gmane.comp.hardware.gps.gpsbabel.devel

patch for linux gui

tsteven4 — 2012-05-19T22:38:39

While my bias is towards the command line I like the preview available 
in the gui which led me to build it on centos 6.

I needed a few changes to make the linux distribution.  Some of these 
parallel the changes in setup.iss from r4096.  These are included in the 
patch.

I need other changes to build on my system, but at least some of these 
are system dependent.
1. Change QTLIBDIR, QTSHDIR.  It would be better to get the necessary 
information from "qmake -query QT_INSTALL_TRANSLATIONS" and "qmake 
-query QT_INSTALL_PLUGINS", at least with qt4, and make appropriate 
changes where these variables are used.
2. comment out libphonon and libaudio.  These are not dynamically loaded 
by the fe when I compile it, and if the ldd|grep|awk returns nothing 
then the script bombs.  I don't know if these are ever needed, or if 
that they just  weren't on my system.
3. for me the plugins were in another path, but see 1. above.
4. with qt4 on my system lrelease is named lrelease-qt4.


I also note that  ../../babelweb

test on skytraq-artificial.gpx failing

Bernd Zeimetz — 2012-05-18T09:31:23

hi,

in the current trunk (and since adding the test
Running testo.d/skytraq.test
--- /tmp/gpsbabel.21809/skytraq-artificial.gpx2012-05-18 11:23:53.131458101 +0200
+++ ./reference/skytraq-artificial.gpx2012-05-18 11:12:50.347645329 +0200
< at >< at > -82,7 +82,7 < at >< at >
 </trkpt>
 <trkpt lat="47.547749767" lon="9.679499575">
   <ele>435.413042</ele>
-  <time>2009-09-10T03:07:53Z</time>
+  <time>2009-09-10T21:20:09Z</time>
   <speed>0.000000</speed>
   <name>TP0009</name>
 </trkpt>
< at >< at > -232,7 +232,7 < at >< at >
 </trkpt>
 <trkpt lat="-30.376946121" lon="-97.665030673">
   <ele>185.882330</ele>
-  <time>2019-04-05T18:25:39Z</time>
+  <time>2019-04-06T12:37:55Z</time>
   <speed>55.555557</speed>
   <name>TP0034</name>
 </trkpt>
ERROR comparing /tmp/gpsbabel.21809/skytraq-artificial.gpx
./reference/skytraq-artificial.gpx

r4178 fixed most of the issues, but these two still show up.
I could try to find out what the reason for that is, but first I would prefer if
somebody ensures that the skytraq-artificial.gpx is valid and what I see i

clang scan-build report of gpsbabel

Bernd Zeimetz — 2012-05-17T12:50:20

Hi,

I've built gpsbabel with clang's scan-build, the current report (which shows 207
bugs) is available at

http://devel.recluse.de/scan-build/gpsbabel/

For now I'll rebuild the sources every night and a new report will be generated,
assuming there is some interest to fix those bugs :)

Cheers,

Bernd

Bad CRC with Qstarz travel recorder EX

Dirk Stichling — 2012-05-09T16:31:09

Hi.

A user of my application myTracks (which uses gpsbabel internally) reports that he has problems with his Qstarz travel recorder EX. The mtk driver of gpsbabel is used for the communication. Reading tracks from the logger works fine as long as the memory of the logger is not too full. But when the memory reaches about 50% gpsbabel reports "bad CRC" errors.

I did some research on the net and found a discussion on that topic (http://old.nabble.com/Error-reading-BT-Q100P-when-more-than-50--td22123960.html).

Now my question is whether this problem has not yet been addressed? If not, is there anything I can do to help solving the problem?

Thanks,
Dirk


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm

Problem with Columbus V900 (with link)

Dirk Stichling — 2012-05-08T17:13:31

Hi.

I have a problem with the following CSV file from a Columbus V900:
http://www.mytracks4mac.com/tmp/V900.CSV

It results in the following assertion:
v900.c:364: failed assertion `line.bas.common.tag == 'C' || line.bas.common.tag == 'V''

When I comment out that assertion everything seems to be fine.

Can anybody check that?

Dirk

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/

Problem with Columbus V900

Dirk Stichling — 2012-05-08T17:08:58

Hi.

I have a problem with the attached CSV file from a Columbus V900. It results in the following assertion:
v900.c:364: failed assertion `line.bas.common.tag == 'C' || line.bas.common.tag == 'V''

When I comment out that assertion everything seems to be fine.

Can anybody check that?

Dirk



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/_______________________________________________
Gpsbabel-code mailing list  http://www.gpsbabel.org
Gpsbabel-code< at >lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/gpsbabel-code

Patch to add "discard" option to track filter

Tom Paton — 2012-05-04T11:07:24

Hi,

This is revisiting something I attempted a few years back, I think
I've got it working this time (with some help from a friend).

The patch adds a "discard" option to the "-x track,merge" filter that
suppresses the "Found track point at lat,lon without time!" error and
instead just drops the points.

I've attached a zip that includes the diff to trackfilter.c (as at svn
revision 4173) and a test script and test files.  I also tried adding
a documentation file, but I'm not sure how to make that work.  Is
anything else needed?

Also, on a side note, I had problems running the test scripts that I
think is probably timezone related: The reference output files didn't
match due to the dates being off by one day (I'm UTC+10). Is that
something there's a workaround for or just something that no one's
bumped into yet?

Kind regards,

Tom
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and

bug with reverse filter and tracks

tsteven4 — 2012-05-03T12:59:56

There is a bug with the reverse filter when applied to tracks.  It 
effects output formats kml and gpx, causing invalid xml to be produced 
due to mismatched tags.  This results because  wpt_flags.new_trkseg is 
not updated when the track is reversed.  The new_trkseg flag should be 
set at, and only at, the start of each track segment, but after reversal 
it is set only at the end of each track segment (which was the start 
before reversal).  I am thinking about a fix but I don't have anything yet.

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/

BUG+PATCH: garmit-fit with extra header data fails

Niobos — 2012-05-01T13:23:22

The FIT files of my "Garmin Edge 200" contain two additional header
bytes. This patch does account for extra header bytes, but ignores them.

Greetings,
Niobos


PS: You should update the README.contrib: it says to make patches with
`cvs`, while the current code is in SVN.

Index: gpsbabel/garmin_fit.c
===================================================================
--- gpsbabel/garmin_fit.c(revision 4172)
+++ gpsbabel/garmin_fit.c(working copy)
< at >< at > -96,6 +96,12 < at >< at >
   if (sig[0] != '.' || sig[1] != 'F' || sig[2] != 'I' || sig[3] != 'T') {
     fatal(MYNAME ": .FIT signature missing\n");
   }
+
+  // Read in rest of header (if any)
+  len -= 12;
+  while( len-- ) {
+    gbfgetc(fin); // throw away unknown header data
+  }
 }
 
 static gbuint8

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include

obsolete files

tsteven4 — 2012-04-28T21:47:39

It appears to me that tpo.README.patch and tpo.testo.patch are obsolete 
and should be deleted.  They were introduced in r1509, Oct 31 2005.

svn delete tpo.README.patch
svn delete tpo.testo.patch



------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/

bad gpx introduced with r4167, r4168, r4170 "Steve from Climber adds line style handling in TPO3."

tsteven4 — 2012-04-28T19:25:56

The reference file LineStyles.gpx is invalid.  The ^K (#xB vertical tab) 
characters are not allowed.  The reference files LineStypes.gpx and 
LineStyles.kml are created by gpsbabel when building the target check:
   I note that r4170 (Steve adds line styling to TPO3 reader. Hacked KML 
writer to discard control characters in a description.)  may have been 
dealing with this issue in the kml output, but it was not dealt with in 
other output xml formats such as gpx.  Furthermore, I believe that some 
of the control characters discarded by the kml "hack" may be legal.  
Note that the ^M (carriage return) characters are not detected as 
invalid in either the gpx or the kml (before r4170 stripped them).  In 
general 3 characters with values less than #x20 (space) are allowed in 
strings, there is a fairly long series of specs that often comes down to 
http://www.w3.org/TR/2000/WD-xml-2e-20000814#charsets which allows |#x9 
| #xA | #xD  (horizontal tab, |line feed, carriage return).||  Perhaps a 
fix can be foun

gpx test patch

tsteven4 — 2012-04-28T01:14:22

This patch fixes a reference test input so that it is valid.  This is a 
reordering of the fix and type elements. The output of the test is not 
currently compared due to issues noted in testo.d/subrip.test.


Index: reference/track/gpx_subsecond-sample.gpx
===================================================================
--- reference/track/gpx_subsecond-sample.gpx(revision 4170)
+++ reference/track/gpx_subsecond-sample.gpx(working copy)
< at >< at > -14,40 +14,40 < at >< at >
   <time>2011-07-02T17:47:25Z</time>
   <speed>13.818100</speed>
   <name>trkpt-2011-07-02T17:47:25.000Z</name>
+  <type>T</type>
   <fix>3d</fix>
-<type>T</type>
 </trkpt>
 <trkpt lat="49.794710000" lon="9.833996000">
   <ele>289.300000</ele>
   <time>2011-07-02T17:47:25Z</time>
   <speed>14.049800</speed>
   <name>trkpt-2011-07-02T17:47:25.000Z</name>
+  <type>T</type>
   <fix>3d</fix>
-<type>T</type>
 </trkpt>
 <trkpt lat="49.794734000" lon="9.833976000">
   <ele>289.300000</ele>
   <time>2011-07-02T17:47:25Z</time>
   <speed>14.161100</speed>

Internal styles conflicting with custom xcsv styles?

Tom Paton — 2012-04-09T10:52:02

Hi,

Just ran into a little issue and I'm wondering if a) i'm doing
something wrong, b) there's a simple work around someone can think of,
or c) it's just a bug...

I'm currently using gpsbabel 1.4.2 and if I read from a file defined
using the xcsv internal styles (e.g. garmin301) and then try to
convert using a custom xcsv style file, the custom style file is
ignored and the output is in the same format as the input.

I suspect anyone familiar with the code can track down the issue
faster than they can reproduce my test case, somewhere around
xcsv.c:566 is my guess.

tom< at >neon:~/dev/gpsbabel/gpsbabel-1.4.2$ cat ~/garmin301.example
Garmin 301 data
Timestamp,Latitude, Longitude, Altitude(ft), heart rate
1280310051,31.158194,121.423790,93.832022F, 0
1280310052,31.158197,121.423798,93.832022F, 0
1280310053,31.158196,121.423859,93.832022F, 0
1280310054,31.158222,121.424141,93.832022F, 0
1280310056,31.158243,121.423981,93.832022F, 0
1280310057,31.158642,121.424347,93.832022F, 0
1280310058,31.158548,121.424095,93.8

Topo line color now works in KML out! Re: enhancement for Nat Geo "Topo" files

Steve — 2012-03-31T19:25:24

Ah-HAH! That was the hint I needed!
I had searched for gb_color, which I didn't find anywhere.

KML output now accurately contains the correct line color and line width
for Topo tracks. Here's an example (turn off the points to see the color):
   http://g.co/maps/f3hkp
I've mapped dashed lines into semi-transparent (regular lines have opacity
of 255=0xff, dashed lines have opacity 0x50) so all the Topo line style
info is there in one form or another. It turns out KML output really does
respect per-track line style and doesn't need command-line options... as
long as the input module loads ->line_width and ->line_color.bbggrr )with
or without ->line_color.opacity).

I'll need your help again to check in the new source. (see other email)

Thanks for staying engaged while I came up the learning curve! It's
daunting to fiddle with code that's been around this many years, but
yesterday's exchange broke the log jam.

Now, about adding color to GPX output... any hints about Garmin extensions
like the depth extensio

enhancement for Nat Geo "Topo" files

Steve — 2012-03-21T16:34:16

New to the list, please excuse if this is the wrong place to post!

I've modified tpo.c to read freehand route line styles (color,
width, dashed or solid) from .tpo files (the tpo3 format). The
code was already in place to find the style block, all I had to
do is reverse engineer that one block and use the style index
that the track-reading block extracted but did not use.

I put style info in the "desc" field of the track since there
doesn't seem to be any direct provision for track styles and
the Topo parser doesn't use the desc field for anything else.
It's visible in GPX output but would need to be put into a
private extension to be read by tools like MapSource that do
allow track colors or destinations like Google Maps that allow
both colors and widths.

Can anyone point me to a central definition for tracks, or
does each format have it's own?

There doesn't seem to be much recent activity on the Topo format,
and this is the first C programming I've done for many years, so
what's the procedure for getti

code/protocol for BG728 ("navibe mercury") logger

Thorben Thuermer — 2012-03-08T03:24:34

Hello,

i've been using this device, reverse-engineered it's protocol,
and wrote some code to interface with it (none could be found on the web),

...which i think i should share, an this seems to be the best place for that.

the BG728 ("navibe mercury") is a gps-based bicycle computer and gps logger.
(it omits the wheel-rotation sensor usually used by bicycle computers,
 and relies solely on gps as it's data-source, even for the speed-display)
it also includes some basic navigation features.
(i did not write code to program POIs/routes, but it could be done.)
as an extra, it contains a temperature sensor (temp data is not logged),
and maybe a digital compass - i didn't yet open the device to verify this.

the manufacturer appers to be http://www.cpss.com.tw/product/?d=37
a good reference page is http://home.icon.co.za/~supercyc/mercury.htm
and it's f.ex. sold by conrad: http://www.conrad.de/ce/de/product/372728
ofcourse the windows software is horrible and partially broken.
(for the map display it embeds IE

Contributing translation

Erik Krause — 2012-02-11T22:45:08

Hello,

I enhanced the German translation for gpsbabelfe_de (gpsbabel_de is 
already perfect). I went through the process described on 
http://www.gpsbabel.org/tips/translate.html
Unfortunately I didn't find a hint how to provide the translated files...

gb_color

tsteven4 — 2012-02-07T00:54:38

Shouldn't we make bbggrr a long to guarantee we get at least 32 bits, or 
perhaps int_least32_t?  I don't know of any problems, but I am preparing 
an enhancement and I want to program defensively.


------------------------------------------------------------------------------
Keep Your Developer Skills Current with LearnDevNow!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-d2d

patch - a bunch of documentation spellingcorrections

tsteven4 — 2012-02-05T21:30:01

Index: xmldoc/filters/options/transform-del.xml
===================================================================
--- xmldoc/filters/options/transform-del.xml(revision 4154)
+++ xmldoc/filters/options/transform-del.xml(working copy)
< at >< at > -1,5 +1,5 < at >< at >
 <para>
-This option, when used in connction with the wpt, rte, or trk options, tells
+This option, when used in connection with the wpt, rte, or trk options, tells
 GPSBabel to delete the source data after conversion.  This is most useful if 
 you are trying to avoid duplicated data in the output.
 </para>
Index: xmldoc/filters/options/discard-sat.xml
===================================================================
--- xmldoc/filters/options/discard-sat.xml(revision 4154)
+++ xmldoc/filters/options/discard-sat.xml(working copy)
< at >< at > -1,4 +1,4 < at >< at >
 <para>
-This option specifies the minimum required number of satelites.
+This option specifies the minimum required number of satellites.
 </para>
 
Index: xmldoc/filters/discard.xml
=============================

more svn mode fixes

tsteven4 — 2012-02-03T02:25:17

I recommend deleting svn:executable on the following files.  This can be 
done by running the attached script "reset_exec".  The log from my run 
is "reset_exec.txt".  I was able to build and test code afterwards, 
although that certainly doesn't attempt to execute everything that 
should be executable.

My strategy was
i) no executable files should be in the reference directory,
ii) no .ico, .png, .cpp, .xml files should be executable.



#!/bin/bash
find reference -type f -executable -exec svn propdel svn:executable {} \;
find . -name "*.ico" -executable -exec svn propdel svn:executable {} \;
find . -name "*.png" -executable -exec svn propdel svn:executable {} \;
find . -name "*.cpp" -executable -exec svn propdel svn:executable {} \;
find . -name "*.xml" -executable -exec svn propdel svn:executable {} \;
property 'svn:executable' deleted from 'reference/expertgps.gpx'.
property 'svn:executable' deleted from 'reference/mapsource.mps'.
property 'svn:executable' deleted from 'reference/mps-empty.mps'.
propert

.FIT conversion for Edge800 No Go

Stefan Skopnik — 2012-02-01T09:19:44

------------------------------------------------------------------------------
Keep Your Developer Skills Current with LearnDevNow!
The most comprehensive online learning library for Microsoft developers
is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
Metro Style Apps, more. Free future releases when you subscribe now!
http://p.sf.net/sfu/learndevnow-d2d_______________________________________________
Gpsbabel-code mailing list  http://www.gpsbabel.org
Gpsbabel-code< at >lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/gpsbabel-code