<dariusz@_NO-SPAM_mnsi.net> wrote: > I am consistently getting a SYS3175 when scrolling down in a post to the last > line, the following shows up in the POPUPLOG.OS2:
> Using V 1.60.cp125 version, has anyone seen this?
No.
Can you open a command prompt and cd to your \pronews directory and run
showhead name.of.the.group.affected
This will spit out some output that contains the filenames of the files that contain the data for that group - e.g.
[I:\pronews]showhead local.test headers.idx contains 64 group entries Read group entries from headers.idx and closed file Successfully opened file .\HEADERS\0235731c.IDX Information from file .\HEADERS\0235731c.IDX for group local.test [snip]
now zip up and send me all files in \pronews\articles\0235731c.* \pronews\headers\0235731c.* (obviously changing the filenames to match the ones returned by showhead on your system).
Followups set to comp.os.os2.mail-news
-- Trevor Hemsley, Brighton, UK Trevor dot Hemsley at ntlworld dot com
<Trevor.Hems...@mytrousers.ntlworld.com> wrote: > On Sun, 9 Aug 2009 01:45:50 UTC in comp.os.os2.bugs, "Dariusz Piatkowski" > <dariusz@_NO-SPAM_mnsi.net> wrote:
> > I am consistently getting a SYS3175 when scrolling down in a post to the last > > line, the following shows up in the POPUPLOG.OS2:
> No.
> Can you open a command prompt and cd to your \pronews directory and run
> showhead name.of.the.group.affected .... > now zip up and send me all files in \pronews\articles\0235731c.* > \pronews\headers\0235731c.* (obviously changing the filenames to match the ones > returned by showhead on your system).
> Followups set to comp.os.os2.mail-news
Email with pertinent debug information sent...thanks for supporting this app!
Figures. I was scrolling to the bottom of _your_ post, Trevor, and it died. Fortunately, not in the middle of my repying to you.
> Can you open a command prompt and cd to your \pronews directory and run
> showhead name.of.the.group.affected
> This will spit out some output that contains the filenames of the files that > contain the data for that group - e.g.
> [I:\pronews]showhead local.test > headers.idx contains 64 group entries > Read group entries from headers.idx and closed file > Successfully opened file .\HEADERS\0235731c.IDX > Information from file .\HEADERS\0235731c.IDX for group local.test > [snip]
[D:\pronews]showhead comp.os.os2.bugs headers.idx contains 30 group entries Read group entries from headers.idx and closed file Successfully opened file .\HEADERS\0206404d.IDX Information from file .\HEADERS\0206404d.IDX for group comp.os.os2.bugs
> now zip up and send me all files in \pronews\articles\0235731c.* > \pronews\headers\0235731c.* (obviously changing the filenames to match the ones > returned by showhead on your system).
On Wed, 12 Aug 2009 01:44:10 UTC in comp.os.os2.mail-news, "Dariusz Piatkowski"
<dariusz@_NO-SPAM_mnsi.net> wrote: > On Sun, 9 Aug 2009 10:57:07 UTC, "Trevor Hemsley" > <Trevor.Hems...@mytrousers.ntlworld.com> wrote:
> > On Sun, 9 Aug 2009 01:45:50 UTC in comp.os.os2.bugs, "Dariusz Piatkowski" > > <dariusz@_NO-SPAM_mnsi.net> wrote:
> > > I am consistently getting a SYS3175 when scrolling down in a post to the last > > > line, the following shows up in the POPUPLOG.OS2:
This doesn't happen here with the files you sent me but maybe I am not scrolling down in the same way that you are (there are several different ways to scroll through a post).
Can you and Al please tell me what method of scrolling you usually use when you see this problem.
-- Trevor Hemsley, Brighton, UK Trevor dot Hemsley at ntlworld dot com
<Trevor.Hems...@mytrousers.ntlworld.com> wrote: > On Wed, 12 Aug 2009 01:44:10 UTC in comp.os.os2.mail-news, "Dariusz Piatkowski" > <dariusz@_NO-SPAM_mnsi.net> wrote:
> > On Sun, 9 Aug 2009 10:57:07 UTC, "Trevor Hemsley" > > <Trevor.Hems...@mytrousers.ntlworld.com> wrote:
> > > On Sun, 9 Aug 2009 01:45:50 UTC in comp.os.os2.bugs, "Dariusz Piatkowski" > > > <dariusz@_NO-SPAM_mnsi.net> wrote:
> > > > I am consistently getting a SYS3175 when scrolling down in a post to the last > > > > line, the following shows up in the POPUPLOG.OS2:
> This doesn't happen here with the files you sent me but maybe I am not scrolling > down in the same way that you are (there are several different ways to scroll > through a post).
> Can you and Al please tell me what method of scrolling you usually use when you > see this problem.
I have an IBM ScrollPoint optical mouse. Using Mouse driver v10.152, dated Jan-03-2005.
I am using the 'scroll stick' on the mouse to scroll through the posts. My mouse object is set to use "Simulated Keys", they are defined as follows (text copy from the HELP screen below), but I do NOT have either one of the options checked, so I actually assume the mouse is sending arrow up/down events to the window:
======== START ============ Scroll type
This function specifies the type of scrolling to be done if enabled. If scrolling is enabled, the user can choose between PM Scroll and Simulated keys scrolling.
None - Scrolling is disabled for all scrollable (wheel, stick, TrackPoint and other three button mice) devices.
Use PM Scroll - Presentation Manager scroll messages are sent to active window when scroll events are detected. Scrolling is determined by window settings. The cursor does not move with the window.
Always send scroll messages - Scroll messages are sent to windows including those that do not have visible scrollbars. Some applications may experience problems with this feature. It should be disabled to have scroll messages sent to windows with visible scrollbars only.
Use Simulated Keys
Use PageUp/PageDn - Defines how fast the cursor changes position vertically. When checked, the scroll rate will change from single line scrolling to multiple line scrolling. Scrolling with an ordinary three button mouse is based upon the acceleration of mouse movement. An ordinary three button mouse is more effective if the scrolling speed is set to fast. When unchecked, the scroll rate remains constant from line to line.
Use Ctrl+Left Arrow/Ctrl+Right Arrow - Defines how fast the cursor changes position horizontally. When checked, the scroll rate will change from single character scrolling to word boundary scrolling. When unchecked, the scroll rate remains constant from character to character. ========== END ===========
On Mon, 17 Aug 2009 21:12:58 UTC, "Trevor Hemsley"
<Trevor.Hems...@mytrousers.ntlworld.com> wrote: > Can you and Al please tell me what method of scrolling you usually use when you > see this problem.
Sure, since it just happened again when I was scrolling through Dariusz' reply to you . . .
I *always* use the <CurDn> key (the "down arrow") to scroll. I've been trying to pay attention the last couple of days, and I've been using the mouse on the vertical scroll bar every so often, to AVOID crashing when scrolling posts, and it never crashes if I use the mouse scroll bar. <CurDn> to the end of the post it looks like. I think the cursor has to actually reach the end of the post for the crash to occur.
On Tue, 18 Aug 2009 03:54:33 UTC in comp.os.os2.mail-news, "Al Savage"
<asav...@iname.com> wrote: > On Mon, 17 Aug 2009 21:12:58 UTC, "Trevor Hemsley" > <Trevor.Hems...@mytrousers.ntlworld.com> wrote:
> > Can you and Al please tell me what method of scrolling you usually use when you > > see this problem.
> Sure, since it just happened again when I was scrolling through Dariusz' > reply to you . . .
> I *always* use the <CurDn> key (the "down arrow") to scroll.
Which seems to be what the two of you have in common as far as I can see.
> I've been > trying to pay attention the last couple of days, and I've been using the > mouse on the vertical scroll bar every so often, to AVOID crashing when > scrolling posts, and it never crashes if I use the mouse scroll bar. > <CurDn> to the end of the post it looks like. I think the cursor has to > actually reach the end of the post for the crash to occur.
I have no idea what's going on. I've taken Darius' files and used those - and he says that he consistently crashes with those. There are only 6 bodies in his group of 2121 headers so it should be a _lot_ easier to debug than your 7MB monster (which I haven't even downloaded because I don't see how I'd find a problem post in among that lot).
One wild idea that occurs to me: what is the date/timestamp/bldlevel of your uconv.dll?
[I:\pronews2]which -al uconv.dll h:\os2\dll\UCONV.DLL [I:\pronews2]bldlevel h:\os2\dll\UCONV.DLL Build Level Display Facility Version 6.10.480 Oct 6 2000 (C) Copyright IBM Corporation 1993-2000 Signature: @#IBM:14.093#@ Unicode Conversion Functions - Copyright IBM Corp., 1994, 1997 Vendor: IBM Revision: 14.93 File Version: 14.93 Description: Unicode Conversion Functions - Copyright IBM Corp., 1994, 1997
[I:\pronews2]dir h:\os2\dll\UCONV.DLL
The volume label in drive H is WSEB. The Volume Serial Number is 6A2A:4015. Directory of H:\os2\dll
4-09-02 5:30p 23331 61 UCONV.DLL
(That's a European date dd-mm-yy)
-- Trevor Hemsley, Brighton, UK Trevor dot Hemsley at ntlworld dot com
Trevor Hemsley wrote: > I have no idea what's going on. I've taken Darius' files and used those - > and he says that he consistently crashes with those. There are only 6 > bodies in his group of 2121 headers so it should be a lot easier to debug > than your 7MB monster (which I haven't even downloaded because I don't see > how I'd find a problem post in among that lot).
Trevor, there used to be a problem in the message handling if a task didn't handle keystrokes properly - too many would crash the thread with an overflow. I ran into this with the speed users who who had set the keyboard repeat rate very high. I had to finally check the cursor position and swallow the navigation keys when the end of the scroll list was reached. Been a while but it seems to me that the actual culprit was way down in the WPS class code and it couldn't keep up with the key strokes.
