<div class="gmail_quote"><div>For now, I suggest just merging the angular renderer changes into the composer redesign branch, keeping the non-caching behavior. I have been very busy the last month or so, but I expect to have some time in about a week to work on it. If others want to do it, go ahead.<br>
I'd like to do some more testing of the speed of the different branches. Ultimately, a combination may be best - caching when working with data, and full rendering when printing.<br>Steven<br> </div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Date: Sun, 10 Feb 2008 16:04:33 +0100<br>From: Marco Hugentobler <<a href="mailto:marco.hugentobler@karto.baug.ethz.ch">marco.hugentobler@karto.baug.ethz.ch</a>><br>Subject: [Qgis-developer] composer redesign branch<br>
To: <a href="mailto:qgis-developer@lists.osgeo.org">qgis-developer@lists.osgeo.org</a><br>Message-ID: <<a href="mailto:200802101604.33706.marco.hugentobler@karto.baug.ethz.ch">200802101604.33706.marco.hugentobler@karto.baug.ethz.ch</a>><br>
Content-Type: text/plain; charset="us-ascii"<br><br>Hi<br><br>I remember from the mailing list that there have been some conflicting changes<br>in renderer branch(now in trunk) and print composer redesign branch. I'd like<br>
to bring these in synchronisation again such that development of composer<br>branch may continue.<br><br>As far as I remember, the conflicting point is the rendering of point symbols.<br>The composer redesign branch does no caching to QImage, but uses QSvgRenderer<br>
for rendering the svg symbols. The advantage is that the resolution is<br>perfect on all output devices, the disadvantage is the increased rendering<br>time (by a factor 4). I wonder what the way to go is and like to discuss this<br>
here. I see two solutions:<br><br>- The solution of trunk is taken because of better performance and adapted<br>such that the resolution of the output device is considered when<br>caching/retrieving point symbols.<br><br>
- The solution of the composer branch is taken and adapted to point symbol<br>rotation to fit with trunk.<br><br>In my opinion, factor 4 is much performance difference and so I'm inclined to<br>prefer the first option. What do you think?<br>
<br>Regards,<br>Marco<br><br>--<br>Dr. Marco Hugentobler<br>Institute of Cartography<br>ETH Zurich<br>Technical Advisor QGIS Project Steering Committee<br></blockquote></div><br>