Hi Pierre,
Thanks for answering my other questions. Just one more question, based on your answer to Alex.
I had a quick look in the Metcard archives and found this snippet:
>>
put "" into DbAuteurs
get shell("echo" && quote & "select distinct auteur__ from citations order by
auteur__" && quote && "| psql -h localhost citalis")
repeat for each line l in line 3 to -3 of it
put word 1 to -1 of l & return after DbAuteurs
end repeat
<<
This is more or less how I imagined you were shelling out to the psql interpreter. Am I correct in thinking this is how you are invoking postgresql?
I can understand Alex's disbelief in this regard. I share his understanding of how things are supposed to work using the combination the PHP Apache module, and persistent connections.
One of my apps is running on Linux and querying Firebird via ODBC. All the queries have configurable debug parameters, so I can actually change the parameter to time each step of the process. So I can isolate how long that app takes to connect and return a result set.
I will see if I can find some details on running Rev 2.01 as a command line application, and I will try to find some time next week to compare the speed of the two methods.
Bernard
| Pierre Sahores <[EMAIL PROTECTED]>
Sent by: [EMAIL PROTECTED] 13/07/2003 01:40
|
To: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]> cc: Subject: Re: About MC/RR applications servers |
On Sun, 2003-07-13 at 01:19, Alex Rice wrote:
> On Saturday, July 12, 2003, at 04:56 PM, Pierre Sahores wrote:
> >
> > Just do tests and you will see as me that this works perfectly, faster
> > for example, than in using ASP's or PHP commands, because the linux
> > bash
> > optimisations, because the psql perfect design to work in command-line
> > pipe mode
>
> This doesn't make any sense to me. I would like to see the shell
> command you are using.
Hum... It's, even for you, freely, available on the Metacard archive
list.
>
> You're saying what you are doing is faster than PHP direct to
> PostgreSQL? PHP direct to PosgreSQL can have a persistent connection
> already open, and PHP (as apache module) is running already in memory?!
>
> Your way: have to open a socket from PHP, launch the shell, then launch
> the psql command line program, which THEN finally connects to the
> database.
Did you learn a little (aka lots) about WebSphere Weblogic or WebObjects
before. Have you any idea about how applications servers works...
>
> It absolutely impossible that your way could be faster.
Perhaps are you too sure of you about how unixes are handeling multiples
processes tasks. Did you ever watch at the processor idle average time
of a linux box ? To the end, you are not alone to think so... even if
my clients are, probably, not only too rich and stupid persons...
>
> Again, why are you even using Revolution instead of just going from PHP
> to PostgreSQL? Call me confused,
Did you ever ask you about the difference it makes to use real
application server instead of just including sql replies in web forms ?
>
> Alex Rice, Software Developer
> Architectural Research Consultants, Inc.
> http://ARCplanning.com
>
> _______________________________________________
> use-revolution mailing list
> [EMAIL PROTECTED]
> http://lists.runrev.com/mailman/listinfo/use-revolution
--
Bien cordialement, Pierre Sahores
Serveurs d'applications & bases ACID SQL
Penser et produire l'avantage comp�titif
_______________________________________________
use-revolution mailing list
[EMAIL PROTECTED]
http://lists.runrev.com/mailman/listinfo/use-revolution
