[Qgis-developer] Reverse Changeset 13499

[Jeremy Palmer]

Hi Jürgen,

Yes we do use symbology-ng. Specifically we use the vectorfieldrender python plug-in and some of the advanced rendering features in symbology-ng. We are not currently using the graduated symbol renderer.

Yes you are correct; we avoid ftools for large datasets. If we need to do queries and analysis of the data we do that at the PostGIS level, and then view those results in QGIS.

We do try to avoid the use of the attribute table, but some of our users still try to load it every now and again. When they do this it puts extra load on the server and of course QGIS become unresponsive for a large period of time. In our build of QGIS I have put a warning message before the attribute table is loaded if the feature count (estimated!) is larger than a defined option. Not the best solution, but it works. If we could only display features in the current canvas view that would be excellent. Is it hard to put this feature back in?

As Chris stated the whole reason why we implemented the estimated statistics option was to allow fast loading of layers into the interface with very large datasets. The caveat when using this option is the extents, geometry types and features counts could be wrong. But the user knows this when choosing the option and can choose to turn it off. If they want more correct information, then they don't have use the option.

Our users do look at the vector layer metadata in the properties dialog to view the URI and provider type. So if the feature count is still going to happen automatically at this point then it will be a big inconvenience. Maybe a solution to this to change the QgsVectorDataProvider interface to have a new method "bool hasEstimatedFeatureCount()" that returns false by default. In the postgres provider this method could be set to true when an estimated feature count is done. Then in the vector properties dialog it could display that the feature count is estimated, and give the option to click a button to do a "correct" count. Or something like that...

I've also noticed that feature count is fetched for progress reporting during layer rendering. This in my opinion is invalid as it's the number of features within the current dataprovider select filter that is more relevant. Maybe this code should just be removed. 

Cheers,
Jeremy

-----Original Message-----
From: Jürgen E. Fischer [mailto:jef at norbit.de] 
Sent: Thursday, 20 May 2010 10:01 a.m.
To: Chris Crook
Subject: Re: [Qgis-developer] Reverse Changeset 13499

Hi Chris,

On Thu, 20. May 2010 at 09:37:20 +1200, Chris Crook wrote:
> So that is the premise on which the estimated metadata option is 
> based.  If you call it a "no non-spatial query" option may that would make more sense?

> I hope that makes it clearer what why we are very keen on reversing or 
> modifying the change!

Well, I'm still interested in figuring out, where we still retrieve feature counts, that we actually don't need - which would be where the change hits you.

As said symbology-ng is a candidate, are you using it?

I imagine ftools is pretty unusable for you kind of dataset, give or take feature count.

You probably don't use the attribute table in those datasets either.  Am I right?

BTW there once was an option to have the attribute show only the features in the current canvas, which I imagine would still be a nice thing for your setup, would it?


Jürgen

-- 
Jürgen E. Fischer         norBIT GmbH               Tel. +49-4931-918175-20
Dipl.-Inf. (FH)           Rheinstraße 13            Fax. +49-4931-918175-50
Software Engineer         D-26506 Norden               http://www.norbit.de

--
norBIT Gesellschaft fuer Unternehmensberatung und Informationssysteme mbH Rheinstrasse 13, 26506 Norden
GF: Jelto Buurman, HR: Amtsgericht Emden, HRB 5502

______________________________________________________________________________________________________

This message contains information, which is confidential and may be subject to legal privilege. 
If you are not the intended recipient, you must not peruse, use, disseminate, distribute or copy this message.
If you have received this message in error, please notify us immediately (Phone 0800 665 463 or info at linz.govt.nz) and destroy the original message.
LINZ accepts no responsibility for changes to this email, or for any attachments, after its transmission from LINZ.

Thank you.
______________________________________________________________________________________________________