gpxviewer

gpx-viewer starts on the wrong side of the world

Bug #525427 reported by Joe Buck
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
gpxviewer
Fix Committed
High
Qball Cow
gpxviewer 0.3.0 "Three is a Magick Number"

Bug Description

I'm using version 0.1.2.

When I start gpx-viewer on a GPX file with a track that's located in California, it starts up with the map positioned on the other side of the word (apparently in Turkmenistan). Perhaps it's dropping the sign of the longitude?

What I would expect is for the tool to start by selecting the most zoomed-in scale that displays the whole track.

If it helps, I can send you the GPX file. Just say where to send it.

Revision history for this message
Joe Buck (jbuck-welsh-buck) wrote :

It's not as simple as reversing the sign of the longitude; I have another file that starts up in Antarctica. After I zoom way out and zoom back in, the display is correct, only the startup position is bogus.

Revision history for this message
Qball Cow (qball-qballcow) wrote :

Yeah it does something odd, it should normally "zoom out" when you select a track so the whole track is visible, but it goes wrong on startup..

Changed in gpx-viewer:
status: New → Confirmed
importance: Undecided → High
assignee: nobody → Qball Cow (qball-qballcow)
Revision history for this message
Joe Buck (jbuck-welsh-buck) wrote :

I debugged a little, and it looks like it might be a libchamplain issue. I'm building gpx-viewer on a Fedora 11 system using the distro-provided libchamplain, which is version 0.3.5, which is old. I trace it through to the champlain_view_ensure_visible call, and it looks like gpx-viewer is passing correct parameters but the map is positioned wrong.

I'll email you (Qball Cow) a GPX file. Please let me know if you see the bug. If you don't, I'll download and build with the latest libchamplain.

Revision history for this message
Joe Buck (jbuck-welsh-buck) wrote : Re: [Bug 525427] Re: gpx-viewer starts on the wrong side of the world

On 02/21/2010 11:22 AM, Qball Cow wrote:
> Yeah it does something odd, it should normally "zoom out" when you
> select a track so the whole track is visible, but it goes wrong on
> startup..
>
>
> ** Changed in: gpx-viewer
> Status: New => Confirmed
>
> ** Changed in: gpx-viewer
> Importance: Undecided => High
>
> ** Changed in: gpx-viewer
> Assignee: (unassigned) => Qball Cow (qball-qballcow)
>
>
 I'm attaching a GPX file that displays the problem. Could you see if you
see this? I'm thinking that it might be because I have an old
libchamplain (0.3.5).

This is the one that starts in Turkmenistan for me.

Thanks.

Revision history for this message
Qball Cow (qball-qballcow) wrote :

With libchamplain 0.5 and current bzr it seems to be working again.

I have had many problems with the 0.4.* champlain, so currently sticking to 0.5 (that isn't released)

Revision history for this message
Qball Cow (qball-qballcow) wrote :

This also works for the attached file.

Changed in gpx-viewer:
status: Confirmed → In Progress
Qball Cow (qball-qballcow)
Changed in gpx-viewer:
status: In Progress → Fix Committed
Qball Cow (qball-qballcow)
Changed in gpx-viewer:
milestone: none → 0.3.0
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.