Issue 152 in osmeditor4android: consecutive identical nodes in ways

[osmedito...@googlecode.com]

Status: New
Owner: ----
Labels: Type-Defect Priority-Medium

New issue 152 by osm52...@googlemail.com: consecutive identical nodes in
ways
http://code.google.com/p/osmeditor4android/issues/detail?id=152

What steps will reproduce the problem?
dunno.

What is the expected output? What do you see instead?
dunno.

What version of the product are you using? On what operating system?
None. I use JOSM and GNU Emacs, if you really want to know.

Please provide any additional information below.

Please make sure not to create ways containing multiple consecutive
references to the same node (e.g. A-B-B-C).

Simple solution: before uploading, iterate through the each way's list of
nodes and discard those which are identical to their immediate predecessor.
However, such a situation should not even arise in the first place, so
something is rotten elsewhere in the code if such a way is found.

Proof that the issue is not hypothetical:

http://www.openstreetmap.org/browse/way/211732673 v1
2013-03-23 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/209292121 v1
2013-03-11 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/209291704 v1
2013-03-11 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/209291337 v1
2013-03-11 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/209290107 v1
2013-03-10 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/209285910 v1
2013-03-10 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/208783472 v1
2013-03-07 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/208783436 v1
2013-03-07 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/208122841 v1
2013-03-03 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/208056273 v1
2013-03-03 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/202441340 v1
2013-01-22 Vespucci 0.8.2r416
http://www.openstreetmap.org/browse/way/202195353 v1
2013-01-20 Vespucci 0.8.2r416
http://www.openstreetmap.org/browse/way/200422454 v1
2013-01-10 Vespucci 0.8.1r402
http://www.openstreetmap.org/browse/way/172253739 v2
2013-03-25 Vespucci 0.8.3r419
http://www.openstreetmap.org/browse/way/121583393 v2
2013-03-19 Vespucci 0.8.3r419

--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings

[osmedito...@googlecode.com]

Updates:
Status: Accepted
Owner: sp8...@gmail.com

Comment #1 on issue 152 by sp8...@gmail.com: consecutive identical nodes in
ways
http://code.google.com/p/osmeditor4android/issues/detail?id=152

Vespucci AFAIK currently doesn't check for such errors. Ir however
shouldn't be all to difficult to stop this from happening.

[osmedito...@googlecode.com]

Updates:
Status: Fixed

Comment #2 on issue 152 by sp8...@gmail.com: consecutive identical nodes in
ways
http://code.google.com/p/osmeditor4android/issues/detail?id=152

Tentative fix in 0.9 branch
http://code.google.com/p/osmeditor4android/source/detail?r=443

[osmedito...@googlecode.com]

Comment #3 on issue 152 by osm52...@googlemail.com: consecutive identical

nodes in ways
http://code.google.com/p/osmeditor4android/issues/detail?id=152

Thanks for the quick fix, Simon. I'm monitoring the creating of newly
created erroneous ways already, so I'll keep an eye on the issue.

[osmedito...@googlecode.com]

Comment #4 on issue 152 by sp8...@gmail.com: consecutive identical nodes in
ways
http://code.google.com/p/osmeditor4android/issues/detail?id=152

Note that the 0.9 branch probably is a couple of weeks away from a release,
I can patch the current trunk if deemed urgent.

[osmedito...@googlecode.com]

Comment #5 on issue 152 by osm52...@googlemail.com: consecutive identical

nodes in ways
http://code.google.com/p/osmeditor4android/issues/detail?id=152

Don't bother. Vespucci is just one out of several editor creating such
broken ways. The main source currently is JOSM, or more precisely its
FastDraw plugin. The bug has been fixed a few weeks ago, but people just
won't update.
I'll keep monitoring http://toolserver.org/~mapjedi/blame.txt . If the
issue has been properly fixed, Vespucci 0.9 (once released) should not
appear there.

In Germany and Austria my robot already cleans up those ways periodically.
In the long run I may extend its scope to worldwide. So while it is
important to have the editor fixed, there is no need to rush anything.

There's a related bug, well known from the Potlatch family, but also
present in Vespucci: ways consisting of only one node. Have a look at
changesets 14972189 15066983 15300985 15310476 15312283 15408916 15486076
15502888.