OpenStreetMap

Govanus has commented on the following diary entries

Post When Comment
On Mapping Golf Courses 8 days ago

the main way is to suck in all the data for series of bounding boxes that cover the site in question then use sort process (often linked to renderstyling files like Mapcss and ealy xml stylesheets. these rejct everything you don't tell them to map in the render (mapcss can in advanceced use not only drill down but do both computation on varibles stored in tags and plot multiple line and area features {see the road lane stylesheet add-on option of a multilane highway to see it rendered with dashed line between each lane on a route widenend to meet the number of lanes}. MapCSS can in most systems if study yhte advance guides work with nested relations withen boolean equations of different tags).

So it shouldn't be too hard with some systems to avoid roads and airports if that is a bit unhelpful complex and reliant on others work you could try a data appoch and treat the xml datafile in a spreadsheet or text editior to sort thourgh using find and replace. If you you have too much to do I not the odd course now and then and need something quicker for baulk work then try reasuching the likes of gawk/awk, grep, perl and other text pocessing systems oftern originating on unix machines but are now availble from some place for a lot of the other popular platforms like MS-Windows, Risc-OS and different types of the Mac-OS's (pre & post X) etc.

Ihope this might be useful.

Oxford routes. 19 days ago

I'm not arogant just after spending overtime on the issue a little bit driven or determined ;) I think the grammer reflected the weriness.

Convince me and I'll be convinced :)

Having read anything pertaining by subject header from 2007 about A-roads and checked the same on the local Cotswolds section that didn't begin before 2009 by focusing on the last year and this. I seems that it was actually agreed that all A roads would either be green Trunks or red primaries with all B-roads coming as Secondaries. Only two people in the cotsowld group seemed to show concern for the high street reming tertairy or below. The only other option I could deduce was that in Dec 2007 when you met in the Jerrico Arms pub you may have discussed it and then not put any minites up on the list...

You also seem to have difficulty keeping the opion that it should remain so lowly rated that you felt that the council would need to change it to meet your point of view.

So it looks like the community of officials and some of the locals are struggling to agree with your stated position. {thats just a judgement of what been presented to me to detmine a judgement on - in the sence of being pusuaded otherwise}.

As nobody wanted to take up the question of tagging of areas that if not messed up by others can work with older systems by making the new ares look distigusable. I think some thought that the logic of useing the same taging for points, ways, areas and relations would be the most logical - but its been shown that two meny things take ways and areas (both held in data the same way as ways cept some additional tagging or tag inferance of area like landuse, ment that wasn't going to fly with roads like it was solved for rivers land use and buildings etc) messy but thats dealing with legacy for you.

I do agree that boycotting for the demotion of a major road just because it works in two halves during the day is a bit too political.

I've to leave for closeing now but I'll look some more in a few sdays at your grooups .

Oxford routes. 21 days ago

Thinks for the messages coming in from so many paths that I'm largely out of sync and in lag to many... I don't read my email as often as some may think so if I map before reading changeset comments might take 2 to 4 days to get noticed! they do get there in the end but they might be out of date by then.

ok so lets try and answer some the flurry of points and counterpoints before we need to erect another memorial to martyrs south of the covered market [this for those not familiar to Oxford under discussion relates to the two memorials marking where Catholics were burned and killed during the civil-war in Christendom some centuries ago].

first I have been walking though these roads for a long time and all through the recent edit phase,(add have been out surveying again today) and have a good feel of the weight of traffic by length as well as number for most parts of the day and night (I usually walk though the site anytime upto about 1-3am in the morning as well as throughout the day and evenings). So unlike other areas I might map more lightly for less in depth info this is a spot I do no and from today have the ref's for every sign in the middle of the road and a proportion of the lamps and signals too, plus I could map fairly easily every manhole cover etc...

So why did I change it from tertiary to Primary? Well several factors coalesced to get it done the week. One I had previously challenged the removal of Ref tags on these roads. As I had seen them in use and felt that that removal had been judgemental and incorrect kind of how people felt about me bumping up the priority of the High street. Incidentally it was the out of date ref:note tag that I referred to as out of date. As it clearly was - stating that the refs where not present when they now were again.

Two the reasoning given in the note didn't really reflect what I saw on the ground which was a busy road dispite the restriction that people try to use when it was justified.

I looked a lot at that confusion issue some have been mentioning about it being a through route and then not and I specificly added the first basic bus_gate barrier tags to help resolve the problem. ..but they didn't make it to many renderings and so people still became confused. thats when I invested 7 hours to mark the barrier in the area:highway system [that's been in the wiki-tagging "manual" in different forms for near 4 years at least.

I had a lot of need for these when mapping the relations between narrow footways and wide roads and before sidewalks became known to me as separatable as I need to make so meny fictious ways to join stuff up even if the side walk made them less than a meter apart the centre of the road could be 5 or 10meters in places so the result was just horrible to look at and dificult to relate to the real landscape at the end. I did my homework on the wiki and tried adopt the tagging being asposed for us to use for this purpose. ..but again not everyone was reading from the same "hymn-sheet" and the discord was driving me nuts with people chngeing my contributions into tag sets that broke renders and routers alike. eventually I looked up what were the latest wiki options and posed it as a question on this diary system http://www.openstreetmap.org/user/Govanus/diary/24021 and then, upon advice from an reader of the entry, to the help questions system https://help.openstreetmap.org/questions/38079/latest-fashions-in-area-highways-and-the-plan-for-wash-common-coming-soon (it does vary between the two versions a bit as I tried to refine it etc). Which promptly failed to get any answers because it was closed prematurly for being a copy and paste of the diary entry. So so much for community help and consensus. So I pick a wiki type and blunderbust it into areas for others tidy or squabble over.

So thirdly with having thought the confusion problem was solved. I looked at the anomaly of such a highly prioritised road that apparently wasn't because you couldn't drive a car through it all day. So I began importing the latest version of the Millennial Transport strategy - lorry part that should how planners wanted lorry and other access 24 hours a day. The road is used as a branching spine for lorry routes approching from opersite side of the barrier. (this process is not yet compleated for the whole city yet but about half might be complete (not such if it needs new tags etc..))

Fourthly all those previously stated reasons came into my head. Now I understand that by vehicles alone it might look less but what if those figures are weighted statistically by size, weight, road-wear, number of occupants in each vehicle and traffic jams of 12-9 buses with some vans thrown in is a bit more substantial than 12-9 cars in a row. Incidentally when I came into the library to write all this and do other stuff I clocked that that the traffic on Longwall backs up from the southern lights by the high street as far back as the Marston Cycleway at the corner by the parks....

So I wasn't impressed that it had been still redemoted for being "confusing" and an anomaly. There's less traffic at the top end of the A4158 (Iffley Road) especially after the junction with the B4495. I also don't believe that so much traffic from the A420 goes straight down the A4158 for instance - a Good chunk finds its way onto the A4165 and the A4144.

Really if its such a massive contention why net let the highways authority sort it out - and yes they picked to make and keep it as a primary now if British B-roads (originally 25% non-local traffic) are tertiary so that the Supped Expressways like the E05/A34 are primary and the other minorised A roads (original 50% or more nationally funded for a lot of non-local traffic) then make some secondary but I loby for the B480 to go to a secondary A-road then. The B480 has so much built around it since the 1930's began. but then it local to non-local traffic not busyness that determined that after all there B road dual carriageways about the land and on the side of southpark in Oxford there is the remains of a white unclassified dual-carriageway Warnford Lane(to many students park in it to work properly now).

So when local people find the tertiary setting an anomaly and very confusing to use the because of it. I might just crack under the evidence and start taking action.

I'm happy to debate it, but I would like to remind renderers that make road maps that they should add barriered streets for the benefit of there users as this is how the whole Transport control system is designed to work around in Oxford at least. The adage is you can drive virtually anywhere but just not through anywhere.

The other big point I reserlution on is an agreead area mapping standard that we all agree on implement or doge and so things and just be mapped in peace and the OSM mapping system improved.

Oxford routes. 22 days ago

thanks for the link. Haveing read it I think they haven't given a substantial rebuful to the above it what I could read there. It seems that just becuse a few buses cycles and pedestrians use it some how dosn't count to be rated highly. The vey high volume of remaining traffic as mentioned above and the factthat the councils are looking to build a tunnel to take just the buses alone shows that it isn't a good enough reason to demote the route.

personal link about 1 month ago

https://www.gov.uk/government/publications/ngec-regional-prime-maps

personal link about 1 month ago

https://www.gov.uk/government/publications/con29m-coal-and-brine-area-polygon-data

personal link about 1 month ago

https://www.gov.uk/coal-mining-records-data-deeds-and-documents#historic-data

Learn-a-tag: driving_side=left/right 5 months ago

Having looked at jackfifield's picture it feels more like a private site as the signs used by the operater are not correct to the highway code.

The problem is that both the one-way and the no entry would apply to an entire carrageway regardless of which side of the road they were placed.

In fairness as its private land the operator could invent there own signage (like at some highway service stations and shopping parking roads, though comprehension is usally best served by a degree of conformity.

In the south of Oxford, UK there is a business estate that had a roundabout at the end of a front drive. As the front drive was laid as an all-purpose dual-carageway (4 lanes in all) the sidewalk that ran around the roundabout went though a very short paved path at the end of a bush (growing down the central reservation) as the plants grew up tall the site operater erected a sign faceing the oncoming traffic approching the roundabout (from the outside the park), that I guess was suppose to warn people about pedestians.

Unfortunatly the sign chosen was a UK version of No Pedestians! So as it faced away from pedestians going to cross it was ignored by them and sent the inverse message to any drivers who could be justified to expect no pedestians crossing beyond that point. Needless to say it was just ignored.

The site is at the rounderbout here (drive comes in from the north):- https://www.openstreetmap.org/#map=19/51.73092/-1.20747&layers=D

Learn-a-tag: driving_side=left/right 5 months ago

Having looked at jackfifield's picture it feels more like a private site as the signs used by the operater are not correct to the highway code.

The problem is that both the one-way and the no entry would apply to an entire carrageway regardless of which side of the road they were placed.

In fairness as its private land the operator could invent there own signage (like at some highway service stations and shopping parking roads, though comprehension is usally best served by a degree of conformity.

In the south of Oxford, UK there is a business estate that had a roundabout at the end of a front drive. As the front drive was laid as an all-purpose dual-carageway (4 lanes in all) the sidewalk that ran around the roundabout went though a very short paved path at the end of a bush (growing down the central reservation) as the plants grew up tall the site operater erected a sign faceing the oncoming traffic approching the roundabout (from the outside the park), that I guess was suppose to warn people about pedestians.

Unfortunatly the sign chosen was a UK version of No Pedestians! So as it faced away from pedestians going to cross it was ignored by them and sent the inverse message to any drivers who could be justified to expect no pedestians crossing beyond that point. Needless to say it was just ignored.

The site is at the rounderbout here (drive comes in from the north):- https://www.openstreetmap.org/#map=19/51.73092/-1.20747&layers=D)

Latest fashions in area highways and the plan for wash common coming soon. 8 months ago

I have thought alot about sidwalk inclusion and find it more useful for the thisgs I do and map wich relate to pedestrian facilities and class seperated roads with borders seperated and mixed cycleways and diversions into footpaths. I find adding to much to a way is going to be a problem going forward as to implement the current parking scheme I already will need to chop some short residential streets about 30-50 times for each offical change in parking allowances along its length If start choping it for other small features its going to become very unweildly and considrably less useful to routing that has to wade though more routeing ways when one cleaner one would have been faster to use. I think relationships connected feature in the areas descibed in the article would help glue all of it back into a logical data entity.

Thinking carefully about backwards compatibleity I concluded that if a system wasn't expecting something new then it wouldn't hurt too much to put a new thing where the old system wasn't expecting it to be on the basis that a system adapting to find the new thing could as happily look for it in the new place when it was in a form an old system could understand its format ie only taged ways nodes and relatiions still.

I'll pass this up when transfer it to the help sys soon i'mout of timetoday...

is this political...? about 1 year ago

Looking at the map as it is on OSM today the La Linea de la Concepcion marked on the west side of the pininsular reclaimed harbour works in Spain running suspicious along close to a boundry into the sea open water in the waters controled by the Gibraltarian Goverment. The site is frequentend oftern by Royal Naval ships and submarines and so the idea of their being nothing going on in the water around there would be week Naval vesels have been to police

indoor and street gazettering about 1 year ago

I think for meny you are probably right. I try to use josm offline (no picture tiles work over) then was used to reworking the alignments online to the arial data, which is where I became unstuck.

I also found little things about needing to get the latest java to run on a generaly not upgradeing machine a niggle to me too.

I've been meaning to write a vector graphic editor for a while so it was good reason to develop a new interface for working in 3D OSM a lot of the time.and also worked simply with cached versions wms tiles

I'm not to sure but I had some difficulties with using the sorting seach in JOSM but I'm now about 700 versions behind as JOSM changes so fast.

In general I think you are right that a new plugin or interface extension to JOSM would solve it for most people. and maybe I'm wrong for tring to foce it to work offline (with no available internet). I use library and work machine to transfer to the OSM system.

Good luck with your mappings,

With kind regards,

Govanus

Night hospital'er about 1 year ago

I've been using the indoor: system to Gazetteer flats in blocks either by outline separated with their corresponding layer/level data or just by entrance nodes were outlines are not clear to me yet.

In basic rendering the view is generally ignored unless you have a renderer that knows how to make a building plan in various possible styles from flat to isometric. big map renderers generally gloss over and ignore these but take the data into a spreadsheet or database and it gives a good postal listing like those that Kelley's used to make. As a data user this is more useful for me than say a number range especially when I try to add office or owner data into the Gazetteer which due to it being in osm allows it to be automatically to custom maps. It also helps navigation when the dwellings aren't sequentially numbered which is actually quite common (from both alterations and odd/even splitting schemes as well as say something that tries to combine floor and unit numbers from to separate ranges. The Gazetteer form just cuts though all of this and makes it straightforward and easy to work with automatically

Josm hasn't been so easy to work with for this needing a layer approach workflow to work and the online editors are very difficult to use with stacked features (think of a 40-story block of flats or halls of residence - the baseplan for most floors in the middle are likely to be very similar with most principle doors and loadbearing walls being in the same horizontal place on each one, with just layer/level/ele/height data to split them up). I have been thinking about writing a new editor to make things easier but I've had so much going on recently that I've surveyed a lot more than I've had time to add to osm lately, especially when I have to work in complex way without online editors, to get the data clear and without unwanted clashes of wrongly joined nodes.

Customised units and dwellings with variant layout features are also easy to handle like this.

-- In general Gazetteering is more useful for data users than large-scale mapmakers though interactive maps could highlight the important sub parts or jump into a building layer mode.

I read that is increasing use of osm to work in 3D on the wiki and the like. Describing where the dentist is on the 4th floor on the northern wing corridor, with these is easy this way.

=

In general I am adhering to the principles elsewhere expounded in OSM that: -the information provide and stored should not restricted or specially overly tailored to a specific purpose like a certain rendering engine or flight-sim such that it less useful for other users purposes. -If you don't like the rendering, change the render settings not the data per say, -if the render can't understand the difference between things add more differentiating information to OSM and not just remove things or worse deliberately remove detail such as for ways so they "store better on system x because no one renders under [arbitrary] zoom level y"!

It can be tough but it helps to make a general-purpose data set that can be used by so many people in different ways.

osm text yet to be done completed and a quick method to add direct from text over 1 year ago

I tried adding them today partly fixed to fit the online data in terms of alignment alas my boss was rushing me to close and I had only just started the new josm version and thought I'd save all edits but only the active was saved my mistake for not having time to read all the closing dialog boxes.

I'd hopped to get the data side fixed in then have time to algien before the upload but this month has been to busy and hadling the foybles of trying upgrade java and josm in time I didn't really have spare didn't help either.

The result is half of the suvey data transcribed and un aligned to exsisting well half still on scrap notes, and the parts uploaded contain good data but have deliberate miss shapes to show where the alignment process with a satimage is required (they arn't currently available offline to me.

the other areas I've been working on is aspley rd and more up cross street mordlen rd some in st clements and layer work in southfield park (I'd have got a lot fin if I needed to finish tax and other finacial things in another sphere).

I will tidy the shocking quadralatuals soon...

Call to map Misery over 1 year ago

I might be able to photo scetch some of the thing visable from the air if that helps though someone would need to visit for the topigraphical details and to verify any offsets that should have been available to handle the genral displacement errors in saterlight picutre tring to show above from a partly sideways view.

osm text yet to be done completed and a quick method to add direct from text over 1 year ago

well that didn't seem to wook either tag confusion I guess.

I had another upload crash afewvhours back over 1 year ago

Sorry for the long reply time I don't always have frequent access time with the internet as I mainly work when I visit a library, though I can also be using, when I get spare time (which is currently not too often), a non-internet connected computer running an old version of josm.

Yes I seem to have been getting various upload problems since September. I think I thought it had to do with new use of Relationships from online editors. So I mentioned it on the diary and someone looked onto it for me and found no records about it in the logs. I changed how demanding my use had been for setting up relationships as I been attempting to create totally new node & way sets with new nesting relationships which seemed to be able to get the uploader tied up trying to hold so many links to checks to things that didn't exist yet on osm (a merge process maybe). I worked into short groups of upload new ways before a separate changeset to make each relationship one by one. This seemed to work ok with the occasional hiccup. More recently I did get a browser struggling to keep up with processing a large up date of about a 130-odd new things made in id-editor and handling sort routines for looking up new entry options. As it was still busy when it tried to save it made the nodes have messed up co-ordinates, I ran out of access time before I could fix them all. The latest on was a bit of a mystery as I did edit in chrome/id on one machine changed to another with similar set up and the OSM server faulted with a error page after 20 min twice. before I had to abandon the session.

So I general I am getting quite used to it not always being a 100% reliable for uploading edit sessions.
Gaussian Processes: reconstructing the surface terrain of the UK from OSM "ele" data over 1 year ago

This is a useful reminder of how diversely OSM data is used I try to include ele a bit more when I have a good source for it.

Tonights work on rail ref's over 1 year ago

bugun to refix but isn't finished yet a shame so much lost.

GeoVRML over 1 year ago

the evolving to version 2 (http://www.ai.sri.com/geovrml/2.0/)