Conference gyro::internet_toolss

    Same image, but with a later timebomb (30 April).
    
    Directory BULOVA::IPS$KITS:
    
    NETSCAPE-EXPORT-ALPHA-V30B4-012697.EXE;1        15202/15204
    NETSCAPE-EXPORT-VAX-V30B4-012697.EXE;1           9059/9060
    
    Total of 2 files, 24261/24264 blocks.

>    NETSCAPE-EXPORT-ALPHA-V30B4-012697.EXE;1        15202/15204

This image is considerably bigger than the one that AFAIK was the previous
image:

NETSCAPE-EXPORT-ALPHA-2021.EXE;1
                       9940  10-OCT-1996 11:34:07.00


Whats up with this?  Did I miss an image between Oct. and now, or did
the executable really grow 50% bigger?   Yikes!

 - Josh

You missed an image.  Most likly the first Nav Gold image (which includes
an html editor).

bjm

>>    NETSCAPE-EXPORT-ALPHA-V30B4-012697.EXE;1        15202/15204
>NETSCAPE-EXPORT-ALPHA-2021.EXE;1
>                       9940  10-OCT-1996 11:34:07.00

Brian is right. You are comparing Navigator 2.02 with 3.0 Gold.

Colin.

	I know the beta gold is no different than the last one, just a
different bomb date. But ever since installing it (today), all I have been
doing is watching the thing crash. I end up having to either give up on using
it altogether or run to another office and use the pc version cuz my vxt had to
be turned off/on before I can use anything. This is the 3rd time today. Has
anyone else had this problem since installing it? 



Glen

I've been using 3.0b4 on my Alpha workstation all week with no problems.  For
me, it behaves just like the previous version.

Paul

oh.... i'm using a vxt....

We are pleased to announce the availability of the first beta version of
Netscape Navigator Gold V3.0 for OpenVMS Alpha with Java. But before you
start running all those cool Java applets, a word of caution. Many Java
applets are by design either CPU-bound or CPU-intensive. This doesn't matter
too much if you are on a single-user workstation, but on a multi-user
system, performance of the entire system may be impacted.

To install the Navigator image plus the Java class libraries. These are
located in BULOVA::IPS$KITS:

  NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297.EXE  17010 blocks
  JAVA_30.                                      1414 blocks

AT THIS TIME ONLY AN ALPHA VERSION IS AVAILABLE. Due to the difference in
floating point formats, we don't yet have a VAX version.

The Java class libraries (JAVA_30.) need to be copied to either of the
[.NETSCAPE] sub-directories in SYS$LIBRARY or your SYS$LOGIN. Alternatively,
if you want to place the class libraries elsewhere, or change the name,
define the logical name NETSCAPE_CLASSPATH to point to the file.

RECOMMENDED SYSTEM-WIDE INSTALLATION
Copy the Navigator image to SYS$SYSTEM. Copy the JAVA_30 file to
SYS$COMMON:[SYSLIB.NETSCAPE] (create this sub-directory if you need to).

RECOMMENDED PERSONAL INSTALLATION
Copy the Navigator image to SYS$LOGIN or wherever you like to keep images.
Copy the JAVA_30 file to the [.NETSCAPE] sub-directory in your SYS$LOGIN.

DO NOT RENAME THE JAVA-30 file. It must be named "JAVA_30." (no file type).

BUG REPORTING - if you find a problem before Friday 7th February, reply to
this note as soon as possible. We have three engineers out at Netscape until
this date and problems reported before we leave stand a good chance of
getting fixed. After Friday, use the IPS-V20 QAR database on TRIFID.

Please include URL location and precise steps to reproduce the problem.
    
The Netscape Team.
                                                                       

Didn't take long:

What I was doing was "getting mail" from a UCX pop server and then 
used the Window...Address Book pull down to try to bring up the address 
book window while it was downloading from the pop server.

The nav window was up at the same time also.

Yes it is reproduceable.

Brian J.

$!
				[ snip ]
$ netscape 
%SYSTEM-F-ACCVIO, access violation, reason mask=04, virtual address=000351D8, PC=80566368, PS=0000001B
%TRACE-F-TRACEBACK, symbolic stack dump follows
 Image Name   Module Name     Routine Name    Line Number  rel PC      abs PC 
                                                        0 80566368    80566368
 NETSCAPE     XP_FILE         XP_FileTruncate       41208 00000954    005FCE24
 NETSCAPE     SPOOL           msg_end_incorpo       48093 00003FDC    004CBBEC
 NETSCAPE     SPOOL           MSG_Incorporate       48154 00004140    004CBD50
 NETSCAPE     MKPOP3          NET_ProcessPop3       54849 00003AB8    003EB8C8
 NETSCAPE     MKPOP3          NET_InterruptPo       54917 00003D10    003EBB20
 NETSCAPE     MKGETURL        net_InterruptAc       74741 0000471C    003AC30C
 NETSCAPE     MKGETURL        net_InternalInt       75038 00004D6C    003AC95C
 NETSCAPE     MKGETURL        NET_InterruptWi       75069 00004E48    003ACA38
 NETSCAPE     XP_CNTXT        XP_InterruptCon       42113 00000A98    005FFCF8
 NETSCAPE     MSGUTILS        msg_InterruptCo       46390 00000178    004F3318
 NETSCAPE     FOLDERS         MSG_Command           49623 000051EC    004F013C
 NETSCAPE     MAILNEWS        fe_msgDoCommand       78230 00005D70    002F7540
 NETSCAPE     MAILNEWS        fe_msg_command        78241 00005DE8    002F75B8
                                                        0 808F1AC0    808F1AC0
                                                        0 80A43B34    80A43B34
                                                        0 80A4153C    80A4153C
                                                        0 80AB04FC    80AB04FC
                                                        0 8099414C    8099414C
                                                        0 8092C9D0    8092C9D0
                                                        0 8092D230    8092D230
                                                        0 8092D8A0    8092D8A0
                                                        0 808FFB58    808FFB58
                                                        0 809002C8    809002C8
                                                        0 8090F3BC    8090F3BC
 NETSCAPE     MOZILLA         fe_EventLoop         107107 000002FC    002FBC8C
 NETSCAPE     MOZILLA         main                 108929 00003240    002FEBD0
 NETSCAPE     MOZILLA         __main                    0 0000006C    002FB9FC
                                                        0 846DC170    846DC170
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=646E682F, PC=80407EE8, PS=0000001B

  Improperly handled condition, image exit forced.
    Signal arguments:   Number = 00000005
                        Name   = 0000000C
                                 00000000
                                 646E682F
                                 80407EE8
                                 0000001B

    Register dump:
    R0  = 000000007EEA8008  R1  = 0000000000000003  R2  = 000000007EE6A940
    R3  = 0000000000000046  R4  = 00000000000000E8  R5  = 000000007EEA8008
    R6  = 000000001000000C  R7  = 000000007FF91FC0  R8  = 0000000000000000
    R9  = 000000000028E7B0  R10 = 0000000000000001  R11 = 0000000000000000
    R12 = 0000000000000000  R13 = 000000000000006F  R14 = 0000000000000000
    R15 = 0000000000C7A5E8  R16 = 000000007EEA88C0  R17 = 5F6E6D79646E682F
    R18 = 0000000000044578  R19 = 000000007EEA8080  R20 = 000000007EEA88C0
    R21 = 001E5FC300000000  R22 = 00000000646E682F  R23 = 0000000000000000
    R24 = 00000000646E682F  R25 = 000000005F6E6D79  R26 = FFFFFFFF80409F40
    R27 = 0000000000000001  R28 = FFFFFFFF805AF7BC  R29 = 000000007EC6EA80
    SP  = 000000007EC6EA80  PC  = FFFFFFFF80407EE8  PS  = 000000000000001B
  MCCARTHY     job terminated at  3-FEB-1997 07:17:11.45

  Accounting information:
  Buffered I/O count:           13451         Peak working set size:  22720
  Direct I/O count:              2194         Peak page file size:   200240
  Page faults:                   1579         Mounted volumes:            0
  Charged CPU time:           0 00:00:55.94   Elapsed time:     0 00:41:42.80

	Colin,

	Why not just put JAVA_30. in a 'normal' system directory
	like SYS$SYSTEM.  We have funny logicals in our multi
	system disk cluster to avoid having multiple copies of
	things like NETSCAPE.EXE yet can not use the same trick
	with JAVA_30..  I can not see the advantage of having this
	separate directory.

	-Pete

	Update to .208..... it just crashed YET again..... but I think I have
pin pointed it a bit. I can search all I want, all day long if I go from link
to link. But if I use the open option, it just stops working at some point.
Sometimes it is right away (like this time), sometimes it will happen if I hit
the back key. But it happens if I use the open option to go to a site, and not
if I go there via altavista or through a link that didn't go through the open
option.



Glen

    Pete,
    
    Its in a "funny" directory because that's where the code is looking for
    it. Remember, this is a port and we don't change code we don't have to.
    However, we anticipated situations like yours, and that's why we added
    the NETSCAPE_CLASSPATH logical. If you want to put the JAVA_30 file
    onto a "folkd" or wherever, you can - just define the
    NETSCAPE_CLASSPATH logical to point to it.
    
    Colin.

    Suddenly there are 5438 new applications for OpenVMS. Check out the
    Java directory at www.gamelan.com and see for yourself.

	I get the following on STAR when I try to run Netscape Navigator copied
from BULOVA::

IXIVIV> sho sym netscape
  NETSCAPE == "$COUPE:[NETSCAPE]NETSCAPE-EXPORT-ALPHA-V30B4-012697.EXE"
IXIVIV> netscape
%LIB-W-ILLRECTYP, illegal record type (9182)
%TRACE-W-TRACEBACK, symbolic stack dump follows
 Image Name   Module Name     Routine Name    Line Number  rel PC      abs PC
                                                        0 80444D64    80444D64
                                                        0 804573E0    804573E0
                                                        0 80457D58    80457D58
                                                        0 8057BD80    8057BD80
                                                        0 805A5470    805A5470
                                                        0 8059D3BC    8059D3BC
 NETSCAPE-EXP MOZILLA         fe_create_pidlo      103425 00003350    002DE370
 NETSCAPE-EXP MOZILLA         main                 102725 00002414    002DD434
 NETSCAPE-EXP MOZILLA         __main                    0 0000006C    002DB08C
                                                        0 83DD8170    83DD8170
%LIB-W-ILLRECTYP, illegal record type (!UL.) in module !AC
%TRACE-W-TRACEBACK, symbolic stack dump follows
 Image Name   Module Name     Routine Name    Line Number  rel PC      abs PC
                                                        0 80457D88    80457D88
                                                        0 8057BD80    8057BD80
                                                        0 805A5470    805A5470
                                                        0 8059D3BC    8059D3BC
 NETSCAPE-EXP MOZILLA         fe_create_pidlo      103425 00003350    002DE370
 NETSCAPE-EXP MOZILLA         main                 102725 00002414    002DD434
 NETSCAPE-EXP MOZILLA         __main                    0 0000006C    002DB08C
                                                        0 83DD8170    83DD8170
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=00000002, PC
=80566578, PS=0000001B
%TRACE-F-TRACEBACK, symbolic stack dump follows
 Image Name   Module Name     Routine Name    Line Number  rel PC      abs PC
                                                        0 80566578    80566578
 NETSCAPE-EXP PRPRF           dosprintf              6550 00000CC4    00590934
 NETSCAPE-EXP PRPRF           PR_vsmprintf           6681 000010C0    00590D30
 NETSCAPE-EXP PRPRF           PR_smprintf            6668 00001058    00590CC8
 NETSCAPE-EXP MOZILLA         fe_create_pidlo      103485 000033FC    002DE41C
 NETSCAPE-EXP MOZILLA         main                 102725 00002414    002DD434
 NETSCAPE-EXP MOZILLA         __main                    0 0000006C    002DB08C
                                                        0 83DD8170    83DD8170

re .216,

> Check out the Java directory at www.gamelan.com and see for yourself.

I went to the Amusements and Games section and Netscape went away.

Paul


%SYSTEM-F-OPCCUS, opcode reserved to customer fault at PC=FFFFFFFF80938A98,
PS=0000001B
%TRACE-F-TRACEBACK, symbolic stack dump follows
  image    module    routine             line      rel PC           abs PC      
                                            0 FFFFFFFF80938A98 FFFFFFFF80938A98
                                            0 FFFFFFFF80911608 FFFFFFFF80911608
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  XFE  fe_add_input
                                       100706 0000000000000FEC 000000000033855C
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  XFE  FE_SetConnectSelect
                                       100829 00000000000012FC 000000000033886C
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  MKHTTP  net_finish_http_connect
                                        55454 00000000000008A8 00000000003C31B8
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  MKHTTP  NET_ProcessHTTP
                                        57775 0000000000004484 00000000003C6D94
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  MKGETURL  NET_ProcessNet
                                        74495 000000000000419C 00000000003ABD8C
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  XFE  fe_stream_callback
                                       100630 0000000000000CFC 000000000033826C
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  XFE  fe_VMS_AppInput
                                       100690 0000000000000F3C 00000000003384AC
                                            0 FFFFFFFF80BA8774 FFFFFFFF80BA8774
                                            0 FFFFFFFF80BA91F0 FFFFFFFF80BA91F0
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  MOZILLA  fe_EventLoop
                                       107188 0000000000000334 00000000002FBCC4
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  MOZILLA  main
                                       108929 0000000000003240 00000000002FEBD0
 NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297  MOZILLA  __main
                                            0 000000000000006C 00000000002FB9FC
 PTHREAD$RTL                                0 000000000004C148 000000000084C148
 PTHREAD$RTL                                0 0000000000030664 0000000000830664
                                            0 FFFFFFFF838DD0D8 FFFFFFFF838DD0D8

re: .217 - please copy the image again. Somehow the image on BULOVA was
corrupt.

re: .218 - Does it always die at www.gamelan.com? What VMS version, MOTIF
version, and hardware do you have?

re .219,

> Does it always die at www.gamelan.com?

I just went back there and pressed the same buttons, but this time it worked. 
Between last time and this time, I changed the HTTP proxy.  Last time, the page
was loading for at least ten minutes before it died.

> What VMS version, MOTIF version, and hardware do you have?

OpenVMS Alpha V7.1 and whatever came with V7.1 (DECwindows server ident DW
V7.1-961126) on a DEC 3000 Model 500.

I'll reply if it happens again.

Paul

I changed proxies back to the slow one, tried it again, and Netscape failed with
the same %SYSTEM-F-OPCCUS error message as before.

Paul

    Brian,

    	Can you tell me what version of VMS you are running? I tried
    reproducing this problem on 7.1 (field test) and everything seemed to
    work fine. 

    	Could you give me a little more detail as to when you pulled down
    the address book window (when it was in the middle of pulling down
    mail, beginning, end, etc), how many messages where being pulled
    down...


    Keith Griffin (trying to reproduce the problem).

I'll send Keith the info off line.

Another interesting change with this baselevel.  The non-default newsrc files
have changed their names. 

It used to be newsrc-nn-nnn-nn-nn.  With this release its
newsrc-name$name$name$com.  where the $ replace the .'s in the newsserver
name (ie the above would mean my non-default newsserver was named
name.name.name.com).

Really screws up your "seen list".

Brian J.

    	Yes, the non-default news files have had a name change. This was
    due to complaints that the old method (using IP addresses) was
    returning the wrong news group names (i.e. if an IP address had more
    than one name associated with it, you could get a different name back).

    	A solution for this problem will be provided with the released
    version of the navigator (i.e. a procedure will be provided to rename
    the NEWSRC-NN-NN-NN-NN to NEWSRC-name$name$name).

    Keith 

If one uses the "beta test" of AltaVista's "Live Topics" feature
( http://altavista.digital.com/cgi-bin/query?moo=yes ) and selects the
Java applet version of LT, the applet only partially displays on VMS and
doesn't appear functional.  (Note - don't give out the Live Topics URL outside
Digital.)

				Steve

I have also reported a problem with one of the clock applets pointed to by 
a previous reply.  It kind of works but does not rotate.

re: new names or news servers:

Better - I wonder how the PC version works because it has just the first part
of the news server in the file ie name.rc for name.foo.bar.com.

bjm

>If one uses the "beta test" of AltaVista's "Live Topics" feature
>( http://altavista.digital.com/cgi-bin/query?moo=yes ) and selects the
>Java applet version of LT, the applet only partially displays on VMS and
>doesn't appear functional.  (Note - don't give out the Live Topics URL outside
>Digital.)

Thanks for the report, Steve, We'll take a look and see if we can figure out
why its failing.

Colin.

    	The process of downloading mail files, or a very large mail
    message, is interrupted when you activate the 'Address Book' window.
    The code that performs the interruption changed from V2.0 to V3.0 of
    the navigator. The problem has been isolated and fixed. Good catch.

    Keith

>>    the navigator. The problem has been isolated and fixed. Good catch.

Like many bugs - it was an accident on both ends (coding and finding the fault
at run time).

bjm

   Two bugs that I can see with the Gold Beta ...

	(my system is VMS 6.2, DEC 3000 m600, UCX 4.1-12 ECO 2)

 
   BUG A
   =====

   When I look at the following URL

	http://www.melee.com/tool/page/ColorCylinder.html

   and compare the behavior between VMS and WNT (running Navigator 3.0)
   I notice that the VMS java seems to lose the update that responds
   to a click in the Autoscan box, when the box already is checked.  With
   Navigator on WNT, I can reliably click in the box and it updates to
   unchecked.  With VMS, the update that clears the box seems to occur
   only about 10% of the time or less.  I infer that the click action
   is detected because the autoscan does stop.


   BUG B
   =====

   This is not a Java bug, but it is a bug with the VMS port.  Something
   is deviant about the handling of background pictures with respect to
   table cells that are colored.  Look at this URL on both WNT (Navigator
   3.0) and the VMS beta:

	http://www.melee.com/

   Perhaps all the gold versions of Navigator have this bug ... I wouldn't
   know because I haven't tried any gold version.  On a Macintosh and W95
   and WNT, the rendered page does not leak portions of the background
   picture into the table cells.  On my VMS box (as specified above), the
   Gold beta is messing up mostly those table cells that have text.  You'll
   have to look to see what I mean.  Some table cells display as expected.

   I hope this doesn't mean that Navigator Gold v3 renders table colors
   and backgrounds differently from Navigator (plain) v3.

   Paul

    Remember that this is a port of the UNIX version of the Navigator, not
    the Windows version. I just tried both of things you described, and the
    same behaviour exists on UNIX.

re Note 3480.231 by STAR::BLAKE:

>     Remember that this is a port of the UNIX version of the Navigator, not
>     the Windows version. I just tried both of things you described, and the
>     same behaviour exists on UNIX.
  
        I had been using Navigator GOLD V3.01 as my main browser on
        Digital UNIX until about a month ago.  It crashed once or
        twice a day.

        My desktop then changed to a Windows NT machine.  Navigator
        GOLD V3.01 crashed once or twice the whole month of January.

        Bob

  re .231

  is Netscape interested in letting the world know which sets of behavior
  they consider standard:

	(a) what happens on Windows/Mac
	(b) what happens on UNIX/VMS

  I know it is foolish to think that Java really will be a portable
  platform, across all possible vendors.  But, if even Netscape can't
  decide for the products it releases, I guess I should buy more of
  Bill Gates' stock.

  Paul 

| <<< Note 3480.232 by LGP30::FLEISCHER "without vision the people perish (DTN 381-0426 ZKO1-1)" >>>


| I had been using Navigator GOLD V3.01 as my main browser on Digital UNIX until
| about a month ago.  It crashed once or twice a day.

	That's what the VMS version does to me now. I switched back to 2.0.
What's weird is the last version of 3.0 didn't crash and the only thing that
changed was the time bomb date.

	Oh.... to give a quick update on the vms version.... the same thing
happens to a friend of mine who uses it. It does seem to do it a lot when there
is a site that has a lot of extras on it... when it is looking for the mdi is
when it does it most often....

re: .225

> If one uses the "beta test" of AltaVista's "Live Topics" feature
> ( http://altavista.digital.com/cgi-bin/query?moo=yes ) and selects the
> Java applet version of LT, the applet only partially displays on VMS and
> doesn't appear functional.

We just found and fixed this one. Turned out to be the same bug as the
revolving digital time bug (and probably some other bugs too). This will
be fixed in the next image (which will be coming back in a day or two).

Colin.

        re: .233
    
    We have reported these Bugs to Netscape.  However, the VMS Netscape
    engineers, 2, can only work on the code for two weeks every 8 weeks. 
    So, for core netscape issues, we will report them to Netscape Engineering 
    as core Motif GUI issues.  However, from time to time we have fixed
    several crashs in VMS, and sent the fixes to Netscape Engineering.  
    So, VMS Netscape shouldn't crash with the same frequency.
    
    >I guess I should buy more of Bill Gates' stock.
    
    I'm glad to see you appreicate our porting efforts in bring netscape to
    OpenVMS.  ;-)
    
    Thank you for your support.
    
    Powell Hazzard
    Netscape Engineering for this week, on the West coast.
    Customer Support Engineer, on the East coast.
    

    re: .235
    
     Glen, 
    
       I need some more letters to be able to solve the puzzle so... I'll would 
    like to buy a vowel.  Or how about a stack trace?  Or how about when
    and what pages?
    
    You might want to try and backup your netscape directory and clean
    out the netscape directory.  Then start the Navigator, does it still
    crash?  Have you tried the New 3.0b5 images?
    
    Thanks
    
    Powell Hazzard

| <<< Note 3480.238 by PEACHS::HAZZARD >>>



| I need some more letters to be able to solve the puzzle so... I'll would
| like to buy a vowel.  

	You have several to choose from! 

| Or how about a stack trace?  Or how about when and what pages?

	It just locks, so there is no stack dump that occurs. And it happens
whenever it wants. :-)  I will look for the page that crashes it and then post
them. 

| You might want to try and backup your netscape directory and clean
| out the netscape directory.  

	I'll try that.

| Have you tried the New 3.0b5 images?

	The 3.0b5 images is for alpha only. So I can't use it. I'm using the
vms 3.0b4 version. The version before that never gave me a problem. This one
just locks up. 


Thanks,


Glen

>
>    I'm glad to see you appreicate our porting efforts in bring netscape to
>    OpenVMS.  ;-)
>
>    Thank you for your support.
>

  Powell, Colin, et al.    I am glad to see that the VMS port is so close
  that my complaints end up being complaints about Netscape's well-known
  quality/consistency standards.

  Paul

    Powell, Colin
    
    how'd he get in here?  :-)

    I must be missing something major here.
    
    I've copied:
    
JAVA_30.;2            				4-FEB-1997 11:30:44.66
NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297.EXE;2	4-FEB-1997 12:18:05.81
    
    The dates are my modification dates. JAVA_30. is in
    SYS$COMMON:[SYSLIB.NETSCAPE]. When I start NETSCAPE, I
    get an error box:
    
    "Netscape is unable to find the file or directory named
    DISK$USER1:[GILLINGS]/24000602 Check the name and try again"
    
    The number is the PID of the process. If I restart Netscape and reuse
    the same Motif Session Manager subprocess, the message goes away.
    
    Once the process starts, it goes compute bound and issues around 5000
    I/Os per second to one of 2 BG devices. UCX says "%UCX-W-NODEVSOCK,
    Device_socket not found"
    
    I can navigate "normally", though perhaps rather slowly. I tried using
    Java on the Altavista page, it just hung the window.
    
    Digital TCP/IP Services for OpenVMS Alpha Version V4.1 - ECO
    Level 3  on a DEC 3000 Model 500 running OpenVMS V6.1. I'm using the
    same setup which was working fine with the AUG-1996 version of
    Netscape. I can still run that version without generating millions of
    I/Os.
         					John Gillings, Sydney CSC

    One issue resolved:
    
> When I start NETSCAPE, I
>    get an error box:
>    
>    "Netscape is unable to find the file or directory named
>    DISK$USER1:[GILLINGS]/24000602 Check the name and try again"
>    
>    The number is the PID of the process. If I restart Netscape and reuse
>    the same Motif Session Manager subprocess, the message goes away.
    
    I was starting Netscape with VUE$NETSCAPE which passes the current
    Motif selection as a parameter to Netscape. Hence the no such file
    message. It just happened that each time I restarted Netscape, I had
    the PID selected.
    
    I haven't been able to find a way to start it which doesn't immediately
    go crazy with I/Os.
    						John Gillings, Sydney CSC

The second beta version of Netscape Navigator Gold V3.0 for OpenVMS Alpha with
Java is now available. If you installed the first beta, please upgrade;
important bug fixes are in this version.

The files are located in in BULOVA::IPS$KITS:

  NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020697.EXE  17020 blocks
  JAVA_30.                                      1414 blocks (unchanged)

AT THIS TIME ONLY AN ALPHA VERSION IS AVAILABLE. Due to the difference in
floating point formats, we don't yet have a VAX version.

The Java class libraries (JAVA_30.) need to be copied to either of the
[.NETSCAPE] sub-directories in SYS$LIBRARY or your SYS$LOGIN. Alternatively,
if you want to place the class libraries elsewhere, or change the name,
define the logical name NETSCAPE_CLASSPATH to point to the file.

RECOMMENDED SYSTEM-WIDE INSTALLATION
Copy the Navigator image to SYS$SYSTEM. Copy the JAVA_30 file to
SYS$COMMON:[SYSLIB.NETSCAPE] (create this sub-directory if you need to).

RECOMMENDED PERSONAL INSTALLATION
Copy the Navigator image to SYS$LOGIN or wherever you like to keep images.
Copy the JAVA_30 file to the [.NETSCAPE] sub-directory in your SYS$LOGIN.

DO NOT RENAME THE JAVA-30 file. It must be named "JAVA_30." (no file type).

BUG REPORTING - use the IPS-V20 QAR database on TRIFID. Please include URL
location and precise steps to reproduce the problem.
    
The Netscape Team.

>>The files are located in in BULOVA::IPS$KITS:
>
>  NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020697.EXE  17020 blocks

Colin,

Must be the timezone difference.  The file is:
  NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020797.EXE  17020 blocks
                                      ^
				      7 not 6

Brian J.

:>>The files are located in in BULOVA::IPS$KITS:
:>
:>  NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020697.EXE  17020 blocks
:
:Colin,
:
:Must be the timezone difference.  The file is:
:  NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020797.EXE  17020 blocks

I don't see either of those names -- I see:

Directory BULOVA::DISK$I18N:[IPS$KITS]

NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697.EXE;1
                       17020/17022     7-FEB-1997 02:49:17.00

I assume this is the right file?

    Sorry about that. Its been a long two weeks. The correct file name is:
    
      NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697.EXE
    
    Colin.

    re .242 (I/O flood)
    
    Tried the version mentioned in .247. Same results :-(
    
    						John Gillings, Sydney CSC

I'm running Navigator V3.0b4 on OpenVMS VAX.  I have it autostart when I log
into my workstation.  Lately I've been getting the following message box when I
log in...

	"While loading a public customization file,
	 the following error occurred:

	 Insufficient privilege or file protection violation."

What is weird is that after I'm logged in and things have had a moment to settle
down, I can restart Netscape and not get the message!

	I have three questions:

	1)  Is this a netscape message box?

	2)  What is the name of the public customization file?

	3)  Why can't the name be displayed as part of the message?

Thanks!
	David

    re: 248
    
      Sounds like something is up with UCX.  Have you been able to try
    another system?  Maybe running a different version of UCX...
    We didn't make any changes in the network connection area, from 3.0b3 to
    3.0b6...
    
    Powell

    re: 249
    
    
     I don't think it is a Netscape Message.  What happens if you start it
    from a decterm?
    
    Powell

	I used to use it on an auto start and got the same problem as well. I
just stopped using it from an autostart. :-)  From a decterm for me it always
started up fine. I have it in my applications section of my vxt now and that
will start it up fine as well. Keeping it going without it crashing is still a
wonder for me.

    re: .242 
    
        The combination seems to be Ucx 4.* and OpenVMS 6.1.
    
        OpenVMS 6.1 and UCX V3.* seems to be working. 
        OpenVMS 6.2 and UCX V4.1 seems to be working...
    
    Powell

    .242 .248
    
      What version of Motif are you running?
    
      @sys$update:decw$versions all
    
    Powell

    re .256: (Powell)                                                    
    
    $ @sys$update:decw$versions all
    DECwindows ident is                     DW V1.2-940307
    DECwindows server ident is              DW X2.0-940825
    DECwindows transport ident is           DW V6.1-940402
    DECwindows xlib ident is                DW V1.2-940307
    DECwindows motif toolkit ident is       DW V1.2-940307
    DECwindows apps ident is                DW V1.2-940307
    DECwindows programming ident is         DW V1.2-940307
    
    I know this is old, but my workstation is the Alpha V6.1 support
    system in the CSC. Maybe when V7.1 hits the streets I'll be able to
    upgrade a notch or 2. We need systems of all versions that customers
    run :-(
    						John Gillings, Sydney CSC

| <<< Note 3480.204 by STAR::BLAKE "OpenVMS Engineering" >>>

| NETSCAPE-EXPORT-VAX-V30B4-012697.EXE;1           9059/9060

	I went to my ISP's homepage today using the above. It hangs for a
while, then it crashes. Except when it crashed it took everything with it. All
window applications died. This is happening a lot with this version. What I
don't understand is why..... the isp url is http://www.ziplink.net



Glen

I had a similar problem (seemed like Netscape was causing my DECwindows server
to crash) but it all went away after a more memory was installed and 
a few runs of autogen...feedback.

I would suggest trying to bump some of the process/system quotas and see if 
the problem goes away.

bjm

| <<< Note 3480.257 by HNDYMN::MCCARTHY "A Quinn Martin Production" >>>


| I would suggest trying to bump some of the process/system quotas and see if
| the problem goes away.

	Ok.... now how would I do that? Talk to the system manager?

	The timing of this was when I got the new version (time bomb set for
April). Before that I could go anywhere and it was fine. 

Re: .249

The error has nothing to do with Netscape or the Performance Advisior display
program as I took them both off the autostart list.  So there is something in
Motif having trouble.  I'll move this discussion to the Motif conference to see
if anyone there can help me.

FWIW I checked all of the public .DAT files for the proper protections.  All
looked fine.  Don't know what brought this error on though...

	Thanks!
		David

Several people have reported problems with the Java version of the OpenVMS
Navigator going compute-bound. This occurs immediately the Navigator is
started up, and does NOT require a Java app to even be loaded. We have been
unable to determine the cause of this, and, so far, have not even been able
to determine under what conditions is happens.

If you are currently using (or trying to use!) the Java version, please reply
to this note with the following information:

  OpenVMS Version:           (from a SHOW SYSTEM)

  MOTIF Version:             (from an @sys$update:decw$versions all)

  Display Transport:         (from a SHOW DISPLAY)

  Display Head (only if display transport is NOT LOCAL):
			     (eg. Win95 running eXcursion, UNIX system)

  Do you go CPU-bound:       (do CTRL/T's and check the CPU number)

If you are not sure if you are running the Java version, click on "Help" and
then "About Netscape". The Java version will show the "Java Compatible" logo
between the Netscape and RSA logos. We only want this feedback for the Java
version.
    

$ show sys
OpenVMS V6.2  on node SITBUL  25-FEB-1997 13:26:24.33  Uptime  3 20:26:52

$ ucx show ver

  Digital TCP/IP Services for OpenVMS Alpha Version V4.1 - ECO Level 2
  on a DEC 3000 - M600 running OpenVMS V6.2

$ @sys$update:decw$versions all
DECwindows ident is                     DW V1.2-940307
DECwindows server ident is              DW T6.2-950505
DECwindows transport ident is           DW T6.2-950505
DECwindows xlib ident is                DW V1.2-940307
DECwindows motif toolkit ident is       DW V1.2-940307
DECwindows apps ident is                DW V1.2-940307
DECwindows programming ident is         DW V1.2-940307

$ show display

    Device:    WSA1:  [exec]
    Node:      0
    Transport: LOCAL
    Server:    0
    Screen:    0

$ netscape
SITBUL::_FTA11: 13:28:57 NETSCAPE- CPU=00:02:11.22 PF=6862 IO=36470 MEM=703
SITBUL::_FTA11: 13:28:59 NETSCAPE- CPU=00:02:11.22 PF=6864 IO=36471 MEM=704
SITBUL::_FTA11: 13:29:00 NETSCAPE- CPU=00:02:11.22 PF=6864 IO=36472 MEM=704

$ show sym netscape
  NETSCAPE == "$SYS$SYSTEM:NETSCAPE-EXPORT-ALPHA-JAVA-V30B5-020297.EXE;1"

OpenVMS version: V7.1
Motif version: DW V1.2-4960827
Display transport: LOCAL
CPU bound?  No

				Steve

Accvio (reproduceable):

Use file:/ syntax to open this file, then click on the one link in it.  It does
not matter if the link exists or not.

And yes, I know I should be using /sys$login/ but an accvio is an accvio.

Brian J.

test.html:

<html>
<head>
<title> cxx.hlp</title>
</head>
<body>
<pre>
</PRE>
<P>Information available on:<BR><DIR>
<li><A HREF="sys$login:cxx_1.HTML#Ref1">CXX</a>
</PRE>
</BODY>
</HTML>


$!
$ SET ON
$ netscape "file:/hndymn_usr$/mccarthy/test.html"
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual address=7EC6EF1F, PC=005FCA58, PS=0000001B
%TRACE-F-TRACEBACK, symbolic stack dump follows
 Image Name   Module Name     Routine Name    Line Number  rel PC      abs PC 
 NETSCAPE     XP_FILE         XP_Stat_VMS           41921 00000168    005FCA58
 NETSCAPE     XP_FILE         XP_Stat_VMS           41974 00000330    005FCC20
 NETSCAPE     XP_FILE         XP_Stat               41899 000000A8    005FC998
 NETSCAPE     MKFILE          net_check_file_       45023 000000A4    003B6124
 NETSCAPE     MKFILE          NET_ProcessFile       46307 000024D4    003B8554
 NETSCAPE     MKFILE          NET_FileLoad          46289 00002438    003B84B8
 NETSCAPE     MKGETURL        NET_GetURL            73941 0000325C    003AB0AC
 NETSCAPE     XFE             fe_GetSecondary      100602 00000CA8    003383E8
 NETSCAPE     XFE             fe_GetURL            100534 00000AD0    00338210
 NETSCAPE     LAY             fe_HandleHREF        102502 000066B4    002F0444
 NETSCAPE     LAY             fe_activate_lin      102583 0000693C    002F06CC
                                                        0 8092C9D0    8092C9D0
                                                        0 8092D230    8092D230
                                                        0 8092D8A0    8092D8A0
                                                        0 808FFB58    808FFB58
                                                        0 809002C8    809002C8
                                                        0 8090F3BC    8090F3BC
 NETSCAPE     MOZILLA         fe_EventLoop         107107 000002FC    002FBC5C
 NETSCAPE     MOZILLA         main                 108929 00003240    002FEBA0
 NETSCAPE     MOZILLA         __main                    0 0000006C    002FB9CC
                                                        0 846DC170    846DC170
  MCCARTHY     job terminated at 26-FEB-1997 12:15:18.27

  Accounting information:
  Buffered I/O count:             456         Peak working set size:  21136
  Direct I/O count:               532         Peak page file size:    75344
  Page faults:                   1108         Mounted volumes:            0
  Charged CPU time:           0 00:00:10.35   Elapsed time:     0 00:01:14.58

    
    
    We have tracked the CPU bound Navigator to a combination with OpenVMS 6.1 
    and OpenVMS V6.1 DECC$SHR.  Installing the ECO ALPACRT09_061 clears up
    the CPU bound Navigator.  I don't know why, but it does...
    
    bulova::ips$kits:ALPACRT09_061.a
    
    Powell

I've received a consistant crash while printing
http://tsavo.zko.dec.com/raven/projects.htm. The same problem happens whether I
print to a file or to a printer. This happens on OpenVMS V7.1 with:

$ @sys$update:decw$versions a
DECwindows ident is                     DW V1.2-4960312
DECwindows server ident is              DW V7.1-961126
DECwindows transport ident is           DW T6.2-961126
DECwindows xlib ident is                DW V1.2-4960312
DECwindows motif toolkit ident is       DW V1.2-4960313
DECwindows apps ident is                DW V1.2-4960312
DECwindows programming ident is         DW V1.2-4960312

Process dump available on request. The output file has nothing but the
"%!PS-Adobe-3.0" header in it.

John Saunders
DECdns Engineering
====================================================================

	Welcome to OpenVMS AXP (TM) Operating System, Version V7.1    

Username: saunders
Password: 

           Property of Digital Equipment Corporation
                >> Digital Internal Use Only <<


    Last interactive login on Tuesday,  4-MAR-1997 15:54:37.14
    Last non-interactive login on Friday,  7-MAR-1997 17:30:10.35
	1 failure since last successful login
.$ set dis/crea/node=wa1knx/tran=local
.$ set process/dump
.$ run sys$system:NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697
\
%SYSTEM-F-ILLEGAL_SHADOW, illegal formed trap shadow, Imask=00000000,
Fmask=00400000, summary=21, PC=000000000039C8E0, PS=0000001B
%TRACE-F-TRACEBACK, symbolic stack dump follows
  image    module    routine             line      rel PC           abs PC      
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697  PS  xl_begin_document
                                        40057 0000000000000260 000000000039C8E0
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697  PRINT  xl_generate_postscript
                                        40375 00000000000002D8 0000000000399058
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697  PRINT  PSFE_AllConnectionsComplete
                                        41085 00000000000020E0 000000000039AE60
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697  MKGETURL  net_CheckForWaitingURL
                                        72659 0000000000000F68 00000000003A8DB8
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697  MKGETURL  NET_ProcessNet
                                        74667 0000000000004510 00000000003AC360
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697  XFE  fe_stream_callback
                                       100630 0000000000000CFC 000000000033843C
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697    
                                            0 000000000032867C 000000000033867C
                                            0 FFFFFFFF807A6774 FFFFFFFF807A6774
                                            0 FFFFFFFF807A71F0 FFFFFFFF807A71F0
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697    
                                            0 00000000002EBC94 00000000002FBC94
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697    
                                            0 00000000002AEB10 00000000002BEB10
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697    
                                            0 00000000002B385C 00000000002C385C
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697    
                                            0 00000000002B3BE8 00000000002C3BE8
                                            0 FFFFFFFF80789A60 FFFFFFFF80789A60
                                            0 FFFFFFFF808124D4 FFFFFFFF808124D4
                                            0 FFFFFFFF80789A60 FFFFFFFF80789A60
                                            0 FFFFFFFF807EC03C FFFFFFFF807EC03C
                                            0 FFFFFFFF807E755C FFFFFFFF807E755C
                                            0 FFFFFFFF8094932C FFFFFFFF8094932C
                                            0 FFFFFFFF8091E38C FFFFFFFF8091E38C
                                            0 FFFFFFFF807C49C0 FFFFFFFF807C49C0
                                            0 FFFFFFFF807C5B3C FFFFFFFF807C5B3C
                                            0 FFFFFFFF80797B58 FFFFFFFF80797B58
                                            0 FFFFFFFF807982C8 FFFFFFFF807982C8
                                            0 FFFFFFFF807A73AC FFFFFFFF807A73AC
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697    
                                            0 00000000002EBC5C 00000000002FBC5C
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697    
                                            0 00000000002EEBA0 00000000002FEBA0
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697    
                                            0 00000000002EB9CC 00000000002FB9CC
 PTHREAD$RTL                                0 000000000004C148 000000000084C148
 PTHREAD$RTL                                0 0000000000030664 0000000000830664
                                            0 FFFFFFFF825C10D8 FFFFFFFF825C10D8
.$ dir *.dmp

Directory SYS$SYSDEVICE:[SAUNDERS]

NETSCAPE.DMP;1      

Total of 1 file.
.$ 


...$ dir *.dmp/siz

Directory SYS$SYSDEVICE:[SAUNDERS]

NETSCAPE.DMP;1         23825

Total of 1 file, 23825 blocks.
.$ logo
  SAUNDERS     logged out at  7-MAR-1997 17:36:21.64

>>I've received a consistant crash while printing
>>http://tsavo.zko.dec.com/raven/projects.htm. 

Worked for me (OpenVMS Alpha V6.2, same Netscape image).  The Netscape image 
is so huge, I'd suggest bumping your process quotas and trying again.

bjm

Ok, I increased PGFLQUO (the only thing I could imagine it running out of), and
tried again. This time, I watched with SDA. It didn't run out of anything:
ASTLM, BYTLM, FILLM, TQELM, PGFLQUO, DIOLM, BIOLM. In fact,
ANALYZE/PROCESS_DUMP/MISC of the dump shows the same thing.

Perhaps someone should look at the dump?

Thanks,
John Saunders
DECdns Engineering

Somehow `illegal formed trap shadow' does not read as a process
quota issue to me.  I can't help on this one, my workstation is 
VAX at VMS 6.2

Bruce

| <<< Note 3480.267 by DRAGNS::SAUNDERS >>>

| Perhaps someone should look at the dump?

	gak!

    John,
    
       Have you  checksumed the image?  
    
    $ checksum/image
    sys$system:NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697.EXE;1
    file SYS$SYSROOT:[SYSEXE]NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697.EXE;1
    image section %D'1' checksum is %X'B4B30DC9'
    image section %D'2' checksum is %X'E9170CC6'
    image section %D'3' checksum is %X'3CF64A49'
    image section %D'5' checksum is %X'406D70ED'
    image header checksum is %X'0023F8F4'
    checksum of all image sections is %X'213F3BAB'
    
    
    Powell
    
    

My checksum matches yours.

Any more ideas?

John

I can't print from Netscape either.  I'm running
NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697.EXE on OpenVMS Alpha V7.1.  Whenever I
attempt to print, I get

%SYSTEM-F-ILLEGAL_SHADOW, illegal formed trap shadow, Imask=00000000,
 Fmask=00400000, summary=21, PC=000000000039C8E0, PS=0000001B

Thanks! I feel validated now.

John Saunders
DECdns Engineering

In fact, both Print and Save As fail in the same way.

John, glad I made you feel less alone.

Paul

I am using an AlphaStation 200 4/100 with 64MB memory and OpenVMS V6.2. I find
Navigator Gold V30B5 to be so intolerably slow that I keep going back to V2.02.

Should I expect performance improvements with the final release?
Are there minimum parameter/quota settings for Navigator Gold?
Am I just SOL with trying to run such a big piece of software on such a
slow, memory constrained system? 

-John

Here are my current process quotas if that helps:

Maxjobs:         0  Fillm:       100  Bytlm:       327680
Maxacctjobs:     0  Shrfillm:      0  Pbytlm:           0
Maxdetach:       0  BIOlm:       200  JTquota:      16384
Prclm:          10  DIOlm:       200  WSdef:          512
Prio:            4  ASTlm:      1024  WSquo:         1024
Queprio:         0  TQElm:        64  WSextent:    260000
CPU:        (none)  Enqlm:      4000  Pgflquo:     600000

I too was expceriencing Netscape running slowly when I first start it and
load my homepage, and anytime I load my homepage, etc.

I thought it was because I have a lot of URL's in my homepage (I just write
out my bookmarks into my homepage).
Well, I was cleaning out files on my account, to get back more diskquota.
Noticed that my HISTORY.DB;1 files was really big, like around 2000 blocks.
I figured I didn't really need history going back a year or whatever that
I've been running Netscape on Alpha. I checked out every dang pulldown menu
and option and found NOTHING on reseting or clearing or deleting this file.
Should there be?  Is there something that I've missed?

So, I deleted the file and started up Netscape.  WOW!  After Netscape came
up (the usual amount of time, maybe 10 seconds or so) the homepage loaded
RIGHT AWAY!  I don't think Netscape is downloading anything faster, but its
sure zippier now when going BACK, to Bookmarks, etc.

Anyone have anything to say about the HISTORY.DB;1 file, care and maintainance
of ?

 - Josh Belkin

	Josh, mine was crawling as well. I just looked at the history thing and
it was up to 5000 blocks! I deleted it and man oh man.... this baby hums!



Glen

>it was up to 5000 blocks! I deleted it and man oh man.... this baby hums!

hmm.. and I thought mine was big at around 2K blocks.   I think I see a
"contest" coming: who has the biggest HISTORY.DB file?, or, actually, 
who has been surfing the most?!

Eager to hear what the INTERNET_TOOLS gurus say about this file...

 Josh

I'm not at work right now but there is an option somewhere that tells 
Netscape how long to remember your visited links for.

Since April 8th 1996, mine has grone to 5152 blocks.  When I get in in the
morning I'll find that damn option and clear some of that space out (I think).

On the PC the option is under General->Followed Links->Expires (never or put
in some delta time)

bjm

... so how do I view it?

I'd like to be able to look at my history just like I can
look at my bookmarks.

Andy V

    I bet if you open about:global you'll see it.  Unfortunately, it won't
    look like your bookmarks.  There was just a discussion about this in
    here recently...
    
    Cheers,
    
    Roy

hey, the NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697.EXE and its predecessor
NETSCAPE-EXPORT-ALPHA-V30B4-012697.EXE;  both give me a pop-up window when
first fired up, saying:

  This is a pre-release copy of Netscape Navigator that will expire on 
  Wed Apr 30 08:00:03 1997  

yada yada yada...

So... ?

thx, Josh

gotta question about clearing the disk cache. 

I have about 9000 blocks of files in my [.cache...] subdirectories.
My disk cache is set to the default 5000 kilobytes, a lot less than 9K blocks?
huh?  Anyway, I just want to clear out the cache and get some quota back.
So I click on Options..Network..Cache..Disk Cache.. Clear Disk Cache Now?, 
figuring it would delete the files.

Nope.  
The [.cache]index.db file goes to 0 blocks, but the 9000 blocks of files 
remain. Actually, something (dunno what) does get deleted, I see only about 
8000 blocks of files.  but why did it not delete all the files? I thought it 
would have done this.   This non-behavior is frustrating.  Am I supposed to 
manually go down into the 32 cache subdirectories and delete the files ?  
(Actually I'd just use CSWING and nuke the whole [.cache...] tree).  

 Josh

re: .282

Yeah, we'll be posting a new kit before this one expires.

re: .283

All the files should go away. Only the directory structure should remain. Check
the dates on the files - maybe they are orphaned files left by some incarnation
of the Navigator the abnormally exitted.

You might want to set the disk cache to zero - you can usually get better
performance that way (on OpenVMS). Then you can just nuke the entire cache tree
to tidy up. 

Colin.

> You might want to set the disk cache to zero - you can usually get better
> performance that way (on OpenVMS).

Are you suggesting always running with the disk cache set to zero?  Meaning no
files are ever stored on disk?

Paul

RE:  Setting disk cache to 0

I've actually done this and noticed significant speed ups.  At
first I did this mainly because my cache area was on my local
disk and that disk was amazingly loud and annoying.

- Rob

>> You might want to set the disk cache to zero - you can usually get better
>> performance that way (on OpenVMS).
>
>Are you suggesting always running with the disk cache set to zero?  Meaning no
>files are ever stored on disk?

Absolutely. It forces the Navigator to use the memory cache instead.

Colin.

    ... as long as your computer's got as much memory as it has disk ...

Netscape's memory cache is now set to 3000 KB.  I've got 128 MB on my
workstation.  Should I increase the memory cache from 3 MB?

Paul

>re: .283

>All the files should go away. Only the directory structure should remain. Check
>the dates on the files - maybe they are orphaned files left by some incarnation
>of the Navigator the abnormally exitted.

Some files are fairly recent (16-APR-1997).  Netscape crashes on me about once
a day with MALLOC errors/dumps. (BTW, what to I do about _this_ ? bump quotas?
which, to what?)

>You might want to set the disk cache to zero - you can usually get better
>performance that way (on OpenVMS). Then you can just nuke the entire cache tree
>to tidy up. 

I'll do this.   any other tricks of the OpenVMS-Netscape_trade to know about?

thx, Josh

Setting the disk cache to 0 has improved performance by an order of magnitude! 
Pages display very quickly now, and there's no more waiting twenty or thirty
seconds after pages paint before links become active!

I assume this is because Netscape Navigator on OpenVMS is ported from the Unix
version, and on Unix opening files is an inexpensive operation.

Unbelievable!

Paul

>I assume this is because Netscape Navigator on OpenVMS is ported from the Unix
>version, and on Unix opening files is an inexpensive operation.

Yes this is a big piece of it, but the most surprising thing is that Netscape 
will only use memory cache *after* disk cache is exhausted.  This is certainly
contrary to what a VMS person would assume.

The reasoning seems to be to store the files in a location that is persistent
across uses of the navigator to prevent the penalty of downloading a file over
a slow network as opposed to giving the best performance when revisiting pages
during a session.

-dale

> I assume this is because Netscape Navigator on OpenVMS is ported from the Unix
> version, and on Unix opening files is an inexpensive operation.

Dale is right. But also on UNIX, you get a lot of caching for free from the
file system, so even by using the "disk cache", you end up caching from memory
most of the time (and its disk-backed so its persistant across invokations,
hence the reason for the disk cache being the primary cache).

Yeah, I know VMS does caching too, but not in the same way as UNIX does. UNIX's
file system is very quick, but they don't make the same guarantees about the
locking, getting the data onto disk, etc as VMS does. This, coupled with VMS's
relatively slow file open/close operations, is why turning the disk cache off
actually improves overall performance.

Colin.

>Some files are fairly recent (16-APR-1997).  Netscape crashes on me about once
>a day with MALLOC errors/dumps. (BTW, what to I do about _this_ ? bump quotas?
>which, to what?)

What's your page file quota? Try at least 60,000 pages. It likes a lot :-)

> I'll do this.   any other tricks of the OpenVMS-Netscape_trade to know about?

Kill of the bookmarks file if its large. This was discussed already in this
notes file. This and other tips are in the FAQ. From the VMS Navigator,
HELP -> Frequently Asked Questions, then click on the third bullet for
OpenVMS FAQs.

Colin.

re:            <<< Note 3480.295 by STAR::BLAKE "OpenVMS Engineering" >>>
                             -< up PGFLQUO to 60k >-

I'm already at 95K.  From sho process/all:

21-APR-1997 10:26:33.51   User: BELKIN           Process ID:   2A802217
                          Node: QUOIN            Process name: "_FTA6:"

Terminal:           FTA6:
User Identifier:    [3TL,BELKIN]
Base priority:      4
Default file spec:  U0:[BELKIN]

Devices allocated:  QUOIN$FTA6:

Process Quotas:
 Account name: 3TL
 CPU limit:                      Infinite  Direct I/O limit:      1024
 Buffered I/O byte count quota:     99808  Buffered I/O limit:     512
 Timer queue entry quota:             100  Open file quota:        300
 Paging file quota:                 95744  Subprocess quota:        20
 Default page fault cluster:           64  AST quota:             1022
 Enqueue quota:                      3000  Shared file limit:        0
 Max detached processes:                0  Max active jobs:          0

Accounting information:
 Buffered I/O count:       461  Peak working set size:       3408
 Direct I/O count:         149  Peak virtual size:          24464
 Page faults:              926  Mounted volumes:                0
 Images activated:          14
 Elapsed CPU time:          0 00:00:01.14
 Connect time:              0 01:07:07.08

Authorized privileges:
 NETMBX    TMPMBX

Process privileges:
 NETMBX               may create network device
 TMPMBX               may create temporary mailbox

Process rights:
 INTERACTIVE
 LOCAL
 GRSPECS
 GRWRSPECS

System rights:
 SYS$NODE_QUOIN

Auto-unshelve: on

Process Dynamic Memory Area
    Current Size (bytes)         57344    Current Size (pagelets)        112
    Free Space (bytes)           33644    Space in Use (bytes)         23700
    Size of Largest Block        32780    Size of Smallest Block           8
    Number of Free Blocks            6    Free Blocks LEQU 64 Bytes        4

There is 1 process in this job:

>I'm already at 95K.  From sho process/all:

Try this:

$ mcr sysgen show virtualpagecnt

re:           <<< Note 3480.298 by STAR::BLAKE "OpenVMS Engineering" >>>
>$ mcr sysgen show virtualpagecnt

what, are you kidding?  to wit:

QUOIN>  mcr sysgen show virtualpagecnt
%DCL-W-ACTIMAGE, error activating image SYSGEN
-CLI-E-IMGNAME, image file QUOIN$DKA0:[SYS0.SYSCOMMON.][SYSEXE]SYSGEN.EXE;
-RMS-E-PRV, insufficient privilege or file protection violation

I don't have no sysprivs for this.  Will try to pester a sysmgr to do this
for me, but I'm already on their sour side because the 100 DPI fonts (or the
75/100 dpi density stuff in some .com file) that I use for my VMS workstation 
somehow got blown away in a huge DECNET? plant-wide crash in NIO last week, 
and they're not sure how to fix it yet, and I'm stuck right now with the
BIG fonts for everything.

 Josh

>> >$ mcr sysgen show virtualpagecnt

OK, try

	$ write sys$output f$getsyi("virtualpagecnt")

ok, tried -.1 and got:

QUOIN> $ write sys$output f$getsyi("virtualpagecnt")
600000

 -Josh

> hey, the NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-020697.EXE and its predecessor
> NETSCAPE-EXPORT-ALPHA-V30B4-012697.EXE;  both give me a pop-up window when
> first fired up, saying:
> 
>   This is a pre-release copy of Netscape Navigator that will expire on 
>   Wed Apr 30 08:00:03 1997  
    
New kits are available from BULOVA::IPS$KITS:

  NETSCAPE-EXPORT-ALPHA-V30B5-041597.EXE;1                  15212/15213
  NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597.EXE;1             17020/17022
  NETSCAPE-EXPORT-VAX-V30B5-041597.EXE;1                     9065/9066

(Also available via FTP from ftp://bulova.zko.dec.com/ips$kits:<name-above>)

For those running the Java version, the class libraries file has not
changed, so you DON'T need to copy it again.

These images should appear on the external OpenVMS web site over the
weekend.
    
> yada yada yada...
    
Saw Seinfeld last night, huh :-)

    I don't want to seem ungrateful, but besides the "yeah, it crashes for
    me, too" and the "it doesn't crash here messages," has anyone gotten
    any help getting the crash when printing from Netscape solved?  Here
    is the dump:

$ netscape
%SYSTEM-F-ILLEGAL_SHADOW, illegal formed trap shadow, Imask=00000000, Fmask=0040
0000, summary=21, PC=000000000039C8E0, PS=0000001B
%TRACE-F-TRACEBACK, symbolic stack dump follows
  image    module    routine             line      rel PC           abs PC
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  PS  xl_begin_document
                                        40057 0000000000000260 000000000039C8E0
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  PRINT  xl_generate_postscript
                                        40375 00000000000002D8 0000000000399058
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  PRINT  PSFE_AllConnectionsComplete
                                        41085 00000000000020E0 000000000039AE60
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MKGETURL  net_CheckForWaitingURL
                                        72659 0000000000000F68 00000000003A8DB8
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MKGETURL  NET_ProcessNet
                                        74667 0000000000004510 00000000003AC360
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  XFE  fe_stream_callback
                                       100630 0000000000000CFC 000000000033843C
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  XFE  fe_VMS_AppInput
                                       100690 0000000000000F3C 000000000033867C
                                            0 FFFFFFFF807A6774 FFFFFFFF807A6774
                                            0 FFFFFFFF807A71F0 FFFFFFFF807A71F0
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MOZILLA  fe_EventLoop
                                       107188 0000000000000334 00000000002FBC94
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  DIALOGS  fe_await_synchronous_url
                                        81443 0000000000004F80 00000000002BEB10
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  DIALOGS  fe_Print
                                        84269 0000000000009CCC 00000000002C385C
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  DIALOGS  fe_print_cb
                                        84325 000000000000A058 00000000002C3BE8
                                            0 FFFFFFFF80789A60 FFFFFFFF80789A60
                                            0 FFFFFFFF808124D4 FFFFFFFF808124D4
                                            0 FFFFFFFF80789A60 FFFFFFFF80789A60
                                            0 FFFFFFFF807EC03C FFFFFFFF807EC03C
                                            0 FFFFFFFF807E755C FFFFFFFF807E755C
                                            0 FFFFFFFF8094932C FFFFFFFF8094932C
                                            0 FFFFFFFF8091E38C FFFFFFFF8091E38C
                                            0 FFFFFFFF807C49C0 FFFFFFFF807C49C0
                                            0 FFFFFFFF807C5B3C FFFFFFFF807C5B3C
                                            0 FFFFFFFF80797B58 FFFFFFFF80797B58
                                            0 FFFFFFFF807982C8 FFFFFFFF807982C8
                                            0 FFFFFFFF807A73AC FFFFFFFF807A73AC
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MOZILLA  fe_EventLoop
                                       107162 00000000000002FC 00000000002FBC5C
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MOZILLA  main
                                       108929 0000000000003240 00000000002FEBA0
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MOZILLA  __main
                                            0 000000000000006C 00000000002FB9CC
 PTHREAD$RTL                                0 000000000004C148 000000000084C148
 PTHREAD$RTL                                0 0000000000030664 0000000000830664
                                            0 FFFFFFFF8269B0D8 FFFFFFFF8269B0D8


The image that I'm running is:

NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597.EXE

and the checksums from that image are:
NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597.EXE;1
image section %D'1' checksum is %X'78A71EC5'
image section %D'2' checksum is %X'E9170CC6'
image section %D'3' checksum is %X'3CF6EBD6'
image section %D'5' checksum is %X'406D70ED'
image header checksum is %X'0023F8F4'
checksum of all image sections is %X'ED2B8938'

I'm running OpenVMS/Alpha V7.1 with Motif V1.2-4.  UAF parameters are:

Maxjobs:         0  Fillm:       200  Bytlm:       256000
Maxacctjobs:     0  Shrfillm:      0  Pbytlm:           0
Maxdetach:       0  BIOlm:       300  JTquota:       4096
Prclm:          10  DIOlm:       300  WSdef:         8192
Prio:            4  ASTlm:       500  WSquo:        32767
Queprio:         0  TQElm:       120  WSextent:     65536
CPU:        (none)  Enqlm:      2000  Pgflquo:     600000

VIRTUALPAGECNT is:

$ write sys$output f$getsyi("virtualpagecnt")
2147483647

The system has 128MB of memory, and all the DECw giblets are:

DECwindows ident is                     DW V1.2-4960312
DECwindows server ident is              DW V7.1-961126
DECwindows transport ident is           DW T6.2-961126
DECwindows xlib ident is                DW V1.2-4960312
DECwindows motif toolkit ident is       DW V1.2-4960313
DECwindows apps ident is                DW V1.2-4960312
DECwindows programming ident is         DW V1.2-4960312

Yeah, setting the disk cache to 0 is a real hoot!

It'd still be nice to see something that will help with the crash issue, or
at least an indication of where it's going awry...

bob

>    I don't want to seem ungrateful, but besides the "yeah, it crashes for
>    me, too" and the "it doesn't crash here messages," has anyone gotten
>    any help getting the crash when printing from Netscape solved?  Here
>    is the dump:

OpenVMS doesn't have a permanent engineering team out at Netscape working on
porting their products (we are in the process of changing this).

We will look into this problem (and all other reported problems) when we next
have engineers at Netscape.

Colin.

Hi!
	I've verified with V2.02 that their site is fine.  When I try to login
to the NETbuyer ( www.netbuyer.com ) webserver with Netscape Gold 3.0B5 on VAX,
Netscape just hangs.  No messages are displayed.  As soon as I click on the
Login button, it's out to lunch.  This is easily reproducible.  They are not
using a secure link, just a form requesting username and password.

	Anyone else see this with any other sites?

	David

I can get to the form requesting User Name and Password on both Alpha and VAX
versions.

	NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597.EXE;1
	NETSCAPE-EXPORT-VAX-V30B5-041597.EXE;1

Paul

Hi Paul!

	I don't get it.  If it works over there, why doesn't it work for me here
in Salem?  Guess I'll have to look into this a little more.  I have the same
version you do.  Although I'm only using the VAX version currently.

	Thanks for checking for me!

	David

    
    re: .304.  Thanks.  At least that's an indication that the voices have
    been heard, and a reason for the situation.  Looking forward to hearing
    back on this.  Just as further info, I did see/receive the Field Test
    announcement, but got it like the day before closure (at the time I was
    on-customer-site).  So it was difficult to do much checking...
    
    bob

Me too. Check you have some reasonable process quotas. See the FAQs for
details.

Colin.

I let it sit last night trying to log in.  Got the following stack dump:


%SYSTEM-F-ACCVIO, access violation, reason mask=01, virtual address=43657950, PC
=003C18C1, PSL=03C00008
%TRACE-F-TRACEBACK, symbolic stack dump follows
module name     routine name                     line       rel PC    abs PC

PRDTOA          mult                             4599      00000041  003C18C1
PRDTOA          pow5mult                         4680      00000064  003C1AB8
PRDTOA          PR_dtoa                          6243      000006B8  003C3304
PRDTOA          PR_cnvtf                         6438      00000022  003C3746
MO_NUM          mocha_NumberToString            15611      00000066  00373EE2
MO_STR          mocha_RawDatumToString          16074      0000026D  00374E8D
MO_STR          mocha_DatumToStringVM           16091      00000031  00374EF5
MOCHA           mocha_Interpret                 16976      00001B78  0036D5C8
MOCHA           Call                            16361      00000350  0036B52C
MOCHA           mocha_Interpret                 17321      00002AC1  0036E511
MOCHA           Call                            16361      00000350  0036B52C
MOCHA           mocha_Call                      16541      000000BF  0036B987
MOCHAAPI        MOCHA_CallMethod                19653      00000083  00369DAF
LM_INIT         lm_SendEvent                    48561      0000006D  0035DEB9
LM_FORM         form_event                      48270      000000A1  0036250D
LM_FORM         LM_SendOnSubmit                 48311      0000001B  003625DF
LAY             fe_HandleHREF                   96393      00000131  000EF6A5
LAY             fe_activate_link_action         96748      00000180  000EFC04
                                                           004F3400  004F3400
                                                           004F385D  004F385D
                                                           004F3D27  004F3D27
                                                           004ECEAB  004ECEAB
                                                           004ED5C4  004ED5C4
                                                           004ED68C  004ED68C
                                                           004EE90B  004EE90B
MOZILLA         fe_EventLoop                    *****      000000BC  000F9C24
MOZILLA         main                            *****      000010F7  000FB6B7
%SYSTEM-F-ACCVIO, access violation, reason mask=05, virtual address=DD01CAAC, PC
=0041D322, PSL=03C00000

  Improperly handled condition, image exit forced.

        Signal arguments              Stack contents

        Number = 00000005                20040000
        Name   = 0000000C                7FE7E56C
                 00000005                7FE7E530
                 DD01CAAC                0042ED55
                 0041D322                7FE7EA86
                 03C00000                00000002
                                         7FE7E950
                                         00000004
                                         7FE7E94C
                                         00000000

        Register dump

        R0 = 0043B100  R1 = 00000008  R2 = 7FE7E52C  R3 = 0000002F
        R4 = 00000000  R5 = 005241F0  R6 = 0043AE08  R7 = 00000010
        R8 = 0058BB00  R9 = 7FE7E954  R10= 00095D30  R11= 7FE7E95C
        AP = 7FE7E4CC  FP = 7FE7E48C  SP = 7FE7E508  PC = 0041D322
        PSL= 03C00000



I'm on my own VAXstation here and have plenty of quota.  I'll check the FAQ
though.  This didn't occur on previous V3.0 beta versions.  Just this one.

	David

I searched the 3480 topic string looking for a FAQ and couldn't find one.  Is
there an installation document around that would tell me what my quotas should
be?

	Thanks!
		David

>I searched the 3480 topic string looking for a FAQ and couldn't find one.  Is
>there an installation document around that would tell me what my quotas should
>be?

From the Navigator (OpenVMS version), click on HELP on the menu bar, and then
select Frequently Asked Questions. From that page select the third bullet, the
one that says OpenVMS something.

Colin.

Colin,
	I didn't realize that was there.  I'll take a look.

	Regards,
		David

For people who are seeing ACCVIO when trying to print, try setting the paper
orientation to landscape. Works for me!

Colin.

The landscape trick doesn't work for me.  Netscape still goes away.

Paul

Same here...  And it also dosen't seem to matter if you're trying to create a
file either.

Chris

    
    Agreed; it dumps here regardless.  Here's the screen data:
    
$ netscape
%SYSTEM-F-ILLEGAL_SHADOW, illegal formed trap shadow, Imask=00000000, Fmask=0040
0000, summary=21, PC=000000000039C8E0, PS=0000001B
%TRACE-F-TRACEBACK, symbolic stack dump follows
  image    module    routine             line      rel PC           abs PC
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  PS  xl_begin_document
                                        40057 0000000000000260 000000000039C8E0
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  PRINT  xl_generate_postscript
                                        40375 00000000000002D8 0000000000399058
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  PRINT  PSFE_AllConnectionsComplete
                                        41085 00000000000020E0 000000000039AE60
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MKGETURL  net_CheckForWaitingURL
                                        72659 0000000000000F68 00000000003A8DB8
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MKGETURL  NET_ProcessNet
                                        74667 0000000000004510 00000000003AC360
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  XFE  fe_stream_callback
                                       100630 0000000000000CFC 000000000033843C
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  XFE  fe_VMS_AppInput
                                       100690 0000000000000F3C 000000000033867C
                                            0 FFFFFFFF807A6774 FFFFFFFF807A6774
                                            0 FFFFFFFF807A71F0 FFFFFFFF807A71F0
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MOZILLA  fe_EventLoop
                                       107188 0000000000000334 00000000002FBC94
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  DIALOGS  fe_await_synchronous_url
                                        81443 0000000000004F80 00000000002BEB10
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  DIALOGS  fe_Print
                                        84269 0000000000009CCC 00000000002C385C
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  DIALOGS  fe_print_cb
                                        84325 000000000000A058 00000000002C3BE8
                                            0 FFFFFFFF80789A60 FFFFFFFF80789A60
                                            0 FFFFFFFF808124D4 FFFFFFFF808124D4
                                            0 FFFFFFFF80789A60 FFFFFFFF80789A60
                                            0 FFFFFFFF807EC03C FFFFFFFF807EC03C
                                            0 FFFFFFFF807E755C FFFFFFFF807E755C
                                            0 FFFFFFFF8094932C FFFFFFFF8094932C
                                            0 FFFFFFFF8091E38C FFFFFFFF8091E38C
                                            0 FFFFFFFF807C49C0 FFFFFFFF807C49C0
                                            0 FFFFFFFF807C5B3C FFFFFFFF807C5B3C
                                            0 FFFFFFFF80797B58 FFFFFFFF80797B58
                                            0 FFFFFFFF807982C8 FFFFFFFF807982C8
                                            0 FFFFFFFF807A73AC FFFFFFFF807A73AC
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MOZILLA  fe_EventLoop
                                       107162 00000000000002FC 00000000002FBC5C
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MOZILLA  main
                                       108929 0000000000003240 00000000002FEBA0
 NETSCAPE-EXPORT-ALPHA-V30B5-JAVA-041597  MOZILLA  __main
                                            0 000000000000006C 00000000002FB9CC
 PTHREAD$RTL                                0 000000000004C148 000000000084C148
 PTHREAD$RTL                                0 0000000000030664 0000000000830664
                                            0 FFFFFFFF8269B0D8 FFFFFFFF8269B0D8

    
    Looks like the ILLEGAL_SHADOW trap is a OpenVMS V7.1 issue.
    see attached.
    
    Thanks
    
    Powell Hazzard
    Netscape Engineering
    
    
    Crossposted in the VMSNOTES and DECC conferences:

    Hello,

    I'm running into a ieee floating point conversion issue with OpenVMS
    Alpha V7.1.  The attached program works under OpenVMS 6.1 and OpenVMS
    6.2, but fails with an ILLEGAL_SHADOW trap error on V7.1.  I'm seeing
    this with a few IEEE applications,  Netscape Navigator, Java JDK For 
    OpenVMS.  I also noticed where a customer has reported the failure with 
    SoftWindows.

    OpenVMS V7.1 Alpha
    DEC C T5.6-001 on OpenVMS Alpha V7.1 and earlier.

    Any Clues with what changed?

    Thanks

    Powell Hazzard

/*

 Problems with ILLEGAL_SHADOW trap on OpenVMS V7.1

How to compile/link:

$ cc/pref=all/float=ieee/ieee=denorm test_ieee
$ link test_ieee

Output on OpenVMS V6.2

$ run test_ieee
double value -0.000000
value should be 0
int value 0

Output on OpenVMS V7.1:

$ run test_ieee
double value -0.000000
value should be 0
%SYSTEM-F-ILLEGAL_SHADOW, illegal formed trap shadow, Imask=00000000, Fmask=0000
0002, summary=03, PC=0000000000020148, PS=0000001B
%TRACE-F-TRACEBACK, symbolic stack dump follows
  image    module    routine             line      rel PC           abs PC
 TEST_IEEE  TEST_IEEE  main              1413 0000000000000148 0000000000020148
 TEST_IEEE  TEST_IEEE  __main               0 0000000000000064 0000000000020064
                                            0 FFFFFFFF852B90D8 FFFFFFFF852B90D8

*/

#include <stdio.h>
#include <math.h>

double *dbl;
unsigned unsigned_value[2];
int ret_value;
main()
{

	unsigned_value[0] = 1;
        unsigned_value[1] = 0x80000000;
	dbl = (double *) &unsigned_value[0];
	printf("double value %f\n",*dbl);
	if (*dbl < 1.0e-307 && *dbl > -1.0e-307)
		printf("value should be 0\n");

	ret_value = *dbl;
	printf("int value %d\n",ret_value);
}