From lechner@cs.uml.edu Sun Apr 10 04:31:26 2005 Received: from saturn.cs.uml.edu (cs.uml.edu [129.63.8.2]) by earth.cs.uml.edu (8.11.6/8.11.6) with ESMTP id j3A8VOw27707; Sun, 10 Apr 2005 04:31:24 -0400 Received: from saturn.cs.uml.edu (localhost [127.0.0.1]) by saturn.cs.uml.edu (8.12.9/8.12.9) with ESMTP id j3A8VOYn382970; Sun, 10 Apr 2005 04:31:24 -0400 (EDT) Received: (from lechner@localhost) by saturn.cs.uml.edu (8.12.9/8.12.9/Submit) id j3A8VNts380846; Sun, 10 Apr 2005 04:31:23 -0400 (EDT) From: Bob Lechner Message-Id: <200504100831.j3A8VNts380846@saturn.cs.uml.edu> Subject: mmulaka lmalladi sboddire checkin attempts FYInfo To: mmulaka@cs.uml.edu (Mallika Mulakaluri) Date: Sun, 10 Apr 2005 04:31:23 -0400 (EDT) Cc: lechner@cs.uml.edu (Bob Lechner), sboddire@cs.uml.edu (Swathi Boddireddy), smalladi@envoyww.com, lmalladi@cs.uml.edu In-Reply-To: from "Mallika Mulakaluri" at Apr 09, 2005 11:00:07 PM X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Status: RO Team mmulaka lmalladi sboddire are having trouble (mmulaka with cvs co, others with cvs ci). To lmalladi, sboddire: (and all of 05s523): I am sorry to have confused you all by assuming -A was valid for checkin.) Removing this sticky non-branch tag requires cvs update -A then (resolve any conflicts and) cvs ci. Unfortunately sometimes the retrieval may come from a totally diferent branch of the rev. tree. During this learning curve, always BACK UP the files you already changed FIRST before checkout from a different branch. Always run cvs update before cvs ci; this is critical after prior revisions are checked in to remerge other concurrent updates. Unfortunately I don't know if removing the old sticky non-branch tag wil retain its 6-digit rev.# and increment it on checkin. So we probably have to learn and use the -j option[s]. I notice there is a new tag genv13sb: 1.4.4.2.2.1 from sboddire that is on the same rev. as my genv12-merge050402: 1.4.4.2.2.1. That means to me that her updates were not (or could not be) checked in - this is merely a retag of my non-branch with a new nonbranch or snapshot tag. (Note: both tages have 6 digits; this means they are points along branch 1.4.4.2.2 (the same 5 most signif. digits without the last rev# digit). Branch 1.4.4.2.2 diverges from rev# 1.4.4.2 which is along prior branch #1.4.4. It is not a branch off the main trunk.) To Mallika: Mallika had a different starting point than partners sboddire and lmalladi, from checking out the main trunk instead of using '-r genv12-merge050316'. (I infer this from her missing CVS/Tag file and I assume the others have CVS/Tag containing Ngenv12-merge050316) It is imperative to get her changes back on the same configuration (tagged genv12-merge050316). I tried to use cvs status in her path, but CVS/Entries is not readable (chmod 755 was used without the -R :-( : ---------- saturn.cs.uml.edu(126)> pwd /nfs/earth/misc/proj3/case/04f522/mmulaka/chgen/src saturn.cs.uml.edu(127)> cvs status gen_pr_init.c cvs status: cannot open CVS/Entries for reading: Permission denied cvs [status aborted]: cannot open CVS/Entries.Log: Permission denied saturn.cs.uml.edu(128)> ----------- Mallika's partners can check in their updates using cvs update -A filename then cvs ci filename (not cvs ci -A). Since their start point was my non-branch tag genv12-merge050316 with a rev# of 6 digits, I should have created a 5-digit branch tag predecessor first. Then, Mallika could merge updates along this branch tag into her working copy with cvs co -j pr_load.c etc.. (See below) This will retrieve the latest leaf node on that branch, and merge it into her working copy. Retrieving the fixed snapshot that I non-branch-tagged (cvs co -r genv12-merge050316 chgen) will NOT merge them. Unfortunately there is no rev# 1.4.4.2.2. There are just these from cvs log gen_pr_load.c: Working file: gen_pr_load.c head: 1.4 branch: symbolic names: genv13sb: 1.4.4.2.2.1 genv12-merge050402: 1.4.4.2.2.1 tag01f522: 1.4.4.2.0.2 genv11: 1.4.4.2 genlog_branch: 1.4.4.1.0.2 ... (It looks like genlog_branch was merged back into tag genv11 but tag 1.4.4.2.0 was never created before tag01f522=1.4.4.2.0.2, just as branch tag 1.4.4.2.2 was not created before rev# 1.4.4.2.2.1. I.e., I didn't pre-tag the branch point before creating the branch.) It may require two steps to merge 6-digit rev#s back onto the genv11 branch tip 1.4.4.2, then merge that branch 1.4.4 back onto the trunk. SBoddireddy has her own tag genv13sb: 1.4.4.2.2.1 with the same rev# as mine: genv12-merge050402: 1.4.4.2.2.1 (from cvs log gen_pr_load.c). It looks like she hasns't checked updates back into CVS either. [Fogel p. 247: cvs checks out modules (directories) not files.] [Fogel: p. 248: Commits to a branch always go on the tips; you cannot commit to a specific rev# on a branch.] Mallika could use cvs update -j genv13sb gen_pr_load.c, to merge sboddire's updates into her working copy. Unfortunately I can't guarantee that this will work since Mallika's version 1.5 is on the main trunk, two forks back from sboddire's. There is another way to merge her backed-up copy changes with changes from sboddire. Because they updated disjoint line# ranges, mmulaka can remove the checked-out lines that sboddire had not changed and paste in the changes from the matching lines in her backup copy. (Any revision history would be wrong. Fortunately chgen doesn't use /* $Log$ */ to include it.) (Be careful - prior checkins probably changed the line numbers). Fogel: Then do chgen update -A (-A works on update and co but not ci). and cvs ci file. But I do not know hether the old sticky non-branch tag that was removed still results in checking in a new 6-digit rev. no. 1.4.4.2.2.2 as successor to sboddire's rev. (cvs log filename will show all tags and the revision# tree structure.). Incidentally, gen_pr_dump.c has this history of revsion#s: genv12-merge050402: 1.3.4.1.4.1 tag01f522: 1.3.4.1.0.4 genv11: 1.3.4.1 genlog_branch: 1.3.4.1.0.2 R Lechner ----------------------------------------------- > From mmulaka@cs.uml.edu Sat Apr 9 23:00:15 2005 > From: "Mallika Mulakaluri" > To: "Bob Lechner" > Subject: RE: mmulaka/hjha checkin of gen_pr_{init.c,add.c,dump.c} > Date: Sat, 9 Apr 2005 23:00:07 -0400 > > Sorry, for the inconvenience caused. I remember doing cvs ci gen*.c ONLY > AFTER I failed trying all the possible options on ci and all other options > specified in your previous mails. Very sorry again :( > > - Mallika > > -----Original Message----- > From: Bob Lechner [mailto:lechner@cs.uml.edu] > Sent: Saturday, April 09, 2005 10:48 PM > To: harshjha82@yahoo.com; hjha@cs.uml.edu; mmulaka@cs.uml.edu; > Mallika.Mulakaluri@sun.com > Cc: bswathi2@yahoo.co.in; Swathi_Boddireddy@student.uml.edu; > sboddire@cs.uml.edu; Scot Frye; mmulaka@cs.uml.edu; > Mallika.Mulakaluri@sun.com; Norm Gross; smalladi@envoyww.com; > lmalladi@cs.uml.edu; harshjha82@yahoo.com; hjha@cs.uml.edu; > kevincy@netscape.net; kyang@cs.uml.edu; Bob Lechner > Subject: mmulaka/hjha checkin of gen_pr_{init.c,add.c,dump.c} > > > HJha's checkin of -r_dump.c failed with sticky non-branch tag. > > MMulaka's checkin of gen_pr-init.c and gen_pr_add.c > has different problems, reviewed later below. > > 1. hjha's commit attempt of pr_dump.c: > ----------------------------------- > > From harshjha82@yahoo.com Sat Apr 9 21:58:44 2005 > > From: harsh jha > > Subject: CVS commit problem. > > To: lechner@cs.uml.edu > > > > Hi professor, > > I've done my part of gen_pr_dump.c file and the > > results are stored in genv12_13/test_run.doc under > > $CASE/04f522/hjha. > > I'm having problems in commiting my changes to CVS how > > do i go about it ? > > > > ERRLOG : > > > > [hjha@mercury src]$ cvs ci -r genv12-merge050402 > > gen_pr_dump.c > > cvs commit: sticky tag `genv12-merge050402' for file > > `gen_pr_dump.c' is not a branch > > cvs [commit aborted]: correct above errors first! > > > > - Harsh > > > ------------------------------ > Did you do cvs update pr_dump.c first? and make sure the report > is U or M not C? > > Do not use > cvs ci -r genv12-merge050402 gen_pr_dump.c > This tag cannot be re-used for your new version because > that would change the tag's snapshot of chgen/src > (its tagged configuration). > > Instead, use > cvs ci -A gen_pr_dump.c > > (and be sure to add a rev. comment (use option -m "..", > or edit into vi which should open up if you do not include -m ...). > > CVS/Tag=Ngenv12-merge050402 is not a branch tag (my mistake). > It tagged the same revision 1.1.2.1 as tag genv11 > from cvs log below. With -A, I expect pr_dump.c will be > committed as rev. 1.1.2.2, successor to rev. 1.1.2.1. > > > ------------------------------------------ > saturn.cs.uml.edu(104)> cvs log pr_dump.c > > RCS file: /usr/proj3/case/gen/base/Master/chgen/src/Attic/pr_dump.c,v > Working file: pr_dump.c > head: 1.1 > branch: > locks: strict > access list: > symbolic names: > genv12-merge050402: 1.1.2.1 > tag01f522: 1.1.2.1.0.2 > genv11: 1.1.2.1 > genlog-branch: 1.1.0.2 > keyword substitution: kv > total revisions: 2; selected revisions: 2 > description: > ---------------------------- > revision 1.1 > date: 1999/05/07 01:56:43; author: jkarner; state: dead; > branches: 1.1.2; > file pr_dump.c was initially added on branch genlog-branch. > ---------------------------- > revision 1.1.2.1 > date: 1999/05/07 01:56:44; author: jkarner; state: Exp; lines: +169 -0 > 99s523 - New sources for metaschema bootstrapping - Karner-Spinney > ============================================================================ > = > saturn.cs.uml.edu(105)> > --------------------------------------------------- > > > ========================================================= > > 2. Mallika's Checkin of gen_pr_init.c and gen_pr_add.c: > -------------------------------------------------- > > mmulaka/chgen has no CVS/Tag files so she apparently checked out > chgen without -r genv12-merge050402. This made her > new version become a main trunk entry 1.5. > I believe but haven't checked (cvs diff -r...) > her update was based on v. 1.4 rather than > the 6-digit rev. 1.4.4.2.2.2 that would have > succeeded the rev. with tag genv12-merge050402 below: > > See cvs log entries below; Mallika's rev. 1.5 > shows no tag, no rev. comment, and a future date! > (violating the CVS Best Practices guide > we ran thru last Tuesday.:-( > ---------------------------------------------- > > saturn.cs.uml.edu(99)> pwd > /nfs/earth/misc/proj3/case/04f522/hjha/chgen/src > saturn.cs.uml.edu(100)> cvs status gen_pr_init.c > =================================================================== > File: gen_pr_init.c Status: Up-to-date > > Working revision: 1.4.4.2.2.1 Sat Apr 2 08:50:44 2005 > Repository revision: 1.4.4.2.2.1 > /usr/proj3/case/gen/base/Master/chgen/src/gen_pr_init.c,v > Sticky Tag: genv12-merge050402 (revision: 1.4.4.2.2.1) > Sticky Date: (none) > Sticky Options: (none) > > -------------------------------------------------------- > > saturn.cs.uml.edu(101)> cvs log gen_pr_init.c > > RCS file: /usr/proj3/case/gen/base/Master/chgen/src/gen_pr_init.c,v > Working file: gen_pr_init.c > head: 1.5 > branch: > locks: strict > access list: > symbolic names: > genv12-merge050402: 1.4.4.2.2.1 > tag01f522: 1.4.4.2.0.2 > genv11: 1.4.4.2 > genlog_branch: 1.4.4.1.0.2 > .... > ---------------------------- > revision 1.5 > date: 2005/04/10 00:46:20; author: mmulaka; state: Exp; lines: +159 -80 > *** empty log message *** > ---------------------------- > > I plan to remerge other chgen/src files on the main trunk > to make the main branch whole again. > > R Lechner > > >