Horton

Re: Report: removing the inconsistencies in our CVS->git conversion

Tom Lane <tgl@sss.pgh.pa.us>

View thread
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Magnus Hagander <magnus@hagander.net>
Cc: Robert Haas <robertmhaas@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2010-09-18T01:29:15Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Applied patch to jdbc from Kim Ho at RedHat, fixing improper handling of empty queries under the V3 protocol

  2. Applied patch for jdbc from Kim Ho at RedHat to more correctly deal with boolean values

  3. Applied patch from Oliver Jowett to improve a buffer sizing.

  4. Code cleanup inspired by recent resname bug report (doesn't fix the bug

  5. Pass shared memory id and socket descriptor number on command line for

  6. Second try at avoiding conflicts with system isblank().

  7. Applied Kris Jurkas patch to fix rollback and SQLException

  8. applied patch from Oliver Jowett

  9. Make the world at least somewhat safe for zero-column tables, and

  10. Add mention of -F not supported by pg_dumpall. Code already disallows it.

  11. Two versions of QueryExecutor, currently only version 2 works 100%

  12. modifications to the way the protocol is handled to be consistent with

  13. Part of Anders Bengtsson's patch to clean up Connection.java

  14. Applied Anders patch to move the startup code out of Connection into StartupPacket

  15. >Added ServerEncoding

  16. > I am backing out this patch. Please resubmit with this corrected. Thanks.

  17. Fixes bug where query hangs if there is an error in the query, went back to

  18. Backpatch array I/O code and documentation fixes, also array slice

  19. fixed QueryExecuter to deal with multiple errors

  20. Repair bad calculation in timetz_izone() which gave erroneous results.

  21. backed out changes for cancel, no need to look for two 'Z' responses

  22. Patch by Nicolas Verger to correctly propogate SQLWarning to the Statement and ResultSet

  23. fixed cancel query bug introduced by patch

  24. Autoconf for jdbc debug changes.

  25. Updated the list of encodings supported to match what the backend now supports

  26. Added some additional comments in the code

  27. Update TODO list.

  28. Remove ConnectionHook.java. No longer used, bad code.

  29. The attached patch removes some old and dead code (and some related

  30. First batch of the tools merged in...

  31. This patch fixes an arrayindexoutofbounds exception that was just

  32. Add ./include/config.h.win32 to RELEASE update list.

I wrote:
> Oh, mystery explained upon comparing the cvs2git.options files.
> I was using Max's file which had this in it:
>     ExcludeRegexpStrategyRule(r'unlabeled-.*'),
> I think I'll rerun without that just to convince myself of what it is
> we're dropping.  But right now it seems that everything is pretty sane.

OK, I found out what we're dropping.  Those "unlabeled" branches each
have a manufactured creation commit that just deletes a lot of files,
plus a single real commit, which all look like this:

  
+ commit 6c23cb5f371f273cade66408f3a888c2f94af293	refs/heads/unlabeled-1.51.2
+ Author: Tom Lane <tgl@sss.pgh.pa.us>
+ Date:   Sat Sep 28 20:00:29 2002 +0000
+ 
+     Make the world at least somewhat safe for zero-column tables, and
+     remove the special case in ALTER DROP COLUMN to prohibit dropping a
+     table's last column.
+ 
+ M	src/test/regress/expected/alter_table.out
+ 
+ commit d554d56827e4337835b0792c2568e8e9fee2d950	refs/heads/unlabeled-1.87.2
+ Author: Tom Lane <tgl@sss.pgh.pa.us>
+ Date:   Sat Sep 28 20:00:28 2002 +0000
+ 
+     Make the world at least somewhat safe for zero-column tables, and
+     remove the special case in ALTER DROP COLUMN to prohibit dropping a
+     table's last column.
+ 
+ M	src/backend/executor/nodeAgg.c
+ 
+ commit 7724eba3a121e5c84da3467fe6f0b176bf3f3d4b	refs/heads/unlabeled-1.59.2
+ Author: Tom Lane <tgl@sss.pgh.pa.us>
+ Date:   Sat Sep 28 20:00:27 2002 +0000
+ 
+     Make the world at least somewhat safe for zero-column tables, and
+     remove the special case in ALTER DROP COLUMN to prohibit dropping a
+     table's last column.
+ 
+ M	src/backend/executor/execTuples.c
+ 
+ commit 6924390f0cfa2aac1308719e51cb4cc2b24e2bc7	refs/heads/unlabeled-1.44.2
+ Author: Tom Lane <tgl@sss.pgh.pa.us>
+ Date:   Sat Sep 28 20:00:26 2002 +0000
+ 
+     Make the world at least somewhat safe for zero-column tables, and
+     remove the special case in ALTER DROP COLUMN to prohibit dropping a
+     table's last column.
+ 
+ M	src/backend/commands/tablecmds.c
+ 
+ commit 83029b0525f3324c30d35fcc9c77d2c301bdf7cf	refs/heads/unlabeled-1.90.2
+ Author: Tom Lane <tgl@sss.pgh.pa.us>
+ Date:   Sat Sep 28 20:00:25 2002 +0000
+ 
+     Make the world at least somewhat safe for zero-column tables, and
+     remove the special case in ALTER DROP COLUMN to prohibit dropping a
+     table's last column.
+ 
+ M	src/backend/access/common/tupdesc.c
+ M	src/backend/parser/parse_target.c
+ 
  commit 6d0d15c451739396851d3f93f81c63a47535bf1e	refs/tags/REL7_4_BETA1
  Author: Tom Lane <tgl@sss.pgh.pa.us>
  Date:   Sat Sep 28 20:00:19 2002 +0000


If you go and look at the CVS history of these files, you'll see that
indeed they all have branches sprouted on that date that are
unreferenced.

This commit occurred during that interval between where we'd initially
branched 7.3 and where we moved it up to head.  I think what happened
was that I tried to back-patch a fix into what was then the 7.3 branch,
and when Marc moved the branch point, these commits all ended up
orphans.  So indeed there's no harm discarding them.  (I always
suspected that what Marc had done wasn't entirely kosher, and yup,
now the chickens are coming home to roost.)

			regards, tom lane