Hi,
I've fixed the off-by-one error in the CVS tree- thanks for the push.
I shouldn't have left it for so long.
As for the graphics problem, it appears to be an ImageMagick issue.
The 6.6.1 version that Curt is running has the problem, but the
6.6.2-6 that I am running does not. The error does
Hi Tom,
Thanks for fixing that for me.
...jerry
KA6HLD
On Mon, Nov 8, 2010 at 5:48 PM, Tom Russo ru...@bogodyn.org wrote:
On Mon, Nov 08, 2010 at 07:20:08PM -0600, we recorded a bogon-computron
collision of the jerut...@gmail.com flavor, containing:
Curt, et al -
Tonite I was attempting to
Hi,
I just committed a fix to CVS that solves the problem with 16-bit
*Magick quantums that I reported in August. With this fix, topOSM
tiles should work even when ImageMagick (or GraphicsMagick) is
configured for 16-bit quantums. Note that in the Ubuntu 10.04
repositories ImageMagick is built to
Hi Eldon,
You don't state which maps you are trying to use, but I'm going to
guess you are trying to use the OSM styles (because of the red areas).
The feature you are looking for does not exist at this point, but it
sounds useful and I'll add it to my list of future implementation
ideas.
In the
Hi Kurt,
I'd like to know more about why the OSM tile cache dosn't work off line for
you. I have verified that the tiles work off line for me and I can't (yet)
see a problem in the code.
If you can send me coordinates, zoom levels, and OSM styles you are using, I
might be able to recreate the
Hi,
I have committed a minor clean up to the OSM tile code that fixes the
problem I reported earlier with topOSM tile refresh. In addition, the
names of the topOSM map files (.geo) have been changed so that they
will layer properly without setting specific layer values, and an
obsolete executable
Hi Kurt,
Which of the OSM maps are you trying to use? The full size maps
probably won't work due to floating point to integer transformations
that yield new center coordinates for the center of the map for even
the most minor changes. However, the tiled OSM maps should (and for me
do) work great
Gluszenia Slawinski
curi...@bwv190.internetdsl.tpnet.pl wrote:
wouldn't it be easier to just switch ... xastir projection
while osm maps are used?
--
On Mon, 16 Aug 2010, Jerry Dunmire wrote:
All the ones I've looked at require some sort of transformation. It
all depends on the projection used
The information is available at rendering time, so it shouldn't be
very difficult. However, I would not want anyone to get their
expectations too high. I don't believe that we will ever get 'slippy'
map type performance because we handle the pixels one at a time.
I won't be able to work on any
Two items to be aware of:
1) Transparency will work only if GraphicsMagick/ImageMagick is
compiled for 8-bit Quantum.
The default for the GraphicsMagick code base is 8-bit, but the
package for OS X appears to get built with 16-bit Quantum.
I have recreated the problem (by rebuilding
Trying to determine why my announcement didn't make it to the list.
...jerry
___
Xastir mailing list
Xastir@lists.xastir.org
http://lists.xastir.org/cgi-bin/mailman/listinfo/xastir
Hi,
I just committed code to CVS the does following:
* fix for '1 of 0' tile download message
* fixed a memory leak in map_OSM.c
* fixed installation of OSM CC icons
* added support for .jpg tiles
* implemented transparency for tiles. No more red blocks for missing
Hi Tom,
I've tried it on Ubuntu with both Graphic and ImageMagic, so I don't
think it is library related.
One area that could easily be a problem is the QuantumDepth. On all of
my systems it is 16. Check /usr/include/magick/image.h (path may
differ) and let me know the QuantumDepth for your
Hi Tom,
Thanks for the log. I don't see any errors either, but I didn't expect
that it would be that easy. I do see a difference between the image
info in your log and mine:
mine: #a7a7a7 XC 1280x1024+0+0 PseudoClass 95c 8-bit 2.810u 0:04
(425.2K pixels/s)
yours: #a7a7a7 XC 1536x1280+0+0
Not a bad idea. As each tile is retrieved it would be scaled to Xastir
dimensions, which would change the tile size to something other than
256x256, and then saved.
I would be happy to help anyone who would like to take on such an improvement.
...jerry
On Mon, Aug 16, 2010 at 4:28 PM,
On Mon, Aug 16, 2010 at 5:52 PM, Jason KG4WSV kg4...@gmail.com wrote:
On Fri, Aug 13, 2010 at 3:36 PM, Jerry Dunmire je...@dunmire.org wrote:
The root of
the problem is that projection mapping that must occur between the
Mercator projection of OSM tiles and the 'un-projected' Xastir view.
Each
No problem. The suggestions are appreciated.
...jerry
On Mon, Aug 16, 2010 at 7:35 PM, Jason Godfrey godfr...@gmail.com wrote:
On Mon, Aug 16, 2010 at 9:32 PM, Jason Godfrey godfr...@gmail.com wrote:
You might want to look at GDAL. It is already an optional library for
xastir. I haven't
Hi Piotr,
Yes, it is slow. As I noted in the README file, performance will not
approach that of tangoGPS, or any other native OSM system. The root of
the problem is that projection mapping that must occur between the
Mercator projection of OSM tiles and the 'un-projected' Xastir view.
Each pixel
Hi Piotr,
I too saw a seg fault when I was first implementing the tiling code
that made me worry that I was destabilizing the system. On further
examination of older versions I found I could recreate the problem but
I have not had a chance to purse it further. It sounds a bit like your
problem
Hi Piotr,
Sounds like you would like the missing tiles to be 'transparent' so
that the tile (or any other graphics) will show through from a lower
level. Correct?
It's a good idea and it fits with the things I'm working on now, but
it will be a bit. I'm trying to figure out the transparency
Hi Piotr,
If you set debugging to 512 (-v 512) the tiles will not be removed
when a read error occurs. Then you can examine the tile file for
possible problems. A common problem (but I don't think it is the one
you are having) is that the host returns a 404 when the tile is not
found on the
2010, Jerry Dunmire wrote:
Hi Piotr,
If you set debugging to 512 (-v 512) the tiles will not be removed
when a read error occurs. Then you can examine the tile file for
possible problems. A common problem (but I don't think it is the one
you are having) is that the host returns a 404 when
Sounds like the problem is bigger than just tiling even OSM support.
Your guess about uclibc sounds reasonable, but I'm not sure where to
recommend you start. Are all of the errors related to png.c?
I'm hoping someone with uclibc experience can jump in here, or maybe
one the Xastir gurus can help
I've used Ubuntu exclusively since 6.06. I upgraded to 10.04 in May
and things are working for me:
xastir (now built from source, but originally from Ubuntu .deb)
soundmodem (built from source)
ax25
linpac
fldigi
I am using audio connections to a 2m HT and to my HF rig and a serial
connection to
Something to watch out for on Ubuntu. During the 7.x series they
changed the default shell from bash to dash. dash is not bash and
scripts that require bash functionality and do not specify
'#!/bin/bash' at the top of the script will fail in wild and wonderful
ways.
Personally, the first thing I
[mailto:xastir-boun...@lists.xastir.org] On Behalf Of Jerry Dunmire
Sent: Friday, July 30, 2010 20:23
These are effective only if libcurl support is enabled:
___
Xastir mailing list
Xastir@lists.xastir.org
http://lists.xastir.org/cgi-bin/mailman
I just committed changes to CVS to improve the OSM tile download and
cache. These are effective only if libcurl support is enabled:
- check for tile updates for tiles that have been cached for more
than 7 days (per OSM tile usage policy).
- use http pipelining (may not help much, since it
regards,
Fred
-Original Message-
From: xastir-boun...@lists.xastir.org
[mailto:xastir-boun...@lists.xastir.org] On Behalf Of Jerry Dunmire
Sent: Tuesday, July 27, 2010 13:11
To: Xastir - APRS client software discussion
Subject: [Xastir] Announcing tile support for OSM
Early
Thanks, Tom. I thought that libcurl should have something like, but I
hadn't found it
yet. Given the command line argument you've provided it should be pretty easy to
figure out now.
...jerry
On Wed, Jul 28, 2010 at 1:30 PM, Tom Hayward esa...@gmail.com wrote:
One item that will be useful at
Brett and Curt,
The only problem is that bulk downloads are strongly discouraged, and
expressly prohibited for some zoom levels, by
the OSM tile servers. See the policy at
http://wiki.openstreetmap.org/wiki/Tile_usage_policy
Of course other tile servers may have different policies, and I know
of
of us could do the same using this server which would take all
the load off the OSM servers.
It's late, so maybe I'm way off here, but it's a thought. I'm sure something
could be figured out.
Brett
KC9MWG
On Wed, Jul 28, 2010 at 7:06 PM, Jerry Dunmire je...@dunmire.org wrote:
Brett
, and the maps are centered where
they're supposed to be, but as expected they're not filling the screen.
That's much less disconcerting than seeing your QTH about 100 miles to the
west of your marker!
Thanks for the explanation
-KC2YWE Robert
On Jul 14, 2010, at 6:21 PM, Jerry Dunmire
Hi Robert,
There is a limit to the size of the returned OSM static map of 2Kx2K.
That OSM map typically must be larger than the window because OSM maps
are Mercator projections and Xastir is not. As a result, the present
OSM implementation limits the requested map size and then tries to
align it
Where do I find info on how to setup a Ridge Radar map?
The code that enables/disables the F4 is dead simple, so there must be
something in the other maps that disables it. When I can recreate the
problem I trace it with GDB and see what's going on.
Also, could you try changing to another key?
No problem, I can debug the code if I can get a setup that causes the problem.
I've got the Ridge Radar maps installed. Now I just have figure out
which to use...
...jerry
On Tue, Jun 29, 2010 at 9:32 PM, Curt, WE7U curt.w...@gmail.com wrote:
On Tue, 29 Jun 2010, Jerry Dunmire wrote:
Where
If you start xastir with -v 512, then the F5 key will report the OSM
zoom level.
I check both osmarender and cycle at Xastir zoom level 4, which
requests OSM zoom level 17. A map was retrieved for osmarender, but
not cycle. I was looking at the San Jose area.
So I concur with Curt- it varies by
Yes, I have started to look at using the tile servers. I expect much better
performance when using cached tiles.
73,
...jerry
KA6HLD
On Mon, Jun 28, 2010 at 12:30 PM, Curt, WE7U curt.w...@gmail.com wrote:
On Sat, 26 Jun 2010, Piotr Gluszenia Slawinski wrote:
i've tried downloading cvs version
I probably put those warnings back in- :-(
There were a large number of #ifdef to get around conflicts for
ImageMagic, but since none of the macros appear in the map_OSM.c and
since I didn't get any warnings when I removed them, I assume they
were dead code left over from the files I started
and external to Xastir.
...jerry
On Mon, Jun 28, 2010 at 4:41 PM, Curt, WE7U curt.w...@gmail.com wrote:
On Mon, 28 Jun 2010, Jerry Dunmire wrote:
Yes, I have started to look at using the tile servers. I expect much
better
performance when using cached tiles.
We could also play the same
Are you using the .osm maps? If so, I'm not surprised. I'm working on
fixes for a number of zooming and scaling problems with OSM maps. Try
the .geo versions before reporting any problems.
73,
...jerry
KA6HLD
On Wed, Jun 23, 2010 at 6:51 PM, David Aitcheson
david.aitche...@gmail.com wrote:
Hi Lee,
All the info on the xserver patch is in the bug report at:
https://bugs.launchpad.net/ubuntu/+bug/581032
73,
...jerry
KA6HLD
On Tue, Jun 22, 2010 at 8:33 AM, Lee Bengston lee.bengs...@gmail.com wrote:
On Tue, Jun 22, 2010 at 12:17 AM, Jerry Dunmire je...@dunmire.org wrote:
For what
For what it's worth-
I am not having any trouble with the right click menu. I click and it
stays open.
I am running Ubuntu 10.04 and the tip of CVS xastir. I did patch my
xserver to fix the mouse stuck in window problem (before I started
building from the CVS). So I guess I have both the fix (for
On Fri, Jun 18, 2010 at 2:57 PM, Garland Sparks gspark...@gmail.com wrote:
every time I try to run 'sudo' it asks me for a [sudo] password and none of
my passwords work.
You should be able to use the password you defined for your account
when you installed Ubuntu. If someone else installed
Hi David,
Thanks for your offer.
The files specify the source and options for various on-line map
images generated from the OpenStreetMap data. The best starting point
to identify each of the 'styles' is this URL:
I am working to generate the numerous OSM_* files from a single
template. It would be easy to do using a gnumake pattern rule
something like this:
%.geo: OSM_template.geo
sed
However, use of pattern rules causes automake to warn that `%'-style
pattern rules are a GNU make extension.
Hi Tom,
OK, but I haven't used a non-gmake many many years, so could you help
me out with the alternative? I'm trying to take a list of styles
(LIST=mapnik,skiing, ...) and use that to transform OSM_template.geo
to OSM_mapnik.geo, OSM_skiing.geo, etc. The sed(1) argument would also
come from the
The CVS archive has been updated to include code that will scale the
OSM bitmap images. This restores the default linear scaling used by
Xastir and 'fixes' the area zoom (no changes to area zoom were needed
and that part of the code has not been changed).
Support for binned zoom levels remains.
If you are using a large display (lucky you) and large window for
Xastir and then the area selected for a zoom is a high aspect ratio,
you may get an OSM map that does not fill the display and is not
aligned with the vector layers. The specific aspect ratio where the
problem occurs depends on your
Slight correction for the tip of CVS at this time (2010/06/14 11:43
PDT)- the file to select is 'OpenStreetMaps.osm'. The extension was
changed from .geo to .osm with the last patch.
Also a bit of warning, the extension and file name is like to change
at least one more time as I integrate linear
Thanks for the report. I kind of expected something like that, but my
first attempt to fix it with the binned levels didn't work so I
decided to postpone further work on fixing the boxed (area) zoom
function until the linear scaling is working. I really expect the zoom
to work just fine (with
Yup, that should do the job.
Thanks,
...jerry
On Fri, Jun 11, 2010 at 11:20 PM, Curt, WE7U curt.w...@gmail.com wrote:
On Fri, 11 Jun 2010, Jerry Dunmire wrote:
I am actually thinking about how to select the other map styles. I'm
not quite sure how yet, but probably by different key words
PM, Andrew Errington
a.erring...@lancaster.ac.uk wrote:
On Fri, June 11, 2010 14:03, Jerry Dunmire wrote:
I hadn't noticed the extension collision. I was looking only at the
Xastir use. Do you have any suggestions for a more appropriate
extension?
Not really. I don't really know what you would
new_image() is declared external at the top of main.c:
main.c:989 extern void new_image(Widget da);
but new_image() is defined in main.c and as far as I can tell, used
only in main.
Shouldn't the declaration be:
void new_image(Widget da);
or possibly:
static void new_image(Widget da);
Am I
...@gmail.com wrote:
On Thu, 10 Jun 2010, Jerry Dunmire wrote:
new_image() is declared external at the top of main.c:
main.c:989 extern void new_image(Widget da);
but new_image() is defined in main.c and as far as I can tell, used
only in main.
That appears to be a left-over from my aborted
and if that doesn't happen soon
enough it's OK to tell me that I should stop.
...jerry
On Thu, Jun 10, 2010 at 12:59 PM, Curt, WE7U curt.w...@gmail.com wrote:
On Thu, 10 Jun 2010, Jerry Dunmire wrote:
Great. I'll clean this up in a future patch.
The 'static' qualifier does not seem
Thanks for the report. Yes, there is a problem with the transition
from linear scaling to
OSM levels when OSM maps are first selected and when they are
deselected. It causes
locations to be off (see the reports above related to being in the
ocean) and constrains
scaling to the OSM levels when they
Curt has committed a new set of patches from me to CVS that should fix
some of OSM problems that have been reported. The new code processes
the scaling/level change when the OSM maps are selected, deselected,
and at startup. The scaling is still binned.
Note, if you are running a previous version
My announcement was incorrect. The new file name has an 's' in it. The
new map file name is:
OpenStreetMaps.osm
Sorry for any confusion I caused.
...jerry
On Thu, Jun 10, 2010 at 9:20 PM, Andrew Errington
a.erring...@lancaster.ac.uk wrote:
On Fri, June 11, 2010 12:23, Jerry Dunmire wrote
will do.
...jerry
On Wed, Jun 9, 2010 at 12:08 PM, Curt, WE7U curt.w...@gmail.com wrote:
On Wed, 9 Jun 2010, Jerry Dunmire wrote:
Would you like me to make the change to add %s format strings and
send it to you as a patch?
If so, should I send the patch to you off line or send
, at 11:55 AM, Curt, WE7U wrote:
Thanks to Jerry Dunmire, KA6HLD, we now have Open Street Map support
in Xastir. Woohoo!
For an interesting place to look, zoom in on a border and see what
differences you can find for roads crossing said border. For the
main Canadian border crossing north of me
Thanks for the feedback. Help is always appreciated Tom and Curt. I'll
try to keep up :-)
It will take me a bit to try out the great example you provided, but
I'll get started.
The OSM code is really just a (simple) modification of the tigermaps
code. The slow speed seems to be all related to
Thanks Tom for the analysis. I'll look into where else calling the
adj_to_OSM_levels()
needs to be called as a short term fix until I can figure out (really
just learn what's already there) a better scaling method.
...jerry
On Wed, Jun 9, 2010 at 2:09 PM, Tom Russo ru...@bogodyn.org wrote:
On
Hi,
I'm new to the group and new to the internals of xastir, but with the
TIGER/Online resource gone I've taken advantage of the open-source for
Xastir to see what I could do.
Google is a problem because their usage policy does not allow caching.
Several other standalone applications (viking
for this
conversation.
73,
...jerry
KA6HLD
On Tue, Jun 8, 2010 at 2:58 PM, Curt, WE7U curt.w...@gmail.com wrote:
On Tue, 8 Jun 2010, Jerry Dunmire wrote:
Google is a problem because their usage policy does not allow caching.
Several other standalone applications (viking comes to mind) have had
with version 1.98 or 1.99.
Lee - K5DAT
On Tue, Jun 8, 2010 at 6:46 PM, Jerry Dunmire je...@dunmire.org wrote:
This is a known problem with the xorg-server for Ubuntu 10.04 (and
others) and possibly nvidia drivers. The problem effects lesstif/motif
applications. Patches are available. See
65 matches
Mail list logo
