[Talk-us] Blame me for JOSM yellowness

[Apollinaris Schoell]

over all this is a complex topic, maybe I should summarize some  
obervations
tiger data is that the quality is from excellent to really bad in  
accuracy
tiger data is old and contains abandoned roads
tiger has no level info, no direction for oneways, no turn  
restrictions, or any other fancy info which we add in osm

all this requires rework for most freeways, but it is good enough for  
lot of residential areas.
As others pointed out the reviewed tag is too crude to mean anything  
we can all agree on. everyone has a different working style and  
interest.
to produce a map the relative positions of roads are good enough. This  
is the way maps have been done in the past. absolute accuracy is not  
important and this way you can beautify a map by shifting some  
features slightly.
for an online map this is still good enough. and it good enough for  
many mappers. I have come across severe errors on mototways and some  
locals had added every tiny area and POI beside it. I guess they want  
to see a beautiful map and don't care about the rest.
If you print such a map it is good enough for orientation. so their  
need is fulfilled.

for me osm is good enough if the data can be used for routing too.  
this requires turn restrictions, correct level, correct oneway  
directions, correct classification, use restrictions, ref tags,   
ideally maxspeed, ....

tiger:reviewed could measure anything in between. simply said one tag  
can not map a multi dimensional requirement and subjective among  
different users

osm doesn't have any metrics for quality or review model. might be  
really useful but it's not there yet.
the minimum requirement for a quality tag system is
define the use model and quality measurement

some examples.

- relative positions. good enough to print
- absolute precision +/- 5m, good enough for most GPS users.
- absolute precision ???? good for high precision GPS. but good enough  
for parcel data?
- navigation for cars, bikes, pedestrians
......

not sure if a project like osm can find enough mappers to add all  
these levels of accuracy. for now I will spend my time to fix the  
basics first. and shouldn't spend so much time on these emails  ;-)



On 25 Apr 2009, at 19:02 , Russ Nelson wrote:

>
> On Apr 24, 2009, at 7:47 AM, Apollinaris Schoell wrote:
>>
>> if it's an option I wouldn't wast a second to write about the pro/ 
>> con.
>> why does anyone try to force users to do it?
>> I have patched josm already but not every user knows how to do it.
>
> Apollinaris came to Sarah's mapping party in the Henry Coe State Park
> today (yay!).  We spoke briefly about this issue.  The problem (and I
> hope Apo will correct me if I mis-represent him) is that
> tiger:reviewed is but the crudest possible metric for data quality (I
> agree with that).  In Apo's mind it is worthless.  Thus, highlighting
> it is also worthless.  I'm open to a discussion of whether it's
> worthless or not, but I AGREE with Apo that if tiger:reviewed is
> worthless, then so is highlighting its presence.  I hope that Apo can
> agree that if we can find a way to make tiger:reviewed=no worth
> removing, then highlighting it for everyone (not an option, not a
> plugin) is the right thing to do.  We did a survey of people who have
> mapped, but not recently, and some of them said that they stopped
> because the "map is complete" in their area.  One of the reasons why I
> suggested this change is to let people know that the tiger data,
> unreviewed, is not complete.
>
> So, the crux of our disagreement is that Apo says that
> tiger:reviewed=no isn't worth editing, while I say that it's worth
> editing.
>
> --
> Russ Nelson - http://community.cloudmade.com/blog - http://wiki.openstreetmap.org/wiki/User:RussNelson
> russ at cloudmade.com - Twitter: Russ_OSM - http://openstreetmap.org/user/RussNelson
>
>
> _______________________________________________
> Talk-us mailing list
> Talk-us at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk-us