Have you seen TagCentral, my proposal from SOTM2010? http://wiki.openstreetmap.org/wiki/SotM_2010_session:_Tag_Central:_a_Schema_for_OSM

Also, Ilya Zverev is working on something sufficiently similar, it may be called TagCentral. I don’t know where this is now, I heard from him in December (2012) http://wiki.openstreetmap.org/wiki/RU:Catalog/Zverik

Have you considered one of the books about OSM, which do indeed collect a lot about OSM in one place, for example… http://www.openstreetmap.info/

@Robert Whittaker: the map data is being updated from OSM from time to time; it’s an OSM XML file which is downloaded via JOSM and the same file is used to propagate changes back up to OSM in exactly the same way as any other editing activity. We expect to update at least the area around any change before making the change, so there’s no merging involved other than the normal way you’d handle editing in general. The rendering is direct from the OSM data; my renderer allows for remembered manual repositioning of captions and icons (it is a sufficiently small area that this is practicable), but there are no modifications or transformations to the actual data - all the university data is in OSM centrally. Of course the renderer doesn’t render everything - it is focussed around University properties which it separates out using the operator, ref and other tag combinations, plus a selective list of third-party features they want to show.

Oh, and I have used and there is provision for alternative names, such as ‘Grad Pad’, but again within the University, not for general locations in Cambridge.

Hmm. It isn’t intended to be a general search tool for pubs. It’s for locating university institutions (and only in the Cambridge area, of course). You won’t get a hit searching for ‘The Eagle’! Cafe should work, but only for University cafes of which there are three or four, but as part of the name, not abstract conceptual terms like ‘coffee’.

Anything within buildings was deemed outside the brief for the project. I agree with you about finding rooms, and in fact it wouldn’t take much technically to add an index entry to identify location at least by building an entrance, but the cost of gathering and entering the information would be large.

@EdLoach it’s a relatively small area (8x8 km^2) that all the tiles are generated in advance, so the access is directly to images, no mod_tile. So server-side caching isn’t relevant.

@Richard, thank you!

The old maps used sections from an Adobe Illustrator file used to create the published paper map, which were indexed by pixel co-ordinates in the index (which was a hand-edited XML file). The origin of the base map seems to be lost in the mists of time: Cambridge University Press claims the copyright.

See June 2012 update at http://www.openstreetmap.org/user/davidearl/diary/17050

@robert: it isn’t updating automatically and unconditionally. There’s a preview step before publication that provides a cross check.

Thanks for that Alexander - I'll contact you directly so we can keep in touch more directly than via diary pages!

The volatility issue is why I chose to include explicit reference IDs to identify buildings, sites and entrances (and it is also the recommended method). And it helps that the University has a ID scheme for sites and buildings already to draw on (I have in effect extended that to the colleges).

The IDs and URIs will be equivalent, so they function just like yours, and like you I expect to be able to produce site maps like your science-area one. I'm not planning on using RDF though, merely linked tables within a conventional database.

Welcome.

Thousands of people read these - all diary entries are available as a fgeed at http://www.openstreetmap.org/diary and those are further aggregated with other useful OSM-related blogs and posts at http://blogs.openstreetmap.org/

David

Better still, have a separate feed for each language which puts posts in other languages through e.g. Google's translation tool before supplying them (but then if you use one of the feed aggregators which runs in a browser, you can translate them on one click anyway on the whole).

Just seen rsh's comment. Yes you can use that method, but you won't need to if modified times are preserved properly. I'll try to write this up.

Hi, I wrote the audio stuff in JOSM.

It does indeed display markers at relevant points. The intent was originally to support continuous audio soundtracks so you didn't have to keep stopping and starting the recorder. There is comprehensive help for this if you look in the help.

For individual snippets, this is less well documented, and not so well tested, but I did work with someone who was recording in this way to get this working. You can import audio as for a single soundtrack but select multiple files. It then uses the timestamps on the file (the modified time = the end of the recording) and applies those to the track. Of course it is vital that the modified time stamps are accurate and preserved in whatever method you use to get them off the recorder. You still need to synchronise the audio layer so the time coincides with the GPS time, and you should check your audio recorder has an accurate clock and if not compensate for it in the calibration setting (see help).

I don't have the means to generate time stamped individual WAV files, so it's hard to test. If you are having trouble, contact me directly and I'll help.

David

Wikipedia claims there was one for every week of the year.

http://en.wikipedia.org/wiki/Whittlesey#Public_houses

Another one has closed down since that list was compiled. There's a few more to collect - we've not done the centre yet.

> It boggles me, however, why no OSM search tool has
> searching for intersections implemented

Perhaps because Nominatim and NameFinder weren't developed by Americans.

The idiom "First and State" to mean the intersection of First Street and State Street is pretty much unknown in British English (perhaps because we tend not to have streets in grid patterns). Likewise omitting "street" etc in this and other contexts isn't widespread because it is confusingly commonplace to have an Orchard Close, Orchard Street, Orchard Avenue and Orchard Road all within a few streets of each other.

Now you say it, it's obvious, but it had never occurred to me as a Brit.

David

> It boggles me, however, why no OSM search tool has searching
> for intersections implemented

Perhaps because Nominatim and Namefinder weren't developed by Americans. The idom "First and State" to mean the interscetion of First Street and State Street does not exist in British English (perhaps because we tend not to jhave block oriented street patterns) - nor is omitting the word "street" in the UK in this and other contexts because it is commonplace for there to be an Orchard Close, Orchard Street, Orchard Avenue and Orchard Road (for example) all in the same town, often close together.

Now you mention it, it is clear, but it wouldn't have occurred to me as a Brit.

David

Peter Miller organised some for Stratford-upon-Avon recently: have a look at the mailing list archives.

I also think this is a good idea. The routes were build around a destination system - about 10 destinations were chosen as the names to go on the signs and all the blue signs reference those and no others. Unfortunately a few key ones didn't materialize because of cross boundary political issues - by village of Fulbourn being the key missing one! These were then also designed to tie up with the Cambridge Cycle Map (http://www.cambridge.gov.uk/public/images/cycmap.pdf) so not only is the signage on the ground well thought out and consistent, it matches what the official map says too. The red squares on the map are the key destinations that appear on the signs.

What we can't do though is copy the routes off the map because it is Crown Copyright. If we were doing the map again now, I'd be suggesting to the authorities that we use OSM, but it predates OSM coverage in Cambridge.

I think the main problem is what to call them, because they aren't numbered. They are, in this respect, more like the Paris metro than the London Underground - Paris names its tube lines by the end points.

There is also a subnetwork of even more local local cycling and walking routes signed on the ground in white on black signs and marked on the map in orange.

Another way of doing this in JOSM is to select an item with the group of tags you want to apply to other objects, Copy (CTRL+C), select the target objects (you may want to create them as you go) and Paste Tags (CTRL+SHIFT+V). The tags of the copied item are applied to the new/newly selected item. If you want to replicate a point and its tags, Paste (CTRL+V) is what you need, and the Paste takes place at the mouse position.