home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!sdd.hp.com!uakari.primate.wisc.edu!doug.cae.wisc.edu!umn.edu!orstnews!ucs.orst.edu!hilmera
- From: hilmera@ucs.orst.edu (Andrew Hilmer)
- Newsgroups: comp.os.os2.misc
- Subject: Trident 800x256 & downloading problems
- Message-ID: <1992Aug20.171356.18187@talon.ucs.orst.edu>
- Date: 20 Aug 92 17:13:56 GMT
- Sender: usenet@talon.ucs.orst.edu (Usenet News admin)
- Organization: University Computing Services - OSU
- Lines: 96
- Nntp-Posting-Host: ucs.orst.edu
-
- Machine: generic asian clone, AMD 386/33, 64kSRAM, 8MbRAM
- 65 meg RLL drive, Supra 2400 internal modem
- Trident 8900c w/1meg
-
- Problems: Difficulty with background downloads.
- Non-fatal corruption of the WPS caused
- by the CMD.EXE shell, esp. in 132x43 mode.
-
- 1. I have not been able to download more than the simplest things
- in the background during downloads. I have tried various communication
- programs, including the bundled applet, Telix, and a few others. I
- use PMqvt to download. I am able to play solitaire and other PM games
- while downloading, but if I start to switch to and from full screen
- sessions, start up emacs, or do some djpegging, I will peek back in on
- PMqvt and it will be recovering from errors, if it hasn't already
- acked. Needless to say, TE/2 (unregistered) is useless for
- downloading, since it dies with only the slightest prodding.
-
- My IO card shouldn't have anything to do with it because my modem is
- on a card. Does drive access with RLL take up so much resource that it
- would do it? Ever since the install disasters, it doesn't seem to be
- THAT slow, compared to what others have said on the net. I am
- clueless.
-
- 2. Originally, I was running the plain 800x600x16 driver, and I
- don't remember having any problems. However, I am not sure that I
- discovered the 132x43 mode until after I installed the virtual desktop
- driver.
-
- My initial ecstasy at finding this driver was short lived when I
- switched to the desktop and found the whole lower half of the
- 1024x1024 screen covered with a strange pattern of corruption. That is
- the LOWER half, which, I assume, points to some hanky panky in the
- lower(?) 512Kb on my video card. The corruption consists of many small
- vertical lines maybe 8-10 pixels high and about 2 wide. They are
- arranged very regularly about 30 pixels apart horizontally and about
- 60 pixels apart vertically (32x64?).
-
- |---------------------|
- | |
- | |
- | |
- | |
- |'''''''''''''''''''''| <-- Like so.
- |'''''''''''''''''''''|
- |'''''''''''''''''''''|
- |'''''''''''''''''''''|
- |---------------------|
-
- The effect happens in this scenario. At the WPS. Start CMD.EXE. Switch
- back to the WPS. A row of these lines will appear at the very bottom.
- More on this row later.
-
- If I run 132x43, this is what will happen. Start CMD.EXE. Switch text
- mode to 132x43. Alt-Esc (or Ctrl-Esc) back to the WPS. WPS has just
- the one line at the bottom. Switch back to the full-screen session.
- Switch back to WPS. Full corruption is now there.
-
- Refresh takes care of any corruption that has not been laid on a
- window. To fix a window requires that it be completely redrawn.
-
- The full screen session is not in any way corrupted.
-
- If I start a Dos session, the WPS is immediately cured until the next
- time I switch to an OS/2 full screen session.
-
- The lowest row of corruption lines are not spaced ~30 pixels apart,
- but are spaced much closer together. They also appear to be finer in
- width. Their vertical spacing away from the lines above them is not
- unique. However, while they look as though the spacing of every fifth
- line is the same as those above, they are offset to the right a few
- pixels, instead of lining up into the columns made by the lines above
- them.
-
- The coloration of this lowest row of lines is different as well. Every
- fifth line is just as randomly colored as all of the other lines. The
- three lines in between each of these, however, are uniformly black.
-
- Also, when I was using the virtual desktop, I noticed that if I
- repeated the switching during the course of work, movement of the
- desktop produced more and more offset sets of corruption until I was
- annoyed enough to start a DOS session to clear it all.
-
- I am now using the plain 800x600x256 driver (I got tired of chasing
- help windows all over hell and gone :-( ) and the same thing happens,
- only it affects the whole desktop.
-
- I may have been remiss not to report this earlier, but I think I am
- making up that error with raw detail. I would like to find out how to
- report this bug to Trident properly. Also, if helpful, I would not
- mind downloading Nikon to see if it can take snapshots of the
- corruption on both the plain and virtual displays.
-
- andy
- --
- hilmera@ucs.orst.edu Why settle for the *lesser* of evils? Cthulu 92!
-