[OSM-dev] New t at h client with multi-layer support

Jon Schlueter jonschl at iriscorp.org
Fri Apr 6 21:50:48 BST 2007


I just got a bunch of warning messages as my t at h client was running in loop mode

[#47 414%] nice "inkscape" -w 4096 -h 256
--export-area=0.000000:439.133372:878.910000:494.025043
--export-png="/tmp//28
769.part.png" "/tmp/output-28769-z16.svg" > /tmp//28769.stdout failed
[#47 500%] Rendering... sh: line 1: 12796 Aborted                 nice
"inkscape" -w 4096 -h 256 --export-area=0.000000:
384.241700:878.910000:439.133372 --export-png="/tmp//28769.part.png"
"/tmp/output-28769-z16.svg" >/tmp//28769.stdout 2>/
tmp//28769.stderr
ERROR
  The following command produced an error message:
  nice "inkscape" -w 4096 -h 256
--export-area=0.000000:384.241700:878.910000:439.133372
--export-png="/tmp//28769.part.
png" "/tmp/output-28769-z16.svg" > /tmp//28769.stdout
  Debug output follows:
  |
  | (inkscape:12796): Gdk-CRITICAL **: gdk_display_list_devices:
assertion `GDK_IS_DISPLAY (display)' failed
  |
  | ** (inkscape:12796): CRITICAL **: void
gnome_canvas_bpath_def_curveto(GnomeCanvasBpathDef*, double, double,
double,
double, double, double): assertion `bpd->moveto_idx >= 0' failed
  |
  | Emergency save activated!
  | Emergency save completed. Inkscape will close now.
  | If you can reproduce this crash, please file a bug at www.inkscape.org
  | with a detailed description of the steps leading to the crash, so
we can fix it.
  | ** Message: Error: Inkscape encountered an internal error and will
close now.
  |
[#47 500%] nice "inkscape" -w 4096 -h 256
--export-area=0.000000:384.241700:878.910000:439.133372
--export-png="/tmp//28
769.part.png" "/tmp/output-28769-z16.svg" > /tmp//28769.stdout failed
[#47 543%] Rendering... sh: line 1: 12802 Aborted                 nice
"inkscape" -w 4096 -h 256 --export-area=0.000000:
329.350029:878.910000:384.241700 --export-png="/tmp//28769.part.png"
"/tmp/output-28769-z16.svg" >/tmp//28769.stdout 2>/
tmp//28769.stderr
ERROR
  The following command produced an error message:
  nice "inkscape" -w 4096 -h 256
--export-area=0.000000:329.350029:878.910000:384.241700
--export-png="/tmp//28769.part.
png" "/tmp/output-28769-z16.svg" > /tmp//28769.stdout
  Debug output follows:
  |
  | (inkscape:12802): Gdk-CRITICAL **: gdk_display_list_devices:
assertion `GDK_IS_DISPLAY (display)' failed
  |
  | ** (inkscape:12802): CRITICAL **: void
gnome_canvas_bpath_def_curveto(GnomeCanvasBpathDef*, double, double,
double,
double, double, double): assertion `bpd->moveto_idx >= 0' failed
  |
  | Emergency save activated!
  | Emergency save completed. Inkscape will close now.
  | If you can reproduce this crash, please file a bug at www.inkscape.org
  | with a detailed description of the steps leading to the crash, so
we can fix it.
  | ** Message: Error: Inkscape encountered an internal error and will
close now.
  |
[#47 543%] nice "inkscape" -w 4096 -h 256
--export-area=0.000000:329.350029:878.910000:384.241700
--export-png="/tmp//28
769.part.png" "/tmp/output-28769-z16.svg" > /tmp//28769.stdout failed
[#47 557%] Rendering...

I know this is only a partial list of the error message....

any thoughts and direction for reproducing this problem?

Jon


On 4/4/07, Frederik Ramm <frederik at remote.org> wrote:
> Hi,
>
>   I've just commited to SVN a new t at h client which supports rendering
> multiple layers from one downloaded data set.
>
> It is currently configured to run the default (old "basemap") and
> maplint layers but I expect that 80n will soon add the bicycle/"outdoor"
> layer he has been preparing.
>
> The layer stuff is controlled from a new config file named, uhm,
> layers.conf.
>
> I hope that some of you "early adopters" find the time to download the
> new client and test it. If you want it to render maplint ONLY, simply
> say "Layers=maplint" in layers.conf. This is what I just did for the
> whole of London; maplint tiles should be visible there now. There was
> a problem where node duplication would occur when a tile was
> downloaded in multiple stripes and concatenated; this prompted me to
> re-write the concatenation code - in typical "parse XML with regexes"
> fashion, but if someone wants to come up with an XSLT based solution
> for that, feel free to change it.
>
> When playing around with layers, please keep in mind that it makes no
> sense to upload layers for which the server is not configured - so if
> you want your very special type of slippy map, talk to OJW first.
>
> I haven't found a way to make the maplint tiles transparent and use
> them as an overlay in the OpenLayers thingy; this should be possible
> somehow so if someone wants to experiment...
>
> Please note that the selection of checks done by maplint and the way
> the results are rendered is just a first shot right now, I believe
> there will be good ideas and improvements in the future. Maybe if we
> find a good colour-coding of the different kinds of problems then we
> can drop textual error messages on the maplint layer altogehter?
>
> Here's the svn comment:
>
> ---
> Multi-layer support. Layers controlled by layers.conf (currently
> built-in: default layer and maplint layer). Empty layers.conf emulates
> old behaviour. All layers rendered off the same downloaded data.
>
> Memory consumption of tilesGen.pl may have increased for complex tiles
> as maplint demanded a more fine-tuned approach to merging of multiple
> stripes.
>
> Version name bumped up to "Essen" even though there's no visible
> changes on the default layer - I felt the large number of changes
> merited that.
>
> TODOs: Make community decisions on what maplint tests should be run
> and how they are supposed to be rendered. This version uses a rather
> strict set of tests. Tests are encoded in tests.xsl which is generated
> by a Makefile from within the original maplint distribution; the
> Makefile and the components from which tests.xsl is assembled are not
> included in tiles at home.
> ---
>
> Bye
> Frederik
>
> --
> Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00.09' E008°23.33'
>
>
> _______________________________________________
> dev mailing list
> dev at openstreetmap.org
> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/dev
>




More information about the dev mailing list