Submitted by Armando on 2015/12/04 13:37
I'm bringing this back up as I found IQ slower recently. IQ's grids display/autorefresh (calculations?). Did Something change in the way grids are refreshed (or calculations performed?)?
 
I just did some tests between 59 and 62, and only found a  slight difference :
 
-V59 : check the followup field : 11 items : 7s
-V62 : check the followup field : 11 items : 7.3s 
 
But often the slowness is very perceptible, so I'm wondering. Maybe it gets slower over time? Not sure yet.
 
Thanks
 

Comments

Humm... you must have lots of things going on in background as checking / unchecking a field in the properties pane when 11 items are selected in near instantaneous...
 

Armando

2015/12/04 16:30

In reply to by Pierre_Admin

[quote=Pierre_Admin]
Humm... you must have lots of things going on in background as checking / unchecking a field in the properties pane when 11 items are selected in near instantaneous...
 
[/quote]
 
I think I'll change my signature for something like "Disclaimer: I have lots of things going on in the background and things are generally not instantaneous." That's a given -- I do have autoassigns. But that's not my point.
 
I'm not saying things are slow in an absolute sense, but in a relative sense. I.e. : slower than before. Not just slow. But maybe that wasn't clear, sorry.
 
 
-------------------------------------------------------
Windows 8.1
Sony Vaio S Series 13 (SVS131E21L)
Ram:8gb, CPU: Intel i5-3230M, 2.6ghz

Pierre_Admin

2015/12/04 16:55

In reply to by Armando

Very funny...
I like the disclaimer.
You may also add "Testing IQ with the most advanced (and complicated) IQBase in the world"  
 
That said, software dev is kind of a danse... 2 steps forward, 1 step backwards. When going forward and adding new features, it is necessary later on to step back and optimize...
 
 

Armando

2015/12/04 23:10

In reply to by Pierre_Admin

I just thought I'd mention since it's clearly perceptible (and pretty annoying at times). But for some reasons, not always; not sure why. Performance wise, there's probably something about autoassigns or equations in general that's a bit on the slow side. Again, I'm not sure what it is.
 
Just for the sake of it, even if  it's not really the point; removing a few useful functions and keeping the essential, I've got these autoassigns and nothing else:
 
 
A:FollowUpDate=followupValue( [TaskActStart] ,  [FollowUp] ,  [FollowupDate] ,  [taskStartDurLock] ,  [Rechour] ) 
A:itemforecolor="&h91CFFF"
A:icon_statut=AddKeyword( [Icon_Statut] , "<img>hourglass_trans</img>")
 
E:Icon_statut=RemoveKeyword( [Icon_Statut] , "<img>hourglass_trans</img>") 
E:itemforecolor=
E:FollowUpDate=
 
 
Checking 11 items takes 2.6 sec. with a 3ghz processor. It's not hugely slow, but it's certainly not fast.
 
 
Of course I could throw all my autoassigns away, but then...
 
 
 
===================================================
The mysterious function written about 5-6 years ago (when I didn,t know that much but still does exactly what I need) goes like this:
 
 
Function followupValue(TaskActStart, FollowUp, FollowUpDate, locked, recHour)
  
  Dim rightHour
  If (Not IsNull(recHour) Or  recHour > 0)  Then rightHour = recHour Else rightHour = 0.25
  
  If followUp = False Then
    followupValue = Null
    Exit Function
  ElseIf locked = True And Not IsNull(followupdate) Then
    followupValue = FollowUpDate
    Exit Function
    
  ElseIf (((TaskActStart >= Now) Or (rightHour >= Now-Int(Now))) And (Int(TaskActStart) - Int(Now) <= 0)) Or (IsNull(TaskActStart) And rightHour >= Now-Int(Now)) Then
    followupValue = Int(Now) + (rightHour)
  ElseIf (TaskActStart < Now Or (rightHour < Now-Int(Now))) Or (Int(TaskActStart) - Int(Now) > 0) Or (IsNull(TaskActStart) And rightHour < Now-Int(Now)) Then
    followupValue = Int(Now) + 1 + (rightHour)
  Else
    If followupValue = Null Then MsgBox "followupValue is null !"
    If followupValue = Empty Then MsgBox "followupValue is empty !"
    followupValue = FollowUpDate
  End If
  
End Function
 
 

(I knew you'd like my disclaimer.)
 
-------------------------------------------------------
Windows 8.1
Sony Vaio S Series 13 (SVS131E21L)
Ram:8gb, CPU: Intel i5-3230M, 2.6ghz

Armando

2015/12/05 15:33

In reply to by Armando

[quote=Armando]
 
Checking 11 items takes 2.6 sec. with a 3ghz processor. It's not hugely slow, but it's certainly not fast.
 
[/quote]
 
Today, exact same items, same equations (see above), sames grids open.:
 
Checking 11 items takes 5.2 sec. It,s not due to any other process running in the background. It just takes longer. A bit like if some cashing file was getting bigger and bigger, and executing slower -- no idea, just trying to describe how it feels. Will see if I can find a quicker version so that I can actually tell when it started.
 
 
-------------------------------------------------------
Windows 8.1
Sony Vaio S Series 13 (SVS131E21L)
Ram:8gb, CPU: Intel i5-3230M, 2.6ghz

Armando

2015/12/05 16:17

In reply to by Armando

closing/reopening IQ brought back the timing to about 4.1 sec.
 
However, changing to 59 consistently brought back the time to 1.8 sec. (Will stick to it for now as the performance change does affect my wTim --  could be that 60a is ok too but am running out of time. : -) )
 
 
-------------------------------------------------------
Windows 8.1
Sony Vaio S Series 13 (SVS131E21L)
Ram:8gb, CPU: Intel i5-3230M, 2.6ghz

Armando

2015/12/17 13:23

In reply to by Armando

[quote=David_H]
Spock: Scotty, I store 74,734,215 items in InfoQube.  I'm noticing .3 nanosecond slowdown in the latest release.
 
Scotty: For God's sake Spock, I'm making it go as fast as I can here!  Any faster and it may tear apart!
 
Spock: I'll be patient but the slowdown is illogical.
[/quote]
 
Eheh! I'll take that as a compliment. I like Spock.

Actually, the difference is much bigger. My first post tells that story : it feels slow, but I can't time it. However, in other experiment, I was able to time the difference (later posts). v59 is usually more than 2x quicker.

don't know if that's a tiny difference for you guys, but for me it's -- how shall I put it -- not that spockish :

[quote=Armando]
closing/reopening IQ brought back the timing to about 4.1 sec.
 
However, changing to 59 consistently brought back the time to 1.8 sec. (Will stick to it for now as the performance change does affect my work --  could be that 60a is ok too but am running out of time. : -) )
 
 [/quote]
 
 
-
Disclaimer: "Testing IQ with the most advanced/complicated IQBase in the world". I.e. slower than average.
Windows 8.1
CPU: Intel i5 2.6ghz

This might not be a priority for every one but it is for me and maybe for those with bigger database dealing with auto-assigns  : - ) . Can,t really work properly with 63, so am using 59.
 
do you know what could be causing the performance hit?

(I can wait till 65, 66 or whatever as 59 is mostly stable, no problem, but won't be able test the intermediary versions thoroughly.)

-
Disclaimer: "Testing IQ with the most advanced/complicated IQBase in the world". I.e. slower than average.
Windows 8.1
CPU: Intel i5 2.6ghz

Pierre_Admin

2015/12/16 15:07

In reply to by Armando

I'm not sure what it can be.
Let me release v63, as many users are waiting for it and I'll then look into this.
I'll compare v59 with the current. Something should come up...
 

Armando

2015/12/16 15:51

In reply to by Pierre_Admin

[quote=Pierre_Admin]
I'm not sure what it can be.
Let me release v63, as many users are waiting for it and I'll then look into this.
I'll compare v59 with the current. Something should come up...
 
[/quote]
 
Thanks Pierre.
(No problem, take your time.)
-
Disclaimer: "Testing IQ with the most advanced/complicated IQBase in the world". I.e. slower than average.
Windows 8.1
CPU: Intel i5 2.6ghz

Spock: Scotty, I store 74,734,215 items in InfoQube.  I'm noticing .3 nanosecond slowdown in the latest release.
 
Scotty: For God's sake Spock, I'm making it go as fast as I can here!  Any faster and it may tear apart!
 
Spock: I'll be patient but the slowdown is illogical.

Pierre_Admin

2015/12/17 11:32

In reply to by David_H

lol
 
I love it !
 

Just having fun Armando, if I was seeing a difference like that it would bother me.  I suspect most of our grids loads instantaneously as we are not using IQ as you are, so we don't notice a difference.  If you want to send me your database ;-), I'll test it for you on a workstation with PCI SSD drives and 64 GB of RAM, and tell you what type of performance I see.

Pierre_Admin

2015/12/17 20:21

In reply to by David_H

Wow, a real beast you have !!!
 

Armando

2015/12/17 23:29

In reply to by David_H

[quote=David_H]
Just having fun Armando, if I was seeing a difference like that it would bother me.  I suspect most of our grids loads instantaneously as we are not using IQ as you are, so we don't notice a difference.  If you want to send me your database ;-), I'll test it for you on a workstation with PCI SSD drives and 64 GB of RAM, and tell you what type of performance I see.
[/quote]
 
I know David, and I like that kind of humor.
 
Nice workstation. For the kind of performance increase I'd be looking for,  I believe CPU "power" is the key. SSD and RAM would probably only affect searching/grid refresh and loading time.
 
-
Disclaimer: "Testing IQ with the most advanced/complicated IQBase in the world". I.e. slower than average.
Windows 8.1
CPU: Intel i5 2.6ghz