Re: I just got it: PostgreSQL Application Server -- a new project. - Mailing list pgsql-hackers

> ...That's one of the reasons I wrote Pl/Java.

More power too you! I'd really like to hear more about this project. Is

http://gborg.postgresql.org/project/pljava/projdisplay.php

your URL?

> In essence, I don't think we disagree on anything.
> The only thing I'm reacting to is the term "app-server" which I
> think is badly chosen. Stored procedures and functions doesn't
> make a database an app-server, no matter what you put in them.

I'm now in complete agreement: app-server doesn't fit.  Do you have any
suggestions?  Would a postgreslet be out of bounds, do you think?

> You can write your own functions in C and thereby get
> all the process control you want. Or if you want to make life
> easier and get a more portable solution (by my standards that is)
> why not use Java?

I admit my almost complete ignorance of how sensitive the postgres backend
is to all the hazards of process-control: is the postgres process REALLY
just another UNIX process?  Can I "exec" on top of it?  Can I fork? Can I
have a child-process using IPC wait for 10 mins for its connected process do
its work without hosing the postmaster with its shared memory locks and all
that?  I've held off any serious development along these lines since I don't
have the time to do heavy code-trawling, that seeming the only way of
obtaining the level of detail necessary to do the job well.

I would most definitely use embedded java if it could do at-minimum SRF's
and spawn processes.  Something similar to SPI for Java would be pretty
useful too, I imagine.

Best Regards,

Carl <|};-)>



-----Original Message-----
From: Thomas Hallgren [mailto:thhal@mailblocks.com]
Sent: Saturday, June 12, 2004 4:24 PM
To: Carl E. McMillin
Cc: pgsql-hackers@postgresql.org; pgsql@mohawksoft.com; Bob; 'Bill Martin';
'Joe Burks'; verbus@hotmail.com
Subject: Re: [HACKERS] I just got it: PostgreSQL Application Server -- a new
project.


"Carl E. McMillin" <carlymac@earthlink.net> writes:
> My name is Carl E. McMillin and I'm still establishing my balance in
> this particular knowledge domain with its nomenclature and entities.
>
Ok, I was thinking more the name behind pgsql@mohawksoft.com ;-)

> But Postgres isn't purely a storage solution; it is not just a place
> to
hang
> your data.  Aren't stored procedures, whether SQL-based or backed by
native
> libraries, very much essential to application-logic performance and
> portability?  Ok, portability may suffer, but they do help
> performance!
>
I agree. Some app logic is best performed in the backend. Especially if the
logic is storage intensitive or deals with advanced storage
constraints/rules. That's one of the reasons I wrote Pl/Java. In essence, I
don't think we disagree on anything. The only thing I'm reacting to is the
term "app-server" which I think is badly chosen. Stored procedures and
functions doesn't make a database an app-server, no matter what you put in
them.

> One of my personal interests is "hybridizing" a strong SQL
> execution-environment such as Postgres with an equally strong
> process-control framework so that components which would normally be
> in
the
> "middle" tier are directly accessible by way of "extensions".  For
instance,
> constructs such as the following would be really useful in some
> bioinformatics-related consulting I'm involved in:
>
> SELECT * FROM get_list_of_hsp_from_blastall('ACGGATTAT', 'H_sapiens');
>
> The function "get_list_of_hsp_from_blastall" takes a primer
> ('ACGGATTAT') and an organism ('H_sapiens') and runs an external
> process called
"blastall"
> to locate "high-scoring pairs" where the primer "aligns well" with the
> organism's nucleotide-sequence (its genome).  This would be a
> relatively trivial exercise if Postgres had a robust framework for
> process control - maybe it does, I haven't gotten many responses
> indicating yea or nay.
>
You can write your own functions in C and thereby get all the process
control you want. Or if you want to make life easier and get a more portable
solution (by my standards that is) why not use Java?

Kind regards,

Thomas Hallgren