[talk-ph] [Fwd: [Fwd: [OpenStreetMap] Re: Re: Re: NGA-GNS uploads]]

Anthony G. Balico anthony.balico at gmail.com
Tue Sep 21 01:32:12 BST 2010

Hi Guys,

Axk and I had a bit of a discussion on GNS.  Your thoughts please?

I find the thread below so cluttered, im sure you'll also find it the
same way.  Sorry about that.

I just thought i should include my first reply so you would know the
start.  This is how it goes:

"Hi Ax,

Yeah, i was bored this past few days, wasn't able to go to office for
some reasons. So out of boredom I just thought of reading about GNS. I
noticed that the GNS is a work in progress and that it is frequently
updated. So I did try to crosscheck with the latest updates with what we
currently have now in Phils.

I was convinced of the corrections (of course not all of 'em is
perfectly true), this is one fine example
http://sautter.com/map/?zoom=15&lat=10.55199&lon=123.95908&layers=00B000TTFFFTFF If you enable both the Mapnik and ToH overlays, you will still notice the former and the corrected points (Mt Lantoan and Danao Peak). Checking these with GoogleEarth base, the corrected spots hit closer to the intended points." 

-------- Forwarded Message --------
From: axk <m-134594-628d8a at messages.openstreetmap.org>
To: anthony.balico at gmail.com
Subject: [OpenStreetMap] Re: Re: Re: NGA-GNS uploads
Date: Mon, 20 Sep 2010 03:26:35 +0100

Hi BlueArrow, 

axk has sent you a message through OpenStreetMap with the subject Re: Re: Re: NGA-GNS uploads:

hey anthony,

got your msg, tnx. i'm on travel d comin 9 days and only have very limited 
internet connectivity from my cellfon, thats why i didnt reply. will do 
once im back. if you want, you can forward our conversation to the mailing 
list, so there is some time for discussion already. regards, and talk to 
you in a week.


------- Original message -------
> From: BlueArrow <m-134593-6a7b43 at messages.openstreetmap.org>
> To: axel+osm at kollmorgen.net
> Sent: '10-09-20,  8:26
> Hi axk,
> BlueArrow has sent you a message through OpenStreetMap with the subject 
> Re: Re: NGA-GNS uploads:
> ==
> Hi Ax,
> How are you?  Sent you a reply from gmail but seems like it didn't reach 
> you.
> Resending my responses here in osm mail.
> Anthony
> On 2010-09-17 16:56:20 UTC axk wrote:
>> On 2010-09-16 07:55, BlueArrow wrote:
>> > On 2010-09-15 12:47:19 UTC axk wrote:
>> >> i noticed that you have been uploading NGA-GNS place files,
>> >> especially this one:
>> >> http://www.openstreetmap.org/browse/changeset/5778632 (i'm mapping
>> >> in the ormoc area). now while i think that extending osm's coverage
>> >> is good, i don't know if using NGA-GNS place names is a good idea:
>> >>
>> >> * http://wiki.openstreetmap.org/wiki/GEOnet_Names_Server#Caveats :
>> >>   + Once a name is put into the database, it is never removed unless
>> >> it is an obvious duplicate. This means that there are many, many
>> >> names that have no modern significance.
>> >>   + The geographical resolution is often very coarse, from
>> >> experience features can often be two to three kilometres from there
>> >> actual locations.
>> >>   + Places where people live are are generally just classifies as
>> >> "PPL", Populated Place. This can be anything from a city to what is
>> >> now just a farm house.
>> >>   + There is a small but significant amount of entries that are
>> >> inaccurate or plain wrong: mistranscriptions, places that are
>> >> listed as in one country but actually somewhere else completely.
>> >>
>> >> * also, 
>> http://wiki.openstreetmap.org/wiki/WikiProject_Philippines/Data_sources#Gu 
>> idelines_for_OSM-PH_bulk_data_import :
>> >>    + All planned imports should be announced to the talk-ph list.
>> >>    + Community-consensus should be made before imports.
>> >>
>> >> * there were some NGA-GNS imports before:
>> >> 
>> http://wiki.openstreetmap.org/wiki/WikiProject_Philippines/Data_sources#Ge 
>> onet_Name_Server
>> >> . however, they were
>> >>   + coarser (towns and cities, islands and islets, mountains, ridges 
>> and
>> >> hills) instead of 60 villages per 0.16 square degrees as in that 
>> changeset
>> >>   + we are (i am) still struggling with correcting these imports, see
>> >> http://www.mail-archive.com/talk-ph@openstreetmap.org/msg00360.html
>> >>
>> >> so, all in all, i think that - unless you double checked the uploaded
>> >> places with a secondary source - these imports should be rolled back.
>> >>
>> >> what do you think?
>> >
>> > [...]
>> >
>> > The other entries, these are the barangays which i thought is a good
>> > way to start the initiative to include in the map, until such time
>> > that it will be corrected on the ground.
>> it is *not* the barangays, but (see above) just "PPL", Populated Places
>> - "which can be anything from a city to just a farm house". plus (also
>> see above), these places might not exist anymore, might be misspelled,
>> might be at the wrong location.
> I stand corrected, didn't intend to generalize all as barangays :)
>> i checked the accuracy of the places in your changeset 5778632 by
>> overlaying it with a PHL barangay shapefile from GADM [1]. have a look
>> at it:
>>    http://osm.kollm.org/nga-gns-accuracy.html
> Wow, this is cool!
>> (the areas with the dotted gray borders and the green names are the
>> barangays from GADM). you will see that there are some matches, but that 
>> for the majority of your imported places, there is no corresponding
>> barangay, it is at the wrong location, it is misspelled, it is
>> duplicate, or in another way inaccurate.
>> [1] see 
>> http://www.mail-archive.com/talk-ph@openstreetmap.org/msg01617.html
>> > I'm afraid I have to disagree on you on rolling back the changes.
>> > The caveats you've highlighted are true to some extent, but not
>> > entirely right to date.
>> from the analysis above, more than half of the imported places are
>> wrong. as there is no way to tell which of them are good and which are
>> bad (we cannot use GADM, see [1]), i don't see any other choice. do you?
> This is the beauty of a collaborated project.  At some point somebody 
> will find a bad data.  Being a mapping wiki, that somebody have the 
> opportunity to correct the data.  This is no different with roads imported 
> from 50k topo maps, most of the time these are too off from where they are 
> supposed to be.  But in some ways it still helps that they are there. 
> Sooner or later these will get better as a result of the work of the 
> community.
> Please don't get me wrong here, Ax, i know where you're getting at.  You 
> worked hard to maintain a high-integrity data in your area, and you find 
> the bad data cluttering the place in a no-sense way.  But i still keep my 
> stand and thoughts that this is a good start to eventually include all 
> barangays (or farm houses, etc.) on the map.  Very fortunate that i picked 
> your area, i'm getting precious informations and ideas from you.  I 
> appreciate you for that.
> What's your thoughts on the mountain peak corrections that i pointed out? 
>  Do you find the gns updates reliable enough?  I'm interested with 
> mountain peaks too, its part of the job.  Correcting these data in OSM is 
> like hitting two birds with one stone, helping for a better OSM and giving 
> me knowledge and background that may be useful to my job.
>> > Announcing this act to talk-ph was missed out i admit.
>> would you mind if i forward our conversation to the list? can't hurt to
>> get some other points of view, can it?
> Sure thing.  I was actually thinking of posting a thread after i replied 
> to your message yesterday.  Just that i got busy with work after.
>> regards,


You can also read the message at http://www.openstreetmap.org/message/read/134594
and you can reply at http://www.openstreetmap.org/message/reply/134594

More information about the talk-ph mailing list