[Talk-GB] I have a philosophical question...

Gareth L o.i at live.co.uk
Sun Jan 13 23:54:04 UTC 2019


It’s easy to forget that OSM is a geographically referenced database of objects. That just happens to be rendered as a map. If you want to know how many post boxes have the George the 5th cypher, you can query the database to find out. If you want to limit that by geographic bounds (say, a city’s limits) then great, you can do that.

Google maps things to keep people in their ecosystem and find more out about where people want to go. They care most about points of interest so they can advertise accordingly, or suggest to those points of interest/businesses that they may want to pay google money so they get seen better than they would do normally. It’s why cities have streetview data barely a year old, but country roads are 7 years out of date. There’s little money in covering it.

Roads/lane markings are getting a lot of attention because of assistive driving system development causing a shedload of money being pumped into that area of interest.

I like adding to OSM as a pedestrian. It’s annoying to walk a circuitous route to a shop only to find there’s an alleyway that will take you there in 10% of the time.

The freely available database allows interesting data to be presented with low barrier to entry. As academia finds, it’s annoying to have to pay to view a journal behind a paywall, when you dunno if it even has the info you want.

A couple weaknesses for the main OSM ‘map’ I’ve found:
Transport routes (particularly buses) change too frequently and would be better as a separate service that is overlaid onto an OSM derived map.

All the different features that people have as priorities to add are valid, but it’s sometimes tricky to figure out the best order to add them in. Why add all the stiles and gates to a field before you have the footpaths added? Or adding bins and benches by roads before defining what kind of pavement/sidewalk it has.*

A couple Strengths of OSM:

Updates are fast. Google/bing are less fast at updating areas, and especially so if in a quiet region. OS only publish a new paper map when the old one has gone out of print, meaning the Leicestershire map is more out of date than the more popular peak district ones.

If you want more info available on xyz feature.. you can add it, and encourage others to do so also.


So, back to the philosophical question: I’d say it’s all important. However, the order that things are best added to the map could be helpful to know. Not in a tollgate “don’t add z until qrstuvwxy has been added first” way, but knowing what additional information is enabled to be added as a result of you adding a certain feature.

*sidewalks are such a nightmare in general in osm with no easy approach on how to best add them, it seems.

Not an easy question!
Gareth

From: BD<mailto:dzidek23 at tlen.pl>
Sent: 13 January 2019 23:10
To: talk-gb at openstreetmap.org<mailto:talk-gb at openstreetmap.org>
Subject: [Talk-GB] I have a philosophical question...

Hi All,

I do make my little contribution to the effort of OSM. Recently I added some data to Mapillary and consider adding more (for the use of other mappers).

After reading someone's OSM profile I started to think and now have some doubts... We (mappers) are concentrating on various areas of the map. Some are dedicated to buildings, some to geographical features others add businesses etc. Can someone explain what is the aim of OSM, are we trying to build a map better than Bing and Google (in towns and cities) or are we planning to create a useful tool for tourists (with paths, places of interests)?

What is the aim of OSM, what should we concentrate on?
Should we map roads for sat-nav or buildings for urban area accuracy? Paths and tourists attractions or schools and electric cars charging points?

many thanks,
dzidek23

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-gb/attachments/20190113/4c401ad0/attachment-0001.html>


More information about the Talk-GB mailing list