NetTalk Central

Author Topic: Nettalk VS CodeChargeStudio  (Read 4559 times)

edycartagena

  • Newbie
  • *
  • Posts: 4
    • View Profile
    • Email
Nettalk VS CodeChargeStudio
« on: September 09, 2007, 10:57:49 AM »
Hi people i'm testing this two tools, in fact i think nettalk web server is great because we can use all the clarion code that we know :D
but i did a test with MySQL one table with 5 fields and 285,000 records.

MY hardware is a HP Server AMD atlhon 64 bits 2GB Ram SATA II hardisk
my test includes 3 aspects.

1.Reorder columns
2.Use navigartion buttons (next,previuos,last,first)
3.-locate records

i tested that points locally

1) NetTalk -> it tooks from 10 to 15 seconds
2) it tooks from 10 to 30 seconds
3) it tooks from 10 to 50 seconds

the same test using code charge php + MySQL tooks
from 1 to 5 seconds. :S

why does it happend?
Does it something related to MySQL Database?


i use primary key, in MYSQL and ClarionDCT

i want to know if is possible to be as fast as a CodeCharge Browse.

Thanks.
« Last Edit: September 09, 2007, 11:31:46 AM by edycartagena »

Bruce

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 11250
    • View Profile
Re: Nettalk VS CodeChargeStudio
« Reply #1 on: September 10, 2007, 03:16:45 AM »
Hi Edy,

The most likely cause of the problem is the key in MySql. I haven't done a lot of testing with MySQL, but I know in MsSQL the presence (or absence) of a key on the column field can make a big difference.

Bear in mind that if the column is a string, then ideally the key should be case insensitive, as this is the sort order that clicking on the column header will trigger.

That said, I suspect this is an area that can, and will, get faster as more options are exposed via the template.

Cheers
Bruce

johncort

  • Newbie
  • *
  • Posts: 6
    • View Profile
    • Email
Re: Nettalk VS CodeChargeStudio
« Reply #2 on: October 10, 2007, 04:18:02 PM »
I'm running a plan vanilla browse against a 200,000 record table - using one of the examples shipped with Nettalk ...

I'm a big proponent for speed - it's one of my top concerns.

The speed for bringing up this browse (page loaded mind you) is less than a second. Hit next, previous, first, and last all take less than a second.

I have a sysid (integer, auto-incrementing) set as my primary key. Clicking on other, non-keyed fields, takes about 2 seconds (very, very reasonable IMHO).

I am running MSSQL.

edycartagena

  • Newbie
  • *
  • Posts: 4
    • View Profile
    • Email
Re: Nettalk VS CodeChargeStudio
« Reply #3 on: October 23, 2007, 07:18:29 PM »
John can you tell us if you test the app locally or with a remote conection. and what type of internet connection do you have?
could you post an online example with that records?

thanks in advance. :D

ccordes

  • Sr. Member
  • ****
  • Posts: 384
    • View Profile
    • Email
Re: Nettalk VS CodeChargeStudio
« Reply #4 on: October 25, 2007, 06:09:25 AM »
I am also running against MSSQL and any one of my browses would need to have 3 tables related. My NetTalk app is running an on older (3yrs) WinXP box(P4,2.4GHz,512mb) and the database is on another machine on the network.
Response times in the US are very good in loading any browse while some over-processed <g> reports take a bit longer.
Clarion, out of the box, isn't the best tool for SQL but its pretty good. It does however allow you quite a bit of flexibility in creating the SQL Statements by using prop:sql.
You might want to trace a session and then run the selects from the log in the Query Analyzer. Take a look at the execution plan. You could also use the Index tuning wizard.

HTH
chris
Real programmers use copy con newapp.exe