Re: [HOT] Could we get more out of our mappers by asking for less?

[Blake Girardot HOT/OSM]

On Mon, Feb 27, 2017 at 11:36 AM, Jo  wrote:
> If you use 'w' in JOSM there is hardly ever a need to delete user's
> contributions. It's really easy to drag the corners where they need to be,
> then finish of with 'q' to square the building.
>
> Polyglot

This is a good tip, I forgot about this one for fixing up mapping.

If I do feel the need to re-map from scratch, I do it on a blank
layer, then merge the layer with the existing mapping layer, then use
Replace Geometry for each building pair (old and new). I estimate it
takes only about 10% longer than just totally re-mapping from scratch,
but that is worth it to me to keep the history and original mapper's
contribution.

Video demonstration of that process: https://youtu.be/Kv5AOmX8M9g

Thanks again Polyglot for reminding us about Improve Way for improving
building mapping.

Cheers
Blake







> 2017-02-26 20:00 GMT+01:00 Russell Deffner :
>>
>> On this topic, there has been concern over validators doing so much
>> deletion of recently mapped objects.  This triggers a few monitoring sites
>> that watch for 'suspicious changesets'; it doesn't mean much in the
>> immediate, but thinking longer term for the OSM project, those algorithms
>> could be used for automatic warnings or blocks.  I think no matter what,
>> we're always going to be dealing with poor mapping, so we should also look
>> at some best practices/rules of thumb for when to delete versus improve a
>> previously mapped feature.  Although I do find myself deleting and
>> re-drawing sometimes, I try to minimize that as there is something to
>> preserving the original mappers contribution.  This can also be an issue for
>> trying to calculate mappers contributions for a mapathon or project if a
>> validator unnecessarily deletes the objects.
>>
>> To quickly touch on project creation; I think roads and
>> settlements/residential areas go well together - they both should be done
>> with larger task squares than for building projects, but not sure it matters
>> too much which one comes first, but for disaster response as an example, I
>> would think roads and settlements are first priority followed by buildings.
>>
>> Happy Mapping,
>> =Russ
>>
>> -Original Message-
>> From: Andrew Buck [mailto:andrew.r.b...@gmail.com]
>> Sent: Sunday, February 26, 2017 10:20 AM
>> To: hot@openstreetmap.org
>> Subject: Re: [HOT] Could we get more out of our mappers by asking for
>> less?
>>
>> > I think this is a different conversation, I would prefer mappers take a
>> > little more time and care, some buildings I've seen mapped have little
>> > relationship to the size or shape of the building.
>> >
>> > Cheerio John
>>
>> I agree on this.  Especially with respect to buildings.  Roads are
>> fairly easy to add refinements to by doing them one at a time and using
>> the replace geometry tool in JOSM, however poorly mapped buildings take
>> longer to clean up than simply deleting them and mapping them fresh.
>>
>> I would rather see a user add 10 buildings that are high quality then
>> 100 or even 1000 that have to be remapped by someone else.  Having to
>> map something that has already been mapped poorly by someone else is
>> really disheartening.  On the one hand you spend your whole time mapping
>> thinking about how your work is duplicating that of someone else, and on
>> the other hand you feel bad about removing their contribution from the
>> database.  So you get hit with it on both fronts.
>>
>> -AndrewBuck
>>
>>
>>
>>
>> ___
>> HOT mailing list
>> HOT@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/hot
>
>
>
> ___
> HOT mailing list
> HOT@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot
>



-- 

Blake Girardot
Humanitarian OpenStreetMap Team, TM3 Project Manager
skype: jblakegirardot
HOT Core Team Contact: i...@hotosm.org
Live OSM Mapper-Support channel - https://hotosm-slack.herokuapp.com/
BE A PART OF HOT'S MICRO GRANTS: https://donate.hotosm.org/

___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] Could we get more out of our mappers by asking for less?

[Jo]

If you use 'w' in JOSM there is hardly ever a need to delete user's
contributions. It's really easy to drag the corners where they need to be,
then finish of with 'q' to square the building.

Same goes for highways (Ctrl-left mouse button to add nodes where needed).
Truly nifty that improve way accuracy tool.

Polyglot

2017-02-26 20:00 GMT+01:00 Russell Deffner :

> On this topic, there has been concern over validators doing so much
> deletion of recently mapped objects.  This triggers a few monitoring sites
> that watch for 'suspicious changesets'; it doesn't mean much in the
> immediate, but thinking longer term for the OSM project, those algorithms
> could be used for automatic warnings or blocks.  I think no matter what,
> we're always going to be dealing with poor mapping, so we should also look
> at some best practices/rules of thumb for when to delete versus improve a
> previously mapped feature.  Although I do find myself deleting and
> re-drawing sometimes, I try to minimize that as there is something to
> preserving the original mappers contribution.  This can also be an issue
> for trying to calculate mappers contributions for a mapathon or project if
> a validator unnecessarily deletes the objects.
>
> To quickly touch on project creation; I think roads and
> settlements/residential areas go well together - they both should be done
> with larger task squares than for building projects, but not sure it
> matters too much which one comes first, but for disaster response as an
> example, I would think roads and settlements are first priority followed by
> buildings.
>
> Happy Mapping,
> =Russ
>
> -Original Message-
> From: Andrew Buck [mailto:andrew.r.b...@gmail.com]
> Sent: Sunday, February 26, 2017 10:20 AM
> To: hot@openstreetmap.org
> Subject: Re: [HOT] Could we get more out of our mappers by asking for less?
>
> > I think this is a different conversation, I would prefer mappers take a
> > little more time and care, some buildings I've seen mapped have little
> > relationship to the size or shape of the building.
> >
> > Cheerio John
>
> I agree on this.  Especially with respect to buildings.  Roads are
> fairly easy to add refinements to by doing them one at a time and using
> the replace geometry tool in JOSM, however poorly mapped buildings take
> longer to clean up than simply deleting them and mapping them fresh.
>
> I would rather see a user add 10 buildings that are high quality then
> 100 or even 1000 that have to be remapped by someone else.  Having to
> map something that has already been mapped poorly by someone else is
> really disheartening.  On the one hand you spend your whole time mapping
> thinking about how your work is duplicating that of someone else, and on
> the other hand you feel bad about removing their contribution from the
> database.  So you get hit with it on both fronts.
>
> -AndrewBuck
>
>
>
>
> ___
> HOT mailing list
> HOT@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot
>
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] Could we get more out of our mappers by asking for less?

[Russell Deffner]

On this topic, there has been concern over validators doing so much deletion of 
recently mapped objects.  This triggers a few monitoring sites that watch for 
'suspicious changesets'; it doesn't mean much in the immediate, but thinking 
longer term for the OSM project, those algorithms could be used for automatic 
warnings or blocks.  I think no matter what, we're always going to be dealing 
with poor mapping, so we should also look at some best practices/rules of thumb 
for when to delete versus improve a previously mapped feature.  Although I do 
find myself deleting and re-drawing sometimes, I try to minimize that as there 
is something to preserving the original mappers contribution.  This can also be 
an issue for trying to calculate mappers contributions for a mapathon or 
project if a validator unnecessarily deletes the objects.

To quickly touch on project creation; I think roads and settlements/residential 
areas go well together - they both should be done with larger task squares than 
for building projects, but not sure it matters too much which one comes first, 
but for disaster response as an example, I would think roads and settlements 
are first priority followed by buildings.

Happy Mapping,
=Russ

-Original Message-
From: Andrew Buck [mailto:andrew.r.b...@gmail.com] 
Sent: Sunday, February 26, 2017 10:20 AM
To: hot@openstreetmap.org
Subject: Re: [HOT] Could we get more out of our mappers by asking for less?

> I think this is a different conversation, I would prefer mappers take a
> little more time and care, some buildings I've seen mapped have little
> relationship to the size or shape of the building.
> 
> Cheerio John

I agree on this.  Especially with respect to buildings.  Roads are
fairly easy to add refinements to by doing them one at a time and using
the replace geometry tool in JOSM, however poorly mapped buildings take
longer to clean up than simply deleting them and mapping them fresh.

I would rather see a user add 10 buildings that are high quality then
100 or even 1000 that have to be remapped by someone else.  Having to
map something that has already been mapped poorly by someone else is
really disheartening.  On the one hand you spend your whole time mapping
thinking about how your work is duplicating that of someone else, and on
the other hand you feel bad about removing their contribution from the
database.  So you get hit with it on both fronts.

-AndrewBuck




___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] Could we get more out of our mappers by asking for less?

[Andrew Buck]

> I think this is a different conversation, I would prefer mappers take a
> little more time and care, some buildings I've seen mapped have little
> relationship to the size or shape of the building.
> 
> Cheerio John

I agree on this.  Especially with respect to buildings.  Roads are
fairly easy to add refinements to by doing them one at a time and using
the replace geometry tool in JOSM, however poorly mapped buildings take
longer to clean up than simply deleting them and mapping them fresh.

I would rather see a user add 10 buildings that are high quality then
100 or even 1000 that have to be remapped by someone else.  Having to
map something that has already been mapped poorly by someone else is
really disheartening.  On the one hand you spend your whole time mapping
thinking about how your work is duplicating that of someone else, and on
the other hand you feel bad about removing their contribution from the
database.  So you get hit with it on both fronts.

-AndrewBuck




signature.asc
Description: OpenPGP digital signature
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] Could we get more out of our mappers by asking for less?

[john whelan]

> I would switch it: Map buildings (and maybe residential areas) first and
then roads. Buildings or residential areas help mapping roads (if a road is
a track or unclassified) while roads are no help when mapping buildings.

I would agree that mapping highways without settlements is more difficult.
If its a building project then mapping the residential highways first makes
it a bit easier and divides the tiles up.

Cheerio John

On 26 February 2017 at 05:11, Florian Niel  wrote:

> I would switch it: Map buildings (and maybe residential areas) first and
> then roads. Buildings or residential areas help mapping roads (if a road is
> a track or unclassified) while roads are no help when mapping buildings.
> Am 26.02.2017 um 03:36 schrieb Vao Matua:
>
> John,
>
> I think there is merit to what you are saying.
>
> The one caution I would suggest is that "If something is worth mapping it
> is worth mapping correctly".  Here in Ethiopia I have spent a lot of time
> adding nodes to roads that were digitized at a coarse scale.  Mapping
> quickly doesn't have to be done crudely.
> There is no troupe of magic mapping elves that clean up the rough mapping.
>
> The next set of HOT projects I will set up will be roads/highways only and
> then buildings only.
>
> Keep mapping,
>
> Emmor
>
> On Sun, Feb 26, 2017 at 1:16 AM, john whelan 
> wrote:
>
>> Just an ideal thought.  It came about as I'm loading in bits of Nigeria
>> by the enamel bucket at the moment then looking for untagged ways and
>> crossing highways.
>>
>> What stands out are blocks where every track and settlement is mapped and
>> blank bits of the map.  I've been attempting to join bits of highway
>> together to reduce the islands.
>>
>> Projects that ask for buildings are different but for projects where we
>> are looking to map settlements and connecting highways if we omitted the
>> paths and tracks leading to fields and rivers and mapped say two or three
>> highways max to a cluster of three huts we might be able to get rid of some
>> of the large gaps in the map and unless someone can think of a reason why
>> tracks and paths leading to fields are essential I think many NGOs could
>> manage without.  In return we could make more settlements and highways
>> available in the same time frame.
>>
>> Part of the problem is each project is looked at separately.  Each
>> project manager only thinks in terms of their own project rather than in
>> terms of how much resources are available overall.  I am aware that some
>> organisations feel they are independent having their own mapathons but to
>> get the quality up you need validation and if you don't believe me just
>> have a look at the map after a mapathon.  Nigeria is full of settlements
>> wrongly tagged, building=residential or yes are amongst the most common
>> incorrect tagging.  A number of small villages have been interconnected
>> with highway=primary etc.  Crossing ways by the hundred.
>>
>> I don't know how you would write the instructions for a such a project
>> but it might help the projects get completed more quickly.  I heard an off
>> the cuff remark about a project recently, "It'll take a year at that level
>> of detail.".   Off the cuff comments are more gut feel than accurate but
>> there was an element of truth in the remark and I had the feeling the data
>> was desired within a month or two.  I would suggest if you want the mapping
>> done quickly reduce the scope a little.
>>
>> Thoughts?
>>
>> Thanks John
>>
>> ___
>> HOT mailing list
>> HOT@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/hot
>>
>>
>
>
> ___
> HOT mailing 
> listHOT@openstreetmap.orghttps://lists.openstreetmap.org/listinfo/hot
>
>
>
> ___
> HOT mailing list
> HOT@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot
>
>
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] Could we get more out of our mappers by asking for less?

[john whelan]

>The one caution I would suggest is that "If something is worth mapping it
is worth mapping correctly".  Here in Ethiopia I have spent a lot of time
adding nodes to roads that were digitized at a coarse scale.  Mapping
quickly doesn't have to be done crudely.
There is no troupe of magic mapping elves that clean up the rough mapping.

I think this is a different conversation, I would prefer mappers take a
little more time and care, some buildings I've seen mapped have little
relationship to the size or shape of the building.

Cheerio John

On 25 February 2017 at 21:36, Vao Matua  wrote:

> John,
>
> I think there is merit to what you are saying.
>
> The one caution I would suggest is that "If something is worth mapping it
> is worth mapping correctly".  Here in Ethiopia I have spent a lot of time
> adding nodes to roads that were digitized at a coarse scale.  Mapping
> quickly doesn't have to be done crudely.
> There is no troupe of magic mapping elves that clean up the rough mapping.
>
> The next set of HOT projects I will set up will be roads/highways only and
> then buildings only.
>
> Keep mapping,
>
> Emmor
>
> On Sun, Feb 26, 2017 at 1:16 AM, john whelan 
> wrote:
>
>> Just an ideal thought.  It came about as I'm loading in bits of Nigeria
>> by the enamel bucket at the moment then looking for untagged ways and
>> crossing highways.
>>
>> What stands out are blocks where every track and settlement is mapped and
>> blank bits of the map.  I've been attempting to join bits of highway
>> together to reduce the islands.
>>
>> Projects that ask for buildings are different but for projects where we
>> are looking to map settlements and connecting highways if we omitted the
>> paths and tracks leading to fields and rivers and mapped say two or three
>> highways max to a cluster of three huts we might be able to get rid of some
>> of the large gaps in the map and unless someone can think of a reason why
>> tracks and paths leading to fields are essential I think many NGOs could
>> manage without.  In return we could make more settlements and highways
>> available in the same time frame.
>>
>> Part of the problem is each project is looked at separately.  Each
>> project manager only thinks in terms of their own project rather than in
>> terms of how much resources are available overall.  I am aware that some
>> organisations feel they are independent having their own mapathons but to
>> get the quality up you need validation and if you don't believe me just
>> have a look at the map after a mapathon.  Nigeria is full of settlements
>> wrongly tagged, building=residential or yes are amongst the most common
>> incorrect tagging.  A number of small villages have been interconnected
>> with highway=primary etc.  Crossing ways by the hundred.
>>
>> I don't know how you would write the instructions for a such a project
>> but it might help the projects get completed more quickly.  I heard an off
>> the cuff remark about a project recently, "It'll take a year at that level
>> of detail.".   Off the cuff comments are more gut feel than accurate but
>> there was an element of truth in the remark and I had the feeling the data
>> was desired within a month or two.  I would suggest if you want the mapping
>> done quickly reduce the scope a little.
>>
>> Thoughts?
>>
>> Thanks John
>>
>> ___
>> HOT mailing list
>> HOT@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/hot
>>
>>
>
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] Could we get more out of our mappers by asking for less?

[Florian Niel]

I would switch it: Map buildings (and maybe residential areas) first and 
then roads. Buildings or residential areas help mapping roads (if a road 
is a track or unclassified) while roads are no help when mapping buildings.


Am 26.02.2017 um 03:36 schrieb Vao Matua:

John,

I think there is merit to what you are saying.

The one caution I would suggest is that "If something is worth mapping 
it is worth mapping correctly".  Here in Ethiopia I have spent a lot 
of time adding nodes to roads that were digitized at a coarse scale.  
Mapping quickly doesn't have to be done crudely.
There is no troupe of magic mapping elves that clean up the rough 
mapping.


The next set of HOT projects I will set up will be roads/highways only 
and then buildings only.


Keep mapping,

Emmor

On Sun, Feb 26, 2017 at 1:16 AM, john whelan > wrote:


Just an ideal thought.  It came about as I'm loading in bits of
Nigeria by the enamel bucket at the moment then looking for
untagged ways and crossing highways.

What stands out are blocks where every track and settlement is
mapped and blank bits of the map.  I've been attempting to join
bits of highway together to reduce the islands.

Projects that ask for buildings are different but for projects
where we are looking to map settlements and connecting highways if
we omitted the paths and tracks leading to fields and rivers and
mapped say two or three highways max to a cluster of three huts we
might be able to get rid of some of the large gaps in the map and
unless someone can think of a reason why tracks and paths leading
to fields are essential I think many NGOs could manage without. 
In return we could make more settlements and highways available in

the same time frame.

Part of the problem is each project is looked at separately.  Each
project manager only thinks in terms of their own project rather
than in terms of how much resources are available overall.  I am
aware that some organisations feel they are independent having
their own mapathons but to get the quality up you need validation
and if you don't believe me just have a look at the map after a
mapathon.  Nigeria is full of settlements wrongly tagged,
building=residential or yes are amongst the most common incorrect
tagging.  A number of small villages have been interconnected with
highway=primary etc.  Crossing ways by the hundred.

I don't know how you would write the instructions for a such a
project but it might help the projects get completed more
quickly.  I heard an off the cuff remark about a project recently,
"It'll take a year at that level of detail.".   Off the cuff
comments are more gut feel than accurate but there was an element
of truth in the remark and I had the feeling the data was desired
within a month or two.  I would suggest if you want the mapping
done quickly reduce the scope a little.

Thoughts?

Thanks John

___
HOT mailing list
HOT@openstreetmap.org 
https://lists.openstreetmap.org/listinfo/hot





___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot


___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

Re: [HOT] Could we get more out of our mappers by asking for less?

[Vao Matua]

John,

I think there is merit to what you are saying.

The one caution I would suggest is that "If something is worth mapping it
is worth mapping correctly".  Here in Ethiopia I have spent a lot of time
adding nodes to roads that were digitized at a coarse scale.  Mapping
quickly doesn't have to be done crudely.
There is no troupe of magic mapping elves that clean up the rough mapping.

The next set of HOT projects I will set up will be roads/highways only and
then buildings only.

Keep mapping,

Emmor

On Sun, Feb 26, 2017 at 1:16 AM, john whelan  wrote:

> Just an ideal thought.  It came about as I'm loading in bits of Nigeria by
> the enamel bucket at the moment then looking for untagged ways and crossing
> highways.
>
> What stands out are blocks where every track and settlement is mapped and
> blank bits of the map.  I've been attempting to join bits of highway
> together to reduce the islands.
>
> Projects that ask for buildings are different but for projects where we
> are looking to map settlements and connecting highways if we omitted the
> paths and tracks leading to fields and rivers and mapped say two or three
> highways max to a cluster of three huts we might be able to get rid of some
> of the large gaps in the map and unless someone can think of a reason why
> tracks and paths leading to fields are essential I think many NGOs could
> manage without.  In return we could make more settlements and highways
> available in the same time frame.
>
> Part of the problem is each project is looked at separately.  Each project
> manager only thinks in terms of their own project rather than in terms of
> how much resources are available overall.  I am aware that some
> organisations feel they are independent having their own mapathons but to
> get the quality up you need validation and if you don't believe me just
> have a look at the map after a mapathon.  Nigeria is full of settlements
> wrongly tagged, building=residential or yes are amongst the most common
> incorrect tagging.  A number of small villages have been interconnected
> with highway=primary etc.  Crossing ways by the hundred.
>
> I don't know how you would write the instructions for a such a project but
> it might help the projects get completed more quickly.  I heard an off the
> cuff remark about a project recently, "It'll take a year at that level of
> detail.".   Off the cuff comments are more gut feel than accurate but there
> was an element of truth in the remark and I had the feeling the data was
> desired within a month or two.  I would suggest if you want the mapping
> done quickly reduce the scope a little.
>
> Thoughts?
>
> Thanks John
>
> ___
> HOT mailing list
> HOT@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/hot
>
>
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot

[HOT] Could we get more out of our mappers by asking for less?

[john whelan]

Just an ideal thought.  It came about as I'm loading in bits of Nigeria by
the enamel bucket at the moment then looking for untagged ways and crossing
highways.

What stands out are blocks where every track and settlement is mapped and
blank bits of the map.  I've been attempting to join bits of highway
together to reduce the islands.

Projects that ask for buildings are different but for projects where we are
looking to map settlements and connecting highways if we omitted the paths
and tracks leading to fields and rivers and mapped say two or three
highways max to a cluster of three huts we might be able to get rid of some
of the large gaps in the map and unless someone can think of a reason why
tracks and paths leading to fields are essential I think many NGOs could
manage without.  In return we could make more settlements and highways
available in the same time frame.

Part of the problem is each project is looked at separately.  Each project
manager only thinks in terms of their own project rather than in terms of
how much resources are available overall.  I am aware that some
organisations feel they are independent having their own mapathons but to
get the quality up you need validation and if you don't believe me just
have a look at the map after a mapathon.  Nigeria is full of settlements
wrongly tagged, building=residential or yes are amongst the most common
incorrect tagging.  A number of small villages have been interconnected
with highway=primary etc.  Crossing ways by the hundred.

I don't know how you would write the instructions for a such a project but
it might help the projects get completed more quickly.  I heard an off the
cuff remark about a project recently, "It'll take a year at that level of
detail.".   Off the cuff comments are more gut feel than accurate but there
was an element of truth in the remark and I had the feeling the data was
desired within a month or two.  I would suggest if you want the mapping
done quickly reduce the scope a little.

Thoughts?

Thanks John
___
HOT mailing list
HOT@openstreetmap.org
https://lists.openstreetmap.org/listinfo/hot