Buggy behavior

[D.R. Wahlgren]

Gang,
I'm experiencing a problem with SPSS.  I suspect the problem lies in
my computer, but I was wondering if anyone here was familiar with it.
I run the same syntax file week after week and it just recently
started to give me problems. Here is just one part of it:

***************************************************************************.
GET
   FILE='C:\NAPAtemp\NAPACHN1.SYS'.
DATASET NAME DataSet1 WINDOW=FRONT.
sort cases by napa_id.

***************************************************************************.
GET
   FILE='C:\NAPAtemp\NAPACHN2.SYS'.
DATASET NAME DataSet2 WINDOW=FRONT.
sort cases by napa_id.

***************************************************************************.
MATCH FILES /FILE='DataSet1'
  /RENAME (startime = d0)
  /FILE='DataSet2'
  /BY napa_id
  /DROP= d0.
EXECUTE.

SAVE OUTFILE='C:\NAPAtemp\napachn_all.sav'
   /COMPRESSED.
DATASET NAME DataSet3 WINDOW=FRONT.

***************************************************************************.

SPSS is randomly changing the assignment of the active data set.  For
example, it will insert a "DATASET ACTIVATE DataSet2." command after
the MATCH FILES COMMAND.  This of course causes the subsequent SAVE
command to save the wrong data file and assign it the wrong dataset
name.

I discovered some other, but related, misbehavior, when I was trying
to pinpoint the above problem.  I stepped through the syntax command
by command, and after I opened the first two data files, DataSet2 was
the active file.  I then selected the output file in the task bar and
saw that the green plus sign moved from the DataSet2 taskbar button
to the DataSet1 taskbar button!

Now, if I clicked on the DataSet2 taskbar button, it would switch
back to the active data file, and I could then switch back to the
output without it changing again.  I was able to repeat this whole
set of behavior a few times.

It does essentially (although not identically) the same things in
SPSS Win v15 and v18.  The problem sometimes disappears as
mysteriously as it appears in one version but not the other, or
sometimes disappears from both versions and they both work fine.
This happened just today. It's supremely frustrating.

Now, for the fine print.  I'm running WinXP as a virtual machine via
Parallels on a Mac.  I realize this is another layer of complication,
but like I said, this exact same syntax file was working flawlessly
for me under precisely these conditions for months until now.  That
suggests that something went awry in SPSS, or Win, or perhaps
Parallels, but I'm not having any other problems with Parallels to
implicate it.

Any ideas?  I've tried as much as I know. I was using files in a
shared directory before, but I've tried them in a local directory on
the C: drive with the same results.  I restarted Windows.  I've
rebooted the Mac.  I've repaired disk permissions on the Mac.  I'm
thinking I may install a new virtual machine under Parallels, with a
clean copy of Win XP and SPSS and see if the problem exists there too.

Thanks,
Dennis

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[D.R. Wahlgren]

Just as a follow up, I've also replicated the misbehavior on a
completely separate comptuer and yet another version of SPSS (v17).
This is on an actual PC running Win XP.

It seems more reliable (but not perfect) if I just run the whole lot
of syntax (the part below plus more that I did not copy here).  If I
run parts of the syntax it throws in random DATASET ACTIVATE commands
(and pastes them into the output file).  And if I click between the
output button in the task bar in between sets of commands, it often
changes the active dataset assignment, but without pasting the
command into the output file.

This has really shaken my confidence in the process, but since it's
happening in 3 different versions of SPSS on two different computers,
I think the only common denominator is my data files (.SYS).  Is
there some sort of corruption that could cause SPSS/Win to do this?

Dennis

>Date:    Tue, 27 Jul 2010 14:40:36 -0700
>From:    "D.R. Wahlgren" <[hidden email]>
>Subject: Buggy behavior
>
>Gang,
>I'm experiencing a problem with SPSS.  I suspect the problem lies in
>my computer, but I was wondering if anyone here was familiar with it.
>I run the same syntax file week after week and it just recently
>started to give me problems. Here is just one part of it:
>
>***************************************************************************.
>GET
>    FILE='C:\NAPAtemp\NAPACHN1.SYS'.
>DATASET NAME DataSet1 WINDOW=FRONT.
>sort cases by napa_id.
>
>***************************************************************************.
>GET
>    FILE='C:\NAPAtemp\NAPACHN2.SYS'.
>DATASET NAME DataSet2 WINDOW=FRONT.
>sort cases by napa_id.
>
>***************************************************************************.
>MATCH FILES /FILE='DataSet1'
>   /RENAME (startime = d0)
>   /FILE='DataSet2'
>   /BY napa_id
>   /DROP= d0.
>EXECUTE.
>
>SAVE OUTFILE='C:\NAPAtemp\napachn_all.sav'
>    /COMPRESSED.
>DATASET NAME DataSet3 WINDOW=FRONT.
>
>***************************************************************************.
>
>SPSS is randomly changing the assignment of the active data set.  For
>example, it will insert a "DATASET ACTIVATE DataSet2." command after
>the MATCH FILES COMMAND.  This of course causes the subsequent SAVE
>command to save the wrong data file and assign it the wrong dataset
>name.
>
>I discovered some other, but related, misbehavior, when I was trying
>to pinpoint the above problem.  I stepped through the syntax command
>by command, and after I opened the first two data files, DataSet2 was
>the active file.  I then selected the output file in the task bar and
>saw that the green plus sign moved from the DataSet2 taskbar button
>to the DataSet1 taskbar button!
>
>Now, if I clicked on the DataSet2 taskbar button, it would switch
>back to the active data file, and I could then switch back to the
>output without it changing again.  I was able to repeat this whole
>set of behavior a few times.
>
>It does essentially (although not identically) the same things in
>SPSS Win v15 and v18.  The problem sometimes disappears as
>mysteriously as it appears in one version but not the other, or
>sometimes disappears from both versions and they both work fine.
>This happened just today. It's supremely frustrating.
>
>Now, for the fine print.  I'm running WinXP as a virtual machine via
>Parallels on a Mac.  I realize this is another layer of complication,
>but like I said, this exact same syntax file was working flawlessly
>for me under precisely these conditions for months until now.  That
>suggests that something went awry in SPSS, or Win, or perhaps
>Parallels, but I'm not having any other problems with Parallels to
>implicate it.
>
>Any ideas?  I've tried as much as I know. I was using files in a
>shared directory before, but I've tried them in a local directory on
>the C: drive with the same results.  I restarted Windows.  I've
>rebooted the Mac.  I've repaired disk permissions on the Mac.  I'm
>thinking I may install a new virtual machine under Parallels, with a
>clean copy of Win XP and SPSS and see if the problem exists there too.
>
>Thanks,
>Dennis
>

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[John F Hall]

Just a thought.  Shouldn't SPSS saved files be *.sav not *.sys?

----- Original Message -----

From: [hidden email]

To: [hidden email]

Sent: Saturday, July 31, 2010 7:31 AM

Subject: Re: Buggy behavior

Just as a follow up, I've also replicated the misbehavior on a
completely separate comptuer and yet another version of SPSS (v17).
This is on an actual PC running Win XP.

It seems more reliable (but not perfect) if I just run the whole lot
of syntax (the part below plus more that I did not copy here).  If I
run parts of the syntax it throws in random DATASET ACTIVATE commands
(and pastes them into the output file).  And if I click between the
output button in the task bar in between sets of commands, it often
changes the active dataset assignment, but without pasting the
command into the output file.

This has really shaken my confidence in the process, but since it's
happening in 3 different versions of SPSS on two different computers,
I think the only common denominator is my data files (.SYS).  Is
there some sort of corruption that could cause SPSS/Win to do this?

Dennis

>Date:    Tue, 27 Jul 2010 14:40:36 -0700
>From:    "D.R. Wahlgren" <[hidden email]>
>Subject: Buggy behavior
>
>Gang,
>I'm experiencing a problem with SPSS.  I suspect the problem lies in
>my computer, but I was wondering if anyone here was familiar with it.
>I run the same syntax file week after week and it just recently
>started to give me problems. Here is just one part of it:
>
>***************************************************************************.
>GET
>    FILE='C:\NAPAtemp\NAPACHN1.SYS'.
>DATASET NAME DataSet1 WINDOW=FRONT.
>sort cases by napa_id.
>
>***************************************************************************.
>GET
>    FILE='C:\NAPAtemp\NAPACHN2.SYS'.
>DATASET NAME DataSet2 WINDOW=FRONT.
>sort cases by napa_id.
>
>***************************************************************************.
>MATCH FILES /FILE='DataSet1'
>   /RENAME (startime = d0)
>   /FILE='DataSet2'
>   /BY napa_id
>   /DROP= d0.
>EXECUTE.
>
>SAVE OUTFILE='C:\NAPAtemp\napachn_all.sav'
>    /COMPRESSED.
>DATASET NAME DataSet3 WINDOW=FRONT.
>
>***************************************************************************.
>
>SPSS is randomly changing the assignment of the active data set.  For
>example, it will insert a "DATASET ACTIVATE DataSet2." command after
>the MATCH FILES COMMAND.  This of course causes the subsequent SAVE
>command to save the wrong data file and assign it the wrong dataset
>name.
>
>I discovered some other, but related, misbehavior, when I was trying
>to pinpoint the above problem.  I stepped through the syntax command
>by command, and after I opened the first two data files, DataSet2 was
>the active file.  I then selected the output file in the task bar and
>saw that the green plus sign moved from the DataSet2 taskbar button
>to the DataSet1 taskbar button!
>
>Now, if I clicked on the DataSet2 taskbar button, it would switch
>back to the active data file, and I could then switch back to the
>output without it changing again.  I was able to repeat this whole
>set of behavior a few times.
>
>It does essentially (although not identically) the same things in
>SPSS Win v15 and v18.  The problem sometimes disappears as
>mysteriously as it appears in one version but not the other, or
>sometimes disappears from both versions and they both work fine.
>This happened just today. It's supremely frustrating.
>
>Now, for the fine print.  I'm running WinXP as a virtual machine via
>Parallels on a Mac.  I realize this is another layer of complication,
>but like I said, this exact same syntax file was working flawlessly
>for me under precisely these conditions for months until now.  That
>suggests that something went awry in SPSS, or Win, or perhaps
>Parallels, but I'm not having any other problems with Parallels to
>implicate it.
>
>Any ideas?  I've tried as much as I know. I was using files in a
>shared directory before, but I've tried them in a local directory on
>the C: drive with the same results.  I restarted Windows.  I've
>rebooted the Mac.  I've repaired disk permissions on the Mac.  I'm
>thinking I may install a new virtual machine under Parallels, with a
>clean copy of Win XP and SPSS and see if the problem exists there too.
>
>Thanks,
>Dennis
>

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Lemon, John S.]

My two pennies / cents worth – “.sys” was the recommended extension way back in 1975 when I started using SPSS

 

Best Wishes

 

John S. Lemon

DIT ( Directorate of Information Technology ) - Student Liaison Officer

University of Aberdeen

Edward Wright Building: Room G86a

 

Tel:  [phone]

 

Diary ( Free / Busy )

 

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of John F Hall
Sent: 31 July 2010 09:20
To: [hidden email]
Subject: Re: Buggy behavior

 

Just a thought.  Shouldn't SPSS saved files be *.sav not *.sys?

----- Original Message -----

From: [hidden email]

To: [hidden email]

Sent: Saturday, July 31, 2010 7:31 AM

Subject: Re: Buggy behavior

 

Just as a follow up, I've also replicated the misbehavior on a
completely separate comptuer and yet another version of SPSS (v17).
This is on an actual PC running Win XP.

It seems more reliable (but not perfect) if I just run the whole lot
of syntax (the part below plus more that I did not copy here).  If I
run parts of the syntax it throws in random DATASET ACTIVATE commands
(and pastes them into the output file).  And if I click between the
output button in the task bar in between sets of commands, it often
changes the active dataset assignment, but without pasting the
command into the output file.

This has really shaken my confidence in the process, but since it's
happening in 3 different versions of SPSS on two different computers,
I think the only common denominator is my data files (.SYS).  Is
there some sort of corruption that could cause SPSS/Win to do this?

Dennis


>Date:    Tue, 27 Jul 2010 14:40:36 -0700
>From:    "D.R. Wahlgren" <[hidden email]>
>Subject: Buggy behavior
>
>Gang,
>I'm experiencing a problem with SPSS.  I suspect the problem lies in
>my computer, but I was wondering if anyone here was familiar with it.
>I run the same syntax file week after week and it just recently
>started to give me problems. Here is just one part of it:
>
>***************************************************************************.
>GET
>    FILE='C:\NAPAtemp\NAPACHN1.SYS'.
>DATASET NAME DataSet1 WINDOW=FRONT.
>sort cases by napa_id.
>
>***************************************************************************.
>GET
>    FILE='C:\NAPAtemp\NAPACHN2.SYS'.
>DATASET NAME DataSet2 WINDOW=FRONT.
>sort cases by napa_id.
>
>***************************************************************************.
>MATCH FILES /FILE='DataSet1'
>   /RENAME (startime = d0)
>   /FILE='DataSet2'
>   /BY napa_id
>   /DROP= d0.
>EXECUTE.
>
>SAVE OUTFILE='C:\NAPAtemp\napachn_all.sav'
>    /COMPRESSED.
>DATASET NAME DataSet3 WINDOW=FRONT.
>
>***************************************************************************.
>
>SPSS is randomly changing the assignment of the active data set.  For
>example, it will insert a "DATASET ACTIVATE DataSet2." command after
>the MATCH FILES COMMAND.  This of course causes the subsequent SAVE
>command to save the wrong data file and assign it the wrong dataset
>name.
>
>I discovered some other, but related, misbehavior, when I was trying
>to pinpoint the above problem.  I stepped through the syntax command
>by command, and after I opened the first two data files, DataSet2 was
>the active file.  I then selected the output file in the task bar and
>saw that the green plus sign moved from the DataSet2 taskbar button
>to the DataSet1 taskbar button!
>
>Now, if I clicked on the DataSet2 taskbar button, it would switch
>back to the active data file, and I could then switch back to the
>output without it changing again.  I was able to repeat this whole
>set of behavior a few times.
>
>It does essentially (although not identically) the same things in
>SPSS Win v15 and v18.  The problem sometimes disappears as
>mysteriously as it appears in one version but not the other, or
>sometimes disappears from both versions and they both work fine.
>This happened just today. It's supremely frustrating.
>
>Now, for the fine print.  I'm running WinXP as a virtual machine via
>Parallels on a Mac.  I realize this is another layer of complication,
>but like I said, this exact same syntax file was working flawlessly
>for me under precisely these conditions for months until now.  That
>suggests that something went awry in SPSS, or Win, or perhaps
>Parallels, but I'm not having any other problems with Parallels to
>implicate it.
>
>Any ideas?  I've tried as much as I know. I was using files in a
>shared directory before, but I've tried them in a local directory on
>the C: drive with the same results.  I restarted Windows.  I've
>rebooted the Mac.  I've repaired disk permissions on the Mac.  I'm
>thinking I may install a new virtual machine under Parallels, with a
>clean copy of Win XP and SPSS and see if the problem exists there too.
>
>Thanks,
>Dennis
>

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

The University of Aberdeen is a charity registered in Scotland, No SC013683.

[D.R. Wahlgren]

>Date:    Sat, 31 Jul 2010 10:20:11 +0200
>From:    John F Hall <[hidden email]>
>Subject: Re: Buggy behavior
>
>
>Just a thought.  Shouldn't SPSS saved files be *.sav not *.sys?

We do our data entry in SPSS DE-II, so the resulting file is .sys.
Yes, it's an old DOS app, but it's still supported in the Windows
version (and you'll see it in the menu of file options in any open
file command via menus).  And my syntax file had been working just
fine with these same data files (albeit fewer cases in them) for
months prior to this.

Dennis

>   ----- Original Message -----=20
>   From: D.R. Wahlgren=20
>   To: [hidden email]=20
>   Sent: Saturday, July 31, 2010 7:31 AM
>   Subject: Re: Buggy behavior
>
>
>
>   Just as a follow up, I've also replicated the misbehavior on a
>   completely separate comptuer and yet another version of SPSS (v17).
>   This is on an actual PC running Win XP.
>
>   It seems more reliable (but not perfect) if I just run the whole lot
>   of syntax (the part below plus more that I did not copy here).  If I
>   run parts of the syntax it throws in random DATASET ACTIVATE commands
>   (and pastes them into the output file).  And if I click between the
>   output button in the task bar in between sets of commands, it often
>   changes the active dataset assignment, but without pasting the
>   command into the output file.
>
>   This has really shaken my confidence in the process, but since it's
>   happening in 3 different versions of SPSS on two different computers,
>   I think the only common denominator is my data files (.SYS).  Is
>   there some sort of corruption that could cause SPSS/Win to do this?
>
>   Dennis
>
>
>   >Date:    Tue, 27 Jul 2010 14:40:36 -0700
>   >From:    "D.R. Wahlgren" <[hidden email]>
>   >Subject: Buggy behavior
>   >
>   >Gang,
>   >I'm experiencing a problem with SPSS.  I suspect the problem lies in
>   >my computer, but I was wondering if anyone here was familiar with it.
>   >I run the same syntax file week after week and it just recently
>   >started to give me problems. Here is just one part of it:
>   >
>   =
>>************************************************************************=
>***.
>   >GET
>   >    FILE=3D'C:\NAPAtemp\NAPACHN1.SYS'.
>   >DATASET NAME DataSet1 WINDOW=3DFRONT.
>   >sort cases by napa_id.
>   >
>   =
>>************************************************************************=
>***.
>   >GET
>   >    FILE=3D'C:\NAPAtemp\NAPACHN2.SYS'.
>   >DATASET NAME DataSet2 WINDOW=3DFRONT.
>   >sort cases by napa_id.
>   >
>   =
>>************************************************************************=
>***.
>   >MATCH FILES /FILE=3D'DataSet1'
>   >   /RENAME (startime =3D d0)
>   >   /FILE=3D'DataSet2'
>   >   /BY napa_id
>   >   /DROP=3D d0.
>   >EXECUTE.
>   >
>   >SAVE OUTFILE=3D'C:\NAPAtemp\napachn_all.sav'
>   >    /COMPRESSED.
>   >DATASET NAME DataSet3 WINDOW=3DFRONT.
>   >
>   =
>>************************************************************************=
>***.
>   >
>   >SPSS is randomly changing the assignment of the active data set.  For
>   >example, it will insert a "DATASET ACTIVATE DataSet2." command after
>   >the MATCH FILES COMMAND.  This of course causes the subsequent SAVE
>   >command to save the wrong data file and assign it the wrong dataset
>   >name.
>   >
>   >I discovered some other, but related, misbehavior, when I was trying
>   >to pinpoint the above problem.  I stepped through the syntax command
>   >by command, and after I opened the first two data files, DataSet2 was
>   >the active file.  I then selected the output file in the task bar and
>   >saw that the green plus sign moved from the DataSet2 taskbar button
>   >to the DataSet1 taskbar button!
>   >
>   >Now, if I clicked on the DataSet2 taskbar button, it would switch
>   >back to the active data file, and I could then switch back to the
>   >output without it changing again.  I was able to repeat this whole
>   >set of behavior a few times.
>   >
>   >It does essentially (although not identically) the same things in
>   >SPSS Win v15 and v18.  The problem sometimes disappears as
>   >mysteriously as it appears in one version but not the other, or
>   >sometimes disappears from both versions and they both work fine.
>   >This happened just today. It's supremely frustrating.
>   >
>   >Now, for the fine print.  I'm running WinXP as a virtual machine via
>   >Parallels on a Mac.  I realize this is another layer of complication,
>   >but like I said, this exact same syntax file was working flawlessly
>   >for me under precisely these conditions for months until now.  That
>   >suggests that something went awry in SPSS, or Win, or perhaps
>   >Parallels, but I'm not having any other problems with Parallels to
>   >implicate it.
>   >
>   >Any ideas?  I've tried as much as I know. I was using files in a
>   >shared directory before, but I've tried them in a local directory on
>   >the C: drive with the same results.  I restarted Windows.  I've
>   >rebooted the Mac.  I've repaired disk permissions on the Mac.  I'm
>   >thinking I may install a new virtual machine under Parallels, with a
>   >clean copy of Win XP and SPSS and see if the problem exists there =
>too.
>   >
>   >Thanks,
>   >Dennis
>   >

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[John F Hall]

All my stuff was *.sys from 1972 to 1993 as well (SPSS-X 4 on a Vax cluster).  SPSS will read them as external files in single primes, but many of my archived files came as *.sys_1 to *.sys_n which displayed as Windows icons.  It was a while before I discoverd that all I had to do was delete the _n bit and/or add a .sav suffix et voilà!  Ditto for *.dat files (which Windows thinks are movies!) 

 

Incidentally, in

 

FILE=3D'C:\NAPAtemp\NAPACHN1.SYS'.

. . .should the 3D be there, or is my syntax rusty??

----- Original Message -----

From: [hidden email]

To: [hidden email]

Sent: Sunday, August 01, 2010 8:49 AM

Subject: Re: Buggy behavior

>Date:    Sat, 31 Jul 2010 10:20:11 +0200
>From:    John F Hall <[hidden email]>
>Subject: Re: Buggy behavior
>
>
>Just a thought.  Shouldn't SPSS saved files be *.sav not *.sys?

We do our data entry in SPSS DE-II, so the resulting file is .sys.
Yes, it's an old DOS app, but it's still supported in the Windows
version (and you'll see it in the menu of file options in any open
file command via menus).  And my syntax file had been working just
fine with these same data files (albeit fewer cases in them) for
months prior to this.

Dennis

>   ----- Original Message -----=20
>   From: D.R. Wahlgren=20
>   To: [hidden email]
>   Sent: Saturday, July 31, 2010 7:31 AM
>   Subject: Re: Buggy behavior
>
>
>
>   Just as a follow up, I've also replicated the misbehavior on a
>   completely separate comptuer and yet another version of SPSS (v17).
>   This is on an actual PC running Win XP.
>
>   It seems more reliable (but not perfect) if I just run the whole lot
>   of syntax (the part below plus more that I did not copy here).  If I
>   run parts of the syntax it throws in random DATASET ACTIVATE commands
>   (and pastes them into the output file).  And if I click between the
>   output button in the task bar in between sets of commands, it often
>   changes the active dataset assignment, but without pasting the
>   command into the output file.
>
>   This has really shaken my confidence in the process, but since it's
>   happening in 3 different versions of SPSS on two different computers,
>   I think the only common denominator is my data files (.SYS).  Is
>   there some sort of corruption that could cause SPSS/Win to do this?
>
>   Dennis
>
>
>   >Date:    Tue, 27 Jul 2010 14:40:36 -0700
>   >From:    "D.R. Wahlgren" <[hidden email]>
>   >Subject: Buggy behavior
>   >
>   >Gang,
>   >I'm experiencing a problem with SPSS.  I suspect the problem lies in
>   >my computer, but I was wondering if anyone here was familiar with it.
>   >I run the same syntax file week after week and it just recently
>   >started to give me problems. Here is just one part of it:
>   >
>   =
>>************************************************************************=
>***.
>   >GET
>   >    FILE=3D'C:\NAPAtemp\NAPACHN1.SYS'.
>   >DATASET NAME DataSet1 WINDOW=3DFRONT.
>   >sort cases by napa_id.
>   >
>   =
>>************************************************************************=
>***.
>   >GET
>   >    FILE=3D'C:\NAPAtemp\NAPACHN2.SYS'.
>   >DATASET NAME DataSet2 WINDOW=3DFRONT.
>   >sort cases by napa_id.
>   >
>   =
>>************************************************************************=
>***.
>   >MATCH FILES /FILE=3D'DataSet1'
>   >   /RENAME (startime =3D d0)
>   >   /FILE=3D'DataSet2'
>   >   /BY napa_id
>   >   /DROP=3D d0.
>   >EXECUTE.
>   >
>   >SAVE OUTFILE=3D'C:\NAPAtemp\napachn_all.sav'
>   >    /COMPRESSED.
>   >DATASET NAME DataSet3 WINDOW=3DFRONT.
>   >
>   =
>>************************************************************************=
>***.
>   >
>   >SPSS is randomly changing the assignment of the active data set.  For
>   >example, it will insert a "DATASET ACTIVATE DataSet2." command after
>   >the MATCH FILES COMMAND.  This of course causes the subsequent SAVE
>   >command to save the wrong data file and assign it the wrong dataset
>   >name.
>   >
>   >I discovered some other, but related, misbehavior, when I was trying
>   >to pinpoint the above problem.  I stepped through the syntax command
>   >by command, and after I opened the first two data files, DataSet2 was
>   >the active file.  I then selected the output file in the task bar and
>   >saw that the green plus sign moved from the DataSet2 taskbar button
>   >to the DataSet1 taskbar button!
>   >
>   >Now, if I clicked on the DataSet2 taskbar button, it would switch
>   >back to the active data file, and I could then switch back to the
>   >output without it changing again.  I was able to repeat this whole
>   >set of behavior a few times.
>   >
>   >It does essentially (although not identically) the same things in
>   >SPSS Win v15 and v18.  The problem sometimes disappears as
>   >mysteriously as it appears in one version but not the other, or
>   >sometimes disappears from both versions and they both work fine.
>   >This happened just today. It's supremely frustrating.
>   >
>   >Now, for the fine print.  I'm running WinXP as a virtual machine via
>   >Parallels on a Mac.  I realize this is another layer of complication,
>   >but like I said, this exact same syntax file was working flawlessly
>   >for me under precisely these conditions for months until now.  That
>   >suggests that something went awry in SPSS, or Win, or perhaps
>   >Parallels, but I'm not having any other problems with Parallels to
>   >implicate it.
>   >
>   >Any ideas?  I've tried as much as I know. I was using files in a
>   >shared directory before, but I've tried them in a local directory on
>   >the C: drive with the same results.  I restarted Windows.  I've
>   >rebooted the Mac.  I've repaired disk permissions on the Mac.  I'm
>   >thinking I may install a new virtual machine under Parallels, with a
>   >clean copy of Win XP and SPSS and see if the problem exists there =
>too.
>   >
>   >Thanks,
>   >Dennis
>   >

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[D.R. Wahlgren]

>Date:    Sun, 1 Aug 2010 13:23:32 +0200
>From:    John F Hall <[hidden email]>
>Subject: Re: Buggy behavior
>
>All my stuff was *.sys from 1972 to 1993 as well (SPSS-X 4 on a Vax =
>cluster).  SPSS will read them as external files in single primes, but =
>many of my archived files came as *.sys_1 to *.sys_n which displayed as =
>Windows icons.  It was a while before I discoverd that all I had to do =
>was delete the _n bit and/or add a .sav suffix et voil=E0!  Ditto for =
>*.dat files (which Windows thinks are movies!) =20
>
>Incidentally, in
>
>FILE=3D3D'C:\NAPAtemp\NAPACHN1.SYS'.
>. . .should the 3D be there, or is my syntax rusty??

The listserv seems to throw in a lot of junk.  I see hundreds of such
things in my digest version.  There are also several equal signs and
other strange bits (like the =E0 and =20 in your message above).

Dennis

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD