Re: [Talk-GB] Importing Website Data

[Paul Gillard]

Being a lone mapper up to now it's been interesting to read everyone's
comments, a good wide range of views.

Brian, to answer your question about priorities, I don't particularly have
any. My thoughts were to choose a selection of different types of company
and if see we can't work out where the low hanging fruit is based on the
responses we get. I'm open to other approaches though.

Rob Nickerson has been in touch about the help that the OSM UK chapter
could offer and we'll get a plan together for how to approach companies. If
anyone has thoughts to add on this then they're very welcome.

I'll come back if and when I get anywhere.

Cheers

On 19 December 2017 at 14:46, Brian Prangle  wrote:

> Paul thank you for suggesting this, it's certainly something as a UK
> community (and I guess more widely)we need to deal with. Unfortunately
> website data can be problematic, as others have already indicated, for us
> to use, but instead we should ask the organisations concerned to provide
> the data in a compatible format. Now that we have a formal body
> incorporated as the UK Local Chapter that should be a good vehicle for
> making the requests. Do you have an suggestions for a priority list? Does
> anyone else have any priorities?
>
> For those who decry the approach of using third party data, preferring
> instead the personally surveyed approach, I echo Ilya's and Warin's
> sentiments: Lloyds TSB demerged in 2013 and we still have 200 instances of
> Lloyds TSB, TA Centres became Army Reserve Centres at about the same time
> and we  still have about 40 instances of TA Centre, Shell purchased Total
> filling stations in 2012, and during the recent validation exercise on
> Shell data, name=Total was the commonest error. What about the the
> wholesale closure and transfer  of Post Offices or the planned closure of
> thousands of BT phone boxes?  We don't have the number of motivated mappers
> to do this,  and expecting evrything to be ground surveyed might be a
> reason why we have such a high attrition rate; so we should emulate the
> rest of society and use automated IT methods to assist us and make our
> lives easier where appropriate. OSM is a balance between IT  imported
> data/automated edits and human ground surveys.
>
> Regards
>
> Brian
>
>
>
> On 18 December 2017 at 13:15,  wrote:
>
>> Hi,
>>
>> Reading all the talk of Walmart and Shell imports recently got me to
>> wondering why we can't be doing more of this kind of thing.
>>
>> If store data can be pulled from directly from a company's public facing
>> website ('store finder' page) is there any reason we can't do such imports
>> without discussion with/permission from the company concerned?
>>
>> Paul
>>
>> ___
>> Talk-GB mailing list
>> Talk-GB@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk-gb
>>
>
>
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Importing Website Data

[Andy Mabbett]

On 21 December 2017 at 10:28, Frederik Ramm  wrote:

> Your approach will ultimately lead to every last
> large-chain pub in England being nicely mapped from afar, whereas the
> independent pub next door has to wait until a mapper comes around.

Did we not have a data set supplied by a trade body, that enabled us
to add every independent cycle shop in the UK?

-- 
Andy Mabbett
@pigsonthewing
http://pigsonthewing.org.uk

___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] OpenRailwayMap

[Jez Nicholson]

They're displaying The Volks Railway from Brighton Palace Pier to the
Marina, so I am happy. :)

On Thu, 21 Dec 2017 18:03 Philip Barnes,  wrote:

> On Mon, 2017-12-18 at 17:55 +, Paul Berry wrote:
>
> A nice little social media boost courtesy of Tim Dunn:
> https://twitter.com/MrTimDunn/status/942751174922555393
>
> There's an Open Railway Map of the world and the level of detail is
> incredible http://www.openrailwaymap.org/  
>
> According to the legend, halts should render differently to stations. Any 
> idea why that isn't happening.
>
>
> Also I see some quite big holes in the usage rendering, but not really enough 
> of a railway geek to fix them. For example am puzzled as to why The Heart
>
> of Wales line is a branch and The Cambrian is a main.
>
>
> Phil (trigpoint)
>
>
>
> ___
> Talk-GB mailing list
> Talk-GB@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-gb
>
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Importing Website Data

[Andy Townsend]

On 21/12/2017 09:15, Ilya Zverev wrote:

...
Or do you think the map does not need imports and that every shop and amenity 
will be mapped without them before they are out of business?



Maybe it's worth having a look at an example.

http://www.openstreetmap.org/way/245390956 is a petrol station in South 
Normanton, not that far from me.  It's not that out of the ordinary.  It 
was already mapped in OSM as a Shell petrol station (as most were) 
before Navads came along.


Here's the data that Navads was able to provide:

 * Postcode
 * Phone number


But here's the data that Navads didn't add and was already in OSM

 * There's a Post Office here
 * There's an ATM
 * There's a Spar convenience store
 * There's a branch of WHSmith


Clearly being able to provide e.g. licence-unencumbered postcode data is 
useful, but if we relied _just_ on the likes of Navads we'd have a 
pretty poor view of what's available here.


Similarly, Mark Goodge's point ("Inaccurate data is worse than missing 
data") is a good one, but in this case we do have a source on the 
changeset (see e.g. http://www.openstreetmap.org/way/69974213 up the 
road and http://www.openstreetmap.org/changeset/54785385 ) so if stuff 
is contradicted by future survey it's clear where it came from and can 
be easily updated.


Best Regards

Andy

___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] OpenRailwayMap

[Philip Barnes]

On Mon, 2017-12-18 at 17:55 +, Paul Berry wrote:
> A nice little social media boost courtesy of Tim Dunn:https://twitter
> .com/MrTimDunn/status/942751174922555393
> There's an Open Railway Map of the world and the level of detail is
> incredible
> http://www.openrailwaymap.org/ 
> 

According to the legend, halts should render differently to stations.
Any idea why that isn't happening.
Also I see some quite big holes in the usage rendering, but not really
enough of a railway geek to fix them. For example am puzzled as to why
The Heart of Wales line is a branch and The Cambrian is a main.
Phil (trigpoint)

___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Importing Website Data

[Colin Smale]

This discussion brings a couple of similar sayings to mind (and there
are many more in the same vein): 

_To Sacrifice The Good On The Altar Of The Perfect_
= and = 

_Perfect is the enemy of good_ [1] 

A dataset will never be perfect. Resisting an import because a small
proportion of the data is likely to be wrong, is a standpoint that can
never be countered. And yet, a more pragmatic assessment may come to a
different conclusion. Is "OSM" more valuable to its users after the
import than before? Is the net result considered an improvement?

Such assessments require a definition of "OSM", its "value" and its
"users" and I don't think there is a clear consensus about this, only a
collection of opinions. 

--colin 

[1] https://en.wikipedia.org/wiki/Perfect_is_the_enemy_of_good 

On 2017-12-21 17:06, Mark Goodge wrote:

> On 21/12/2017 15:49, Frederik Ramm wrote: Hi,
> 
> On 21.12.2017 16:13, Mark Goodge wrote: My vision of OSM is a movement which 
> places its users first, by
> providing the maximum utility possible for those who look at the maps.
> That means maximising the quantity, accuracy, relevance and timeliness
> of the data. 
> That is certainly a valid approach that many will subscribe to. The
> goals you mention will sometimes have to be weighed against each other
> ("is a large amount of inaccurate data better than a small amount of
> accurate data", "is it good to add this bulk data which is unlikely to
> be cared for by mappers and hence will soon lack in timeliness" etc) but
> on the whole they're a good selection.

Data has to be accurate. Inaccurate data is worse than missing data
(although, it must be noted that imprecise is not the same as
inaccurate, and we can tolerate a reasonable amount of imprecision
provided it is not misleadingly imprecise).

> I think that relevance plays a big role, and commercial players tend to
> claim that concept for themselves ("tell us something about you so we
> can display ads that are relevant to you") - in my view, a pub in your
> town is not "relevant" because the chain operating it thinks that it
> should be, but because the locals find it relevant.

As far as a map is concerned, something is relevant if it is there. Even
if only one person a year actually wants to know it's there :-)

Mark

___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Importing Website Data

[Mark Goodge]

On 21/12/2017 15:49, Frederik Ramm wrote:

Hi,

On 21.12.2017 16:13, Mark Goodge wrote:

My vision of OSM is a movement which places its users first, by
providing the maximum utility possible for those who look at the maps.
That means maximising the quantity, accuracy, relevance and timeliness
of the data.


That is certainly a valid approach that many will subscribe to. The
goals you mention will sometimes have to be weighed against each other
("is a large amount of inaccurate data better than a small amount of
accurate data", "is it good to add this bulk data which is unlikely to
be cared for by mappers and hence will soon lack in timeliness" etc) but
on the whole they're a good selection.


Data has to be accurate. Inaccurate data is worse than missing data 
(although, it must be noted that imprecise is not the same as 
inaccurate, and we can tolerate a reasonable amount of imprecision 
provided it is not misleadingly imprecise).



I think that relevance plays a big role, and commercial players tend to
claim that concept for themselves ("tell us something about you so we
can display ads that are relevant to you") - in my view, a pub in your
town is not "relevant" because the chain operating it thinks that it
should be, but because the locals find it relevant.


As far as a map is concerned, something is relevant if it is there. Even 
if only one person a year actually wants to know it's there :-)


Mark

___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Importing Website Data

[Frederik Ramm]

Hi,

On 21.12.2017 16:13, Mark Goodge wrote:
> My vision of OSM is a movement which places its users first, by
> providing the maximum utility possible for those who look at the maps.
> That means maximising the quantity, accuracy, relevance and timeliness
> of the data.

That is certainly a valid approach that many will subscribe to. The
goals you mention will sometimes have to be weighed against each other
("is a large amount of inaccurate data better than a small amount of
accurate data", "is it good to add this bulk data which is unlikely to
be cared for by mappers and hence will soon lack in timeliness" etc) but
on the whole they're a good selection.

I think that relevance plays a big role, and commercial players tend to
claim that concept for themselves ("tell us something about you so we
can display ads that are relevant to you") - in my view, a pub in your
town is not "relevant" because the chain operating it thinks that it
should be, but because the locals find it relevant.

I do concede that I've mapped many a business in random places I've been
to without asking the locals if it is relevant to them though.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"

___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Importing Website Data

[Brian Prangle]

Eloquently put as ever Frederick. I share your lament about the seemingly
unattainable vision, which I happen to share. I would dearly love there to
be 10 times the number of consistently active mappers in the UK, but it
seems to be an order of magnitude more difficult ot persuade people to
maintain a map and "fill in the gaps" rather than build out a map from
scratch. I also share your  view of the increasing grip that large
corporations have on our social and political life, much to our detriment.

However we share a space with them and if they have data we can use then we
should use it: it won't be perfect but it's in the interests of
corporations for their customers and suppliers to be able to find their
locations so they'll be expending some effort to achieve a high quality (
and Tom I've seen inside enough large corporations not be naive about their
level of quality, but as Warin says some data is better than no data  as
long as we  validate it as much as we can.). Also the population at large
who might  use our map will want to find these large corporate outlets, so
we need to make the effort for them.

 I'm concerned that rate of change in the UK's High Streets with closures
of banks, pubs, chain stores,phone boxes and almagamations,takeovers,
brand-name changes will overwhelm us and degrade the reputation we have
laboured to establish. I've certainly lost the will to keep up with this
consistently.  I'm almost convinced by Frederick's argument that we
shouldn't map any of this, but just the base information of building outine
and address and building name where it is prominent, just as Ordnance
Survey does. In mapping this level of detail we seem to have made  a rod
for our own backs.

Regards

Brian

On 21 December 2017 at 10:28, Frederik Ramm  wrote:

> Hi,
>
> On 21.12.2017 10:15, Ilya Zverev wrote:
> > Frederik and Tom, please explain what has been wrong with the last
> import, and why osm_conflate + cf_audit tools used for it (conflation +
> community validation) still do not attain the required quality for OSM
> contributions?
>
> Your tools are certainly best on the market. Still I wish they had not
> been written at all.
>
> > How would you build a process for importing large batches of business
> chains?
>
> I am 100% against importing large batches of business chains. That data
> has no place in *my* vision of OSM; it should reside in a separate open
> database which anyone can choose to add to their local data set if and
> when needed.
>
> My vision of OSM is that of a grassroots movement that makes its own
> decisions about the data. We decide what the fuel station around the
> corner is called - not the marketing department of the fuel station
> chain. Corporate interests already dictate enough of our everyday lives;
> I don't want Shell to "help" me show the "right" information about their
> fuel stations.
>
> A lot is wrong with capitalism; large chains are one of those things.
> Small independent shops, booksellers, or pubs *already* face
> difficulties against the marketing and purchasing power of big
> corporations. Your approach will ultimately lead to every last
> large-chain pub in England being nicely mapped from afar, whereas the
> independent pub next door has to wait until a mapper comes around. You
> might say "well one pub mapped in town is better than no pub", but this
> is not my opinion; I would have wished for OpenStreetMap to be a map
> made by the people, not a map made by corporations large enough to hire
> SEO companies to manage their online presence.
>
> I am totally aware that "my" vision of OSM is a romantic dream and that
> ultimately, OSM is going to be just another venue for big money to
> battle it out, but please allow me to at least talk about this romantic
> dream from time to time.
>
> As I said, part of "growing up" is probably to toss your romantic dreams
> an accept reality for what it is. I'm totally ok with you being more
> grown up than me.
>
> Bye
> Frederik
>
> --
> Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"
>
> ___
> Talk-GB mailing list
> Talk-GB@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-gb
>
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Importing Website Data

[Frederik Ramm]

Hi,

On 21.12.2017 10:15, Ilya Zverev wrote:
> Frederik and Tom, please explain what has been wrong with the last import, 
> and why osm_conflate + cf_audit tools used for it (conflation + community 
> validation) still do not attain the required quality for OSM contributions?

Your tools are certainly best on the market. Still I wish they had not
been written at all.

> How would you build a process for importing large batches of business chains? 

I am 100% against importing large batches of business chains. That data
has no place in *my* vision of OSM; it should reside in a separate open
database which anyone can choose to add to their local data set if and
when needed.

My vision of OSM is that of a grassroots movement that makes its own
decisions about the data. We decide what the fuel station around the
corner is called - not the marketing department of the fuel station
chain. Corporate interests already dictate enough of our everyday lives;
I don't want Shell to "help" me show the "right" information about their
fuel stations.

A lot is wrong with capitalism; large chains are one of those things.
Small independent shops, booksellers, or pubs *already* face
difficulties against the marketing and purchasing power of big
corporations. Your approach will ultimately lead to every last
large-chain pub in England being nicely mapped from afar, whereas the
independent pub next door has to wait until a mapper comes around. You
might say "well one pub mapped in town is better than no pub", but this
is not my opinion; I would have wished for OpenStreetMap to be a map
made by the people, not a map made by corporations large enough to hire
SEO companies to manage their online presence.

I am totally aware that "my" vision of OSM is a romantic dream and that
ultimately, OSM is going to be just another venue for big money to
battle it out, but please allow me to at least talk about this romantic
dream from time to time.

As I said, part of "growing up" is probably to toss your romantic dreams
an accept reality for what it is. I'm totally ok with you being more
grown up than me.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"

___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Importing Website Data

[Tom Hughes]

On 21/12/17 09:15, Ilya Zverev wrote:


Frederik and Tom, please explain what has been wrong with the last import, and 
why osm_conflate + cf_audit tools used for it (conflation + community 
validation) still do not attain the required quality for OSM contributions?


I wasn't commenting on any particular import just on the general principles.

I was merely trying to point out that the view that people often have of 
"official" data as somehow perfect is often far from the truth. I've 
heard enough real world stories of databases inside companies to know 
just how far from reality it can be.


People often imagine these things as perfectly curated and fully 
normalised and standardised when the reality is often that they're 
maintained as an excel spreadsheet by this weeks intern.



How would you build a process for importing large batches of business chains? 
Can I improve something in my tools, or should I build something better from 
scratch?


Well I probably wouldn't because it doesn't especially interest me and I 
have no commercial reason for wanting to do so. Plus I know that it's an 
extremely hard problem.


It's quite true that an import may well be better than nothing where 
things haven't been mapped or aren't being actively maintained, but it's 
equally true that an import that includes updating existing objects may 
sometimes make things worse, and I don't know how you can tell when you 
are making an object worse.


That's what makes it so hard.

That said I will certainly agree that what you're doing is far better 
than what many companies trying to get their clients locations into OSM 
have done in the past.


Tom

--
Tom Hughes (t...@compton.nu)
http://compton.nu/

___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] Importing Website Data

[Ilya Zverev]

Again, with sarcasm.

Frederik and Tom, please explain what has been wrong with the last import, and 
why osm_conflate + cf_audit tools used for it (conflation + community 
validation) still do not attain the required quality for OSM contributions?

How would you build a process for importing large batches of business chains? 
Can I improve something in my tools, or should I build something better from 
scratch?

Or do you think the map does not need imports and that every shop and amenity 
will be mapped without them before they are out of business?

Ilya

> 21 дек. 2017 г., в 10:20, Frederik Ramm  написал(а):
> 
> Hi,
> 
> On 19.12.2017 16:20, Tom Hughes wrote:
>> Which is exactly what everybody said about OSM when it started - that it
>> couldn't possibly work and there'd never be enough people.
>> Pretty sure we proved them wrong.
> 
> Perhaps that's what people mean when they say "OSM has to grow up" -
> that we need to finally replace our youthful enthusiasm with a more
> resigned "the grown-ups were right after all" attitude.
> 
> Luckily, SEO companies are here to help, to enable even more people to
> "engage with brands" through our map. Win-win!
> 
> Bye
> Frederik
> 
> -- 
> Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"
> 
> ___
> Talk-GB mailing list
> Talk-GB@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-gb


___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb