New Map, Bad GPX format?

[+bensemailis]

Yesterday I downloaded GPX files, uploaded them to my garmin and it worked fine.

 

Today with the new map style, when I download them, it downloads them as "waypoints" rather then traditional Cache files.

 

Anyone have a fix for this?

[+noncentric]

Could you be more specific about how you downloaded the  .gpx  files?

Was there any difference in the method you used yesterday vs today?

[+WolfHH]

If I download a GPX file from the Search-Map (www.geocaching.com/play/map?...), the content is different than what you get from the Cache-Page (www.geocaching.com/geocache/GC....) or from pocket queries.

 

For example, the GPX from the Search-Map does not start with an xml processing instruction. The attributes elements are text-only (the xml-elements have no id or inc attributes). There are more differences like milliseconds vs. timezone in times elements and so on. Looks like a new implementation?

[+The A-Team]

9 hours ago, WolfHH said:

If I download a GPX file from the Search-Map (www.geocaching.com/play/map?...), the content is different than what you get from the Cache-Page (www.geocaching.com/geocache/GC....) or from pocket queries.

 

For example, the GPX from the Search-Map does not start with an xml processing instruction. The attributes elements are text-only (the xml-elements have no id or inc attributes). There are more differences like milliseconds vs. timezone in times elements and so on. Looks like a new implementation?

Yeah, I just compared a pair of GPX files for the same cache (after making the map one more readable with new lines and tabs) and there are a lot of differences in the map GPX, including (but not limited to):

• Missing <?xml...> tag in the map GPX - CRITICAL ISSUE
• Different <gpx...> tag
• Missing <author> tag
• Missing <bounds> tag
• Different <url>
• Missing Groundspeak extension version in <Groundspeak:cache>
• PR code rather than ID number in <Groundspeak:owner id=...> and in logs for <Groundspeak:finder id=...>
• No "id" or "inc" attributes on <Groundspeak:attribute> tags
• Missing "html" attribute on <Groundspeak:short_description> and <Groundspeak:long_description> tags

Not to mention many other small differences like the capitalization of "true" and "false", etc.

 

Something has gone horribly wrong somewhere along the line, whether that be some test code making it into production or a replacement GPX generator's output not being validated or tested.

[+bensemailis]

Thank you everyone for your responses 

 

On 4/23/2019 at 12:59 AM, noncentric said:

Could you be more specific about how you downloaded the  .gpx  files?

Was there any difference in the method you used yesterday vs today?

 

Same method, but I found that the GXP file is now different downloading from the list, as apposed to the map.

 

On 4/23/2019 at 3:49 AM, WolfHH said:

If I download a GPX file from the Search-Map (www.geocaching.com/play/map?...), the content is different than what you get from the Cache-Page (www.geocaching.com/geocache/GC....) or from pocket queries.

 

For example, the GPX from the Search-Map does not start with an xml processing instruction. The attributes elements are text-only (the xml-elements have no id or inc attributes). There are more differences like milliseconds vs. timezone in times elements and so on. Looks like a new implementation?

 

Correct. This is the difference that was throwing me off. My older Garmin cant handle the GPX file without the processing.

 

19 hours ago, The A-Team said:

Yeah, I just compared a pair of GPX files for the same cache (after making the map one more readable with new lines and tabs) and there are a lot of differences in the map GPX, including (but not limited to):

• Missing <?xml...> tag in the map GPX - CRITICAL ISSUE
• Different <gpx...> tag
• Missing <author> tag
• Missing <bounds> tag
• Different <url>
• Missing Groundspeak extension version in <Groundspeak:cache>
• PR code rather than ID number in <Groundspeak:owner id=...> and in logs for <Groundspeak:finder id=...>
• No "id" or "inc" attributes on <Groundspeak:attribute> tags
• Missing "html" attribute on <Groundspeak:short_description> and <Groundspeak:long_description> tags

Not to mention many other small differences like the capitalization of "true" and "false", etc.

 

Something has gone horribly wrong somewhere along the line, whether that be some test code making it into production or a replacement GPX generator's output not being validated or tested.

 

I agree. So for now I am just downloading from the list, after finding them on the map. A Little more work, but its what works.

[topografix]

I'm glad to see others are reporting this. I'm one of the original authors of the GPX standard (along with Jeremy Irish), and I just came here to try to report that the new change is creating invalid GPX files (see below). In addition to breaking Garmin receivers, this is causing my software (EasyGPS and ExpertGPS) to treat the caches as waypoints, because all of the additional cache data is getting ignored because of the invalid Groundspeak schema.

 

If anyone from Geocaching HQ reads this, you're creating broken GPX files, and you need to (as a minimum)

add back the missing the <?xml...> tag at the top

replace 

xmlns:Groundspeak="Groundspeak"

with 

xmlns:Groundspeak="http://www.Groundspeak.com/cache/1/0"

Edited April 30, 2019 by topografix

[+thebruce0]

This should absolutely be reported as a BUG.  The GPX data should follow the standard schema if there isn't a new version being implemented.

[+mjoniak]

Hello,

additionally, BaseCamp shows cache ID (GC69GV2) instead of the name (Vodarenska vez Bechovice). The same issue for both methods of GPX download. All caches are recognized as Waypoints, not as a cache. I downloaded caches a two weeks ago without any issue, with correct data and correctly displayed in BaseCamp. 

[+The A-Team]

Bump

 

It's now been several months, and this is still an issue. The <?xml> tag has since been added, but there are still various inconsistencies and there's something about the GPX files downloaded from the map that prevents my browser from detecting the file type correctly.