[Qgis-user] qgis extremly slow on grass data. More info

[Jarek Jasiewicz]

Hi
I tried to gather more information about the problem to create ticket, 
and it seems a different problem. Map was created with Bovrey 
sharpening. It is Uint16 and its color table has 2050 entities.

First I trim the size of the map to 853x943 (it is not large data set), 
but the problem still exist.

I tried to export that map to UInt16 GtIff to attach it, and I cannot, I 
recive empty (null bitmap) file. I can it still export to Int16Gtiff as 
gray without the problem

the warnings I recive on console from qgis:

Warning: QPainter::end: Painter not active, aborted
Warning: QPainter::begin: Cannot paint on a null pixmap
Warning: QPainter::end: Painter not active, aborted
Warning: QPixmap::scaled: Pixmap is a null pixmap

So It looks rather like gdal problem with that particular map (Uint16 
with full color table)

If I remove that color table to more simple (for example grey) 
everything works fine. So If I send the sample dataset problem may not 
be reproducable.

To reproduce this problem I did as following:

In grass I used region as before: 853x943


r.surf.random output=random min=0 max=32100
r.mapcalc int_random = int(random)
g.copy rast=int_random,int_random_col
r.colors map=int_random_col color=random


And next I add data to qgis. With map named "random" there were no 
problems. With map with "random_col" where I assigned random pallete 
time of reading was over 5 minut (after that I kill qgis)

So if someone more experienced can reproduce that problem as I described 
and distinguish it is gdal or qgis problem.


greetings

Jarek

configuration:
Ubuntu 8.04 i386, grass 6.4.svn, gdal 1.5.3 with grass plugin
Qt 4.3.4
RAM 2GB, CPU AMD4800+
last svn up compilation of qgis and grass: 12.11.2008
gdal ins't svn




Tim Sutton pisze:
> Hi Jarek
>
> Could you open a ticket in trac for this and if possible attach a sample
> dataset that replicates the issue. Be sure to include also the
> configuration details from your message below.
>
> Thanks
>
>
>