Could we have this fixed asap for binary distributions? This is
a really important one in practical terms, and of general interest. (For
example, I have "real projects" split
in different qgs files because of this problem.)
Agus
Andreas Neumann wrote:
Hello,
I'd like to inform you that Jürgen provided a fix for the problem, which
is now in trunk. Thank you Jürgen!
The problem really was the primary key detection, which was checked with
every project load and took very long for complex views/tables. Now the
pkey check is done with the initial layer loading and stored in the
project file. When loading the layer it seemse like it first starts the
primary key detection with the stored key attribute from the project files
and stops searching for further keys after it confirmed that the suggested
key works - at least this is how I understand the current behavior in
trunk - Jürgen - please correct me if I am wrong.
Anyway - the project that previously loaded in 3-4 minutes now loads in
approx. 10 seconds.
I hope that this can still make it into the 1.1 release, as this is a big
problem for people with more complex tables/views.
Thanks,
Andreas
On Fri, April 24, 2009 10:59 am, Paolo Cavallini wrote:
Andreas Neumann ha scritto:
Here is what I did so far:
I saved the project as a different name. So far this did not speed
things
up. I did a project file comparison in the OxygenXML editor. There are
only minor differences, like
Very detailed report, thanks a lot Andreas!
Hopefully jef will be able to understand what's going on.
If I can help with testing, just let me know.
All the best.
--
Paolo Cavallini: http://www.faunalia.it/pc
begin:vcard
fn:Agustin Lobo
n:Lobo;Agustin
org:Institut de Ciencies de la Terra "Jaume Almera" CSIC
adr:;;Lluis Sole Sabris s/n;Barcelona;;08028;Spain
email;internet:agustin.l...@ija.csic.es
url:http://www.ija.csic.es/gt/obster
version:2.1
end:vcard
_______________________________________________
Qgis-user mailing list
Qgis-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-user
