[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference bulova::decw_jan-89_to_nov-90

Title:DECWINDOWS 26-JAN-89 to 29-NOV-90
Notice:See 1639.0 for VMS V5.3 kit; 2043.0 for 5.4 IFT kit
Moderator:STAR::VATNE
Created:Mon Oct 30 1989
Last Modified:Mon Dec 31 1990
Last Successful Update:Fri Jun 06 1997
Number of topics:3726
Total number of notes:19516

3439.0. "debugger command window embedding <cr> on edit" by CSC32::B_BRADACH () Mon Oct 08 1990 16:56

    Using the decwindows interface to the debugger, if you edit a command
    line in the command window, hit the <cr> in the middle of the text, the
    <cr> gets entered into the simple text widget.
    
    I entered a reply in the debug notes conference on this question, as
    you can see the response is that this is a bug in decwindows. Since a 
    simple text widget will by deafault put a <cr> into the text I am not
    sure why someone thinks that this is a decwindows bug.  Anyone
    have any ideas????  If not, I will submit an spr on this.
    
    thanks,
    bernie
    csc/cs
    
    =======================  extracted from debug conference ================
    
            <<< TURRIS::TURRIS$DUA18:[NOTES$LIBRARY]DEBUG.NOTE;4 >>>
                                   -< DEBUG >-
================================================================================
Note 694.0        COMMAND window and command line processing ?s        6 replies
CSC32::M_TURNER                                      19 lines  16-OCT-1989 09:36
--------------------------------------------------------------------------------
    It appears that the behavior of the decwindows flavor of the debugger
    in command mode is inconsistent with the expected behavior. 
    Specifically, if I bring up the COMMAND window, and at a the DBG>
    prompt enter a command in which I make a mistake, then recall the
    command with the up arrow and position to the character in error, and
    at this point press RETURN, only the characters from beginning of the
    line to the current cursor position are accepted, rather than the whole
    line, which is the way everything else works.  It doesn't seem like
    proper behavior for me to have to reposition to the end of line again
    before pressing the RETURN key so that the whole command will be
    accepted.
    
    Also, is there a way to have the COMMAND window come up automatically
    when the debugger comes up, rather than having to go into customize and
    select the SHOW COMMAND option?
    
    Thanks for the help,
    
    Mark
================================================================================
Note 694.1        COMMAND window and command line processing ?s           1 of 6
TLE::ROUTLEY "Kevin Routley - VMS DEBUG"              9 lines  16-OCT-1989 15:45
--------------------------------------------------------------------------------
re .0:

The cursor position bug is a problem in DECWindows, and we have been 
promised that it has been fixed in DW V2.

To get the command region to come up by default, create a DBG$INIT file with
the command DISPLAY PROMPT in it.

Kevin
================================================================================
Note 694.2        COMMAND window and command line processing ?s           2 of 6
CSC32::M_TURNER                                       8 lines  17-OCT-1989 09:45
                       -< Not fixed yet in current FT. >-
--------------------------------------------------------------------------------
    I just tried this again using the debugger that comes with the T5.3-462
    VMS system and the same problem still exists on this.  I hope it will
    be fixed in the next fieldtest version that comes out!  That would be
    great.  
    
    Thanks for the DISPLAY PROMPT suggestion.
    
    Mark
================================================================================
Note 694.3        COMMAND window and command line processing ?s           3 of 6
CSC32::M_TURNER                                       7 lines  23-OCT-1989 10:09
                           -< Any further comments? >-
--------------------------------------------------------------------------------
    Could someone comment on the fact that this problem is still around
    with the DECwindows version 2.0 (FT) that comes with the T5.3-462
    system?  Do we know that indeed it has been/will be fixed in V2.0 of
    DECwindows?  Are we talking 5.3 of VMS or a 'future' release?
    
    Thanks
    Mark
================================================================================
Note 694.4        COMMAND window and command line processing ?s           4 of 6
TLE::ROUTLEY "Kevin Routley - VMS DEBUG"              6 lines  23-OCT-1989 16:29
--------------------------------------------------------------------------------
I do not know.  All I have been told is that this bug was fixed in DW V2.1
(release date unknown) and may have already been fixed in V2.  The customer
will probably have to wait until V5.3 FRS to see if its truly fixed.

I'm sorry, but this one is out of our hands ...
Kevin
================================================================================
Note 694.5        COMMAND window and command line processing ?s           5 of 6
CSC32::B_BRADACH                                      8 lines   5-OCT-1990 17:01
                       -< still can't re-edit commands >-
--------------------------------------------------------------------------------
    I just got a call about this, VMS5.4 and the command interface is still
    putting the <cr> into the command.  Any ideas when/if this will be
    fixed????.
    
    bernie
    
    
    
================================================================================
Note 694.6        COMMAND window and command line processing ?s           6 of 6
TLE::ROUTLEY "Kevin Routley - VAX C RTL"              5 lines   8-OCT-1990 12:25
--------------------------------------------------------------------------------
re .5:
We don't know.  If you are interested in an answer, the DECWindows folks are
the ones to contact, since its (as far as we know) a DECWindows bug.

Kevin
T.RTitleUserPersonal
Name
DateLines
3439.1PSW::WINALSKICareful with that VAX, EugeneMon Oct 08 1990 18:438
RE: .0

>   Anyone have any ideas????

Yes.  Contact the DECwindows developers directly so that you get a definitive
answer.  Don't rely on this or other NOTES conferences.

--PSW
3439.2spr on the wayCSC32::B_BRADACHTue Oct 09 1990 10:574
    Thanks an SPR for this problem will be generated.
    
    bernie
    csc/cs
3439.3Caught in the vortexTOOLEY::B_WACKERTue Oct 09 1990 11:4612
>Yes.  Contact the DECwindows developers directly so that you get a definitive
>answer.  Don't rely on this or other NOTES conferences.

Should we use an SPR which takes 6 months to 2 years for a reply, a 
CLD which is faster but heavy handed, or CSSE who probably won't EVER 
reply?  I'm sure the customer will be favorably impressed when the 
answer comes after the DEC world is on MOTIF and it doesn't make any 
difference, anyway!

Paul, please come skiing this winter and visit us.

Bruce
3439.4PSW::WINALSKICareful with that VAX, EugeneTue Oct 09 1990 16:5715
RE: .3

I said *directly*.  The developers have electronic mail addresses.

The main point is that answers in NOTES conferences aren't always either timely
or accurate, since they represent opinions or conjectures of whoever responds
to the question, not necessarily the official answer from the development group.
I think it's especially dangerous to rely on what's said in NOTES conferences
for problems such as this, where it's not clear in which component (DEBUG or
DECwindows) the problem lies.

Customers pay good money for support services and deserve official answers,
not WAGs.

--PSW
3439.5Direct not really acceptable nowTOOLEY::B_WACKERTue Oct 09 1990 18:0213
>The developers have electronic mail addresses.

But I don't see THEM inviting our queries.  Until they do then we 
really can't afford to take the chance of alienating them.

>Customers pay good money for support services and deserve official
>answers, not WAGs.

I couldn't agree more, but notes is the best we've got now.  How you'd 
like to volunteer to talk to engineering for us or get them to share 
your feelings about direct contact? ;-)

Bruce
3439.6LESLIE::LESLIESynclaviered NoterWed Oct 10 1990 05:106
    CSSE don't ever reply? Documented examples offline to me please.
    
    thanks
    
    andy
    csse/vms
3439.7Please use the QAR systemR2ME2::VANGILDERJim V., DECwindows ToolkitsWed Oct 10 1990 11:1839
Sorry, we (the  DECwindows toolkit developers) do read this notes
conference, but we don't always have time to respond, particularly
when we're near a major baselevel.  Since the DEBUG developers said
this had been reported to the toolkit team and we said we'd fix it,
I assumed it was on our list, or already fixed in some baselevel.
However, I can't find it in our list of open QARs.  Assuming this
is still a problem, please enter a QAR for it, and we'll look at it.
Keep in mind, though, that at this point, the XUI toolkit is
functionally frozen, and we're only addressing major problems.  In
particular, if the same problem also exists in the Motif toolkit, 
we may decide NOT to fix it unless OSF agrees it's a bug.

This is one of the problems with "contacting the developers directly".
Sending mail to one of the developers has the chance of getting lost
or forgotten, particularly when project leadership changes hands.  The
volume of MY mail, at least, is high enough that I've got a two or 
three month backlog of mail messages I should go back and look at
again, but the 'crisis du jour' always gets priority.

QARs get sent directly to me.  SPRs and CLDs get the CSSE support
people involved, and may not get to the developers for some time.
So QARs are the most efficient method of entering problem reports.


Problems with the DECwindows toolkit which are discovered internally
should be reported using the QAR system on TRIFID.

If you don't have a QAR account, you can do the following:

$ SET HOST TRIFID

Username: QAR_INTERNAL
Password: QAR


The QAR system will list the available databases.  Please use either
DECW-V3-INTERNAL (if you're running the DECwindows V3 baselevels) or
DECWINDOWS-IFT (for any older versions).

3439.8If support channels are broken - get them FIXED!!!IO::MCCARTNEYJames T. McCartney III - DTN 381-2244 ZK02-2/N24Wed Oct 10 1990 11:2431
RE: Support problems.

>> Should we use an SPR which takes 6 months to 2 years for a reply, a 
>> CLD which is faster but heavy handed, or CSSE who probably won't EVER 
>> reply?  I'm sure the customer will be favorably impressed when the 
>> answer comes after the DEC world is on MOTIF and it doesn't make any 
>> difference, anyway!

If the CLD and SPR process is badly broken, as is generally believed these
days, then why not take a different tack to the problem and see if you can't
get that FIXED! Surely your management would like to hear about the problems
you're having in getting answeres -- we have the same problems getting the
questions. The disconnect isn't between engineering and the front-line support
groups but (in my opinion) between the support groups and their backup channels.

The only way to get this fixed is to let your management know that it is 
a critical issue. Scream LOUDLY if you have to, but don't just sit there.


>> But I don't see THEM inviting our queries.  Until they do then we 
>> really can't afford to take the chance of alienating them.

I hope you don't believe this! I've personally corresponded with several 
groups to whom I've introduced myself in the mail message. On every occasion,
bar none, I've gotten a reasonable response -- believe it or not, people still
try to do the right thing around here. Some times the answer has been "Submit
a QAR -- that really is a problem." but the reason is that people track the 
official problem reporting channels.


James
3439.9LESLIE::LESLIEAndy Leslie, CSSE/VMSWed Oct 10 1990 12:004
    I agree, have been contacted offline and will persue this vigorously.
    
    
    /andy/
3439.10How about questions?TOOLEY::B_WACKERWed Oct 10 1990 12:2048
>The only way to get this fixed is to let your management know that it is 
>a critical issue. Scream LOUDLY if you have to, but don't just sit there.

Actually, we've all taken turns screaming loudly for the four years 
(and four managers!) since I've been here, all to no avail.  CSSE
isn't (and probably couldn't be) staffed to do the job, but they won't
give up the responsibility, so that remains the "official" channel. 
It probably won't change until a few customers go to K.O.  I'm pretty
sure the U S Area CSC head brought it up at the software woods
meeting, but haven't seen any change.  Part of the problem is that 
very few of our managers know anything about software.  Most come from 
field service or manufacturing.

>>> But I don't see THEM inviting our queries.  Until they do then we 
>>> really can't afford to take the chance of alienating them.

>I hope you don't believe this!

Yes, and no.  Unfortunately, there is about 800 people in this 
building doing software support.  If the channels were open without 
controls on this end then we could definitely paralyze engineering 
with the volume of requests.  I do correspond and call developers with 
whom I've developed a relationship and try to be the channel for the 
center to communicate with them.  If we just opened it up I frankly don't 
think we would exercise enough restraint.  Of course, we could set up 
people to channel it, but that hasn't been done on any widespread or 
formal level.

>believe it or not, people still try to do the right thing around here.

I agree with that, but, especially with a project as large as DW, it is
hard to know who to contact.  Between our turnover and yours the 
informal approaches are pretty vulnerable.

From what Vic said it sounds like the best approach might be to submit 
a QAR.  They are tracked and assigned to the proper person.  Is it 
OK for us to use the QAR system for backup query support?  Generally, 
we've been told QARs are for internal bug reporting only, but it seems
like it might be the most efficient mechanism for customer questions
we can't answer. 

Note, that the spr system appears to be improving.  At least I know of 
one SPR that got to engineering 3 days after we were through with it.  
For non-critical problems it's probably ok.  CLDs do a pretty good 
job for the critical ones.  Our big hole now is query support and we'd 
welcome any suggestions you have on how to do that more efficiently.

Bruce
3439.11we have tried to change the process - in vainCSC32::B_BRADACHWed Oct 10 1990 16:5251
    re: .7
    Jim:
    	Thank you for the very helpful information.  
    
    re:  responses regarding CSSE.
    
    The following is one of the canned responses that we receive from
    CSSE when we follow the process. I have removed the name of the person
    who sent this.
    
From:	VMSSG::VMSINQ       29-SEP-1989 11:17:05.04
To:	CSC32::B_BRADACH
CC:	
Subj:	RE: MULTIPLE APPLICATION CONTEXTS GETTING ERROR.

Bernie,

Your call has been received and logged under number 1501.

We will respond to your request as quickly as possible.  Our
priorities for addressing problems are: 1) CLD's, 2) SPR's
and 3) Query Support.

We appreciate your patience and will make every effort to respond
to your request.

Regards,

=====================================================================
    
    Please note the order in which queries from the csc are responded
    to!!!!
    
    It doesn't take a whole lot of thought process to figure out that
    if our queries were answered in a timely manner, and before sprs that
    maybe we could avoid generating some of these sprs!!!  WOW, What
    a concept.
    
    As Bruce stated above our mangement and the center management are 
    totally and very painfully aware that the process doesn't work. We
    continually complain about it, believe me.
    
    There are some people at the CSSE who have been extremely helpful,
    namely, Andy Mermell, and Dave Haberland concerning decwindows/motif
    and various related programming issues.
    
    bernie
    csc/cs
    decwindows programming/language support team