"external indices" ...

I haven't worked with GiST, although I have been curious from time to
time. Just never had the time to sit, read, and try out the GiST system.

On my text search system (FTSS) I use functions that return sets of data.
It make be easier to implement that than a GiST.

Basically, I create a unique ID index on a table of data. Create my
external index. When a search is done, I return a set of unique IDs and
ranks.

Marc,

On Tue, 8 Feb 2005, Marc G. Fournier wrote:

I believe that this is what Oleg et al tap into with the tsearch2 stuff, no?
I have someone asking me about it, and want to make sure that I'm telling him
the right answer ... is this what GiST is? And, if so, what is a *good* doc
for me to point them at to get up to speed with it? Is what we have in the
docs the best place, or is there something someone else has written that gets
into it even more?

I'm a little bit confused, because I couldn't catch your meaning :)

Thanks ...

----
Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email: scrappy@hub.org Yahoo!: yscrappy ICQ: 7615664

---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?

http://archives.postgresql.org

Regards,
Oleg
_____________________________________________________________
Oleg Bartunov, sci.researcher, hostmaster of AstroNet,
Sternberg Astronomical Institute, Moscow University (Russia)
Internet: oleg@sai.msu.su, http://www.sai.msu.su/~megera/
phone: +007(095)939-16-83, +007(095)939-23-83

On Tue, 8 Feb 2005, Oleg Bartunov wrote:

Marc,

On Tue, 8 Feb 2005, Marc G. Fournier wrote:

I believe that this is what Oleg et al tap into with the tsearch2 stuff,
no? I have someone asking me about it, and want to make sure that I'm
telling him the right answer ... is this what GiST is? And, if so, what is
a *good* doc for me to point them at to get up to speed with it? Is what
we have in the docs the best place, or is there something someone else has
written that gets into it even more?

I'm a little bit confused, because I couldn't catch your meaning :)

tsearch2 is an external indexing method, no? external in that it isn't
intergrated as part of PostgreSQL, but is added on ... with the 'add on'
being through the GiST framework ... right ?

----
Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email: scrappy@hub.org Yahoo!: yscrappy ICQ: 7615664

"Marc G. Fournier" <scrappy@postgresql.org> writes:

I believe that this is what Oleg et al tap into with the tsearch2 stuff,
no? I have someone asking me about it, and want to make sure that I'm
telling him the right answer ... is this what GiST is?

Better ask what he means by "external" exactly. I think I've heard some
people use the phrase "external index" to mean any index that's not
actually part of the table itself; which makes all our index types
"external". (As a counterexample, imagine a table that is itself
organized as a btree.)

regards, tom lane

Summary:

"Currently used with Oracle databases thru their Extensible
Indexing Cartridge.

What we are interested in is finding out whether or not we can use our
technology directly with PostgreSQL. That would involve, from a overly
simplistic view, allowing Postgres to use external indexes."

Does that help any?

On Tue, 8 Feb 2005, Tom Lane wrote:

"Marc G. Fournier" <scrappy@postgresql.org> writes:

I believe that this is what Oleg et al tap into with the tsearch2 stuff,
no? I have someone asking me about it, and want to make sure that I'm
telling him the right answer ... is this what GiST is?

Better ask what he means by "external" exactly. I think I've heard some
people use the phrase "external index" to mean any index that's not
actually part of the table itself; which makes all our index types
"external". (As a counterexample, imagine a table that is itself
organized as a btree.)

regards, tom lane

----
Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email: scrappy@hub.org Yahoo!: yscrappy ICQ: 7615664

On Tue, 8 Feb 2005, Marc G. Fournier wrote:

On Tue, 8 Feb 2005, Oleg Bartunov wrote:

Marc,

On Tue, 8 Feb 2005, Marc G. Fournier wrote:

I believe that this is what Oleg et al tap into with the tsearch2 stuff,
no? I have someone asking me about it, and want to make sure that I'm
telling him the right answer ... is this what GiST is? And, if so, what
is a *good* doc for me to point them at to get up to speed with it? Is
what we have in the docs the best place, or is there something someone
else has written that gets into it even more?

I'm a little bit confused, because I couldn't catch your meaning :)

tsearch2 is an external indexing method, no? external in that it isn't
intergrated as part of PostgreSQL, but is added on ... with the 'add on'
being through the GiST framework ... right ?

I think, that guy asked about if tsearch2 is an external program which
running as a separate process.

----
Marc G. Fournier Hub.Org Networking Services (http://www.hub.org)
Email: scrappy@hub.org Yahoo!: yscrappy ICQ: 7615664

---------------------------(end of broadcast)---------------------------
TIP 4: Don't 'kill -9' the postmaster

Regards,
Oleg
_____________________________________________________________
Oleg Bartunov, sci.researcher, hostmaster of AstroNet,
Sternberg Astronomical Institute, Moscow University (Russia)
Internet: oleg@sai.msu.su, http://www.sai.msu.su/~megera/
phone: +007(095)939-16-83, +007(095)939-23-83

"Marc G. Fournier" <scrappy@postgresql.org> writes:

"Currently used with Oracle databases thru their Extensible
Indexing Cartridge.
What we are interested in is finding out whether or not we can use our
technology directly with PostgreSQL. That would involve, from a overly
simplistic view, allowing Postgres to use external indexes."

Does that help any?

Well, I spent a little time googling on this, and dug through Oracle
documents that seem to have been written by marketroids who think
"componentize" is a verb :-( ... but by the time my brain went numb,
I hadn't actually seen anything that hasn't been in Postgres for years.

What you need to make clear to them is that their existing code base
isn't likely to be worth a dime --- if they don't mind recoding, the
functionality is probably all there, but there will not be any
compatibility of details. There are no standards covering this stuff.

regards, tom lane