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 |
Is there any difference in the hardware and software used to drive the VAXstation 3100's 8-plane GPX configuration and that used to drive the VAXstation II's 8-plane GPX configuration? Obviously, the CPU and a variety of other things are different, but I'm getting what appears to be rather slow performance on my 3100 when changing bunches of colors through XStoreColors() where the color table entries that I'm modifying are 'random'. The exact scenario is with colormap animation. The code that I have works fine on a VAXstation II/GPX 8-plane using the default colormap or a private colormap. It also works fine on my 3100/GPX 8-plane using a private colormap, but not with the default colormap. Running everything synchronously with the server display allows me to see that when I request the XStoreColors(), that the colors seem to change rather slowly. That is, the transition of the two display 'frames' is very much visible. The before and after appearance are correct, however (first one frame is visible, then the other frame is visible). In short, everything works fine, just not quickly enough. So what is there that's different about the 3100's use of color tables that prevents it from functioning at speed in this particular case? I've even sorted the pixels that I'm changing so that they are in ascending pixel order (but they're not consecutive pixels). John
T.R | Title | User | Personal Name | Date | Lines |
---|---|---|---|---|---|
1409.1 | STAR::MCLEMAN | Jeff McLeman, VMS Development | Mon Sep 11 1989 13:17 | 10 | |
I know that there is a big difference in how the hardware colormap works. The Qbus GPX can be updated using MOVC3 instructions. The VS3100 (and VS200) color option's DAC (digital to analog converter) is different. The entries are not aligned as such to use the same code. This must be done with a loop, which is somewhat slower. At least that is how it was done on the VWS side of things, cause I wrote the driver support. But the DECwindows driver does the same thing. Jeff | |||||
1409.2 | Thanks, Jeff | EPIK::BUEHLER | Oh good. I like bustlin' | Mon Sep 11 1989 15:46 | 7 |
So you're saying that 'to the best of your knowledge' this is an artifact of the hardware on the 3100 and that I'm out of luck? Is there nothing that you know of that I can do to accelerate the loading of colors? John | |||||
1409.3 | STAR::MCLEMAN | Jeff McLeman, VMS Development | Mon Sep 11 1989 16:08 | 2 | |
I can't think of one off hand. | |||||
1409.4 | DECWIN::FISHER | Burns Fisher 381-1466, ZKO3-4/W23 | Mon Sep 11 1989 18:33 | 6 | |
For what it's worth, the server does some additional work when the colormap updates are not contiguous, cuz it can package up the updates into a single driver packet as long as they are. Burns |