Обсуждение: commentds on redhats new database
hello: i thought id find out what people think about this: http://www.zdnet.com/zdnn/stories/news/0,4586,2778706,00.html?chkpt=zdnn_rt_latest which basically points out that redhat will be making their own database based off of postgres7.1 my main concern is that this could cause some sort of splintering of the code base could this potentially happen? what other concerns could there be?
On Wed, Jun 20, 2001 at 03:28:34PM -0700, clayton cottingham wrote: > hello: > > i thought id find out what people think about this: > http://www.zdnet.com/zdnn/stories/news/0,4586,2778706,00.html?chkpt=zdnn_rt_latest > > which basically points out that redhat will be making their own database > based off of postgres7.1 Quoting: The Red Hat Database, as it is likely to be called, is basedon PostgreSQL 7.1. Included in the package will be Red HatInstaller,which will aid users in quickly installing thedatabase, robust transaction support and advanced lockingcapabilities.It will comply with SQL92, ODBC and JDBC APIs. > > my main concern is that this could cause some sort of splintering of the > code base I'll be interesting to see how their offering differs from GreatBridge. And if their going to claim SQL'92 compliance, have they done all the SCHEMA support work? ;-) > > could this potentially happen? Sure, we're BSD, so they're under no obligation to share any code changes with us. Haven't noticed many new @redhat.com email address over on HACKERS, though, so I wonder if they've done much work on the core. > > what other concerns could there be? > Getting bug reports from code we can't look at. Ross
"Ross J. Reedstrom" <reedstrm@rice.edu> writes: >> could [a fork] potentially happen? > Sure, we're BSD, so they're under no obligation to share any code changes > with us. I really doubt that RedHat is silly enough to want to fork the code. They haven't forked Linux, gcc, gdb, etc, but have been upstanding contributors to all those projects; why would they do it differently with Postgres? > Haven't noticed many new @redhat.com email address over on HACKERS, > though, so I wonder if they've done much work on the core. Some folk at RedHat's Toronto offices are gearing up to contribute work; I was up there to talk with them just a couple weeks ago. They've not been real visible yet though ... (hey Patrick et al: if you're reading this, it's past time to introduce yourselves). regards, tom lane
On Thu, Jun 21, 2001 at 01:32:10AM -0400, Tom Lane wrote: > "Ross J. Reedstrom" <reedstrm@rice.edu> writes: > >> could [a fork] potentially happen? > > > Sure, we're BSD, so they're under no obligation to share any code changes > > with us. > > I really doubt that RedHat is silly enough to want to fork the code. > They haven't forked Linux, gcc, gdb, etc, but have been upstanding > contributors to all those projects; why would they do it differently > with Postgres? Hey, he did say _potentially_. I agree that it's not at all _likely_. > > > Haven't noticed many new @redhat.com email address over on HACKERS, > > though, so I wonder if they've done much work on the core. > > Some folk at RedHat's Toronto offices are gearing up to contribute work; > I was up there to talk with them just a couple weeks ago. They've not > been real visible yet though ... (hey Patrick et al: if you're reading > this, it's past time to introduce yourselves). Sounds cool. The more the merrier, I say. Ross
Hi, Does plpgsql in 7.1.2 has some change in implementation?? Since I have a plplsql function which works in pg-7.0, however, in pg-7.1.2, sometimes doesn't work, I found sometimes, I generate more then 800 pg_sorttemp???.??? file each has ~430k. which causes our disk full so fast(when we have mutilpel client call it). Here it is: CREATE FUNCTION getstats12(text,date) RETURNS text AS ' declare rater alias for $1; string text:=''''; c char(1):='',''; svid int4:=0; rec record; dnorat int4:=0; drat int4:=0; ddel int4:=0; dunrch int4:=0; wnorat int4:=0; wrat int4:=0; wdel int4:=0; wunrch int4:=0; dhours float4; whours float4; drph int2:=0; wrph int2:=0; nthday int2:=0; start timestamp; last timestamp; finish timestamp; begin SELECT dayofweek($2) INTO nthday; finish:=$2+''23:59:59''::time; last:=finish-''1day''::timespan; start:=($2-nthday)::timestamp; FOR rec IN SELECT urlinfo.id,ratedon,cid FROM urlinfo,ratings_by_serial WHERE urlinfo.id=ratings_by_serial.id and ratedby=rater and ratedon between start and finish UNION SELECTuinfo2.id,ratedon,cid FROM uinfo2,rbs2 WHERE urlinfo.id=ratings_by_serial.id and ratedby=rater andratedon between start and finish LOOP IF rec.id<>svid THEN IF rec.ratedon>last THEN IFrec.cid= -1 THEN dnorat:=dnorat+1; ELSE drat:=drat+1; END IF; ELSE IF rec.cid= -1 THEN wnorat:=wnorat+1; ELSE wrat:=wrat+1; END IF; END IF; END IF; svid:=rec.id; END LOOP; --raise notice ''urlinfo done''; FOR rec IN SELECT deleteddate FROM deleted WHERE allocatedto=rater and deleteddate between start and finish LOOP IF rec.deleteddate>last THEN ddel:=ddel+1; ELSE wdel:=wdel+1; END IF; END LOOP; --raise notice ''deleted done''; FOR rec IN SELECT checkdate FROM unreachable WHEREcheckedby=rater and checkdate between start and finish LOOP IF rec.checkdate>last THEN dunrch:=dunrch+1; ELSE wunrch:=wunrch+1; END IF; END LOOP; --raise notice ''unreachabledone''; wnorat:=wnorat+dnorat; wrat:=wrat+drat; wdel:=wdel+ddel; wunrch:=wunrch+dunrch; SELECT getratinghour(rater,$2,$2)INTO dhours; SELECT getratinghour(rater,start::date,$2) INTO whours; IF dhours>0 THEN drph:=round(drat/dhours); END IF; IF whours>0 THEN wrph:=round(wrat/whours); END IF; string:=string||drat||c||wrat||c||dnorat||c||wnorat||c|| ddel||c||wdel||c||dunrch||c||wunrch||c|| dhours||c||whours||c||drph||c||wrph; return string; end; ' LANGUAGE 'plpgsql'; Thanks. Jie LIANG St. Bernard Software 10350 Science Center Drive Suite 100, San Diego, CA 92121 Office:(858)320-4873 jliang@ipinc.com www.stbernard.com www.ipinc.com