Обсуждение: an other provokative question??
Relational database pioneer says technology is obsolete http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9034619 kindlt explain how?? sincerely siva
> -----Original Message----- > From: pgsql-general-owner@postgresql.org [mailto:pgsql-general- > owner@postgresql.org] On Behalf Of volunteer@spatiallink.org > Sent: Thursday, September 06, 2007 12:33 PM > To: pgsql-general@postgresql.org > Subject: [GENERAL] an other provokative question?? > > Relational database pioneer says technology is obsolete > http://www.computerworld.com/action/article.do?command=viewArticleBasic& ar > ticleId=9034619 > > kindlt explain how?? This bit is a hint: "Column-oriented databases -- such as the one built by Stonebraker's latest start-up, Andover, Mass.-based Vertica Systems Inc. -- store data vertically in table columns rather than in successive rows." Mr. Stonebraker's company sells column oriented databases. So of course the other methods must be "obsolete". It actually is a good idea for some operations. Database warehouses seem to benefit from that storage scheme. All of the database systems that I know of that use this column-oriented scheme are in-memory database systems. I don't know if Mr. Stonebraker's is also. There is at least one open source database that uses columns to store the data: http://monetdb.cwi.nl/
Dann Corbit wrote: > All of the database systems > that I know of that use this column-oriented scheme are in-memory > database systems. I don't know if Mr. Stonebraker's is also. KDB+ (http://kx.com/) is column-oriented and has both on-disk and in-memory capabilities http://kx.com/faq/#6 . It's around since 1998 and both column and row oriented databases are still around so I think it'd be more fair to say both are mature technologies.
volunteer@spatiallink.org writes: > Relational database pioneer says technology is obsolete > http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9034619 > kindlt explain how?? There are several spins relevant to this: 1. He's trying to sell His New Thing, and it certainly makes good copy to say "your old stuff is obsolete - buy our new stuff!" 2. There are problems with SQL which cause many to want something better. The thing is, SQL isn't forcibly particularly "relational;" there is a purist view which says that it definitely *isn't*. http://en.wikipedia.org/wiki/RDBMS#Current_Usage They suggest that people interpret "relational" as implying: - System supports having collections of tables - System supports certain relational operators that work, in some contexts, on these tables. SQL provides that sort of thing, but is not as expressive as people would like. In effect, SQL has gotten hobbled so many ways over the years that people seem to find it easier to say "relational == obsolete" than to try to explain that what they're trying to do is perhaps *more* faithful to the theoretical relational model than the existing products. -- let name="cbbrowne" and tld="linuxdatabases.info" in name ^ "@" ^ tld;; http://cbbrowne.com/info/nonrdbms.html Rules of the Evil Overlord #14. "The hero is not entitled to a last kiss, a last cigarette, or any other form of last request." <http://www.eviloverlord.com/>
"Dann Corbit" <DCorbit@connx.com> writes: >> Relational database pioneer says technology is obsolete >> http://www.computerworld.com/action/article.do?command=3DviewArticleBasic&articleId=3D9034619 > This bit is a hint: > "Column-oriented databases -- such as the one built by Stonebraker's > latest start-up, Andover, Mass.-based Vertica Systems Inc. -- store data > vertically in table columns rather than in successive rows." > Mr. Stonebraker's company sells column oriented databases. So of course > the other methods must be "obsolete". I don't see anything in there where Stonebraker says that relational DBs are obsolete. What he suggests is that column-oriented storage might beat row-oriented storage for a lot of modern applications. He might be right (I'm sure not going to bet against the guy who started Postgres) but this has not got anything to do with the concept of a relational database. It's an implementation detail --- maybe a pretty fundamental one, but in principle you could build a DB either way and no user could see a semantic difference. Count on a reporter to overstate the argument ... regards, tom lane
On Sep 6, 2007, at 10:54 PM, Tom Lane wrote: > "Dann Corbit" <DCorbit@connx.com> writes: >>> Relational database pioneer says technology is obsolete >>> http://www.computerworld.com/action/article.do? >>> command=3DviewArticleBasic&articleId=3D9034619 > >> This bit is a hint: >> "Column-oriented databases -- such as the one built by Stonebraker's >> latest start-up, Andover, Mass.-based Vertica Systems Inc. -- >> store data >> vertically in table columns rather than in successive rows." > >> Mr. Stonebraker's company sells column oriented databases. So of >> course >> the other methods must be "obsolete". > > I don't see anything in there where Stonebraker says that > relational DBs > are obsolete. What he suggests is that column-oriented storage might > beat row-oriented storage for a lot of modern applications. He > might be > right (I'm sure not going to bet against the guy who started Postgres) > but this has not got anything to do with the concept of a relational > database. It's an implementation detail --- maybe a pretty > fundamental > one, but in principle you could build a DB either way and no user > could > see a semantic difference. > I'm curious, given that Postgres wasn't even an SQL-centric database when the original project ended, how much of the current Postgres code base still contains code from the original project before the incorporation of SQl rename to PostgreSQL? Erik Jones Software Developer | Emma® erik@myemma.com 800.595.4401 or 615.292.5888 615.292.0777 (fax) Emma helps organizations everywhere communicate & market in style. Visit us online at http://www.myemma.com
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 09/06/07 22:54, Tom Lane wrote: > "Dann Corbit" <DCorbit@connx.com> writes: >>> Relational database pioneer says technology is obsolete >>> http://www.computerworld.com/action/article.do?command=3DviewArticleBasic&articleId=3D9034619 > >> This bit is a hint: >> "Column-oriented databases -- such as the one built by Stonebraker's >> latest start-up, Andover, Mass.-based Vertica Systems Inc. -- store data >> vertically in table columns rather than in successive rows." > >> Mr. Stonebraker's company sells column oriented databases. So of course >> the other methods must be "obsolete". > > I don't see anything in there where Stonebraker says that relational DBs > are obsolete. What he suggests is that column-oriented storage might Does "column-oriented storage" mean that all of the COLUMN_A values for all 200 million rows are stored together on adjacent pages? If so, then doing aggregates (the bread and butter of DW) *would* seem to be faster. But b-tree leaf that points to "a record" would need num_cols pointers instead of one pointer. Very messy. And large. Definitely a niche product. > beat row-oriented storage for a lot of modern applications. He might be > right (I'm sure not going to bet against the guy who started Postgres) > but this has not got anything to do with the concept of a relational > database. It's an implementation detail --- maybe a pretty fundamental > one, but in principle you could build a DB either way and no user could > see a semantic difference. - -- Ron Johnson, Jr. Jefferson LA USA Give a man a fish, and he eats for a day. Hit him with a fish, and he goes away for good! -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFG4OF4S9HxQb37XmcRAtQeAKCGqjOcdmT6ccrbMy/JDOURjYItSACfVu7/ AEdP1gbDPK/MNwCVlCb1IAg= =PD28 -----END PGP SIGNATURE-----
Erik Jones <erik@myemma.com> writes: > I'm curious, given that Postgres wasn't even an SQL-centric database > when the original project ended, how much of the current Postgres > code base still contains code from the original project before the > incorporation of SQl rename to PostgreSQL? You can still find a lot of code in the current CVS that has obvious ancestry in Postgres v4r2. I think there might not be too many lines that have never been changed at all, but nobody who could read C would have any problem detecting the relationship. Elein might have more to say on the point... I'm just a newbie. regards, tom lane
On Fri, 7 Sep 2007, Ron Johnson wrote: > Definitely a niche product. Stonebraker's commentary was unfortunately spun by the ComputerWorld columnist. I hope people followed the link to his actual blog entry at http://www.databasecolumn.com/2007/09/one-size-fits-all.html where his arguement is that the idea of one database approach always being right just isn't true anyway. With that mindset, every technology is a niche product of sorts; just the size of the niche varies. Given past history of this project and its relation to Stonebraker, I was tempted earlier today to suggest that the Postgres vs. PostgreSQL renaming argument be dropped in favor of renaming the database "Horizontica". -- * Greg Smith gsmith@gregsmith.com http://www.gregsmith.com Baltimore, MD
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 09/07/07 01:37, Greg Smith wrote: > On Fri, 7 Sep 2007, Ron Johnson wrote: > >> Definitely a niche product. > > Stonebraker's commentary was unfortunately spun by the ComputerWorld > columnist. Tech journalist morphing reality to make a provocative story? Say in ain't so! > I hope people followed the link to his actual blog entry at > http://www.databasecolumn.com/2007/09/one-size-fits-all.html where his > arguement is that the idea of one database approach always being right > just isn't true anyway. With that mindset, every technology is a niche > product of sorts; just the size of the niche varies. I read something similar by him a few months ago. Very interesting. > Given past history of this project and its relation to Stonebraker, I > was tempted earlier today to suggest that the Postgres vs. PostgreSQL > renaming argument be dropped in favor of renaming the database > "Horizontica". - -- Ron Johnson, Jr. Jefferson LA USA Give a man a fish, and he eats for a day. Hit him with a fish, and he goes away for good! -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFG4T/TS9HxQb37XmcRAj8gAKC+IeBhW9+0qlrRSiUbVwQH/EiQlACfWxVm Y4uTvkWRDog+W2vPTrULXmQ= =bKYs -----END PGP SIGNATURE-----
> Relational database pioneer says technology is obsolete The headline is grossly misleading; the article is only somewhat less misleading. The actual blog entry: - Does not claim that relational databases per se are obsolete; - Claims that the idea of a "single one size fits all" implementation of the physical storage strategy for a relational db is obsolete; - Does not claim that column-oriented storage is better for all use cases; - Does claim that column-oriented is vastly superior for data warehouses; - Further claims that for all use cases there exists some specialized storage strategy that will perform better than the standard row-oriented strategy. -- Scott Ribe scott_ribe@killerbytes.com http://www.killerbytes.com/ (303) 722-0567 voice
Greg Smith wrote: > the idea of one database approach always being right > just isn't true anyway. Or one software development approach or one application runtime approach or ... . One of the major contributions of the relational model compared to the myriad network model implementations prevalent at the time was that the relation model was a straightforward approach whose theory was easily grasped and applied. The model made *people* more productive; certainly, network implementations could run circles around the early relational implementations, so it didn't make database servers more productive. Any alternative DBMS approach would have to provide a similar level of human productivity. Given relational's staying power over the intervening 25+ years (with many more people available to research alternatives and much more available computing power), that appears to be a significant challenge. -- Guy Rouillier
-----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 > Given past history of this project and its relation to > Stonebraker, I was tempted earlier today to suggest that > the Postgres vs. PostgreSQL renaming argument be dropped > in favor of renaming the database "Horizontica". That name could be confusing - what about the people out there who would be wondering if we'd support SQL or not? The name should definitely be "HorizonticaSQL" - -- Greg Sabino Mullane greg@turnstep.com PGP Key: 0x14964AC8 200709081522 http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8 -----BEGIN PGP SIGNATURE----- iD8DBQFG4vagvJuQZxSWSsgRA7/JAJ46OKu/LsIHaP895z2wxIBAXFuQ8ACghTZG xQTGPm5UrsPXEi6Sm8tdvcI= =vLvO -----END PGP SIGNATURE-----
On 9/7/07, Greg Smith <gsmith@gregsmith.com> wrote: > ... renaming the database "Horizontica". Following the naming convention, wouldn't it be "Horizonta"?
On 9/8/07, Greg Sabino Mullane <greg@turnstep.com> wrote: > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: RIPEMD160 > > > > Given past history of this project and its relation to > > Stonebraker, I was tempted earlier today to suggest that > > the Postgres vs. PostgreSQL renaming argument be dropped > > in favor of renaming the database "Horizontica". > > That name could be confusing - what about the people out there > who would be wondering if we'd support SQL or not? The name > should definitely be "HorizonticaSQL" Hey! Some of us just figured out how to pronounce PostgreSQL properly.
Greg Sabino Mullane wrote: >> ...in favor of renaming the database "Horizontica". > > ...should definitely be "HorizonticaSQL" Surely that should be capitalized "HorizonticASQL", no.
On 9/12/07, Ron Mayer <rm_pg@cheapcomplexdevices.com> wrote: > Greg Sabino Mullane wrote: > >> ...in favor of renaming the database "Horizontica". > > > > ...should definitely be "HorizonticaSQL" > > Surely that should be capitalized "HorizonticASQL", no. I can just see it now. I've been using Horizont and .... followed by 100 messages of "It's HorizonticASQL or Horizontica, but not horizont...
On 9/9/07, Scott Marlowe <scott.marlowe@gmail.com> wrote: > Hey! Some of us just figured out how to pronounce PostgreSQL properly. Postgreh SeeQuell?? /me ducks -- Please don't top post, and don't use HTML e-Mail :} Make your quotes concise. http://www.american.edu/econ/notes/htmlmail.htm
howto check column than row?? create table horizontal (id integer, first varchar(8), last varchar(8)); insert into horizontal values ('1', 'Jack', 'Ja'); insert into horizontal values ('2', 'Jill', 'Ji'); insert into horizontal values ('3', 'Mary', 'Ma'); select * from horizontal where true; +----+-------+-------+ | id | first | last | +----+-------+-------+ | 1 | Jack | Ja | | 2 | Jill | Ji | | 3 | Mary | Ma | +----+-------+-------+ (3 rows) # create table vertical (id varchar(8), one varchar(8)); insert into vertical values ('first', 'Jack'); insert into vertical values ('last', 'Ja'); alter table vertical add column two varchar(8); update vertical set two = 'Jill' where id = 'first'; update vertical set two = 'Ji' where id = 'last'; alter table vertical add column three varchar(8); update vertical set three = 'Mary' where id = 'first'; update vertical set three = 'Ma' where id = 'last'; select * from vertical where true; +-------+-------+------+-------+ | id | one | two | three | +-------+-------+------+-------+ | first | Jack | Jill | Mary | | last | Ja | Ji | Ma | +-------+-------+------+-------+ (2 rows) # select * from horizontal where last like 'J%' order by last; +----+-------+-------+ | id | first | last | +----+-------+-------+ | 1 | Jack | Ja | | 2 | Jill | Ji | +----+-------+-------+ (2 rows) howto get flip result from vertical?? sincerely siva -------- Original Message -------- Subject: [GENERAL] an other provokative question?? From: volunteer@spatiallink.org Date: Thu, September 06, 2007 3:32 pm To: pgsql-general@postgresql.org Relational database pioneer says technology is obsolete http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9034619 kindlt explain how?? sincerely siva