[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

3339.0. "Another Logicraft Problem" by WRKSYS::GOLDBERG (Marshall R. Goldberg, Workstations) Thu Sep 13 1990 18:59

    Here is a message I have received from our friends at Logicraft.
    Any comments on their observations?
    
    Marshall
    
    
From:	DECUAC::"decvax!logicraft!porshe.dnet!mark" 12-SEP-1990 17:42:25.13
To:	decwrl!BILBO.dec.com!"decvax!decuac!wrksys.enet!goldberg" 
CC:	
Subj:	QAR material? 

CRITICAL...URGENT...QAR...CRITICAL...URGENT...QAR...CRITICAL...URGENT...QAR...

	When we running 386Ware on VAXstation 3100's (either SPX or GPX)
we frequently get the following messages:

	Xlib: sequence lost (0x11dee > 0x1def) in reply type 0x5!
	Xlib: sequence lost (0x1291e > 0x291f) in reply type 0x5!
	Xlib: sequence lost (0x18094 > 0x8095) in reply type 0x6!
	Xlib: sequence lost (0x23ef5 > 0x13ef6) in reply type 0x6!
	Xlib: sequence lost (0x24b96 > 0x14b97) in reply type 0x6!

Notice the pattern: When the server is trying to determine whether this is
		    the correct message in the sequence, it is incrementing
		    the high word rather than the low word of the sequence
		    number.  It should be 0x1dee + 1 not 0x1dee + 0x10000.
		    It is like the message does not contain the right
		    information for X to process the byte/word ordering
		    correctly.

	Could you please post this on the notes file and let me know what
type of response you get.  Our DOS sessions die after awhile since the server
is getting confused.

CRITICAL...URGENT...QAR...CRITICAL...URGENT...QAR...CRITICAL...URGENT...QAR...


					Mark.

T.RTitleUserPersonal
Name
DateLines
3339.1PSW::WINALSKICareful with that VAX, EugeneThu Sep 13 1990 20:307
If it's QAR material, then you should QAR it.  Actually, since this is from a
customer site, you should SPR or CLD it.

This is far too serious a problem to be left to informal channels (i.e., NOTES).
There are official product support channels for this sort of thing.  Use them.

--PSW
3339.2WRKSYS::GOLDBERGMarshall R. Goldberg, WorkstationsFri Sep 14 1990 02:4113
    Paul,
    
    Logicraft is a customer and a CMP. We help them technically and are
    currently working to DDH their products. (Digital Distributed Hardware)
    Before filing a QAR, SPR, or CLD, I prefer to run their report by
    the cognoscenti here in this notes file. This is especially true since
    I don't know what impact this problem has. In the past, such a quick
    review has saved us all lots of grief. Logicraft is pursuing their
    formal channels as well. I received a letter from Logicraft's support
    person in Colorado tonight.
    
    Marshall
    
3339.3STAR::HARDYFri Sep 14 1990 14:0020
>Notice the pattern: When the server is trying to determine whether this is
>		    the correct message in the sequence, it is incrementing
>		    the high word rather than the low word of the sequence
>		    number.  It should be 0x1dee + 1 not 0x1dee + 0x10000.
>		    It is like the message does not contain the right
>		    information for X to process the byte/word ordering
>		    correctly.
    
    	The increment of the high word is due to the fact that sequence
    numbers in replies are 16 bit words, and they are maintained as 32 bit
    words by Xlib.  Xlib determines that the sequence number received in a
    reply is less than expected (0x1dee < 0x1def).  Before reporting an
    error, it checks whether an increment of the high word would explain
    the sequence loss (effectively compensating for the 16 bit reply field
    overflowing).  In this case, it didn't.
    
    	QAR it.  You'll probably need to run a wiretap program (Xliddy) if
    you've got it.
    
    			Sam
3339.4WRKSYS::GOLDBERGMarshall R. Goldberg, WorkstationsTue Sep 18 1990 13:347
    Thanks Sam.
    
    Since the CMP/ISV is a Digital customer under full service contracts,
    they will elevate the issue as a CLD.
    
    Marshall