T.R | Title | User | Personal Name | Date | Lines |
---|
433.1 | Use CTRL/B | CESARE::EIJS | All in 1 Piece | Wed Apr 08 1992 11:17 | 15 |
|
Iwao,
VIEW TRACE is equivalent to CTRL/B. You wil see up to 1000 lines of
TRACE. The TRACE output will then indicate what can be done next:
Trace to Terminal
******************************************************************************
Trace information is discarded on returning to the menu (using RETURN/EXIT).
To prevent this, press GOLD F at the bottom to retain up to 1000 lines.
******************************************************************************
Ciao,
Simon
|
433.2 | | IOSG::ECHRISTIE | Eileen Christie | Thu Apr 09 1992 10:11 | 5 |
| Hi Iwao,
What in particular did you find difficult about the trace log? We're
currently eagerly gathering suggestions for a possible future release!
-eileen
|
433.3 | | SAC::JOYCE | Greasing the baseball bat... | Thu Apr 09 1992 11:02 | 18 |
|
Just a few comments on the new tracing facilities in V3.0.
Overall I find it a major improvement. The layout is a lot
more readable and easier to understand.
But, I do find it a bit of a pain sometimes that I can't see the trace
as things are happening. It is sometimes useful to see the trace in the
context of what the user is seeing.
I also find it annoying that DCL trace *does* come out immeadiately
when the rest does not. This makes it very hard to work out what is
going on. I can appreciate, however, the technical problems involved -
DCL tracing is only "set verify" after all (I think).
Andy
|
433.4 | Re-direct to SYS$OUTPUT | UTRTSC::BOSMAN | We're just sugar mice in the rain | Thu Apr 09 1992 11:21 | 9 |
| Andy,
� But, I do find it a bit of a pain sometimes that I can't see the trace
� as things are happening. It is sometimes useful to see the trace in the
� context of what the user is seeing.
You can re-direct the output to SYS$OUTPUT in the LOG field.
Sjaak.
|
433.5 | 2 questions | TKOV50::NAKAYAMA | I am OLYMPIC MAN ! | Fri Apr 10 1992 10:45 | 38 |
|
Thanks my friends .....
> Re:4
>
> You can re-direct the output to SYS$OUTPUT in the LOG field.
It's good information for me. I wanted to know this information !
> Re:2
>
> What in particular did you find difficult about the trace log? We're
> currently eagerly gathering suggestions for a possible future release!
Now, I am using ALL-IN-1 V3.0 everyday. And I have become quite
accustomed to see the Trace Log.
But, Now I have one question.
+--------------------------------------------------------------+
| Trace Flag Argument Form |
| |
| _ FLOW (xxxxxxxxxxxxxxxxxx) _ SCRIPT _______________ |
| _ SCRINP (xxxxxxxxxxxxxxxxxx) _ SYMBOL _______________ |
| _ IO (xxxxxxxxxxxxxxxxxx) _ FORM _______________ |
| | |
|
Q1) What is this field ? <--------------------+
and How use this field ?
Q2) Which manual explain about "How to use Trace Log" ?
Thanks your help.
Iwao
|
433.6 | Used for filtering | CESARE::EIJS | All in 1 Piece | Fri Apr 10 1992 14:03 | 23 |
|
Iwao,
> _ FORM _______________
These fields can be used to indicate which 'specific' forms, Scripts,
symbols etc. you want to include in the TRACE. E.g. indicating:
_ SCRIPT *CM_*________
means that the TRACE will only TRACE Scripts starting with 'CM_' and
not any others. It's a possibility to filter information.
Look at the documentation for:
Guide: Chapter 25, Problem solving
APR II: SET_TRACE function
That is, if you have the V3.0 docset.
Ciao,
Simon
|
433.7 | Confusing messages | GLOVES::ALLERTON | Steve Allerton 343-0205 | Wed May 20 1992 00:55 | 31 |
|
It seems that the SCRIPT filter isn't effective when there's a lot of
named data associated with the scripting language, as is the case at the
Customization Management main menu.
I've also been experiencing an annoying problem with VIEW TRACE :
After several trace operations, targeting the terminal memory buffer, I'll
eventually be unable to view the trace output that is written there...
Pressing CTRL/B returns the message -
"You can't use CTRL/B at this point"
Pressing CTRL/A returns the message -
"You can't use CTRL/A at this point"
Can anyone offer a guess as to why this happens?
Also, when I test an element in Customization Management, and the
element generates an error, the message returned is
"Error occurred during the DO mode script test.."
This is even the case when I'm testing a form that generates an error.
So if I'm testing a form, and the form calls a DO script, but its the
form that generates the error, the message that's returned leads me
to believe (wrongly) that the error is being generated from the
script that gets called. I hope I'm making sense...
Steve
|
433.8 | script tracing is pretty dumb | IOSG::ECHRISTIE | Eileen Christie | Wed May 20 1992 11:23 | 10 |
| Script tracing includes named data which doesnt support filtering .. so I agree,
it adds a lot of noise. I'll add a suggestion to the long list for a PFR.
If you are looking for something specific amongst the noise though, use the
text serach facility (type character string and hit FIND / GOLD FIND)
RE the problems with the CTRL keys, do any other CTRL keys work at that point?
What menu are you on when you try it?
-eileen
|
433.9 | Customized form DEFAULT | GLOVES::ALLERTON | Steve Allerton 343-0205 | Wed May 20 1992 15:34 | 6 |
|
Ooops...customized version of form DEFAULT was it.
Embarassed,
Steve
|
433.10 | The message is used in several tests... | CESARE::EIJS | All in 1 Piece | Wed May 27 1992 11:45 | 29 |
|
Steve,
> Also, when I test an element in Customization Management, and the
> element generates an error, the message returned is
>
> "Error occurred during the DO mode script test.."
Unfortunately I cannot give a plausible reason why this is the way it
is. When testing a BLP the message will refer to 'Error during MERGE
operation...' which is correct, but testing elements which call one of
the following test procedures will all return the message you
mentioned, together with an additional message (use GOLD/W) which
indicates the real problem:
CM_TEST_COM.SCP
CM_TEST_FDL.SCP
CM_TEST_FRM.SCP
CM_TEST_SCRIPT.SCP
The fact that the forms calls a DO procedure has nothing to do with the
final message. CM tests an element of a specific element type, and
that's it.
Sorry, but it looks like SPR time.
Ciao,
Simon
|