AutoIt v3.3.9.13 Beta

[Jos]

  On 7/31/2013 at 7:51 PM, Digisoul said:

 

Thanks, I was using old version of Autoit3Wrapper, because I did not faced such issue in 3.3.9.12 Beta.

Even after updating the new version from "beta_SciTE4AutoIt3" link, still getting the error, console showed:

!> Aut2exe.exe ended errors because the target exe wasn't created, abandon build. (C:\Users\***\AppData\Local\Temp\~AU3lobqafz.exe)rc:9999

Show us an snippet with all Directives that demonstrates this behaviour so I can check what is happening,

EDIT: Also show me all SciTE output pane information so I can check that too.

Jos

Edited July 31, 2013 by Jos

[Digisoul]

  On 7/31/2013 at 7:54 PM, Jos said:

Show us an snippet with all Directives that demonstrates this behaviour so I can check what is happening,

I changed the temp location as you suggested but still getting error.

Scite Output:

>Running:(3.3.9.13):C:\Program Files (x86)\AutoIt3\Beta\aut2exe\aut2exe.exe  /in "D:\Projects\GC\DB_daemon.au3" /out "C:\Users\***\~AU3bkgjvjz.exe" /nopack /icon "D:\Projects\GC\hyperlink_blue.ico" /comp 2

!>00:54:25 Aut2exe.exe ended errors because the target exe wasn't created, abandon build. (C:\Users\***\~AU3bkgjvjz.exe)rc:9999

>Exit code: 0    Time: 5.039
 

Directives

#Region ;**** Directives created by AutoIt3Wrapper_GUI ****
#AutoIt3Wrapper_Icon=hyperlink_blue.ico
#AutoIt3Wrapper_Outfile=Daemon.exe
#AutoIt3Wrapper_UseUpx=n
#AutoIt3Wrapper_Res_Comment=GC DataBase Daemon
#AutoIt3Wrapper_Res_Fileversion=1.2.11.0
#AutoIt3Wrapper_Res_LegalCopyright=©cADD 2013
#AutoIt3Wrapper_Res_Field=CompanyName|cADD®
#AutoIt3Wrapper_Res_Field=Company Name|cADD®
#AutoIt3Wrapper_Res_Field=Orignal File name|Daemon.exe
#AutoIt3Wrapper_Res_Field=Product Name|Daemon
#AutoIt3Wrapper_Res_Field=Product Version|1.2.11.0
#EndRegion ;**** Directives created by AutoIt3Wrapper_GUI ****

Edited July 31, 2013 by Digisoul

[trancexx]

  On 7/31/2013 at 7:51 PM, Digisoul said:

 

Thanks, I was using old version of Autoit3Wrapper, because I did not faced such issue in 3.3.9.12 Beta.

Even after updating the new version from "beta_SciTE4AutoIt3" link, still getting the error, console showed:

!> Aut2exe.exe ended errors because the target exe wasn't created, abandon build. (C:\Users\***\AppData\Local\Temp\~AU3lobqafz.exe)rc:9999

Do you have AV running and which one if so?

I had issues on one system when some version of Aut2Exe used %temp% directory for some PE temp files. I reported that to Jon and was under impression that he changed the location. I haven't had chance to test latest version on that particular comp but very likely I would have similar failure you do.

...But maybe it isn't related and it's really wrapper issue.

Edited July 31, 2013 by trancexx

[Jos]

All looks good to me so agree with trancexx that the first thing to check your AV messing stuff up.

[Digisoul]

  On 7/31/2013 at 8:09 PM, Jos said:

All looks good to me so agree with trancexx that the first thing to check your AV messing stuff up.

Thanks, my own protection software was causing the issue.

  On 7/31/2013 at 8:01 PM, trancexx said:

Do you have AV running and which one if so?

I had issues on one system when some version of Aut2Exe used %temp% directory for some PE temp files. I reported that to Jon and was under impression that he changed the location. I haven't had chance to test latest version on that particular comp but very likely I would have similar failure you do.

...But maybe it isn't related and it's really wrapper issue.

%temp% directory was not causing that issue, because even after updating the temp path I was still getting the error; My protection software was unable to recognize the exe, that's why I was getting that error. Thanks for your help.

Edited July 31, 2013 by Digisoul

[Ascend4nt]

Digisoul, what kind of antivirus or other protection are you running?  Seems odd that it would silently delete or deny access to executables.

[Digisoul]

  On 7/31/2013 at 10:18 PM, Ascend4nt said:

Digisoul, what kind of antivirus or other protection are you running?  Seems odd that it would silently delete or deny access to executables.

Not AntiVirus, that was experimental project and I forget to stop the service.

[czardas]

Finally I found the charger for my netbook (XP SP3) and installed this. Sorry for not testing earlier, but my PC needs some TLC and is currently not a good test machine. So far I compiled a mus++ snippet and it worked like a charm. I need to play around with this for a while. Thanks to all involved and especially thanks to Jon for his endless effort and dedication both now and over the years, without you none of this would exist. Thanks again to trancexx for your inspirational ideas and contributions.

One further thing I want to say. It saddens me to see discord on the forum - which can be hard to ignore sometimes. I have every confidence that the next stable release will be every bit as good as people say and I hope differences of opinion (much of which is beyond my technical understanding) can be resolved without any discord. That's all I wanted to add.

Now back to the important technical discussion.

[FireFox]

In the second example of the DllStructGetPtr, the int -1 is casted into a float but the result is -1.#QNAN, shouldn't it be -1.00 ?

If I try with 1 I get 1.40129846432482e-045, though I expect 1.00.

Tested with the latest stable/beta versions.

Edit: Direct example :

#include <Constants.au3>
 
Example2()
 
Func Example2()
    ; Assign a Local variable a structure with the definition of an int.
    Local $tSTRUCT1 = DllStructCreate("int")
 
    ; If an error occurred display the error code and return False.
    If @error Then
        MsgBox(BitOR($MB_SYSTEMMODAL, $MB_ICONHAND), Default, "Error in DllCall, Code: " & @error)
        Return False
    EndIf
 
    ; Assign a Local variable the tSTRUCT1 structure except that the elements will be got according to the new definition.
    Local $tSTRUCT3 = DllStructCreate("float", DllStructGetPtr($tSTRUCT1, 1))
 
    ; If an error occurred display the error code and return False.
    If @error Then
        MsgBox(BitOR($MB_SYSTEMMODAL, $MB_ICONHAND), Default, "Error in DllCall, Code: " & @error)
        Return False
    EndIf
 
    ; Set the data of the first element (int) in the $tSTRUCT1.
    DllStructSetData($tSTRUCT1, 1, -1)
 
    ; Display the different data types of the same data
    MsgBox($MB_SYSTEMMODAL, "", _
            "int: " & DllStructGetData($tSTRUCT1, 1) & @CRLF & _
            "float: " & DllStructGetData($tSTRUCT3, 1))
 
    ; Release the resources used by the structures.
    $tSTRUCT1 = 0
    $tSTRUCT3 = 0
EndFunc   ;==>Example2

Edited August 2, 2013 by FireFox

[trancexx]

  On 8/2/2013 at 1:02 PM, FireFox said:

In the second example of the DllStructGetPtr, the int -1 is casted into a float but the result is -1.#QNAN, shouldn't it be -1.00 ?

 

If I try with 1 I get 1.40129846432482e-045, though I expect 1.00.

 

Tested with the latest stable/beta versions.

 

Edit: Direct example :

No it shouldn't. You are probably confused by the type casting that happens because you don't see that in AutoIt considering it's not strong-typed language.

What you see here is the same behavior as you would see in c++ (if you are familiar) when using reinterpret_cast casting operator if you would reinterpret pointer of one type as pointer of another type and dereference it later. In this particular case the memory space of the (int) structure is interpretted as the float type. Integer with binary representation of -1 is the same in memory as float with value of... whatever is displayed.

Edited August 2, 2013 by trancexx

[FireFox]

AH. Thank you for the explanation

I thought it was a "direct cast" like this:

int memdata=-1;
printf("%f", (float)memdata);

Edited August 2, 2013 by FireFox

[trancexx]

^^ It would be this:

int memdata = -1;
float fff = *(reinterpret_cast<float*>(&memdata));
printf("%f", fff);

[FireFox]

That's right 

Although the result is not the same for 1, it returns 0.00000 and not 1.40....

Edited August 2, 2013 by FireFox

[Mat]

-1 is all bits set, including the most significant fraction bit, so the result is quiet NaN (if the msb for the fractional part was zero it would be a signalling NaN).

1 would only have the lowest bit set. Because none of the exponent bits would be set it's actually a denormalized float. No exponent also means exponent will be at its smallest. -127 for a float, so the number you are looking at is very small (probably why you are seeing zero).

[FireFox]

  On 8/2/2013 at 2:26 PM, Mat said:

-1 is all bits set, including the most significant fraction bit, so the result is quiet NaN (if the msb for the fractional part was zero it would be a signalling NaN).

 

1 would only have the lowest bit set. Because none of the exponent bits would be set it's actually a denormalized float. No exponent also means exponent will be at its smallest. -127 for a float, so the number you are looking at is very small (probably why you are seeing zero).

Makes all sense now ! Thank you too

 

Forgot the e-045 which is important, so it was wrong in my head (0 <> 1.40 and not 0 <> 0.0000...140..)

Edited August 2, 2013 by FireFox

[Melba23]

Jon,

Following a stupid typo in a script I have discovered that the following is considered valid in 3.3.9.14 - it errors in 3.3.9.4 and earlier:

Global $sSelection
$sSelection &+ "\"
ConsoleWrite('@@ Debug(' & @ScriptLineNumber & ') : $sSelection = ' & $sSelection & @crlf & '>Error code: ' & @error & @crlf) ;### Debug Console

It should of course have read "&=".

M23