Обсуждение: pgsql/src/interfaces/odbc/windev README.txt bi ...
CVSROOT: /cvsroot
Module name: pgsql
Changes by: inoue@postgresql.org 02/01/10 21:50:01
Added files:
src/interfaces/odbc/windev: README.txt bind.c bind.h
columninfo.c columninfo.h
connection.c connection.h convert.c
convert.h dlg_specific.c
dlg_specific.h drvconn.c environ.c
environ.h execute.c info.c iodbc.h
license.txt lobj.c lobj.h md5.c
md5.h misc.c misc.h multibyte.c
multibyte.h notice.txt odbcapi.c
odbcapi30.c options.c parse.c
pgapifunc.h pgtypes.c pgtypes.h
psqlodbc.c psqlodbc.h psqlodbc.rc
psqlodbc.reg psqlodbc_win32.def
qresult.c qresult.h resource.h
results.c setup.c setup.rul socket.c
socket.h statement.c statement.h
tuple.c tuple.h tuplelist.c
tuplelist.h win32.mak win_md5.c
Log message:
Add a directory to save the changes until 7.3-tree is branched.
inoue@postgresql.org writes: > Log message: > Add a directory to save the changes until 7.3-tree is branched. This seems quite silly. If the code doesn't belong to the 7.2 release it shouldn't be in the 7.2 tree. Why don't you just make a branch and merge it after 7.3 is branched? -- Peter Eisentraut peter_e@gmx.net
inoue@postgresql.org writes:
> Add a directory to save the changes until 7.3-tree is branched.
Why exactly has this been added on the main (7.2) branch?
regards, tom lane
Peter Eisentraut wrote: > inoue@postgresql.org writes: > > > Log message: > > Add a directory to save the changes until 7.3-tree is branched. > > This seems quite silly. If the code doesn't belong to the 7.2 release it > shouldn't be in the 7.2 tree. Why don't you just make a branch and merge > it after 7.3 is branched? I have to say I agree with Peter. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026
Peter Eisentraut wrote: > > inoue@postgresql.org writes: > > > Log message: > > Add a directory to save the changes until 7.3-tree is branched. > > This seems quite silly. If the code doesn't belong to the 7.2 release it > shouldn't be in the 7.2 tree. Why don't you just make a branch and merge > it after 7.3 is branched? Yes it's similar. However is it allowed to make such a branch in the first place ? regards, Hiroshi Inoue
Tom Lane wrote: > > inoue@postgresql.org writes: > > Add a directory to save the changes until 7.3-tree is branched. > > Why exactly has this been added on the main (7.2) branch? When could we release 7.2 without breaking off our work ? I've stopped the work about psqlodbc for about 2 months. However beta has been quite long and I'm quite disugusted with minding the release cycle. regards, Hiroshi Inoue
>> Why exactly has this been added on the main (7.2) branch?
> When could we release 7.2 without breaking off
> our work ?
I'm pretty annoyed about the slow pace of the 7.2 cycle, also.
But I don't think it'll help any to start dumping stuff into CVS
that is not directly related to pushing 7.2 out the door. We need
to stay focused on getting this release out.
regards, tom lane
Hiroshi Inoue <Inoue@tpf.co.jp> writes:
> Peter Eisentraut wrote:
>> Why don't you just make a branch and merge
>> it after 7.3 is branched?
> Yes it's similar. However is it allowed to make such a
> branch in the first place ?
Sure, CVS supports that. We have not done it in the past in the
Postgres project, but there's not a reason why you couldn't do it
if you wanted.
However: the only reason I can see for putting stuff into CVS is for
cooperative development. Is there anyone else who will be working with
you on the 7.3 ODBC code before we make the "official" branch for 7.3
development? If not, you aren't buying anything by putting this stuff
in CVS now, rather than keeping it in your own private files until 7.3
development starts. All you're doing is creating confusion among other
developers. (And bloating the official 7.2 tarball, when it comes out)
regards, tom lane
Tom Lane wrote: > > Hiroshi Inoue <Inoue@tpf.co.jp> writes: > > Peter Eisentraut wrote: > >> Why don't you just make a branch and merge > >> it after 7.3 is branched? > > > Yes it's similar. However is it allowed to make such a > > branch in the first place ? > > Sure, CVS supports that. What I meant was if PostgreSQL team allows it as the policy ? > However: the only reason I can see for putting stuff into CVS is for > cooperative development. The only but very significant reason I can see about psqlodbc CVS is that people can see (nearly) up to date source at the official place. People can compile it by themselves and try it. There aren't so many reasons that the released version is more stable than that of cvs as for psqlodbc. The longer(unfortunately beta has always long) I keep my changes privately, the less the cvs stuff has its meaning. regards, Hiroshi Inoue
Hiroshi Inoue <Inoue@tpf.co.jp> writes:
>> Sure, CVS supports that.
> What I meant was if PostgreSQL team allows it as the
> policy ?
Personally, I'd prefer to see development branches off the mainline
than dumping stuff into mainline that doesn't really belong there.
I don't think there's been any "official policy" issued either way;
but if core wants to vote on it, that's how I'd vote.
> The only but very significant reason I can see about psqlodbc
> CVS is that people can see (nearly) up to date source at the
> official place. People can compile it by themselves and try it.
A fair point. Other projects than this would likely already have
a CVS branch going for the-next-version. Historically we've not
done that, so as to avoid having to double-patch bug fixes into
two separate branches. But this release cycle is going so darn
slow that I'm half inclined to agree that making a 7.3 branch now
would be better.
[ Or, we could get off our collective rears and get 7.2 out the
door. Hiroshi is not at fault, but some of us on this side of
the Pacific are ... ]
regards, tom lane
> > The only but very significant reason I can see about psqlodbc > > CVS is that people can see (nearly) up to date source at the > > official place. People can compile it by themselves and try it. > > A fair point. Other projects than this would likely already have > a CVS branch going for the-next-version. Historically we've not > done that, so as to avoid having to double-patch bug fixes into > two separate branches. But this release cycle is going so darn > slow that I'm half inclined to agree that making a 7.3 branch now > would be better. > > [ Or, we could get off our collective rears and get 7.2 out the > door. Hiroshi is not at fault, but some of us on this side of > the Pacific are ... ] The only problem with a 7.3 branch now is double-patching, but I am willing to do it for my stuff and the stuff I apply. Our post-final .0 release has very few patches these days so branching is not as big a deal, _and_ some of us are fulltime so can do the double-patching. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026