[Tilesathome] Experimental Batik Support

Gert Gremmen Administrator at ce-test.info
Wed Nov 14 14:09:39 GMT 2007


Got batik rendering.
I had to install the right java package first,
but my repository server was directed to 
a dutch server, instead of the main server.
abd the dutch one did not contain java-6

Anyway, a new problem appeared :

rendering of a tile is finished really quik (1 second)
and creates 1 empty tile.
The maplint process seems to run longer, but creates
a lot of empty files. Not much better.

Any clue  ???

Regards,

Ing. Gert Gremmen
-----Original Message-----
From: Brent Easton [mailto:b.easton at exemail.com.au] 
Sent: 2007-11-14 11:18 AM
To: Gert Gremmen; Brent Easton; Damian Sulewski;
tilesathome at openstreetmap.org
Subject: RE: [Tilesathome] Experimental Batik Support

Hi Gert,

Sounds like a class path problem, or possibly a library version
compatibility problem. I run XP, I am guessing you are on linux.

I did a bit of google search and found this:

> I'va installed the libgcj6-dev package!!! And it works!!

Does that make sense?

Regards,
Brent.


*********** REPLY SEPARATOR  ***********

On 14/11/2007 at 10:25 AM Gert Gremmen  wrote:

>Finally I am able to test the batik in t at h
>
>Got an error about 
>
>"cannot load AWT toolkit"
>
>
>I followed the batik installation instructions,
>and that seems ok.
>
>Any Clue ??
>
>
>Regards,
>
>Ing. Gert Gremmen
>
>
>Thanks Damien,
>
>I am getting the same thing, I will look into it. I haven't seen that
on
>any other tile.
>
>Regards,
>Brent.
>
>*********** REPLY SEPARATOR  ***********
>
>On 10/11/2007 at 1:28 PM Damian Sulewski  wrote:
>
>>Hi,
>> i rendered 
>>http://tah.openstreetmap.org/Browse/?x=2132&y=1360&z=12&layer=tile
>>
>>with batik.
>>
>>Could it be that i see a line at the top and the bottom of the tile?
>>
>>I did a "svn up" before rendering, so everything should be up to date.
>>Batik version is 1.7 on an ubuntu 7.10 system.
>>
>>Greetings,
>>
>>Damian
>>
>>Am Samstag, den 10.11.2007, 09:57 +1100 schrieb Brent Easton:
>>> After successful testing using Batik as a replacement for Inkscape,
I
>>have added Batik as an option for rendering in tiles at home.
>>> 
>>> This is still somewhat experimental as I only have a 1GB windows
>>machine. I need assistance to push it a bit further. I am especially
>>interested in people trying with 2GB or more of RAM installed and on
>>linux. Please just run tiles manually and check the results.
>>> 
>>> The default behaviour of t at h is unchanged, it will continue to use
>>Inkscape as before. To play with Batik you need to do the following:
>>> 
>>> 1. Download the latest version of Batik and install it somewhere. I
>have
>>been testing with the 1.7beta1 version. Batik just consists of a bunch
>of
>>Java libaries (.jar files), just copy them into a convenient
directory.
>>Make sure you also have the batik-rasterizer.jar file.
>>> 
>>> 2. Add some parameters to your tilesathome.conf file:
>>> 
>>>   a) Turn Batik rendering on:
>>> 
>>> Batik=1   
>>> 
>>>   b) Set the maximum size of the JVM to be allocated. You want to
set
>>this as large as possible to cope with complex tiles. On windows XP
the
>>maximum you can set is 1300M. If you set it too high, java will not
>start
>>up. (Try java -Xmx1400M to test).
>>> 
>>> BatikJVMSize=1300M
>>> 
>>>   c) Point to the location of Batik
>>> 
>>> BatikPath=c:\tilesAtHome\batik\batik-rasterizer.jar
>>> 
>>> 3. Make sure you have the latest Osmarender.xsl installed.
>>> 
>>> That's it.
>>> 
>>> This just uses the standard rasterizer program which does have one
>major
>>disadvantage - If there is an error in the generated SVG, it just
>throws
>>and exception and fails. However, batik-rasterizer is open source and
>we
>>can modify this if testing is successful. There may also be scope to
>tweak
>>the behaviour to match OSM processing and gain some performance
>>improvements. The current procedure is to load up the same SVG
multiple
>>times to render each strip, we may be able to to load the SVG only
once
>>and render all of the required strips in one go.
>>> 
>>> Looking forward to hearing results,
>>> 
>>> Thanks,
>>> Brent.
>>> 
>>> 
>>> ____________________________________________________________
>>> Brent Easton                       
>>> Analyst/Programmer                               
>>> University of Western Sydney                                   
>>> Email: b.easton at uws.edu.au
>>> 
>>> 
>>> _______________________________________________
>>> Tilesathome mailing list
>>> Tilesathome at openstreetmap.org
>>> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/tilesathome
>>
>>
>>_______________________________________________
>>Tilesathome mailing list
>>Tilesathome at openstreetmap.org
>>http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/tilesathome
>>
>>
>>-- 
>>No virus found in this incoming message.
>>Checked by AVG Free Edition. 
>>Version: 7.5.503 / Virus Database: 269.15.26/1120 - Release Date:
>9/11/2007 9:26 AM
>
>
>____________________________________________________________
>Brent Easton                       
>Analyst/Programmer                               
>University of Western Sydney                                   
>Email: b.easton at uws.edu.au
>
>
>_______________________________________________
>Tilesathome mailing list
>Tilesathome at openstreetmap.org
>http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/tilesathome
>
>
>-- 
>No virus found in this incoming message.
>Checked by AVG Free Edition. 
>Version: 7.5.503 / Virus Database: 269.15.31/1128 - Release Date:
13/11/2007 11:09 AM


____________________________________________________________
Brent Easton                       
Analyst/Programmer                               
University of Western Sydney                                   
Email: b.easton at uws.edu.au





More information about the Tilesathome mailing list