box of OS/2 or Windows has a virtual TCP/IP stack, too. http://www.wattcp.com is the only one that I know of, but I have never seen a ssh client that uses wattcp. I will have a look. I imagine it would be possible to port cvs to DOS using djgpp (http://www.delorie.com/djgpp), This seems reasonable, yes. but you would still be stuck at the connectivity level - DOS has no native networking support. Yes, I know. That's the reason I mentioned ssh in the original post. There is no real risk in mixing DOS and windows environments (especially up to w98), since they share a common base (filesystem, line endings). Well, I know, as I have used it that way before. In fact, the DOS project started using a RCS implementation for DOS. As the project has all files in one directory, this is not that bad. As I wanted to have all sources on a central server, which is backed up regularly, I moved all RCS files into a central CVS repository. Although I can share the sandbox between Windows and DOS, there are many manual steps involved, which are error prone: 1. Check out a sandbox on Windows, put it into a Windows share 2. Copy the files from the Windows share to DOS (via network) 3. Work on the files If I want to check in, I have to do: 4. Copy the files back to the Windows share 5. commit 6. perform steps 2 und 3 from above Although it might be possible to checkout directly into a share from DOS, I do not like that idea. I do not trust the Windows TCP/IP stack that much, as it is rather old. | You might run into problems with long filenames, but you probably | already have encountered (or willfully avoided) those. Well, as this project started on plain DOS, it only uses features available on it. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS for DOS?
Hello Mark, * On Wed, Nov 17, 2004 at 06:02:26PM -0800 Mark D. Baushke wrote: SSHDOS has an SSH, SCP, SFTP and telnet client for DOS. http://sshdos.sourceforge.net/ Ok, this sounds promising. PuTTY is an SSH1+SSH2 implementation for Windows (not MS-DOS only). Yes, I know. CVS (from cvshome.org) has a cvs.exe that should work in the DOS command window of a Windows box. I do not know how much MS Windows support is required of if you really mean a MS-DOS only environment or a DOS window on a Windows machine. I want to run it on a machine which uses MS-DOS 6.22, nothing more. On cvshome, I only found the Win32 versions. (Menu: ccvs/binaries/...), nothing DOS only. I do not need a Win32 version, as I have a full cygwin environment here up and running on my main machine. Currently, I'm checking out for DOS on Win32, but this is really error prone. I would like to have a DOS only solution. CVSNT (from cvsnt.org) has a both a 'graphical' interface and a command-line backend for Windows environments. Yes, I know. Arthur Barret advertises it very often here. ;-) Please summarize your experience to help other folks in the future. Of course, I will. I think I will try to compile CVS with djgpp, and report whether I had success or not. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS for DOS?
Hello, * On Thu, Nov 18, 2004 at 10:22:16AM + Steve McIntyre wrote: I would like to have a DOS only solution. I've got to ask the question - why? As I told before, I cannot compile this project on Win32, OS/2 or the like. The compiler simply does not work, because its DOS extender does not. Checking out in Win32 and copying over to DOS is error prone, as you must exactly remember which step to do when. If you forget anything, your local changes might get lost, and I do not like that. ;-) Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS for DOS?
Hello, * On Fri, Nov 19, 2004 at 12:10:47AM -0600 jsWalter wrote: Don't let the path definitions below throw you. This *is* on a Windows box (XP and 2k3 server infact) Well, the OP (from me) clearly stated DOS. That is, MS-DOS x.yy. I do not understand what your posting has to do with this? http://web.torres.ws/walters_way But this is for Windows, not DOS. Furthermore, the CVS part - the thing I was asking about - is still missing. I should have stated it more clearly before: I have CVS and SSH set up - on my Windows machine, as well as an my Linux machine. In fact, I even compiled CVS on my own for both environments. I do not need a solution for Windows. I need a solution for MS-DOS, but there seems not to be any. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS for DOS?
Hello Todd, * On Thu, Nov 18, 2004 at 09:17:08AM -0500 Todd Denniston wrote: sneakernet. Yes, I fear it's the way I have to go. ;-( Anyway, in my environment, I mostly use a DOS-Box inside of VMWare, so using physical media would not be that reasonable, would it? I rarely go to a real DOS box, mostly for testing purposes (especially timing). This is mostly done with the binaries only (compiling is slow like hell on that machine). if you don't the linux vfat driver will make your file Why should I use vfat on Linux? Wouldn't it make better sense to use FAT only, as MS-DOS cannot handle lower-case names and/or long filenames, anyway? BTW: After having a look into DJGPP, I found that there is already a CVS version inside of it. Unfortunately, it is some CVS 1.9 or CVS 1.10 (both options are there), and there is no network support in. I doubt it would make much sense to try to port it over to DOS, as the potential user base would not be that big. Because of this, I think I will try something different, taking into account your suggestions here. Stay tuned! ;-) Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
CVS for DOS?
Hello, 1. As we all know, it is always recommended to generate a CVS sandbox at the system on which the development should occur. That is, if we use Windows, use a Windows client, if we use some kind of Unix, use that client. 2. Furthermore, it is considered bad practice to share sandboxes between systems. Now, I have the problem that I need access from a MS-DOS machine to a remote CVS repository. Anywhere, using google, I did not find any version of CVS for DOS (nor did I find ssh, which would be needed for remote access, but this is another story). Currently, I checkout the sandbox in Windows and move it to DOS via Windows shares. It works, but it surely violates 1. and 2. Of courses, Windows line endings and DOS line endings are the same, so 1. might not be that big a problem, but 2. is one. Furthermore, it is very error prone. So, I ask you if you know of any DOS version of CVS (binary), or how I could build that? If anyone knows how to find a working ssh client, this would help me, too. BTW: Now, it is no option to not use DOS at all, but to use the DOS box of Windows, or some emulators on Linux. The compiler environment uses some custom DOS extender, which is not compatible with all DOS boxes I know of. Running a real DOS on VMWare is an option, and this is in fact what I am doing. Any input is highly appreciated, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Need shell script to check out
Hello, * On Mon, Nov 01, 2004 at 01:53:36PM +0530 Antony Paul wrote: I need different versions for each file. There are no tags. Congratulation for this task. (Just kidding) I would suggest not to check the files out this way, but to tag them, one by one, with a unique, but the same tag TAG. This way, you could use the same set of files laters, just specifying a cvs co -r TAG. Do you think you will not need this again in the future? Well, it seems you (or someone else) thought this way in the first place, too. Now, you see how true this assumption was (or not). The problem is I need to retype the password for each file. I want the script to manage the typing of password. You do not tell us which kind of access method you are using. If you're using ssh, then you might want to look at ssh-agent and ssh-add, which does exactly what you want to achieve. HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Changing the name of a branch tag
Hallo Bernd, * On Tue, Oct 26, 2004 at 09:15:45AM +0200 Bernd wrote: 1. Using Admin cvs admin -nBRANCH-TAG:NEW-BRANCH-TAG Try it the other way around: cvs admin -n NEW-BRANCH-TAG:BRANCH-TAG Gruß, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS:Embedding information with keywords
Hello, * On Thu, Oct 21, 2004 at 01:25:28AM -0700 Dizel wrote: /* $magic_keyword: file library.c,v1.22 2004/08/21 ...$ */ Is any keyword like this $magic_keyword$ in CVS? Did you try $Id$? CVS understands all RCS keywords. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: how to un-add an added but uncommitted file?
Hello Marko, * On Wed, Oct 13, 2004 at 01:23:15PM +0200 marko wrote: BUT, I'd like to preserve the file itself! Is there a way of un-adding a file which got by accident added without deleting it?? Rename the file, do what you have done above, and rename it back. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Equivalent to cvs log -rBASE::HEAD?
Hello, * On Tue, Oct 05, 2004 at 12:52:40PM +0200 Slawomir Nowaczyk wrote: I know the code I have in my workspace. But then I want to make cvs update. I would like to know, however, what is going to change - and not just filenames, but the log entries... for all the commits which happened between my last update and now. Something like cvs log -rBASE::HEAD looks perfect, and has only one single flaw: doesn't work. Well, what about cvs diff -uN -r HEAD? I have not tested if it works at all, and it would not give you the log but the complete diff, but it might be a partial solution to your problem? Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS to Subversion
Hello, * On Fri, Sep 24, 2004 at 08:36:09AM +0200 Mario Scheer wrote: Hi, there isn't a problem with subversion, it works really fine. But my problem is to convert a cvs repository using cvs2svn. Then do what Max Bowsher told you: Go to the cvs2svn mailing list. http://cvs2svn.tigris.org/servlets/ProjectMailingListList Hint: The users mailing list is exactly what you want (have a look at the bottom) HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Checkout./export a copy to /var/www
Hello, * On Tue, Sep 21, 2004 at 08:20:45AM +0200 Fabian Cenedese wrote: The checkout is done from the server, so uses Linux line endings. No matter if you can do it directly or copy the stuff to a Windows PC this is not good. You should checkout the stuff on the system where you want to use it. Maybe a timed action (with 'at' or so) can help you by periodically updating an existing sandbox. If nothing is new not much will happen. Another possibility (if ssh is running on the Windows box): Let commitinfo trigger a script which starts an ssh session on the Windows box, which itself synchronizes the sandbox. You can use similar scheme if ssh is not available, but another way to inform Windows. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Broken mail addresses (was: CVS protocol)
Hell Robin, * On Fri, Sep 17, 2004 at 11:03:48AM + Robin Rosenberg wrote: [valuable information!] I would have written this via PM. Anyway, since your mail is not valid, I ask you here: Could you please stop posting with a broken mail address? I do not have a problem if a mail address is not read, but I have a problem if a mail address is broken (your address contains two @ signs). This problem is not only that my mailer stumbles on this address. I would be very thankful. Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Cannot access ^M/CVSROOT
Hello, * On Wed, Sep 15, 2004 at 09:54:41AM -0400 David Robison wrote: I'm using CVS from the cygwin distribution on Windows XP SP2. [...] Cannot access /src/master^M/CVSROOT No such file or directory You have compiled CVS yourself, didn't you? AFAIR, 1.11.17 is not available from cygwin itself. Do you use LF or CR/LF as default? These types of errors occurs if you use cygwin if LF mode. Just try to install again with CR/LF, recompile CVS again, and your problem should be gone. I once had almost the same problem, and the above advice helped me. HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Retrieving CVS modules information
Hello Ivan, * On Wed, Sep 01, 2004 at 06:50:50AM -0700 Ivan Ivanov wrote: Do you mean that we should describe manually each module in CVSROOT/module? Yes, that's the way it is supposed to be. But we currently have more that 50 modules and more are coming? Doesn't CVS automatically tracks when a new module is imported and adds it to this file? No, CVS does not track this automatically. In fact, CVS does not talk of a module unless there is a line in CVSROOT/modules for it. Can it be set up to do so? Not that I know of. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Question about RCS files
Hello, * On Thu, Aug 26, 2004 at 12:20:14PM +0200 K. Posern wrote: But I need an exact copy of the version a CLIENT has in its /SOMEWHERE/sv dir on the SERVER (so without the rcs stuff). Have a look at C.3.5.2 of your manual: https://www.cvshome.org/docs/manual/cvs-1.11.17/cvs_18.html#SEC175 HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: [OT ?] *WIN*CVS setup questions
Hello, * On Wed, Aug 18, 2004 at 12:23:03AM -0700 New Bie 123 wrote: :pserver:[EMAIL PROTECTED]:/cvsroot/freedos [...] What are the following in the above : --- - Path - Host address (IP / DNS name) I'm not sure if I understand you correctly. The format is: :access method:login name@hostname:path So, in the above example, the following holds true: access method = pserver login name= anonymous host = cvs.sourceforge.net path = /cvsroot/freedos HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: mobile CVS
Hello Geoff, * On Tue, Jul 20, 2004 at 04:57:30PM -0400 Geoff Beier wrote: I've handled this in the past by simply using a standard ssh client to connect to the appropriate server and port, and forwarding port 22 on localhost. *grmph* I use that approach myself for all kinds of stuff with ssh, but I never thought about using this for CVS. Thank you very much for giving the blockhead (me) a whack on the back of the head. ;-) Thank you very much, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: mobile CVS
Hello Stephen, * On Tue, Jul 20, 2004 at 01:15:45PM -0700 Stephen Carville wrote: I use dfferent repository paths Internally: cvs -f -d :ext:[EMAIL PROTECTED]:/var/cvs checkput -P module Externally: cvs -f -d :ext:[EMAIL PROTECTED]:/var/cvs checkout -P module Well, these -d override whatever is written in CVS/Root, don't they? Anyway, I do not like to be forced to use a different command line parameter every time I'm accessing a repository. I find Geoff's solution much more handy. Anyway, thank you very much for your suggestion, too. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
mobile CVS
Hello, I got a question on using CVS (what else?). I'm using a laptop which accesses a remote repository via ssh. Unfortunately, depending on whether I'm inside the network which contains the repository, or outside, I have to use different ssh settings. The machine name, as well as the used port change. I cannot access the repository server with the inside settings from outside, and vice versa. Currently, I help myself by utilizing a ~/.ssh/config file, which contains two entries, namely, CVS-local and CVS-remote, with the correct settings for each case. Depending of whether I'm in or out, I rename one of both to CVS, and only use the ssh server name CVS when working with CVS. This works, but is annoying. I ask myself if there is another way to perform this? I will ask this on an ssh mailing list, too, but there might be some ideas here which do not depend on ssh, but are cvs related. Ideally, there would be a solution which would work automatically. I already thought about writing a skript which renames the appropriate entry depending on my own IP address. An option I used before using ~/.ssh/config was to change CVS/Root in every directory, but this was rather annoying. Any thoughts or experiences on this? Thanks, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Changes made to project
Hello, * On Wed, Jul 14, 2004 at 11:28:36AM -0400 Murrgon wrote: Is there a way to determine if there have been any commits or adds done to a project since a specific date or tag? I would test it with cvs diff -uN or cvs rdiff, and giving the appropriate tags or dates as parameters. HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: determine branch's ancestor?
Hello, * On Wed, Jul 07, 2004 at 08:56:50AM -0400 Jim.Hyslop wrote: Most of the work on FEATURE touches new files or files that haven't changed in months, so the branch tags RELEASE and FEATURE both point to the same versions of the files (the last HEAD revision), so i can't use that to help me determine the ancestor. A difference which makes no difference is no difference (attributed variously to philosopher Bertram Russell, philosopher William James, Ralph Waldo Emerson, and Commander Spock). Pick one, tag it as the base, and move on. What about the following case: There are thousands of files in this module, which are common to FEATURE and RELEASE, and all - except one! - or the same. How can he distinguish this case from the case where all files are the same? I think that was his question in the first place, at least, it is how I understood it. Tyler, if this is your problem, then I don't have another solution but go through the whole repository and check for the files where the tags are different. Some skripting would be required. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Filename case-sensitivity
Hello, * On Tue, Jun 29, 2004 at 10:27:43AM +0100 Brendan Knowlton wrote: Is there anything I can do in the repository itself (running on a HP/UX server) to change the name of this file directly? Or some clever Entries or repository hacking that would do the trick? Yes, rename the file into lowercase on the server. As you already mentioned, the CVS/Entries file will not show this change. You can either - manually edit the CVS/Entries file (not recommended!), or - cvs release -d your sandbox, and check out a fresh new copy of it. I assume http://lists.gnu.org/archive/html/info-cvs/2004-06/msg00343.html does not apply to you, does it? There, I only can fix this issue by changing the CVS/Entries file by myself after doing a fresh checkout. HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Windows with Dir/ and dir/
Hello, I'm replying to an old mail (and answer) from me, because I took the time to generate a skript which shows the behaviour I mentioned (and it still happens, even with 1.11.17). My first posting on this can be found on http://lists.gnu.org/archive/html/info-cvs/2004-04/msg00201.html My problem is that if I have a directory Dir/, and I rename it to the name dir/, cvs has problems with this on a Windows box. * On Mon, Apr 19, 2004 at 11:10:42AM -0400 Derek Robert Price wrote: Spiro Trikaliotis wrote: Anyway, if I do a checkout (again, with -P) from the Windows machine, I get the following files: module/Dir/1 module/Dir/2 but the files 1 and 2 have the contents of the files which are in module/dir/1 and module/dir/2! [...] You shouldn't have this problem if you upgrade your server to 1.11.15. No, this does not fix this issue. The attached bash-skript uses an empty module named testproj to show the issue. You can run it from a Linux box to generate such a module (you have to answer y two times for the release -d), and see what happens. The following tests were all done using self-compiled plain-vanilla cvs 1.11.17 on a Linux box and a Windows box (client); the server was always used remotely via ssh, and it was always located on the Linux box. If I checkout the testproj from the Linux box and want to release it again, I get: [EMAIL PROTECTED]:~/6$ cvs co testproj cvs checkout: Updating testproj U testproj/bla cvs checkout: Updating testproj/Old cvs checkout: Updating testproj/old U testproj/old/testfile [EMAIL PROTECTED]:~/6$ ls -l testproj/ insgesamt 12 -rw-r--r--1 me me 9 21. Jun 15:36 bla drwxr-xr-x2 me me4096 21. Jun 15:51 CVS drwxr-xr-x3 me me4096 21. Jun 15:51 old [EMAIL PROTECTED]:~/6$ cvs release -d testproj You have [0] altered files in this repository. Are you sure you want to release (and delete) directory `testproj': y If I do the same on a windows machine (full cygwin installation), I get: [EMAIL PROTECTED] ~/cvstest $ cvs co testproj cvs checkout: Updating testproj U testproj/bla cvs checkout: Updating testproj/Old cvs checkout: Updating testproj/old U testproj/old/testfile [EMAIL PROTECTED] ~/cvstest $ ls -l testproj/ total 1 drwxr-xr-x+ 2 me None0 Jun 21 15:50 CVS drwxr-xr-x+ 3 me None0 Jun 21 15:50 Old -rw-r--r--1 me None 10 Jun 21 15:36 bla [EMAIL PROTECTED] ~/cvstest $ cvs release -d testproj cvs server: WARNING: global `-l' option ignored. cvs update: Old/testfile is no longer in the repository You have [0] altered files in this repository. Are you sure you want to release (and delete) directory `testproj': y cvs release: no such directory: testproj Have a look at the line cvs update: Old/testfile is no longer in the repository. Furthermore, you will recognize that the Linux machine has the (correct!) old/ directory, while the Windows box named it Old/ with an uppercase O. If I edit testproj/Old/CVS/Repository and replace Old/ with old/, all works as expected. So, obviously, cvs seems to get the traces to the wrong directory. Because of this, not only release does not work, but even a commit to the files in these directories is not possible. In fact, I have to manually edit the CVS/ files to get it to work from the windows box. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ #! /bin/sh cvs co testproj cd testproj cat bla % junk... % cvs add bla mkdir Old cat Old/testfile % just some junk % cvs add Old cd Old cvs add testfile cd .. cvs commit -m - 1st version, Old/ cvs tag v1 cd Old mv testfile .. cvs rm testfile cd .. cvs rm Old cvs commit -m - 2nd version, after removing cvs tag v2 cd .. cvs release -d testproj cvs co -P testproj cd testproj/ mkdir old/ cvs add old cd old/ cat testfile % just some junk % cvs add testfile cd .. cvs commit -m - 3rd version, with old/ cvs tag v3 cd .. cvs release -d testproj ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Can one user develop concurent versions ?
Hello Adrian, * On Thu, Jun 17, 2004 at 05:16:23AM -0700 Adrian Constantin wrote: I work on my project from two computers (work and home) plus one version which I keep on the company's server where I seldom work. So I have three checked out copies of the project, using the same user. I do this myself very often. I wonder if sometimes I get out of sync with my copies so that I have concurent versions of the project, all under my username, will cvs recognize the situation ? Do you ask beforehand, or do you have any problems? I have not yet to encounter any problems with this setup, cvs works good. In fact, it does not seem to be interested in the user account at all. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Release command doesn't delete my working copy
Hello Anand, * On Tue, Jun 15, 2004 at 12:35:52AM +0200 Anand Graves wrote: Then I ran the command: cvs release -d I have never used release this way as you, but I always go to the directory above and release with cvs release -d directoryname and it works quite well. I assume you could even run cvs release -d * to release anything (but the top level directory). Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: strange cvs diff problem with pcl-cvs and viewcvs
Hello Fabian, * On Tue, Jun 15, 2004 at 08:40:32AM +0200 Fabian Braennstroem wrote: it is me again. Is this question to stupid or does nobody have an idea? I believe nobody would have an idea. For the rest: Fabian already asked on other mailing lists, without success. So, if anyone has an idea, don't think he'll ask on another list where it is more appropriate if I don't answer, but answer to help him. ;-) Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Exporting only changed files from CVS
Hello Jonathan, * On Wed, Jun 09, 2004 at 07:33:17PM -0700 Jonathan Hurshman wrote: I am relatively new to CVS, having used Source Safe and StarTeam in the past. I am using CVS to track web development projects. I would like to be able to export from CVS into a directory where a non-CVS version of the application already resides, and have CVS only update new or changed files. Is this even possible? My approach (I assume there are only text files, that is, files that can be used with patch): Whenever you make the non-CVS version, make sure you make a tag in your CVS. For example, I assume the last non-CVS-version is tagged rel_1. Now, if you want to generate a new one, run the following: $ cd CVS-version of your project $ # I assume you already have committed every change $ cvs tag rel_2 # make the tag for the next upgrade $ cd non-CVS version of your files $ cvs rdiff -uN -r rel_1 -r rel_2 cvsmodulename|patch -p1 After this, your non-CVS version should be up-to-date BEWARE! Make a dry run before doing that on the production server! Futhermore, you might want to put the rdiff into a separate file before applying patch. HTH, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Any CVS tricks for creating patch files?
Hello Larry, * On Wed, Jun 09, 2004 at 04:12:47PM -0400 Larry Jones wrote: Given that rdiff doesn't even *have* a -N option, what would you have the documentation say?!? I don't know what others think about it, but I assume that a command cvs rXXX is just a function like cvs XXX, but I do not need a working sandbox to do it. I have never thought that this could not be true. In fact, I did not even know that there is also a cvs patch. It already says that it creates a patch to bring an old release up-to-date with the new release -- creating files that didn't previously exist is obviously an essential part of that process. Obviously, yes. Perhaps we should consider splitting patch and rdiff into separate commands on the feature branch: patch would continue to have the current behavior, rdiff would provide all the usual diff options. Well, now that I know the difference between rdiff and diff, I have no problems. In fact, rdiff -u or diff -uN is exactly what I need. My only concern is that other people might be confused like I was. For me, there is no need to change the rdiff behaviour. Now, for me, it is not even essential to change to docu on rdiff, but there might be new users which might like that one. Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Any CVS tricks for creating patch files?
Hello Larry, * On Tue, Jun 08, 2004 at 05:40:14PM -0400 Larry Jones wrote: [regarding -N for diff and rdiff] It's not just useful, it's essential. The option doesn't exist because it's always assumed for rdiff! Ok, I just tested it, you are right. Well, I looked at the documentation (info cvs, the online manual) but could not find it documented anywhere that -N is always assumed with rdiff. Isn't that a documentation hole? Regards, Spiro. -- Spiro R. Trikaliotis http://www.trikaliotis.net/ http://www.viceteam.org/ ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: newbie question on initializing repository
Hello, * On Mon, Jun 07, 2004 at 05:34:02PM -0400 Larry Jones wrote: However, I now believe its not that simple, as the binary files(only) must be initialized with the -kb option. Is that correct? If so, any suggestions on the simplest/easiest to go about this? If you can distinguish the binary files from the text files by their names (e.g., *.gif files are always binary), you can use the CVS wrappers facility to automatically handle the files correctly To add to this: If the text files are Unix-style (LF only), I also had success in importing everything as binary, and changing the text files to text afterwards with cvs admin. If the text files are Windows style, doing a dos2unix on them, and then treating it as in the above case, works, too. Best regards, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Any CVS tricks for creating patch files?
Hello, * On Mon, Jun 07, 2004 at 05:46:20PM -0400 [EMAIL PROTECTED] wrote: To create a patch, I could do this: % cvs co -d buggy -r Release_0_1 MyProj % cvs co -d fixed -r Release_0_1_Bugfixes MyProj % LC_ALL=C TZ=UTC0 diff -Naur buggy fixed Release_0_1_Patch ...but I wonder if there are some CVS tricks to streamline the process further. $ cvs co -r Release_0_1_Bugfixes MyProj $ cd MyProf $ cvs diff -uN -r Release_0_1 ../Release_0_1_Patch I think (but I have not tested it) that $ cvs rdiff -uN -r Release_0_1 -r Releae_0_1_Bugfixes Release_0_1_Patch should work, too. Regards, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Info-cvs Digest, Vol 19, Issue 17
Hello Bret, * On Tue, Jun 08, 2004 at 01:50:36PM -0400 Bret A Cooper wrote: Thanks for your reponse. It prompts a couple more questions. When you went back and ran the cvs admin command, did you have to run it once for every file you wanted to change? Yes, this has to be run for every file to be changed. Anyway, you can use wildcards, so a cvs admin -kk *.c *.cc *.h makefile and so on works very good. Even if you forget to change a file to text mode, it does not do much harm, as you can even do this afterwards. Also, would it work equally well to put all the files in as regular files and the run cvs admin later to change the binary files? It depends upon your binary files. If 1. your client runs Windows (CR/LF instead of LF), or 2. you binary files contain any of the RCS keywords ($Id$, $Log$, etc), 3. not sure if there are other conditions which could do havoc to your binary files, chances are very good your binary files will get corrupted, and you will not be able to recover them from CVS. Thus, checking everything in as binary is a more defensive solution than the opposite. Regards, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: Any CVS tricks for creating patch files?
Hello Jim, * On Tue, Jun 08, 2004 at 09:03:27AM -0400 Jim.Hyslop wrote: Spiro Trikaliotis wrote: $ cvs rdiff -uN -r Release_0_1 -r Releae_0_1_Bugfixes What does the -N flag do? It isn't documented. forget it, it is non-existant with rdiff, while it exists with diff. So it is my fault. with diff (not cvs diff), the option -N tells diff to treat files which are non-existant in one directory as if it was there, but empty. This way, the diff is complete. If you do not specify -N, diff only tells you that some file was missing, but you don't see the contents of it. As I said, this option is missing with cvs rdiff, while it exists with cvs diff (cvs 1.11.16). Is this difference by intention, or is it accidentially? I would find it very usefull for rdiff. The following simple skript demonstrates this: -- #! /bin/bash export CVSROOT=~/test.cvsroot/ cvs init mkdir ~/test.cvsroot/test/ cvs co test cd test cat f1 % This is file 1 % cvs add f1 cvs commit -m First version cvs tag v1 cat f2 % This is file 2 % cvs add f2 cvs commit -m 2nd version cvs tag v2 rm f1 cvs rm f1 cvs commit -m 3rd version cvs tag v3 cvs diff -u -r v2 ../v2_v3.diff cvs diff -u -r v1 ../v1_v3.diff cvs diff -uN -r v2 ../v2_v3_N.diff cvs diff -uN -r v1 ../v1_v3_N.diff -- Regards, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS corrupts binary files ...
Hello, * On Sat, Jun 05, 2004 at 08:38:15PM -0700 Gianni Mariani wrote: * Peter Connolly wrote: Too dificult to set up, I think Shouldn't cvs have a list of binary file types preinstalled in the cvswrappers ? I agree, it should. I second that ! I did 3 years ago. http://www.mail-archive.com/[EMAIL PROTECTED]/msg09098.html I tend to disagree: It should not! Which extensions are binary files? Is .au a binary file? I know .au as (Sun?) audio file, but I've also seen a project where .au had the meaning of additional user or something like that, and was a text file. The same could be true for other extensions. What about .doc? .doc is not necessarily a Word file, especially not on old project. So, my conclusion is: Only the user is aware what type each file is. Checking in a text file with -kb (from a Windows machine) is something which many administrators do not like, either. If you have so much fear about binary files, why don't you put * -kb into your cvswrappers, and declare any text file explicitly? This way, you cannot miss the binary files. Best regards, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: back up of cvs repository
Hello, * On Fri, Jun 04, 2004 at 07:31:19AM +0100 Ramanuj Singh wrote: How to take the backup of cvs repository. The back up has to be taken up on tapes. 1. Make sure no-one accesses your repository for the whole time! 2. Make a tarball of your cvsroot: tar czf tarname cvsrepodirectory 3. Re-allow accesses to your repository 4. Put the tarball on tape. If the repository is quite large, you might want to try to store just a diff over a previous one on a daily basis, and store the whole repository only once a week or so. I have not tried that, but IMHO, this should work. Best regards, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://lists.gnu.org/mailman/listinfo/info-cvs
Re: CVS complete newbee
Hello, * On Wed, May 26, 2004 at 02:23:07AM -0700 Rolf Bergius wrote: :pserver:[EMAIL PROTECTED]:local:C:\Programme\GNU\repository Is this repository local (i.e., is 192.168.40.4 your own machine?) or remote? If it is local (and you're using cygwin), try: :local:/cygdrive/c/Programme/GNU/repository (warning: This string is case sensitive, so check that Programme, GNU and repository are written exactly as on your system!). If it is really remote (with pserver), I don't know how this line has to be since I don't use pserver. ;-) Best regards, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/info-cvs
Re: binary files bad idea? why?
Hello, * On Thu, May 20, 2004 at 09:50:13AM -0400 Jim.Hyslop wrote: Spiro Trikaliotis wrote: This is no problem from my experience if the initial check-in was done from a Unix (LF-) based system, but it is a problem if it was done from a DOS (CR/LF-) based system. There is also a remote possibility that the binary file might _happen_ to contain what CVS thinks is a keyword, such as $Id$. Chances are pretty slim, but it _could_ happen. Well, since CVS handles these keywords on checkout, not on commit, this should not be a problem: Just change the file to binary (cvs admin -kb) and do an update (cvs update), and you have the original file again. Best regards, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/info-cvs
Re: Checkout from a Linux-based repository with DOS line-endings
Hello Marko, * On Tue, May 18, 2004 at 01:34:00PM +0200 marko wrote: Well, but, perhaps I misunderstand you, I do cross that border! Yes, you do, and I'm aware of this. ;-) And that's the trouble. My repository (created and maintained on linux' cvs) uses LF, but I'd like to check the source out in CR+LF, which obviously doesn't work so easily... Need to try what Doug recommended. Let WinCVS work on a copy of the repository directly and hope it would take care of the LF-CR+LF conversion. Why not use cygwin? Don't you have access to the repository through cygwin? I use it myself, and the only problems I'm having with this is when there are Directories in the repository which differ only in case, that it, something like foo/ and Foo/. Another option I see, which requires a little bit of work by hand: 1. Checkout a new working copy on Linux, which you take home. 2. Make a copy of that working copy, in a new directory, so you have two versions. I'll name the two directories cvs.comp/ and cvs.work/ in the sequel. 3. Make sure you don't touch cvs.comp/ in any way. Do all your work in cvs.work/ ! 4. Editing files in cvs.work/ should work with most Windows editors, as most ones can handle LF as well as CR/LF, but they store only in CR/LF format. This does not apply to Notepad (Editor) in NT/2000/XP; if you want to use that, make sure you convert your files to CR/LF before (for example, with the help of WordPad, or unix2dos from cygwin). 5. If you've done all your work, there comes the tricky part: a. make a diff through your two directories: $ diff -urN cvs.comp/ cvs.work/ diff-file (this has to be a Unix-style diff, for example, from Cygwin. Do not use the diff from Windows, as that tools is totally different) b. Now, convert that diff-file to back to Unix format: $ dos2unix diff-file (if dos2unix is installed on your Windows box, for example, from Cygwin. If not, perform this step back at work, on your Linux machine where this is hopefully installed, just before applying the patch command from step 6) c. carry the diff-file back to work. I would even suggest you take cvs.comp/ and cvs.work/ with you, in case something goes wrong and you have to apply some changes manually (something which was never necessary for me). 6. Back at work (Linux), apply that diff-file to your checked-out working copy (which you created at 1): $ patch -p1 diff-file 7. Now, your working copy at work is synchronized with your working copy at home, and you can do what you are used to do. One note: This does not work if you have binary files in the repository, as diff does not handle binary files. Anyway, since it's not a good idea to store binaries under cvs control in most cases, this should not be a problem. If it is, just make copies of them and apply them manually in step 6. I've done something very similar before, so I know it works (if I did not forget something important in my explanation). HTH, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/info-cvs
Re: Checkout from a Linux-based repository with DOS line-endings
Hello Marko, * On Wed, May 19, 2004 at 10:03:16AM +0200 marko wrote: What concerns cygwin: that might be a good try! I have to test that as well. But what troubles me is the mangling of CaSeS in file names. Windows behaves here kind sort of unpredictable... Most Windows editors, at least on Windows NT, 2000 or XP with NTFS, preserve the case, so that is not the problem. I'm using this setup without problems - despite the fact the long time before (!), there was a case-only renaming in the repository which I have to handle by hand by editing the files in CVS/. Regards, Spiro. -- Spiro R. Trikaliotis I'm subscribed to the mailing lists I'm posting, http://www.trikaliotis.net/ so please refrain from Cc:ing me. Thank you. ___ Info-cvs mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/info-cvs