[OSM-talk] clean gpx tracks

[Lambert Carsten]

On Thursday 25 September 2008 18:01:36 Karl wrote:

> The GPX tracks are intended to show the basis for the ways and other data
> that is in the database, so I think one motivation for timestamps hearkens
> back to a desire to "show your work" to defend the source of OSM data
> against potential future claims of copyright infringement. In other words,
> with timestamps, it's more plausible that it was collected with an actual
> GPS receiver, instead of mocked up into GPX from some "tainted" source
> (with a license not compatible with OSM). Obviously timestamps could be
> synthesized (and I think there are even scripts that will do it for you if
> you want to upload your timestamp-less GPX tracks to OSM), but anyway,
> that's one reason I seem to recall why timestamps are required.

I understand that reasoning but it is not enough to impose the timestamps 
requirement in my view. Anybody who is going to go through the trouble to 
create fake gps tracks most likely has enough motivation to create fake time 
stamps, enter copyrighted material directly. The rule doesn't protect 
opentstreetmap in any way. If I upload a track and it is public anyone with 
concerns knows where to find me and ask me questions. And I can answer them 
because like most mappers I take pictures and geotag those. 
At the same time this rule is motivating people to change their data causing 
all data to be questionable. Chances are programs to change create etc. gps 
data will get more sophisticated and easier to use.

On Thursday 25 September 2008 18:20:36 Andy wrote:
> On Thu, Sep 25, 2008 at 5:01 PM, Karl Newman <siliconfiend at gmail.com> wrote:
> > The GPX tracks are intended to show the basis for the ways and other data
> > that is in the database, so I think one motivation for timestamps
> > hearkens back to a desire to "show your work" to defend the source of OSM
> > data against potential future claims of copyright infringement. In other
> > words, with timestamps, it's more plausible that it was collected with an
> > actual GPS receiver, instead of mocked up into GPX from some "tainted"
> > source (with a license not compatible with OSM). Obviously timestamps
> > could be synthesized (and I think there are even scripts that will do it
> > for you if you want to upload your timestamp-less GPX tracks to OSM), but
> > anyway, that's one reason I seem to recall why timestamps are required.
>
> Yes, that's the reason. Whoever wrote that stuff on the wiki page was
> making it up (shock horror).
>
> It's a simple hurdle to discourage using one of the many
> trace-over-X-maps websites and then uploading the end result into OSM.

It is also a hurdle to upload at all for some, and to upload accurate (cleaned 
up) data for others. Is it worth that? 

> The stuff about using timestamps to work out speeds is inaccurate.
>
> Lambert, it's really worth your while not to believe everything that
> you read on the wiki, the quality of stuff there is often particularly
> poor. Nor even believe everything on the mailing lists, this warning
> (perhaps?) included!

Of course I don't 'believe everything', but who knows maybe the speed thing is 
accurate (perhaps?), depending on my belief of course. :-)

No, I am just really concerned about this (seemingly untraceable decision 
making) and worst of all I think it is a bad decision. And the wiki text is 
all I could find on this issue.

Maybe I am going about this the wrong way and instead I should put in a 
proposal to get this 'rule' changed with the appropriate  discussion and 
voting?

Lambert