odd PH/VMS logical name behaviour

Lorry Litman LLitman@exchange.hsc.mb.ca
Thu, 11 Mar 2004 08:20:08 -0600


This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C40773.F4E37F80
Content-Type: text/plain;
	charset="iso-8859-1"

I have found the same result testing on PH7.10.G1 under both VMS7.2-2 and
VMS7.3-1.
It appears that we are also using the VMS default of case_blind and from the
result of the test I gather PH is not.
As pointed out, it does make a difference if you use double quotes when
defining the logical name. i.e. $define "test" quoteval is different from
$define test noquoteval
However, I assume because case_blind is on if you do $show logical t* you
will only see the last test logical name that you defined. If you change the
case i.e $def "Test" quoteval, and $def test noquoteval, and do $show
logical t* you will see both.
So it does appear that VMS does distinguish between the two on the define
side but not on the translate side.
To deassign the above you would have to $deas "test" and $deas test.
I've never used a double quote around a logical name so I can't say if this
behaved this way under previous versions of PH & VMS.
 
Lorry
 
 
-----Original Message-----
From: Hamilton, Allison [mailto:Allison.Hamilton@Cognos.COM]
Sent: Wednesday, March 10, 2004 3:09 PM
To: 'Lloyd, Gavin'; Chris Sharman
Cc: powerh-l@lists.sowder.com
Subject: RE: odd PH/VMS logical name behaviour



I'm trying to do too many things at once here - never name the logical name
the same as the value or the heading will confuse you.... sigh. You are
correct, if you define the logical name in mixed case, we upshift it and
will not see it since we don't do a case_blind search with trnlnm. If I use
Gavins test case I see the value set by the unquoted gav only.  Sorry for
the confusion.  

Allison Hamilton 

-----Original Message----- 
From: powerh-l-admin@lists.sowder.com 
[ mailto:powerh-l-admin@lists.sowder.com
<mailto:powerh-l-admin@lists.sowder.com> ]On Behalf Of Lloyd, Gavin 
Sent: Wednesday, March 10, 2004 3:50 PM 
To: Chris Sharman; Hamilton, Allison 
Cc: powerh-l@lists.sowder.com; doug.moffatt@cognos.com 
Subject: RE: odd PH/VMS logical name behaviour 


We've got OpenVMS V7.1-2 and I tried your code and it reported blank in 
quiz (i.e. did not get logical value). 

FYI - All our logicals are declared without the speechmarks (e.g. DEFINE 
LOG VAL not DEFINE "LOG" "VAL") which uppercases everything - so no 
problems. 

Did a small test as well if that helps: 

<--------------------- 
DCL> def gav gavval 
DCL> def "Gav" gavval2 
DCL> sho log G*   

(LNM$PROCESS_TABLE) 

  "GAV" = "GAVVAL" 
  "Gav" = "GAVVAL2" 
----------------------> 

Regards, 
Gavin Lloyd - FM Global. 


-----Original Message----- 
From: powerh-l-admin@lists.sowder.com 
[ mailto:powerh-l-admin@lists.sowder.com
<mailto:powerh-l-admin@lists.sowder.com> ] On Behalf Of Chris Sharman 
Sent: 10 March 2004 17:11 
To: Hamilton, Allison 
Cc: powerh-l@lists.sowder.com; doug.moffatt@cognos.com 
Subject: Re: odd PH/VMS logical name behaviour 


Hamilton, Allison wrote: 
> Chris,  I think this may be a difference in behaviour between 7.2 and 
> 7.3 of OpenVMS because I tried this in 7.10G3, 7.10G5 and 8.30D4 (all 
on 
> 6.2 of OpenVMS) and 8.40D on 7.2-1 and 8.40D on 7.2-2 of OpenVMS and 
> they retrieve the value of Test correctly.  I don't have a 7.3 OpenVMS 

> system available to me yet, so I can't test it there.  Perhaps there 
is 
> a setting on the logical or in your environment that would affect the 
> default behaviour of the logical translation function (trnlnm) that is 

> available in 7.3??? 
> 
> I know that OpenVMS is moving more and more into the Posix world which 
> requires case sensitivity, so more things behave somewhat differently. 

> I don't know if this one is an intentional change or not on the part 
of 
> OpenVMS.  Usually they don't change the default behaviour of things so 

> drastically, but who knows. 

Curious. Nothing I'm aware of - DCL has always uppercased everything, 
and still does, unless you enable the mixed-case behaviour which goes 
with the ODS5 file system, which we haven't. 

> Prior to 7.3 retrieval using uppercase always retrieved the correct 
> value due to the basically uppercase or case insensitive nature of the 

> VMS environment.  PowerHouse is built with this assumption currently, 
> and generally upcases just about everything coming in to ensure it can 

> find it. 

sys$trnlnm (and lib$get_logical) have always had a LNM$M_CASE_BLIND 
option, but it has never been the default behaviour, and still isn't. 
DCL f$trnlnm offers case-blind by default, and case-sensitive 
optionally. The older (obsolete since about V5) sys$trnlog and 
associated lib$ routine may do case-blind by default - I don't recall. 

I don't think VMS has ever seen a non-compatible system service change 
like you suggest. 
Anyone got access to VMS 7.3 and an earlier version to confirm/deny ? 

> Allison Hamilton 
> Senior Software Engineer 
> Cognos Inc. 
> Any opinions expressed within this message are mine alone and my 
> employer cannot be held accountable for them  :-} 
> 
> -----Original Message----- 
> From: powerh-l-admin@lists.sowder.com 
> [ mailto:powerh-l-admin@lists.sowder.com
<mailto:powerh-l-admin@lists.sowder.com> ]On Behalf Of Chris Sharman 
> Sent: Wednesday, March 10, 2004 6:30 AM 
> To: PowerH List 
> Subject: odd PH/VMS logical name behaviour 
> 
> 
> I've got a mixed-case logical name: 
> $ def "Test" "MyTest" 
> 
> $ quiz 
> Qz> def mytest char*10 = getsystemval("Test") 
> Qz> rep mytest 
> 
> Reports it as blank - presumably it doesn't find it. 
> Defining it as uppercase works fine. 
> 
> VMS logical names can be defined mixed-case (but are usually 
> uppercase). They can be fetched either case-sensitive or case-blind. 
> 
> Powerhouse would appear to be uppercasing the supplied string, then 
> doing a case-sensitive search - not the optimum behaviour ! Be nice if 

> it used the string I gave it, and probably set the case-blind flag to 
> avoid breaking most existing stuff. 
> 
> VMS 7.3, PH 7.10G3 


----------------------------------------------------------------------- 

Any views expressed in this message are those of the sender and not 
necessarily those of CCA Group.  The unauthorized use, disclosure, 
copying or alteration of this message is forbidden.  The contents of 
this message may be confidential and/or privileged, copyright CCA Group 
and are intended solely for the use of the individual or entity to whom 
they are addressed.  Whilst this message has been scanned, CCA Group 
cannot guarantee that it is virus free or compatible with your systems 
and accepts no responsibility for any loss or damage arising from its 
use. The recipient is advised to run their own anti-virus software. If 
you receive this message in error please contact 
postmaster@ccagroup.co.uk immediately, destroy any copies and delete it 
from your computer systems. 


= = = = = = = = = = = = = = = = = = = = = = = = = = = = 
Mailing list: powerh-l@lists.sowder.com 
Subscribe: "subscribe" in message body to powerh-l-request@lists.sowder.com 
Unsubscribe: "unsubscribe <password>" in message body to
powerh-l-request@lists.sowder.com 
http://lists.sowder.com/mailman/listinfo/powerh-l
<http://lists.sowder.com/mailman/listinfo/powerh-l>  
This list is closed, thus to post to the list you must be a subscriber. 

This message may contain privileged and/or confidential information.  If you
have received this e-mail in error or are not the intended recipient, you
may not use, copy, disseminate or distribute it; do not open any
attachments, delete it immediately from your system and notify the sender
promptly by e-mail that you have done so.  Thank you.


------_=_NextPart_001_01C40773.F4E37F80
Content-Type: text/html;
	charset="iso-8859-1"

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<TITLE>RE: odd PH/VMS logical name behaviour</TITLE>

<META content="MSHTML 6.00.2800.1400" name=GENERATOR></HEAD>
<BODY>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=409244713-11032004>I have 
found the same result testing on PH7.10.G1 under both VMS7.2-2 and 
VMS7.3-1.</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=409244713-11032004>It 
appears that we are also using the VMS default of case_blind and from the result 
of the test I gather PH is not.</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=409244713-11032004>As 
pointed out, it does make a difference if you use double quotes when defining 
the logical name. i.e. $define "test" quoteval is different from $define test 
noquoteval</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN 
class=409244713-11032004>However, I assume because case_blind is on if you do 
$show logical t* you will only see the last test logical name that you defined. 
If you change the case i.e $def "Test" quoteval, and $def test noquoteval, and 
do $show logical t* you will&nbsp;see both.</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=409244713-11032004>So it 
does appear that VMS does distinguish between the two on the define side but not 
on the translate side.</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=409244713-11032004>To 
deassign the above you would have to $deas "test" and $deas 
test.</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=409244713-11032004>I've 
never used a double quote around a logical name so I can't say if this behaved 
this way under previous versions of PH &amp; VMS.</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN 
class=409244713-11032004></SPAN></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN 
class=409244713-11032004>Lorry</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN 
class=409244713-11032004></SPAN></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN 
class=409244713-11032004></SPAN></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN 
class=409244713-11032004></SPAN></FONT><FONT face=Tahoma size=2>-----Original 
Message-----<BR><B>From:</B> Hamilton, Allison 
[mailto:Allison.Hamilton@Cognos.COM]<BR><B>Sent:</B> Wednesday, March 10, 2004 
3:09 PM<BR><B>To:</B> 'Lloyd, Gavin'; Chris Sharman<BR><B>Cc:</B> 
powerh-l@lists.sowder.com<BR><B>Subject:</B> RE: odd PH/VMS logical name 
behaviour<BR><BR></DIV></FONT>
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">
  <P><FONT size=2>I'm trying to do too many things at once here - never name the 
  logical name the same as the value or the heading will confuse you.... sigh. 
  You are correct, if you define the logical name in mixed case, we upshift it 
  and will not see it since we don't do a case_blind search with trnlnm. If I 
  use Gavins test case I see the value set by the unquoted gav only.&nbsp; Sorry 
  for the confusion.&nbsp; </FONT></P>
  <P><FONT size=2>Allison Hamilton</FONT> </P>
  <P><FONT size=2>-----Original Message-----</FONT> <BR><FONT size=2>From: 
  powerh-l-admin@lists.sowder.com</FONT> <BR><FONT size=2>[<A 
  href="mailto:powerh-l-admin@lists.sowder.com">mailto:powerh-l-admin@lists.sowder.com</A>]On 
  Behalf Of Lloyd, Gavin</FONT> <BR><FONT size=2>Sent: Wednesday, March 10, 2004 
  3:50 PM</FONT> <BR><FONT size=2>To: Chris Sharman; Hamilton, Allison</FONT> 
  <BR><FONT size=2>Cc: powerh-l@lists.sowder.com; doug.moffatt@cognos.com</FONT> 
  <BR><FONT size=2>Subject: RE: odd PH/VMS logical name behaviour</FONT> 
</P><BR>
  <P><FONT size=2>We've got OpenVMS V7.1-2 and I tried your code and it reported 
  blank in</FONT> <BR><FONT size=2>quiz (i.e. did not get logical value).</FONT> 
  </P>
  <P><FONT size=2>FYI - All our logicals are declared without the speechmarks 
  (e.g. DEFINE</FONT> <BR><FONT size=2>LOG VAL not DEFINE "LOG" "VAL") which 
  uppercases everything - so no</FONT> <BR><FONT size=2>problems.</FONT> </P>
  <P><FONT size=2>Did a small test as well if that helps:</FONT> </P>
  <P><FONT size=2>&lt;---------------------</FONT> <BR><FONT size=2>DCL&gt; def 
  gav gavval</FONT> <BR><FONT size=2>DCL&gt; def "Gav" gavval2</FONT> <BR><FONT 
  size=2>DCL&gt; sho log G*&nbsp;&nbsp; </FONT></P>
  <P><FONT size=2>(LNM$PROCESS_TABLE)</FONT> </P>
  <P><FONT size=2>&nbsp; "GAV" = "GAVVAL"</FONT> <BR><FONT size=2>&nbsp; "Gav" = 
  "GAVVAL2"</FONT> <BR><FONT size=2>----------------------&gt;</FONT> </P>
  <P><FONT size=2>Regards,</FONT> <BR><FONT size=2>Gavin Lloyd - FM 
  Global.</FONT> </P><BR>
  <P><FONT size=2>-----Original Message-----</FONT> <BR><FONT size=2>From: 
  powerh-l-admin@lists.sowder.com</FONT> <BR><FONT size=2>[<A 
  href="mailto:powerh-l-admin@lists.sowder.com">mailto:powerh-l-admin@lists.sowder.com</A>] 
  On Behalf Of Chris Sharman</FONT> <BR><FONT size=2>Sent: 10 March 2004 
  17:11</FONT> <BR><FONT size=2>To: Hamilton, Allison</FONT> <BR><FONT 
  size=2>Cc: powerh-l@lists.sowder.com; doug.moffatt@cognos.com</FONT> <BR><FONT 
  size=2>Subject: Re: odd PH/VMS logical name behaviour</FONT> </P><BR>
  <P><FONT size=2>Hamilton, Allison wrote:</FONT> <BR><FONT size=2>&gt; 
  Chris,&nbsp; I think this may be a difference in behaviour between 7.2 
  and</FONT> <BR><FONT size=2>&gt; 7.3 of OpenVMS because I tried this in 
  7.10G3, 7.10G5 and 8.30D4 (all</FONT> <BR><FONT size=2>on </FONT><BR><FONT 
  size=2>&gt; 6.2 of OpenVMS) and 8.40D on 7.2-1 and 8.40D on 7.2-2 of OpenVMS 
  and </FONT><BR><FONT size=2>&gt; they retrieve the value of Test 
  correctly.&nbsp; I don't have a 7.3 OpenVMS</FONT> </P>
  <P><FONT size=2>&gt; system available to me yet, so I can't test it 
  there.&nbsp; Perhaps there</FONT> <BR><FONT size=2>is </FONT><BR><FONT 
  size=2>&gt; a setting on the logical or in your environment that would affect 
  the </FONT><BR><FONT size=2>&gt; default behaviour of the logical translation 
  function (trnlnm) that is</FONT> </P>
  <P><FONT size=2>&gt; available in 7.3???</FONT> <BR><FONT size=2>&gt; 
  </FONT><BR><FONT size=2>&gt; I know that OpenVMS is moving more and more into 
  the Posix world which</FONT> <BR><FONT size=2>&gt; requires case sensitivity, 
  so more things behave somewhat differently.</FONT> </P>
  <P><FONT size=2>&gt; I don't know if this one is an intentional change or not 
  on the part</FONT> <BR><FONT size=2>of </FONT><BR><FONT size=2>&gt; 
  OpenVMS.&nbsp; Usually they don't change the default behaviour of things 
  so</FONT> </P>
  <P><FONT size=2>&gt; drastically, but who knows.</FONT> </P>
  <P><FONT size=2>Curious. Nothing I'm aware of - DCL has always uppercased 
  everything, </FONT><BR><FONT size=2>and still does, unless you enable the 
  mixed-case behaviour which goes </FONT><BR><FONT size=2>with the ODS5 file 
  system, which we haven't.</FONT> </P>
  <P><FONT size=2>&gt; Prior to 7.3 retrieval using uppercase always retrieved 
  the correct</FONT> <BR><FONT size=2>&gt; value due to the basically uppercase 
  or case insensitive nature of the</FONT> </P>
  <P><FONT size=2>&gt; VMS environment.&nbsp; PowerHouse is built with this 
  assumption currently, </FONT><BR><FONT size=2>&gt; and generally upcases just 
  about everything coming in to ensure it can</FONT> </P>
  <P><FONT size=2>&gt; find it.</FONT> </P>
  <P><FONT size=2>sys$trnlnm (and lib$get_logical) have always had a 
  LNM$M_CASE_BLIND </FONT><BR><FONT size=2>option, but it has never been the 
  default behaviour, and still isn't.</FONT> <BR><FONT size=2>DCL f$trnlnm 
  offers case-blind by default, and case-sensitive </FONT><BR><FONT 
  size=2>optionally. The older (obsolete since about V5) sys$trnlog and 
  </FONT><BR><FONT size=2>associated lib$ routine may do case-blind by default - 
  I don't recall.</FONT> </P>
  <P><FONT size=2>I don't think VMS has ever seen a non-compatible system 
  service change </FONT><BR><FONT size=2>like you suggest.</FONT> <BR><FONT 
  size=2>Anyone got access to VMS 7.3 and an earlier version to confirm/deny 
  ?</FONT> </P>
  <P><FONT size=2>&gt; Allison Hamilton</FONT> <BR><FONT size=2>&gt; Senior 
  Software Engineer</FONT> <BR><FONT size=2>&gt; Cognos Inc.</FONT> <BR><FONT 
  size=2>&gt; Any opinions expressed within this message are mine alone and 
  my</FONT> <BR><FONT size=2>&gt; employer cannot be held accountable for 
  them&nbsp; :-}</FONT> <BR><FONT size=2>&gt; </FONT><BR><FONT size=2>&gt; 
  -----Original Message-----</FONT> <BR><FONT size=2>&gt; From: 
  powerh-l-admin@lists.sowder.com </FONT><BR><FONT size=2>&gt; [<A 
  href="mailto:powerh-l-admin@lists.sowder.com">mailto:powerh-l-admin@lists.sowder.com</A>]On 
  Behalf Of Chris Sharman</FONT> <BR><FONT size=2>&gt; Sent: Wednesday, March 
  10, 2004 6:30 AM</FONT> <BR><FONT size=2>&gt; To: PowerH List</FONT> <BR><FONT 
  size=2>&gt; Subject: odd PH/VMS logical name behaviour</FONT> <BR><FONT 
  size=2>&gt; </FONT><BR><FONT size=2>&gt; </FONT><BR><FONT size=2>&gt; I've got 
  a mixed-case logical name:</FONT> <BR><FONT size=2>&gt; $ def "Test" 
  "MyTest"</FONT> <BR><FONT size=2>&gt; </FONT><BR><FONT size=2>&gt; $ 
  quiz</FONT> <BR><FONT size=2>&gt; Qz&gt; def mytest char*10 = 
  getsystemval("Test")</FONT> <BR><FONT size=2>&gt; Qz&gt; rep mytest</FONT> 
  <BR><FONT size=2>&gt; </FONT><BR><FONT size=2>&gt; Reports it as blank - 
  presumably it doesn't find it.</FONT> <BR><FONT size=2>&gt; Defining it as 
  uppercase works fine.</FONT> <BR><FONT size=2>&gt; </FONT><BR><FONT 
  size=2>&gt; VMS logical names can be defined mixed-case (but are usually 
  </FONT><BR><FONT size=2>&gt; uppercase). They can be fetched either 
  case-sensitive or case-blind.</FONT> <BR><FONT size=2>&gt; </FONT><BR><FONT 
  size=2>&gt; Powerhouse would appear to be uppercasing the supplied string, 
  then </FONT><BR><FONT size=2>&gt; doing a case-sensitive search - not the 
  optimum behaviour ! Be nice if</FONT> </P>
  <P><FONT size=2>&gt; it used the string I gave it, and probably set the 
  case-blind flag to </FONT><BR><FONT size=2>&gt; avoid breaking most existing 
  stuff.</FONT> <BR><FONT size=2>&gt; </FONT><BR><FONT size=2>&gt; VMS 7.3, PH 
  7.10G3</FONT> </P><BR>
  <P><FONT 
  size=2>-----------------------------------------------------------------------</FONT> 
  </P>
  <P><FONT size=2>Any views expressed in this message are those of the sender 
  and not</FONT> <BR><FONT size=2>necessarily those of CCA Group.&nbsp; The 
  unauthorized use, disclosure,</FONT> <BR><FONT size=2>copying or alteration of 
  this message is forbidden.&nbsp; The contents of</FONT> <BR><FONT size=2>this 
  message may be confidential and/or privileged, copyright CCA Group</FONT> 
  <BR><FONT size=2>and are intended solely for the use of the individual or 
  entity to whom</FONT> <BR><FONT size=2>they are addressed.&nbsp; Whilst this 
  message has been scanned, CCA Group</FONT> <BR><FONT size=2>cannot guarantee 
  that it is virus free or compatible with your systems</FONT> <BR><FONT 
  size=2>and accepts no responsibility for any loss or damage arising from 
  its</FONT> <BR><FONT size=2>use. The recipient is advised to run their own 
  anti-virus software. If</FONT> <BR><FONT size=2>you receive this message in 
  error please contact</FONT> <BR><FONT size=2>postmaster@ccagroup.co.uk 
  immediately, destroy any copies and delete it</FONT> <BR><FONT size=2>from 
  your computer systems.</FONT> </P><BR>
  <P><FONT size=2>= = = = = = = = = = = = = = = = = = = = = = = = = = = =</FONT> 
  <BR><FONT size=2>Mailing list: powerh-l@lists.sowder.com</FONT> <BR><FONT 
  size=2>Subscribe: "subscribe" in message body to 
  powerh-l-request@lists.sowder.com</FONT> <BR><FONT size=2>Unsubscribe: 
  "unsubscribe &lt;password&gt;" in message body to 
  powerh-l-request@lists.sowder.com</FONT> <BR><FONT size=2><A 
  href="http://lists.sowder.com/mailman/listinfo/powerh-l" 
  target=_blank>http://lists.sowder.com/mailman/listinfo/powerh-l</A></FONT> 
  <BR><FONT size=2>This list is closed, thus to post to the list you must be a 
  subscriber.</FONT> </P>
  <P><FONT size=2>This message may contain privileged and/or confidential 
  information.&nbsp; If you have received this e-mail in error or are not the 
  intended recipient, you may not use, copy, disseminate or distribute it; do 
  not open any attachments, delete it immediately from your system and notify 
  the sender promptly by e-mail that you have done so.&nbsp; Thank 
  you.</FONT></P></BLOCKQUOTE></BODY></HTML>

------_=_NextPart_001_01C40773.F4E37F80--