<?xml version="1.0" encoding="utf-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
	>
<channel>
	<title>Comments on: MapServer 5.0 Agg Support: Disappointing</title>
	<atom:link href="http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/feed/" rel="self" type="application/rss+xml" />
	<link>http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/</link>
	<description>Ramblings of a GIS Hacker</description>
	<pubDate>Mon, 13 Feb 2012 00:10:48 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.7.1</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: Steve Lime</title>
		<link>http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/comment-page-1/#comment-18693</link>
		<dc:creator>Steve Lime</dc:creator>
		<pubDate>Mon, 30 Jul 2007 22:35:49 +0000</pubDate>
		<guid isPermaLink="false">http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/#comment-18693</guid>
		<description>Try removing all your TRANSPARENCY ALPHA lines. They seem to be causing at least part of the problem. They aren't needed with AGG and should be ignored, but all the recent testing has been without them so this problem was missed. In my tests with your data the improvement is markedly better. 

Polygon fills are still showing as grey but I bet that's a cut and paste error.

Check out http://maps.dnr.state.mn.us/mapserver_demos/boston/boston.html for examples. I'll add a toggle AGG/GD switch when I get the chance.

This is why you have betas, pre-release or whatever you want to call them.

Steve</description>
		<content:encoded><![CDATA[<p>Try removing all your TRANSPARENCY ALPHA lines. They seem to be causing at least part of the problem. They aren&#8217;t needed with AGG and should be ignored, but all the recent testing has been without them so this problem was missed. In my tests with your data the improvement is markedly better. </p>
<p>Polygon fills are still showing as grey but I bet that&#8217;s a cut and paste error.</p>
<p>Check out <a href="http://maps.dnr.state.mn.us/mapserver_demos/boston/boston.html" rel="nofollow">http://maps.dnr.state.mn.us/mapserver_demos/boston/boston.html</a> for examples. I&#8217;ll add a toggle AGG/GD switch when I get the chance.</p>
<p>This is why you have betas, pre-release or whatever you want to call them.</p>
<p>Steve</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Steve Lime</title>
		<link>http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/comment-page-1/#comment-18690</link>
		<dc:creator>Steve Lime</dc:creator>
		<pubDate>Mon, 30 Jul 2007 20:28:40 +0000</pubDate>
		<guid isPermaLink="false">http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/#comment-18690</guid>
		<description>BTW Need access to the symbolset and fontset files to recreate things exactly.

Steve</description>
		<content:encoded><![CDATA[<p>BTW Need access to the symbolset and fontset files to recreate things exactly.</p>
<p>Steve</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Steve Lime</title>
		<link>http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/comment-page-1/#comment-18689</link>
		<dc:creator>Steve Lime</dc:creator>
		<pubDate>Mon, 30 Jul 2007 19:55:43 +0000</pubDate>
		<guid isPermaLink="false">http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/#comment-18689</guid>
		<description>Appreciate the tests and the great test case. Would be nice to get feedback through the normal project channels (mapserver-dev or trac) rather than having to track down blog entries...

Steve</description>
		<content:encoded><![CDATA[<p>Appreciate the tests and the great test case. Would be nice to get feedback through the normal project channels (mapserver-dev or trac) rather than having to track down blog entries&#8230;</p>
<p>Steve</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Zak James</title>
		<link>http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/comment-page-1/#comment-18688</link>
		<dc:creator>Zak James</dc:creator>
		<pubDate>Mon, 30 Jul 2007 14:02:21 +0000</pubDate>
		<guid isPermaLink="false">http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/#comment-18688</guid>
		<description>You should make a comment about the degraded output in the bug so the developer can take a look at your maps.

I'm not sure about the colour problem since you're definitely rendering in 24 bit. I suspect that the gd antialiasing and transparency is interfering with agg. Can you try commenting 
    transparency alpha
and
        antialias true

out of your water layer?
 
The term 'Pre-release' was my poor attempt at humour. I'm not sure of the official status of the release.</description>
		<content:encoded><![CDATA[<p>You should make a comment about the degraded output in the bug so the developer can take a look at your maps.</p>
<p>I&#8217;m not sure about the colour problem since you&#8217;re definitely rendering in 24 bit. I suspect that the gd antialiasing and transparency is interfering with agg. Can you try commenting<br />
    transparency alpha<br />
and<br />
        antialias true</p>
<p>out of your water layer?</p>
<p>The term &#8216;Pre-release&#8217; was my poor attempt at humour. I&#8217;m not sure of the official status of the release.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: crschmidt</title>
		<link>http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/comment-page-1/#comment-18686</link>
		<dc:creator>crschmidt</dc:creator>
		<pubDate>Mon, 30 Jul 2007 04:20:12 +0000</pubDate>
		<guid isPermaLink="false">http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/#comment-18686</guid>
		<description>I don't call something 'pre-release' until I'd be happy with it shipping, assuming no one reported any bugs. It's possible that MapServer is more lenient with names, in which case I'm being overly harsh. 

Updated to SVN trunk, and added that as a layer to the agg-demo.html file. As you can see, it's actually less impressive, though it does fix the label rendering, which is good.

I'm already using "IMAGEMODE RGB" -- does that not make 24 bit pngs? `identify` says 'PNG 1024x1024 DirectClass 817kb 0.230u 0:01', which I think means 'not 8 bit'. Am I wrong?</description>
		<content:encoded><![CDATA[<p>I don&#8217;t call something &#8216;pre-release&#8217; until I&#8217;d be happy with it shipping, assuming no one reported any bugs. It&#8217;s possible that MapServer is more lenient with names, in which case I&#8217;m being overly harsh. </p>
<p>Updated to SVN trunk, and added that as a layer to the agg-demo.html file. As you can see, it&#8217;s actually less impressive, though it does fix the label rendering, which is good.</p>
<p>I&#8217;m already using &#8220;IMAGEMODE RGB&#8221; &#8212; does that not make 24 bit pngs? `identify` says &#8216;PNG 1024&#215;1024 DirectClass 817kb 0.230u 0:01&#8242;, which I think means &#8216;not 8 bit&#8217;. Am I wrong?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Zak James</title>
		<link>http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/comment-page-1/#comment-18685</link>
		<dc:creator>Zak James</dc:creator>
		<pubDate>Mon, 30 Jul 2007 02:45:21 +0000</pubDate>
		<guid isPermaLink="false">http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/#comment-18685</guid>
		<description>Hmm, the comment system ate my angle brackets but FORMATOPTION “PALETTE=” needs to reference the text file of reserved colours.</description>
		<content:encoded><![CDATA[<p>Hmm, the comment system ate my angle brackets but FORMATOPTION “PALETTE=” needs to reference the text file of reserved colours.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Zak James</title>
		<link>http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/comment-page-1/#comment-18684</link>
		<dc:creator>Zak James</dc:creator>
		<pubDate>Mon, 30 Jul 2007 02:43:23 +0000</pubDate>
		<guid isPermaLink="false">http://crschmidt.net/blog/archives/239/mapserver-50-agg-support-disappointing/#comment-18684</guid>
		<description>Well, it is pre-release software :)

If you look closely at your output you'll see that the 'jaggies' are only present in areas where agg features overlay the gd-rendered background map colour. The lines drawn over the grey water are very nicely antialiased. There are workarounds for this, but a very recent patch by Thomas Bonfort may have fixed this - I haven't tested it yet. See http://trac.osgeo.org/mapserver/ticket/2173

"Make Output Pretty" hasn't been implemented yet, but you can fix your colour problems by either using 24 bit output or by adding    

FORMATOPTION "PALETTE_FORCE=TRUE"
FORMATOPTION "PALETTE="

to the AGG OUTPUTFORMAT block. The palette file is a text file with one line for each colour you want to reserve in the output image's 8-bit index. (e.g. 0,0,255 would reserve blue). The problem you're seeing is that there are too many colours in your output due to the pretty antialiasing so the water colour from the mapfile won't fit in the palette.

I agree that the lack of improvement to text rendering is a disappointment, but I think it's safe to say this will be remedied in the future.

Zak</description>
		<content:encoded><![CDATA[<p>Well, it is pre-release software <img src='http://crschmidt.net/blog/wp-includes/images/smilies/icon_smile.gif' alt=':)' class='wp-smiley' /> </p>
<p>If you look closely at your output you&#8217;ll see that the &#8216;jaggies&#8217; are only present in areas where agg features overlay the gd-rendered background map colour. The lines drawn over the grey water are very nicely antialiased. There are workarounds for this, but a very recent patch by Thomas Bonfort may have fixed this - I haven&#8217;t tested it yet. See <a href="http://trac.osgeo.org/mapserver/ticket/2173" rel="nofollow">http://trac.osgeo.org/mapserver/ticket/2173</a></p>
<p>&#8220;Make Output Pretty&#8221; hasn&#8217;t been implemented yet, but you can fix your colour problems by either using 24 bit output or by adding    </p>
<p>FORMATOPTION &#8220;PALETTE_FORCE=TRUE&#8221;<br />
FORMATOPTION &#8220;PALETTE=&#8221;</p>
<p>to the AGG OUTPUTFORMAT block. The palette file is a text file with one line for each colour you want to reserve in the output image&#8217;s 8-bit index. (e.g. 0,0,255 would reserve blue). The problem you&#8217;re seeing is that there are too many colours in your output due to the pretty antialiasing so the water colour from the mapfile won&#8217;t fit in the palette.</p>
<p>I agree that the lack of improvement to text rendering is a disappointment, but I think it&#8217;s safe to say this will be remedied in the future.</p>
<p>Zak</p>
]]></content:encoded>
	</item>
</channel>
</rss>

